|
Eager Space | Videos | All Video Text | Support | Community | About |
|---|
2023 looks to be an exciting year in space. How well can you predict what will happen?
Play the space prediction challenge, and if you're the best, you might win a prize of nominal value...
Grab a piece of paper and a pen to write down your answers, and we'll get started.
We'll start with starship
#1 - the number of successful orbital launches, where success means "starship gets into whatever trajectory SpaceX was aiming for"
#2 - the number of superheavy landings at the launch site, where success means "the booster could be reused with a reasonable amount of maintenance"
#3 - the number of successful orbital landings of starship, where successful means a soft landing at the designated launch site.
#4 - the number of full stack launches from pad 39A in Florida
2023 promises launches of rockets old and new.
Write down your predicted successful flight counts for each of these launchers
#5 - Falcon Heavy
#6 - Vulcan
#7 - SLS
#8 - Ariane 6
#9 - Terran 1
#10 - RS1
#11 - Electron
Number 12
Electron might be the second rocket featuring reflown boosters.
How many electron flights will feature boosters that were recovered and refurbished?
There are two suborbital companies that may fly humans in 2023.
13: How many crewed flights will Virgin Galactic's SpaceShipTwo complete in 2023?
14: And how many crewed flights will blue origin's new shepard complete in 2023?
15: Rocket lab is building a new engine known as Archimedes to be used on the neutron launcher.
Will Rocket lab test a full version of their Archimedes engine at Stennis space center in Mississippi in 2023?
Answer yes or no
16: Blue Origin has been making progress on New Glenn.
Will New Glenn flight hardware will be place on the pad at launch complex 36 on cape Canaveral space force station?
Flight hardware means hardware that could actually perform a launch. It could be a flight booster with engines or a full stack.
Pathfinders do not count.
17: The number of successful moon landings...
Lunar lander success will be measured by successful contact with the lander at least one day after the landing. Here are the entrants...
Japan's ispace has the Hakuta-R lander. Intuitive machines has their Nova-C lander, which current has two missions planned for 2023, and Astrobotic has their Peregrine lander.
Both the Nova-C and Peregrine are part of NASA's commercial lunar payload services program, or "clips"
That's four possible landings.
And we are just getting started...
The Indian space agency ISRO will take another shot at landing their Vikram lander, which crashed in a previous mission.
Japan's space agency JAXA will send their SLIM lander.
And finally, Roscosmos will send their luna-glob, or luna 25 lander.
That's 7 possible landings. How many do you think will be successful?
To play the game, submit your answer in a comment to this video
The comment should be in the following format, with the number of the question followed by your guess. If it's not in this format, I may decide that it's invalid.
For numeric questions, your score is the difference between your guess and the actual answer. You guess 4, the answer was 2, you score two points.
For yes/no questions a correct answer is 0 points, and incorrect answer is 1 point.
Lower scores are better.
"Flight" means a successful flight - the vehicle was delivered into the intended trajectory. Partial success counts as a failure.
I may need to make some judgement calls about success. The decision of the judge is final.
Comments submitted after January 31st are invalid.
If you enjoyed this video, please put $10 on red. C'mon baby, daddy needs a new lunar rover.
What are gravity losses and how do they impact rocketry?
We're going to be diving into some of physics of rockets, but it's fairly simple and you don't need to understand all the details.
Let's say we are launching a Falcon 1...
We know that gravity is pushing down on the rocket at 9.81 meters per second squared.
At T=0, we fire up the engine and it starts producing thrust that pushes the rocket up. What happens?
We need to go into a bit of physics...
Remembering Newton's first law of motion, force equals mass times acceleration. In this case we know the force of the engine and the mass of the rocket, so we want to rewrite that equation as acceleration equals force / mass
Let's say that our engine produces 250 kilonewtons - or 250,000 newtons - of force and the rocket weighs 26,700 kilograms. Plug that into the formula, and we find that the engine will be producing 9.36 meters per second squared of acceleration.
9.36 is less than 9.81, so gravity is pushing down harder than the engine is pushing up. And nothing happens.
The first rule is that the upwards acceleration from the engine must be more than the downward acceleration from gravity to get off the launch pad.
Luckily, the Merlin 1A engine produces 343 kilonewtons of thrust, giving an acceleration of 12.85 meters per second squared. That is more than 9.81, so the rocket will move.
We can figure out the acceleration by subtracting the gravity from the upward acceleration, and we get 3.03 meters per second squared. That is the amount the rocket actually accelerates, the rest of the acceleration from that engine is lost to gravity. That is what a gravity loss is.
How have things changed 1 second into the flight? The rocket will be lighter because of burned propellants, but how much lighter?
That depends on the mass flow rate, which is the amount of propellant burned every second.
There are two different ways of calculating mass flow rate.
In the first method we calculate the mass flow rate from other factors. We will be talking about exhaust velocity; if you want to learn more, here's a link to a video on exhaust velocity.
It is simply the thrust in newtons divided by the exhaust velocity in meters per second. You can determine the exhaust velocity by multiplying the ISP by g, or 9.81 meters per second squared. I'll be doing the math here with detailed units so that you can see how it works.
Using the raptor as an example, it has a thrust of 1800 kilonewtons or 1800000 newtons, though that's hard to nail down as the engine is under development. It has an Isp of 333 at sea level.
Doing some math and cancelling out the units, and we determine that a raptor with 1800 kilonewtons of thrust uses 551.6 kilograms of propellant each second.
This equation is often used in reverse; if you measure the thrust of an engine and the mass flow rate, the exhaust velocity and Isp can be determined.
In the second method we use information about the rocket.
If we know how much propellant the rocket holds and how long the engines burn, we can come up with an average mass flow rate.
The Falcon 9 first stage holds 411,000 kg of propellant, and on a Starlink launch it burns for 158 seconds.
It therefore averages a mass flow rate of 2600 kilograms per second.
Now that we understand mass flow rates, we can answer our question about what happens after one second.
Based on the 343 kilonewton thrust and the Isp of 255, the mass flow rate for the Merlin 1A is 137 kilograms per second, and therefore the mass is reduced from 26,700 kilograms to 26,563 kilograms.
That increases the acceleration to 12.91 meters per second squared and the net acceleration to 3.1 meters per second squared.
This process will continue as more fuel is burned off. At 152 seconds, the last of the fuel is burned.
At that point, the mass is only 5876 kilograms, giving an acceleration of 58.37 meters per second squared.
In addition, the rocket has reached an altitude of 100 km, and that is far enough away to reduce the gravity from 9.81 meters per second squared to 9.51 meters per second squared. Do the math, and our net acceleration is 48.85 meters per second squared, or about 5 Gs of acceleration
That's how the math works. There are a few factors I ignored:
First, engines are often throttled down to a reduced thrust level to reduce the stress on the structure of the vehicle. This is very common before "Max Q", where the stress from drag is the highest.
Second, engines get more efficient as the move out of the atmosphere. For the Merlin 1A, the thrust will increase from 343 kilonewtons to 409 kilonewtons in vacuum and the actual acceleration will be around 6 Gs
Third, engines are sometimes throttled down at the end of their burn to keep the acceleration from being too high for the payload or vehicle.
Here's a chart that shows what is going on, with the red section showing the gravity losses. You need to pay the gravity tax before you can get any useful work accomplished.
If we add up the total gravity loss for this booster flight, it turns out to be 1489 meters per second.
Is there a way to reduce gravity losses?
We can switch to the Merlin 1C engine, which produces 423 kN, or 23% more thrust than the Merlin 1A. That nearly doubles the net acceleration.
Here we can see the effect. The Merlin 1C burns the propellant faster so it's done in only 125 seconds and only wastes 1216 meters per second on gravity losses. That is great.
Then some bozo comes along and says, "you know, we can upscale the rocket to hold 23% more propellant and take advantage of the higher thrust to launch bigger payloads", and that extends the burn time back to where it was and you end up losing about the same amount to gravity losses.
The summary here is that gravity losses are just something you have to deal with.
Orbital flight adds more complexity.
This is a classical thought experiment known as Newton's Cradle. It involves shooting a cannonball horizontally from a cannon on a very high mountaintop.
What we see is that if the cannonball has a lot of horizontal velocity, it will take some time before it will impact the earth because the earth curves away from it.
In this right image, we see if that horizontal velocity is high enough, the rate at which gravity pulls the cannonball closer to the earth is exactly balanced by the rate at which the earth curves away. That is what it means to be in orbit.
How does this relate to gravity losses?
This is a velocity chart of a Falcon 9 on the Crew 1 mission to the international space station.
The black line shows the total velocity; you can see that it starts slow but increases as the first stage burns off fuel, then flattens out as the second stage takes over.
We can add two more lines to show the vertical and horizontal components.
The blue line shows the vertical velocity; note that the first stage builds up a lot of vertical velocity, and then that vertical velocity bleeds away from gravity losses as the launch continues. This makes sense as the goal when launching to a specific altitude is to reach that altitude with zero vertical velocity. The first stage also builds up some horizontal velocity, but not that much.
The second stage then primarily handles adding enough horizontal velocity to get into orbit.
Knowing what we know about Newton's cradle - that the faster we go the more the earth curves away - how does that effect gravity losses?
This chart shows the horizontal velocity in green and the gravitational constant in yellow. The gravity starts at 9.8 and goes down to 9.2 as the launcher gets farther away from the earth - that is the same effect shown in the previous charts. It's a small effect, only about a 6% reduction.
Now we can add in the effect based on velocity.
The red line shows the effective gravity loss - the amount of acceleration we need to stay at the same altitude. As we go faster - as we near orbital speed - it reduces significantly.
For a flight into orbit, we would expect that the red line would go all the way down to zero. But for crewed flights, the capsules are typically launched so their speed is just a little bit less than orbital speed - about 250 m/s less - so that if there are problems with the capsule's thrusters, the capsule will reenter naturally.
That's the story on gravity losses
(read)
A huge thank you to Declan Murphy for FlightClub.io, a wonderful reference if you want to dig in deeper into these sorts of issues.
When the rocket is tilted, part of the acceleration is up and part of it is to the side.
In this diagram, gravity is pulling down at 9.5 meters per second squared, and the rocket is thrusting forward at 15.8 meters per second squared.
Doing a bit of math, we discover that the rocket is accelerating up at 13.7 meters per second squared and to the side at 7.9 meters per second squared. The upwards acceleration is countered by gravity, producing a net upwards acceleration of only 4.2 meters per second squared.
Before we dive into things, I have two links to share with you.
The first is my video on Space Shuttle abort modes, as it's very useful to understand what the options were for the shuttle.
The second is a truly wonderful book by Rand Simberg titled "Safe is not an option - How a futile obsession with getting everyone back alive is killing our expansion in space".
His main point is that we know that death is a possibility in many human endeavors - driving, flying a private plane, skydiving - and also within many jobs - commercial diving, fishing, research on antarctica, and the military. And we, as a society, have learned to deal with it. That's not to say we don't try to make those activities safer, merely that risk is inherent in life and many of these things are worth doing.
But somehow space - and particularly NASA - has decided that astronauts are different and that we cannot take risks. Perhaps the best example of this is the cancellation of the final Hubble servicing mission in 2004 because of safety concerns. The Hubble is one of NASA's priceless assets, and after significant lobbying and a change in director - some of the lobbying by astronauts who understood both the risk and the benefit - it was reinstated.
The ebook is only $5, and it's well worth your time. So go read it now, and I'll wait.
All done? Okay, lets get started.
If you look around, you will find a lot of explanations why airlines don't have parachutes for passengers that explain how impractical it would be and how most accidents wouldn't provide time to use the parachutes.
All those points are true.
But they largely view parachutes as pointless, and they are wrong in that.
Here's a video of how airplane evacuations are supposed to happen, during an evacuation test of the Airbus A380.
Now ask yourself, what would happen if 10% of those people were wearing bulky parachutes and tried to make their way off the plane?
Aviation disasters are very well studied, and we know that the time it takes to get off the plane can be the difference between life or death. We also know that passengers do not follow instructions, there are documented cases where people have died because they inflated their life vests inside the plane and could therefore not get out the exits that were slightly under water.
Parachutes would kill passengers who otherwise would have lived.
The point here is that safety changes can be good and they can be bad. We can express this in numerical terms.
We can look at how much better our mitigation is, multiply it by the chance of that scenario coming up, and get an estimate of the safety improvement
We can look at the problems our mitigation might cause, multiply it by the chance of that scenario, and get an estimate of the safety loss
Applying this to our parachute scenario with some made-up numbers, let's assume that the parachute can save 50% of the people that would otherwise die, and the chance of that scenario is one in one thousand. That gives us an improvement of half of one in one thousand, or one in two thousand.
Let's assume that the scenario where parachutes slow down evacuation only results in 10% extra deaths and the chance of that is one in one hundred, which would results in a reduction in safety of one in one thousand.
The point being that safety losses in other scenarios can outweigh the safety gains in the scenario you are trying to address.
Let's use SLS as an example, with the orion capsule on top, and some made-up numbers.
SLS is 99% successful and blows up 1% of the time, so the survival rate is 99%.
Take the Orion capsule and add a launch escape system to it. Let's assume that system is 90% successful.
So, we can take the 1% chance of needing the launch escape system and the 90 % success rate when we need it, and figure out that we get an increase in 0.9% in the survival rate, pushing us up to 99.9% total.
Abort systems are great.
Now lets look at the non-abort scenario. For reentry and landing to succeed, the abort system needs to be jetissoned from the capsule.
Let's say that works 98% of the time and fails 2% of the time.
We can take that two percent chance of failure times the 99% of the time an abort isn't needed, and that will lead to a loss of the crew 1.96% of the time, reducing our overall survival rate to 97.94%
Abort systems are terrible.
We end up with a very important point
This leads to an important general principle:
Small failure rates in common scenarios can exceed big improvements in rare scenarios
And the follow on:
Abort systems must be extremely reliable in the non-abort case
This is one reason the Crew Dragon does not a launch escape tower - the tower failing is no longer a possible scenario
Here we have our space shuttle stack.
In a normal mission, both the solid rocket boosters and the three main engines are ignited on the launch pad. The solids burn for about 2 minutes, and the main engines burn for a little more than 8 minutes.
At that point, we reach Main Engine Cut Off, or MECO, and we're in orbit.
The aborts are all about what happens when something doesn't go quite right. The first thing to note is that there is no abort if the solid rocket boosters fail.
The aborts are all based on one or more main engines failing or underperforming, and the options depend upon when the engine failures happen. The abort options and timings depend upon the destination of the shuttle and the amount of payload it is carrying. The chart that I'm showing is from STS-116, which was a mission to the international space station.
We'll start by looking at the cases where a single engine fails, starting at the end of the launch and working our way back.
If the engine failure happens after about 6.5 minutes, there is no abort required. The remaining engines will burn longer and the mission proceeds normally.
If the engine failure happens from 4.5 minutes to 6.5 minutes, the orbiter can not generate enough speed to get into the desired orbit but it can get into a temporary orbit, so the Abort To Orbit - or ATO - abort option is used. This option allows time to decide what to do.
If the orbit is close to the desired - or nominal - orbit, the Orbital Maneuvering System engines can be used to raise the orbit to the nominal orbit, or at least one that is good enough for the goals of the mission.
If continuing the mission is not possible, the flight is converted to the Abort Once Around - or AOA - option. The shuttle gets into orbit and then reenters before completing an entire orbit. Depending on the orbit that the shuttle was aiming for, they would land at Edwards Air Force base in california, Holloman air force base in new Mexico, or back at the launch site in florida.
If the engine fails from 2.5 minutes to 4.5 minutes, the shuttle cannot generate enough speed to reach orbit, so the transoceanic abort landing - or TAL - option is chosen. Depending on the orbital inclination of the mission, different landing sites were used.
A mission to Hubble would be to 28.5 degrees, so their TAL landing site would be Banjul International Airport in Gambia.
STS-116 was to the international space state would be to 51.6 degrees, so their prime TAL landing site was Zaragosa Air Force Base in Spain, with possible backup sites in France, Spain, and Morocco, and the capability to land at other emergency landing sites.
TAL aborts can also be used for systems failures unrelated to engines, such as propellant leaks, cabin leaks, or cooling inssues.
If an engine fails in the first 2.5 minutes of flight, there isn't enough energy to do a TAL abort, so the return to launch site - or RTLS - option is chosen.
Getting back to the launch site is complicated. The orbiter is currently moving away from the launch site, so it will need to perform a powered pitcharound with the engines running so it is pointing back towards the launch site to kill the velocity towards the launch site. Then it will perform a powered pitch down to get the nose at the proper angle for gliding.
At that point - in the highlighted box - the shuttle needs to be at the proper altitude, the proper velocity, the proper angle and be pointed at the proper direction. The main engines can then be shut off, the orbiter detached from the external tank, and the glide to the runway started.
To make it a little more complicated, it's not safe to separate from the external tank if the external tank has more than 2% of its fuel remaining, so it must also use up its fuel before it reaches that point. That may require it to do a maneuver called "lofting", where it flies a high flight path to waste fuel.
It's not clear how feasible RTLS was. In 1980, STS-1 commander John Young said, "RTLS requires continuous miracles interspersed with acts of God to be successful". RTLS was improved after that, but it still remained iffy.
With two engines out, we are into what are known as contingency abort. Quoting, "Contingency abort procedures are executed when multiple main engines fail or suffer a performance degradation that results in the loss of all other intact abort options.
There are both nominal and transoceanic abort landing options, but they have shrunk and there is no longer an Abort to orbit option as the shuttle can't generate enough speed to make it to orbit.
DRP stands for Droop Guidance. There is a requirement that the shuttle stay above 265,000 feet (80 km) at high speeds to keep the external tank from overheating, and if the computers predict this will happen, they will enter droop guidance. The computer will rotate the shuttle into the "stand on the tail" attitude to minimize the sink rate. In this case, TAL should be achieved, but droop guidance may enable at other times.
If both engines fail before 5.5 minutes, we will enter ECAL or BDA abort. ECAL stands for "Eastern Coast of North America" (the NASA docs define that with a straight face, and basically is a serious of airports where a shuttle might be able to land. Because this flight is to the ISS, the ground path of the flight is up the easter seaboard and therefore ECAL is the contingency option of choice; if the launch was more to the east, an abort to Bermuda (or BDA) would be the only option. If this situation is detected, the software will command an unguided 45 degree yaw (turn) in the direction of the landing site; the is intended to make it easier for the shuttle to reach the landing site. The nose will be pointed up to prevent the vehicle from sinking too quickly.
If a two-engine failure occurs during this period, the crew will consult their documentation to choose an appropriate ECL site.
Finally, if two engines fail in the first two minutes, RTLS will be initiated immediately and proceed in the same manner as 1 engine out.
There are other flight rules that may lead to an abort; these happen when a system fails that would compromise part of the mission.
If the orbital maneuvering system cannot function, orbit cannot be achieved, so a TAL orbit will be used
If the auxiliary power units or hydraulics are failing, either RTLS or TAL will be chosen based on getting the orbiter back as soon as possible.
If there is a cabin leak, the cryogenics are failing, or the freon cooling is failing, the same rule applies.
If the two main electric busses are failing, the rules say to do RTLS but my guess is that it's either RTLS or TAL
Finally, if one of the windows breaks, reentry is no longer possible so RTLS is the only option.
Finally, we reach the 3 engines out aborts. These are the most speculative and depend significantly on exactly when the failure happens, what the payload is, etc.
There are small nominal, Transoceanic abort landing, and ECAL/BDA options if the failures occur after 6 minutes.
Just before that we have what is known as a black zone.
The NASA documentation says, "Black zones are regions along the ascent trajectory that may not be survivable".
Black zones exist when:
Exceeding 470 knots airspeed during entry due to the inability of the flight control system to maintain control
Exceeding 4.2 g during entry, due to limits on the strength of the Orbital maneuvering system pods
Three engines fail when the SRBs are firing. This may lead to failure at the external tank and SRB attach points, poor separation dynamics (the ET may hit the orbiter), or a center of gravity beyond limits that makes the orbiter uncontrollable.
Experiencing high dynamic pressure during coast, too much for the reaction control system to be able to deal with
Control surface forces exceed their maximum or the wing or tail become too hot.
You can read all of those as "the orbiter loses control and breaks apart"
This first black zone is because the orbiter dives after the loss of engine thrust and is going too fast when it levels out.
Interestingly, there is a section earlier in flight where ECAL/BDA may be possible as there is less speed and therefore the flight levels out.
There is a black zone in the first 2 minutes if all three main engines fail, as none of the RTLS maneuvers can be accomplished. Further, there are black zones if the remaining engines fail for the same reason.
That covers all of the ascent aborts.
Let's talk about reentry.
Nearly all of reentry is a black zone; there is no abort possible if something goes wrong.
Originally, Columbia had ejection seats for the commander and pilot, but that was only a solution for two astronauts and they took those out after the first few flights.
After challenger, they added a bail out option. If you get to 50,000 feet and you are more than about 55 nautical miles from the landing site, you won't make it and you can bail out. Or if you get there but there's some reason to think you can't land, you can bail out.
On this chart, the red box shows the time you have to prepare to bail out by flattening out the glide and turning on the autopilot. And the green shows the 90 seconds you have to get out.
Here's a simplified sequence:
At 50,000 feet, call for bailout, flatten out the glide with wings level at 185-190 knots and engage the autopilot.
At 40,000 feet, the cabin is vented to equalize pressure so the hatch can opened
At 30,000 feet, the side hatch is jettisoned and the crew exits the vehicle.
NASA Administrator Bill Nelson, in a congressional hearing in May of 2022...
Cost plus has certainly been problematic in Artemis, but are fixed price contracts the answer?
Listening to the current messages coming out of NASA, one would think so, but the reality is considerably more complex and that's what we're going to explore.
We'll start by talking about the plague that is cost plus...
Cost plus is simple.
The contractor tracks their cost, and the government pays the contractor that amount of money.
The contractor gets their profit from one of three sources.
A fixed fee awarded at the completion of a project, an award fee awarded periodically as work is done based on the contractor's performance, or an incentive fee based on cost, performance, or delivery date targets. These fees can be mixed together if desired.
They do not get their profit as a percentage of how much they spend - that is known as "cost plus a percentage ", and it has been illegal in government contracts for decades. There is some benefit to contractors in higher costs, but it does not result in increased fee payments.
The government guide on cost plus says that these fee structures should be used in a way that is "sufficient to provide motivation for excellence in performance".
Cost plus does have overhead, so why would we want to use it?
If there are significant unknowns in the project - if the scope or risk factors are unknown, it's generally not feasible to figure out how much a contract will cost and therefore come up with a fair fixed price. And that risk means that there is a higher chance of failure with fixed price.
Also, if there's little commercial opportunity for the product or technology, there's less incentive for a business to take on the risk of a fixed price contract.
The main cost plus program in NASA in the last decade or so was - and still is - the SLS Rocket and Orion Capsule.
I am not going to delve into the details of how cost plus relates to SLS and Orion in this video, but I do have a few comments...
Congress, NASA, and the contractors had a specific goal in mind for SLS and Orion.
Cost plus was the contracting method they used to achieve their goal.
Cost plus was not the cause of what happened.
It is disingenuous to blame a contract type for a situation when it was the congressional requirements, the detail of the contracts, and the oversight of them that was problematic.
If you look at Bill Nelson's history, you will find that he was one of the main drivers behind SLS and Orion.
I'm feeling a rant coming on, so it's time to move on to our real topic.
Which is fixed price contracting...
Firm fixed price simply means the company agrees to supply a product or service at a specified price. The most obvious use of fixed price is when a product is on the market.
It's possible to add incentives to fixed price contracts if you care about something other than price.
The government gives the following advice on the use of fixed price:
(read)
If fixed price doesn't work, then it's appropriate to use cost plus. That's why cost plus contracting exists.
There are two very common variants of fixed price contracts.
The first is indefinite delivery, or "we don't know when we'll need it". NASA's contract with SpaceX for crew flights to ISS is indefinite delivery because NASA doesn't know at the time of the contract exactly when they will need them.
The second is indefinite quantity, or "we don't know how many we need". There is generally a base quantity in the contract with options for up to a maximum quantity at specified prices. If you see contracts that say "up to $3.2 billion" that number is for the maximum quantity.
These contracts specify enough units so that the contractor can provide a good price but do not constrain the buyer to a specific schedule.
Contracts that have both of these will say "Fixed price, IDIQ".
To order and schedule delivery, the government will send the contractor a task order.
Now that we've covered the basics, let's look at fixed price at NASA.
We should start by looking at the success of previous fixed price contracts.
I asked my twitter followers what is NASA's best example of success with fixed price contracts. I got a great reply that is clearly the right answer but isn't one of my choices...
The correct answer is NASA's launch services program, which is how NASA's science payloads find a ride into wherever they are going.
It meets the "risk can be predicted with a high degree of certainty" requirement very well, and as a program, it just does what it is supposed to do. Nobody talks about it because the program is such an obvious use of fixed price contracting and competition.
I instead want to talk about the flashy programs.
My followers thought that commercial crew was the best fixed price contract, with commercial cargo coming in second.
As is often the case with my polls, it was a trick question, because neither commercial cargo nor commercial crew were fixed-price contracts, at least not in the development phase. Neither were they cost plus contracts.
Which doesn't make sense, as those are the only two contract types permitted under government regulations.
The explanation requires a journey back into history.
In 1957, the Soviet Union launched the Sputnik 1 satellite on top of their Sputnik rocket, derived from their R7 missile. This led to a lot of concern in the US, known as the sputnik crisis.
Congress wanted to do something to help the US catch up, and they passed the National Aeronautics and Space Act of 1958, which was signed by president Eisenhower on October first.
Congress wanted the US to move fast and was worried that following the usual budget rules would slow NASA down, and they therefore granted NASA a sort of budgetary superpower to bypass those rules and come up with their own rules.. You can think it as a "what if we just threw money at the problem?" option.
That power was defined in the space act, and using it is therefore called a "space act agreement"
If NASA comes across a situation where they believe that Fixed Price or Cost Plus contracts will not be effective, they can choose to use a space act agreement instead and create their own contract terms.
Like all superpowers, it requires great responsibility on NASA's part and therefore there's a defined process to use it, but it is a very useful tool when deployed effectively.
Since commercial cargo was first, we'll start there...
In the early 2000s, NASA had a problem - the shuttle was retiring in the early 2010s, and there was no US replacement to take cargo or crew to the space station that NASA had spent so many billions of dollars building. If NASA didn't come up with a solution, they would be paying for the Russian progress and Soyuz vehicles to fly both astronauts and cargo.
NASA was hoping that the constellation program would be that solution, but it wasn't clear when it would show up...
... and in line with Congress' mandate for NASA to use commercial solutions, NASA spun up a program known as the commercial orbital transportation system, or COTS. Congress kicked in $500 million to fund it.
NASA initially selected SpaceX and rocketplane Kistler as the two participants, introducing the now-familiar concept of "two provider redundancy". Kistler was unable to meet the private financing goals in the first year, and they were replaced with Orbital Sciences.
Of the $500 million allocated, $278 million was budgeted to SpaceX for Falcon 9 and Dragon, and $170 million to Orbital for their Antares rocket and Cygnus resupply module.
The contracts were structured with a series of milestones and a payout associated with each milestone. If that sounds like a firm-fixed price contract, you're on the right track, but NASA knew at the outset that this was a hard problem, and they chose to do these contracts as space act agreements.
SpaceX had a goal date of September 2009, and Orbital aimed for late 2010 because of their later start.
By late 2009, it became clear that neither of the companies were going to make their target dates, and - more importantly - they wouldn't be able to do it for the agreed-upon amounts. SpaceX simply did not have the money, and the $170 million orbital would get simply wasn't worth the effort.
Luckily, NASA, the advisory commissions, and congress all agreed, and a further $300 million showed up in 2010 - in the same legislation that created SLS - and the program could continue, with both contractors achieving success by 2014.
The program ended up a huge bargain for NASA, with their total investment at $821 million, less than the shuttle yearly budget for upgrades, and the per-kilogram costs for cargo to ISS were quite a bit cheaper than shuttle...
They were less successful for the two contractors, as they ended up putting a lot of their own money into the program, with SpaceX investing $454 million and Orbital investing $590 million. Orbital was an established company and they would happily earn their investment back in flying cargo to ISS, but SpaceX did not have significant cash reserves, and it was only a combination of Musk's Silicon Valley contacts, the launch backlog that existed at the time, and the constant work of Gwynne Shotwell that kept them afloat. It was a close thing.
NASA talks a lot about private/public partnership and I think that it's generally a good thing, but I think there needs to be an equitable balance. Both SpaceX and Orbital invested a lot of money to develop capsules that are only used for ISS resupply because a commercial market does not exist.
SpaceX did get Falcon 9 out of it and that has been a huge benefit to both SpaceX and NASA, but Antares only flies ISS resupply missions.
There was, of course, a fixed price program known as commercial resupply services for the actual resupply flights, and that seems like a appropriate place to use fixed-price.
The problem was that NASA really wanted to start supply flights early to avoid relying on the Russians, and that meant those contracts were awarded in 2008, and that meant SpaceX and Orbital needed to put in bids for flights of vehicles where they could not adequately determine their costs. Orbital had been working on the program less than a year, and SpaceX had yet to successfully reach orbit with Falcon 1
In that environment, fixed price was not appropriate, but that's what NASA decided to do. Orbital bid a price that would let them earn back their investment, and SpaceX significantly underbid because they really really needed the contract.
SpaceX would later get dinged for raising their prices significantly for the second set of CRS contracts, but they were actually just adjusting their prices based on their actual costs.
My overall evaluation of COTS is that it was NASA's flexibility to go beyond firm fixed price that made it possible, and they got lucky that SpaceX showed up.
Commercial crew also used space act agreements.
In 2010, CCDev 1 gave out a total of $50 million to 5 companies to work on things related to commercial crew.
In 2011, CCDev 2 expanded this considerably, with $270 million to 4 companies.
In 2012, the CCiCap program gave out $1.1 billion for 3 companies to create a fully-thought-out plan for how they would do commercial crew. This was a *lot* of money - the $440 million that SpaceX received was more than they received during COTS for developing both Falcon 9 and cargo dragon.
NASA's choices were in line with NASA's traditional development approach - do a *lot* of planning up front before you start so that your execution will be fast and clean.
For the development program - known as CCtCap - NASA decided to go with a full fixed price approach.
Let me bring up the acquisition guidance again...
Was the risk minimal or easily predicted? The US hadn't developed a crewed vehicle since shuttle in the 1970s, so nobody in the industry had recent experience.
NASA did a study after COTs, and this is one of their lessons learned (read)
Was commercial crew a cutting edge area where markets aren't currently developed?
I would say yes, and even a number of years later, there is only a tiny market for commercial crew transportation.
Which means that fixed price isn't really the appropriate choice.
The contracts went to Boeing for $4.2 billion for their Starliner capsule, though they convinced NASA to cough up another $287 million along the way. SpaceX got $2.6B for Crew Dragon.
SpaceX finished their test flights in May of 2020. This was later than NASA hoped, but it's important to remember that congress underfunded commercial crew significantly in the first few years, and the NASA "Crew Rating" process that both contractors needed to follow didn't even exist at the start of the program.
In the summer of 2023, Starliner has yet to fly crew, and 2024 seems to be the earliest opportunity.
In retrospect, I think this result should have been easy to predict. SpaceX had the relevant experience from cargo dragon and they understood how to work under fixed price contracts.
Boeing didn't have recent capsule experience nor were they used to working under fixed price approaches.
From the perspective of what NASA hoped to accomplish, they got one real success and one incomplete. People like to make fun of Starliner and Boeing certainly bears most of the blame, but this project has turned into a significant problem for them and they have already spent $900 million more than they planned. Fixed price limited the cost from the NASA perspective, but NASA still has spent a lot of time and money on Starliner without any return.
Time for the big boys, the Artemis lunar landers.
Here's how the architecture works.
SLS tosses the Orion capsule on the way to the moon, and then the Orion brakes itself into near rectilinear halo orbit around the moon.
The lander gets into orbit, does whatever it needs to get fueled for the trip, and is sent on the same trip as Orion, getting into orbit along with Orion.
The astronauts transfer to the lander, land on the moon, stay for at least 6 days, then they return to orbit, the astronauts get back in orion and head for home.
What the lander has to do is by far the hardest thing that has ever been done in crewed spaceflight, quite a bit harder than what was done with Apollo.
Based on what happened in commercial cargo and crew, how would you structure the contracts?
NASA decided that the appropriate contract structure is... Fixed price, with SpaceX getting $2.9 billion and Blue origin getting $3.4 billion in a later contract.
NASA is in a budgetary crunch. Wayyy back in 2010 when SLS and Orion were designed to fill in the shuttle-sized hole in the budget, Artemis didn't exist and therefore it was fine to spend the majority of the budget on those two programs.
But now, NASA has actually said they are going back to the moon, and they just *do not* have the money to do it right. They plan to spend $6.3 billion on these two landers, pretty much the same they spent on two capsules to take astronauts to the ISS.
So they are depending on the fact that there are two billionaires who are interested enough in the moon to put in governmental amounts of their own money.
Because of the way the economics work out, instead of NASA being the driver of this program, they are now beholden to two companies that will do whatever they want to do. There's no schedule that NASA can hold them to, and it's possible that either SpaceX or Blue Origin will get tired of the project and just stop working with NASA.
And that's assuming current NASA budgets, and there's a possibility that NASA will see deep cuts in the next budget deal.
NASA had been working on a new lunar spacesuit known as the xEMU for years, but after $420 million spent over a decade, they decided to go commercial.
The contract for the lunar space suit went to Axiom Space, a company with no previous experience in space suits, but one of two companies that expressed interest. The other company is Collins aerospace, and they got the contract to develop a new orbital space suit.
This contract has unknown risk in an area with unknown commercial potential, and therefore the contract that they chose was "fixed price" for $3.5 billion.
It's at this point that I realized something. This is a perfect situation for either a cost plus or space act agreement to both share the uncertainty and risk with NASA and because it's hard to imagine when a commercial market for lunar suits might show up, but "cost plus" is a bad word at NASA and for some reason space act agreements aren't allowed either.
I have no idea where the $3.5 billion number came from - it's clearly a case of "let's throw some money at it".
The contract is an indefinite delivery, indefinite quantity contract, which also makes no sense. That what you would use to buy resupply flights, not to do research and development.
NASA has recently sent Axiom the first task order. It's not to actually buy a spacesuit, it's $228 million to spend on developing a spacesuit. It looks exactly like a space act agreement, but it's hiding inside a fixed price contract. I'm really surprised that the NASA inspector general and the General Accounting Office aren't all over these contracts, as it looks like a very obvious attempt to circumvent the fixed price regulations and invent a new kind of contract, one that NASA can already do with space act agreements. The obvious reason to do this is that space act agreements require high-level review, and fixed price contracts do not. And everybody knows that fixed price is better.
It's clear that neither NASA nor Axiom have any idea what the suits will actually cost. That's not a dig at Axiom - nobody has designed lunar suits since the mid 1960s, and I don't expect them to be able to cost things up front.
It is a dig at NASA, who should really know better.
Which brings us to the NASA commercial lunar payload services program, or clips.
The idea of CLPS is pretty simple - NASA needs to do exploration and science on the lunar surface before the Artemis landings, and rather than create a lander and probe themselves, they will farm it out to commercial companies that will provide the lander and perhaps a rover for the NASA science instruments.
There many companies involved in this program - intuitive machines, firefly, astrobotics, blue origin, and a bunch of others.
The program sounds great, conceptually. But there are a few issues.
China has been knocking it out of the park, with three successful lunar landings in the last decade.
In 2019, the Beresheet lander by a private Israeli company attempted to land. It crashed.
Later that year, the Vikram lander from India attempted to land. It crashed.
In 2023, the private Kakuto-R lander from Japan attempted to land. Are you surprised to find out that it crashed?
The reality is that navigating to the moon and successfully landing is a really hard thing to do, and failures are not surprising.
The contracting approach of the CLPS program is like the Spacesuit contract.
Each company has a fixed price indefinite delivery, indefinite quantity contract, but they are truly indefinite - there is no defined amount of money or delivery. They are just containers for task orders that the company might win.
In May of 2019, NASA gave out the first three task orders, all for landings on the moon, to astrobotics, intuitive machines, and orbit beyond. Orbit beyond dropped out 2 months later, not an especially auspicious beginning to the program. Astrobotics got $79.5 million for their first mission, and Intuitive machines got $77 million.
Those prices are the end-to-end prices; the provider is responsible for developing and building the lander and obtaining launch services to get their lander to the moon. The idea is that the contractor can find other commercial payloads to fly and help defray the costs.
Astrobotic's first lander will fly on the first flight of Vulcan, which presumably gets them a discount on launch costs for their first flight. Intuitive machines will fly on Falcon 9.
I see no way that the economics can work out for these companies. Subtract the cost of the Falcon 9 for the intuitive machines, and they have something like $20 million left. That has to be enough to develop their lander, do all their testing, build the lander and operate it on the journey and while on the moon. If they can sell payloads, they'll get a little more money, but it won't be a ton. It's not clear to me what happens when - not if - but when these companies run out of money. Is there a "here's a little more money task order"?
NASA is putting high-value science experiments on these vehicles to guide their Artemis planning, but given the demonstrated difficulty of landing on the moon, this will not end well.
NASA has had great success with programs where they pick the two best companies out of all that are interested, throw them some development money to get up and running, and then go from there.
I'm utterly at a loss to explain why they have chosen this approach for clips.
To return to the original question - whether fixed price contracts are the answer - apparently the answer from NASA is "yes, because you can play tricks to make them look like cost plus contracts or space act agreements".
There is a role for all the contract types in NASA contracting, depending on the type of project they are doing.
If you enjoyed this video, (read)
Is Artemis a good idea or a bad idea?
Welcome to Artemis Yea or Nay?
Or How I learned to stop worrying and love the artemis program.
Links:
Why we have the SLS - https://www.planetary.org/articles/why-we-have-the-sls
I'll start by confessing that I'm a bit surprised at my opinion on this topic.
I've been pretty clear on what my opinion of SLS and Orion is in a number of my videos.
But I've recently been working on two topics.
The first is "what will it take to cancel SLS and Orion?", and the second topic is "Is the lunar gateway a good idea?".
As often happens, I worked on both of them for a while and put them aside.
Then in early December, I was listening to Anthony Colangelo's excellent Main Engine Cutoff podcast, and he was talking to Casey Dreier from the planetary society, and that led to an article that Casey wrote in early August.
And let me explain my new perspective.
I'm going to start with what seems to be a very common assertion about the Artemis program.
This assertion is a failure on an Artemis mission will lead to the program being cancelled.
And some also assert that the cancellation will lead to SLS and Orion being replaced by Starship.
What evidence is there that this would happen?
We have three historical examples to look at:
In 1967, three astronauts were killed in the Apollo 1 fire during a test on the pad. Congress held hearings to investigate the causes and were unhappy about the poor management of Apollo but allowed the program to continue.
In 1986, the space shuttle challenger blew up during ascent and 7 astronauts were killed. NASA commissioned an investigational panel, and congress held hearings. Congress was unhappy about the decisions that NASA made when deciding to fly on that cold January day, but allowed the program to continue.
In 2003, the space shuttle Columbia broke up on reentry and 7 astronauts were killed. NASA commissioned an investigational panel, and congress held hearings. It was no surprise that congress was unhappy about the decisions that NASA made in managing the foam-shedding problem, but allowed the program to continue.
I think we've established what to expect from congress when NASA programs kill astronauts.
At this point, some of you are saying that Columbia led to the cancellation of shuttle. I believe that is an overly-simplistic view; the seeds of cancelling shuttle came from the Vision for Space Exploration program announced by president bush in 2005, and it is was the fact that there wasn't enough budget to support both shuttle and the constellation program the led to the cancellation.
For those of you who disagree, I'll only note that shuttle flew 22 times over 7 years after Columbia. If that's a cancellation, it's certainly a very protracted one.
The reality is that the support of congress for NASA programs is not significantly influenced by the success of the program or deaths that occur during the program.
But perhaps Apollo and the Space Shuttle were more popular with congress than SLS and Orion are?
At this point, some of you are saying that Columbia led to the cancellation of shuttle. I believe that is an overly-simplistic view; the seeds of cancelling shuttle came from the Vision for Space Exploration program announced by president bush in 2005, and it is was the fact that there wasn't enough budget to support both shuttle and the constellation program the led to the cancellation.
For those of you who disagree, I'll only note that shuttle flew 22 times over 7 years after Columbia. If that's a cancellation, it's certainly a very protracted one.
The reality is that the support of congress for NASA programs has not been significantly influenced by failures or deaths that occur during the program.
But perhaps Apollo and the Space Shuttle were more popular with congress than SLS and Orion are?
Let's look at some articles that have been written about SLS during its development...
And how did congress react?
You will not be surprised that congress held hearings and allowed the program to continue. But they actually did more than that...
If we look at the money side,
Congress has appropriated more money for SLS than NASA requested every single year since the start of the program. That is despite the reports of poor contractor performance and years of delays, despite changes in administrations, and despite a global pandemic. In total, they've given $3.7 billion extra beyond the NASA requests.
The pattern is similar for Orion - the money appropriated exceeds the request every year except for the last 2, and since 2012, that has added up to an extra $1.5 billion.
Money from SLS and Orion go to contractors and subcontractors across the country. This is one of the reasons that there is widespread support for the program in congress - one of the roles of congresspeople is to support things that are good for their constituents.
To summarize, congress loves the artemis program.
When I was working to figure out a scenario where congress cancels SLS and Orion, I couldn't come up with a reasonable one. It just isn't going to happen.
What we need to realize is that SLS and Orion are here to stay. Not in the "next 5 years" timeframe, but in the "10 years" timeframe, and perhaps much longer. Shuttle ran *30* years.
It's certainly true that SLS and orion is a hugely expensive vehicle that cannot launch very often.
But it's *our* hugely expensive vehicle that cannot launch very often.
I was originally working to answer the question "Are SLS and Orion good?", but it turns out that that's not a very interesting question.
The interesting question to ask is, "since they are operational, what should we do with them?"
What can we do with the capability of launching Orion to lunar orbit, and when SLS Block 1B comes along, Orion plus something else heavy to that same orbit?
It turns out that NASA has a plan and I am now convinced it's a pretty good one.
The Artemis accords are an attempt to extend the agreements that worked so well for the international space station to exploration of the moon and beyond.
On the surface, they are merely about agreeing to a common framework that guides space exploration.
That's good, because it makes the barrier to entry low; over half of the signatories of the accords have no real space program. So why are they interested in agreeing to accords that apply to things they aren't doing?
The real point is for them to join the artemis club.
Traditionally, the space club was controlled by NASA and its close partners, the Russians, and more recently the Chinese.
The artemis accords create a bigger club - nations that don't have a space program have a well-defined way to collaborate with others on the kind of space program that works for them.
There's also something equally important going on.
Being part of the artemis club and participating in missions will become a matter of national pride; it's much more difficult to cancel a mission where the US is collaborating with other countries than one that we are doing by ourselves.
But there's a problem... The current Artemis lunar architecture has only two components.
There's SLS and Orion, which is dominated by the US, with an assist from the european space agency, who supplies the orion service module.
And there's the Starship Lunar lander, a purely US project.
What is needed is a international project that the members of the Artemis club can meaningfully contribute to.
Bring on the Lunar Gateway...
Gateway is a *small* space station, composed of three main modules
There is the US-created power and propulsion element, which - not surprisingly - provides power to the gateway and propulsion to keep it in the proper orbit.
There is the habitation and logistics outpost module, produced by Northrop Grumman in cooperation with the Japanese Space Agency JAXA and the European Space Agency ESA. The module is based on the Cygnus resupply module currently in use with the international space station; you can see from this render how small the gateway modules are in comparison to the ISS modules; they are 3 meters in diameter instead of the ISS 4.5 meters.
These first two modules will be launched by a Falcon heavy in November of 2024.
Added to that is the international habitat module from JAXA and ESA, launched on the first SLS block 1B rocket, currently expected to be Artemis 4.
Canada will supply the newest version of their robotic arm, the Canadarm 3.
Next up is the European Esprit module, which provides refueling, infrastructure, and communications. The communications part will launch with the PPE and HALO modules in 2024, and the remainder is scheduled for 2028.
Logistics will be provided by the Dragon-XL, though that contract is still a bit up in the air, and also by the HTV-XG, an updated version of the HTV resupply module used with the ISS.
And of course it is expected that the human landing system will dock with gateway and it will be used to transfer astronauts from Orion to the lander. This will initially be the lunar variant of starship, and perhaps later a second lander from the winner of NASA's current SLD program.
Which leaves us with the airlock.
The airlock was originally going to be provided by ROSCOSMOS, but in 2020 they decided that the Artemis program was "too US-centric" and withdrew.
That left the airlock unbuilt, but NASA announced in December of 2022 that it will be provided by the united arab emerites space agency, which is the first large contribution to artemis by a country that was not part of ISS.
The important part of gateway is that it provides a long-lived destination. It will be the central US point for missions to the moon.
SLS can carry people to gateway, and with block 1B it can carry people and gateway modules or supplies to gateway.
Gateway is not currently envisioned as a continuously crewed station the way ISS is, because the current flight rate of SLS and Orion only supports short missions. This is analogous to the situation with shuttle, where for many years it was the only US launcher to get to the ISS, but then it retired and Crew Dragon and Starliner came along. One can easily see a similar thing happen with gateway in the future.
If you are going to build a lunar space station, you need to choose an orbit for it.
Here's a nice little chart that explores different orbits.
The most obvious one to use is low lunar orbit, the orbit used by Apollo. It's a non-starter in this case because Orion doesn't have the delta-v to get into and out of this orbit - see my video on the story of orion to find more details - but low lunar orbit isn't a great place for a station.
The moon has a number of mass concentrations - or mascons - and those perturb - or mess up - low lunar orbits. You can get back to the orbit you want using thrusters, but that requires fuel. The low lunar orbit takes about 50 meters per second of delta v per year to stay in the proper orbit.
Another issue for a space station is that you would like to use it for communication with the surface and perhaps for remote operation of robotic systems, but a station in low lunar orbit can only communicate with a small portion of the lunar surface at any one time.
These are significant disadvantages for a station in low lunar orbit, even if orion could get there.
https://en.wikipedia.org/wiki/Gravitation_of_the_Moon#/media/File:MoonLP150Q_grav_150.jpg
There are some other options that Orion could maybe get to.
But the ones we care about are the ones Orion can get to. Of these three, the near rectilinear halo orbit is the best choice; it's decently close to low lunar orbits for landing and it has much lower delta-v requirements to stay in the proper orbit.
And because the orbit is very elliptical, it spends most of its time with a great view of the southern hemisphere of the moon and is only out of communication for the short time when it is close to the moon.
And that's why NASA chose that orbit for the gateway.
Why should you love artemis?
SLS and Orion are here to stay, and Artemis makes effective use of their capabilities
Lunar gateway allows for meaningful international participation in exploration programs and a common forum for other space-related activities. This makes it more likely that the things that we want to happen will happen.
Lunar gateway provide a destination that more effective technologies can aim for. Want to do lunar commercial crew? Provide a way to get to gateway.
And finally, you are already getting starship, and NASA is paying $4 billion for one test flight and two crewed flights.
And that's how I learned to stop worrying and love the artemis program
A quick channel announcement - Teller's Tachyon Tablets has asked me to make some merch available for their product, so you can get a very attractive sticker or a t shirt with the tablet box on the front and the "Climbing Spaceman" on the back.
If you go to the channel page you can find them on the "store" tab.
If you enjoyed this video, please watch Doctor Strangelove
We'll start by talking about Mars.
The first stage of our trip to Mars starts in low earth orbit and goes to lunar transfer orbit. It takes about 3100 meters per second of delta v, and it's pretty quick - just a day or two. The next stage is from lunar transfer to mars transfer. It's a little less than 480 meters per second of delta V.
Both of these are generally handled by the launch vehicle, for a total of 3600 meters per second of delta V.
Once we reach this point, the gravity of Mars will accelerate the spacecraft towards the planet. Most missions scrub off this excess velocity using aerobraking, and spend a small amount of delta v - say 200 meters per second - slowing down for the landing. Simple, and pretty cheap in terms of delta v - one of the reasons NASA has done so many missions to Mars.
Checkout, refueling
Draw delta-v map between LEO and mars plus showing the travel times.
This not only makes your job harder, it's a bit like driving to the hotel at the airport to spend the night before you head off to Hawaii.
How does that change if we stop by the lunar gateway on the way to Mars?
We will still start by going to the lunar transfer point, and then we will head into the near rectilinear halo orbit used by the gateway. That will take about 450 meters per second of delta v, and because we are slowing down, that delta v can only be provided by the spacecraft.
After doing whatever we wanted to do at the Gateway, we need to head back to the lunar transfer point, spending another 450 meters per second of delta v.
From there, it's off to Mars, with the delta-v of 480 meters per second, all supplied by the spacecraft, and then the same aerobraking approach to mars.
If we look at the totals, this approach is easier for the launcher, with only 3120 meters per second required. But it's much, much worse for the spacecraft, going from 200 meters per second all the way up to 1130 meters per second.
The original spacecraft could get by with only about 7% of its mass devoted to fuel.
The gateway visitor version needs about 40% of its mass devoted to fuel. That means it's quite a bit heavier. And since the overall spacecraft is heavier the reduction in the required delta v from the launch vehicle is likely eaten up by the extra weight.
What is the impact of these differences?
Here's one example. It assumes we are starting from low earth orbit to make the example simpler, though in most cases missions would not stop off in low earth orbit.
We are going straight to mars, and our spacecraft needs 200 meters per second of delta v. Its total mass is 40 tons. Of that 40 tons, only 3 tons are devoted to propellant, and there are 37 tons to devote to everything else. This is good.
Our earth departure vehicle is 15 tons dry and carries 70 tons. It can send our spacecraft out with 3620 meters per second of delta v, which is enough to get us to Mars.
If we stop by gateway, our spacecraft now needs 1580 meters per second of delta v. To get that out of our current spacecraft, we need to add 23 tons of propellant, pushing our mass up to 63 tons total. Our launch vehicle only needs to give us 3120 meters per second of delta v to get to mars, but unfortunately the huge increase in mass of the spacecraft means it can only give us 2851 meters per second. That's nto enough to get to Mars.
One argument for stopping by gateway is that it's a short trip and the crew can verify that their spacecraft is behaving itself before heading on the 6 month journey to Mars.
This is true. But there's nothing magic about gateway. If we want that option we can simply use the same approach with a launch to lunar transfer orbit by the launch vehicle. If something isn't right, just turn around and head back home. If things are fine, do another burn to head off to Mars.
That skips the 900 meters per second of delta V to get to gateway and back, dropping the spacecraft delta v down to 680 meters per second.
NASA is planning on going back to the moon with their Artemis project, with the first landing planned for Artemis III, no earlier than 2025.
And they've chosen an architecture - a way of using different vehicles to get people to the moon and back - that is a bit different.
The best way to understand the Artemis architecture is to contrast it to the Apollo architecture and the short-lived constellation architecture from 2005.
I'll talking in terms of "delta v", which is the amount of energy it takes to get from one location to another in space. If you want more on delta v, see my "planning a solar system road trip" video.
The first step is low earth orbit. It takes roughly 9400 meters per second of delta v to get there from the surface.
The next step is to get to the capture escape point. From the earth it's "uphill" to the capture / escape point - it takes energy to get there. A little more than 3000 meters per second. On the Apollo missions, NASA used the mighty Saturn V to lift the command, service, and lunar modules all the way to the capture/escape point. The normal rule with spacecraft is to get the launch vehicle to do as much of the heavy lifting as possible.
The maneuver to get from earth orbit out to this point is known as "trans-lunar injection", and the amount of payload that a given rocket can send to this point is known as "TLI payload". It's considerably less than the LEO payload because the rocket needs to provide an extra 3000 meters per second of velocity. The Saturn v could lift 140 tons into low earth orbit, but it could only send about 43 tons on a trip to the moon.
Once the spacecraft hits this point, it is "downhill" to the moon - the moon's gravity will accelerate it towards the moon.
The next goal in Apollo is to get into low lunar orbit at roughly 100 kilometers above the surface. The spacecraft will need to slow down to do this, and that will take about 900 meters per second of delta-v.
A key point here is that the launch rocket can do all the work to get to the capture/escape point, but the rest of the work has to be done by the spacecraft.
The service module engine is used to slow down the spacecraft to get into lunar orbit.
And finally, the lunar module goes down and lands on the moon. It needs to use its engine to slow down to land, and that takes at least 1900 meters per second of delta v, though Apollo budgeted 2100.
Coming back is reversing the process; 1900 meters/second of delta v is required to get back to low lunar orbit. The top half of the lunar module - the ascent stage - detaches and climbs back into orbit. This approach was used because the round trip to the surface requires about 4000 meters per second of delta v and that is *really* hard to do, and using two stages makes it much easier, just as two stages makes it much easier to get to orbit from the earth's surface.
In orbit, the astronauts transfer back to the command/service module, it heads uphill to the capture / escape point, and then it's downhill back to earth.
The Apollo approach was just barely possible with 1960s technology, and NASA had to work very hard to make the spacecraft light enough to be successful.
Constellation was the next architecture, developed in the early 2000s. There were different variations proposed - this is the simplest one.
An Ares V launches an earth departure stage into low earth orbit, along with a lander known as Altair on top.
An Ares I launches an Orion command and service module. They dock in orbit, and the earth departure stage provides the energy to toss the spacecraft to the capture / escape point.
The *lander* slows the stack down to get into low lunar orbit. Then - just like apollo - the lander goes down and lands, and when the exploration is done the ascent stage returns to orbit. It docks with orion, and the orion engine tosses it to the Capture/escape point and it coasts back home.
Orion only has about 1000 meters per second of delta v. It's either slowed down by the lander or - in an alternate three launch proposal - the lander and orion have individual earth departure stages that put them into low lunar orbit.
Why is the Orion command and service module so much less capable than the Apollo one? That's a bit of a long story that I cover in my video "the story of orion".
The constellation program was cancelled in 2010 and this architectural approach was abandoned.
Now we move to the artemis architecture...
The SLS rocket tosses the Orion capsule and service module to the capture/escape point.
But now we have a problem. The Orion capsule was designed for constellation, and in that architecture somebody else did the work to put Orion into lunar orbit. Orion can either get into low lunar orbit or out of low lunar orbit, but it can't do both.
What we need is an alternative orbit with lower energy requirements, and NASA has chosen a Near Rectilinear Halo orbit.
When NASA first announced this choice it had everybody immediately heading to their favorite search engine, because it wasn't a commonly used orbit.
It has to do with the L1 and L2 lagrange points in the earth moon system. You might have heard about langrange points in reference to the james web space telescope, which orbits around the L2 point in the sun earth system. It's about 1.2 million kilometers farther out than the moon, so far out of this picture.
With respect to the L1 and L2 points in the earth moon system, there are four near rectilinear halo orbits - one each for each lagrange point, and one mostly north of the moon and one mostly south of the moon.
These orbits are highly elliptical; the low point is 1600 km above the moon, and the high point is 68,000 km. For earth orbits, we would call that the "Perigee" and "apogee", but because it's the moon, they would be referred to as "Perilune" and "Apilune".
NASA chose this orbit because it's a low energy orbit - much easier to get into and out of than low lunar orbit.
It's a fairly stable orbit - it doesn't take a lot of fuel to stay in this orbit.
It has a great view of the south pole, which means you can use a spacecraft in this orbit as a radio relay between the south pole and the earth.
In terms of delta-v, the near rectilinear halo orbit NASA chose is roughly halfway to low lunar orbit, and the orion can get into and out of that orbit. Problem solved!
Now we need a lander. The first Artemis flights to land on the moon will use a lunar version of SpaceX's Starship. It will be launched into orbit, refueled in orbit, and then journey out to the capture / escape point, and brake itself into the same orbit as orion.
Two of the astronauts on Orion will transfer to starship, and land on the moon. When they are done, starship returns to the NRHO orbit, the astronauts go back to orion, and orion tosses them back to the capture escape point and then they coast home.
These three approaches place very different demands on their vehicles.
For apollo, the command service module will brake the module plus the lander into low lunar orbit, and then send the command service module back to earth after the lunar excursion. It needs a delta v of 1800 meters per second. The lander goes from low lunar orbit to the surface and back, and needs about 4000 meters per second of delta v.
The lunar lander is a two stage vehicle that masses 16,400 kilograms and about 70% of that mass is fuel.
For constellation, the CSM only needs to send itself back to earth, for only 900 meters per second of delta V. In this version, the delta v to get into low earth orbit comes from the lander, and adding that to the delta v to get to the surface and back bumps the total requirement up to 4900 meters per second.
The altair lander was a two stage design massing 45,800 kg, and was 68% fuel.
For artemis, the CSM needs to brake into the near rectilinear halo orbit and send itself back to earth, for a total delta v of about 900 meters per second.
The HLS lander has a difficult job; it needs to brake into NRHO and then go to the surface and back. That's about 5350 meters per second. Starship is a monster single-stage design, with a mass of 1,376,000 kg, and 87% of it is fuel. And it needs all of it, because it also needs to get from low earth orbit to the capture/escape point, which takes another 3050 meters per second bringing the total to 8400 meters per second.
Even if you build a lander that doesn't need to get to the capture escape point by itself, the 5350 meters per second requirement for the lander makes it really challenging to design. The "national team" led by blue origin proposed a three stage design to get the performance they needed; one to get to capture escape and then a two stage landing vehicle. It was a little like the earth departure stage approach used with constellation.
There are a number of downsides to the choice of the near rectilinear halo orbit.
The elliptical orbit limits the landing and departure times to about once a week.
The astronauts are committed to stay for a full orbit, with no earlier options for emergencies.
Both the lander and Orion need to support an extra week in case they cannot hit the first ascent window.
It's also much harder to design the lander, and therefore the lander is very expensive.
Why not build a more capable Orion?
Answering that question requires a bit of a digression...
In the 2010 legislation that directed NASA to build SLS - and keep working on Orion - congress gave NASA a goal of flying SLS by the end of 2016. NASA had the solid rocket boosters, it had the engines for the core stage, and it had the orion capsule.
What it needed was a second stage...
NASA really wanted the earth departure stage from the Constellation architecture, but it unfortunately used two J-2X engines, which were not fully developed. NASA feared that the development would take too long and push SLS beyond the 2016 goal. The earth departure stage was out...
The quickest way to get a second stage was to use one that already existed. NASA looked at existing upper stages, including the one used on the Delta IV.
The delta second stage is cleverly called the "delta cryogenic second stage". NASA decided to adopt that stage so that they could meet their goal, renaming it the "interim cryogenic propulsion stage". I have no idea why they decided they needed the work "propulsion" in the name, but they did.
The delta cryogenic second stage is a fine stage but is grossly undersized for a rocket as big as SLS; the tanks are too small and it only uses a single RL-10 engine so it is low in thrust. But it allowed NASA to say they had a solution that would meet the deadline.
NASA still wanted a bigger stage, and they ultimately designed what is now called the "Exploration Upper Stage" and looks pretty much like a upsized version of the delta second stage, with bigger tanks and 4 RL-10 engines. NASA has so far spent $1.5 billion on this stage.
Two different upper stages gives us two SLS versions.
The block 1 uses the delta cryogenic second stage - sorry, the interim cryogenic propulsion stage - and has a TLI payload of 27 tons.
The block 1B uses the exploration upper stage, and has a TLI payload of 38 tons. As I said, the delta cryogenic stage was very undersized.
There is also - perhaps - a block 2 at some undetermined time in the future.
Artemis 1 through 3 will launch on SLS block 1, and therefore have a TLI payload of 27 tons.
Our launcher can only do 27 tons to the moon.
The orion command module masses a little more than 10 tons. The service module is 4,000 kg dry and carries 8,000 kg propellant, so its total mass is a little more than 13000 kg. The total mass of both is 15,300 kilograms dry, 23,900 fueled, so close to the 27 ton limit. Orion could be a *little* heavier, but even if they just added propellant, it wouldn't be enough to use a low lunar orbit architecture.
NASA is therefore stuck with the NRHO; SLS block 1 simply cannot toss a spacecraft based on Orion that carries enough propellant to do low lunar orbit.
Of course I want to compare Orion to Apollo in more detail.
The Apollo command module is 5600 kg to orion's 10,400 kg, or only about 54% of the mass. It's true that Orion carries 4 astronauts to Apollo's 3 and is designed for longer duration missions, but given that materials like carbon fiber didn't exist for Apollo, nor did advanced electronics, it's disappointing that Orion has such a high mass.
The Apollo service module is much heavier than Orion's but it packs in double the propellant of the Orion service module. Overall, it has a delta v of 2720 meters/second, or twice as much as Orion.
Could NASA do an Orion 2.0?
Upsize the service module to the same diameter as Orion and stretch it a bit, and you can very likely find enough delta-v to switch Artemis over to a low lunar orbit architecture.
Will we see this? No, because of a single word.
That word is "gateway", or more specifically, "lunar gateway".
But that is a topic for a future video....
If you enjoyed this video, please write a 1-act play based on the mythology of Orion the Hunter
I've been reading a lot of good questions in my video comments, and those present a bit of a problem.
I like to answer questions, but if I write an answer in comments, it generally only benefits one person, while content in videos might be viewed by dozens...
So I therefore decided to do a question and answers video.
Here are the kinds of questions you might ask.
You might ask what I think of company x, or to do an analysis of something.
You might ask why an organization does one thing and not another.
You might request a video on a specific topic.
Or you may ask something else...
To add a question, just comment on this video. One question per comment, shorter questions are better. Questions that are liked are more likely to be answered.
Blue Origin Patent
https://patents.google.com/patent/US20230211900A1/en
NASA paper on active cooling
https://ntrs.nasa.gov/citations/19930003270
Rocket Nozzle Types - an Ex Rocket Man's take on it
https://exrocketman.blogspot.com/2023/02/rocket-nozzle-types.html
Video link after - Space - you know rocket nozzles.
Today we're going to talk about Blue Origin's project Jarvis, also apparently known as project clipper.
It's a project to put a reusable second stage on Blue Origin's New Glenn Rocket.
For a long time I had resisted doing a video on this project because Blue Origin has been their usual secretive selves and so little was known, but I found a source of information around what is probably project Jarvis. Or at least was at one time.
AFAICT, the word "Jarvis" first came out in this article by Eric Berger in July of 2021.
It was accompanied by a picture of this test tank that was seen at Blue Origin's launch complex 36 in Florida.
And wow if it doesn't look a lot like Starship SN4.
It does seem like Blue Origin is playing on hard mode - a reusable second stage is very challenging for a company that has yet to reach orbit. But I thought about it a bit more...
New Glenn is a strangely large rocket. The published specs say that with first stage reuse, it can do 45 tons to low earth orbit and 13.6 tons to geosynchronous transfer orbit.
That puts it in the heavy lift rocket class, and the payloads it carries will be pretty similar to those carried by Falcon Heavy. Falcon Heavy doesn't fly very often; most payloads are smaller and can be carried by Falcon 9. But there's no smaller option for New Glenn.
That means it's probably decent for carrying constellation satellites, but it's too big for a lot of other payloads. A very strange size - Falcon Heavy only ended up that size accidentally.
But if you added in a reusable second stage - one that cost 50% of the payload - that puts the payload down into a more interesting range, with payload a bit bigger than Falcon 9 in reusable mode.
That suddenly makes it a much more interesting rocket - a medium lift fully reusable rocket likely beats a heavy lift partially reusable rocket.
What is blue origin planning to do?
I think the general assumption was that Blue Origin was planning on taking the first stage of new glenn, taking the starship design and shrinking it down by about 25%, and putting it on top of the booster.
The fact that their test tank looked so much like a starship test tank appeared to be confirmation.
But I don't think that is the plan...
Unlike SpaceX, Blue Origin does believe in rocket patents, and here's an interesting one for a reusable upper stage rocket with an aerospike engine.
It's important to remember that there are many reasons to file patents - including to confuse your competitors - and therefore this is going to be speculative.
But it's the only detailed data I've been able to find, so we'll end up either somewhat enlightened or somewhat confused...
One of the pictures in the patent application shows the reusable stage "100" on top of a first stage which is very obviously New Glenn.
And here's what the stage looks like. The patent doesn't really talk about payload fairings so it's not clear what their plan is there; the second stage is the lower part, the one labelled 110.
The engine is a what is known as a truncated aerospike.
Small rocket nozzles - they call them thrust modules - are pointed inwards at an angle so that the curve of the aerospike functions as one side of the nozzle and the other side is created by the atmosphere.
This aerospike shape would be a poor choice for reentry as it would be hard to keep cool, so they have further truncated the spike. Whether it still counts as an aerospike is an open question.
Here are some other drawings that make the design clearer. The patent mentions a "plurality of thruster modules" and further details that there might be anywhere from 20 to 150, or even numbers out side of that range.
There are holes in the heat shield - shown in red - that can be used to inject more gas or propellant to keep the center heat shield cool enough, and it may also have active cooling underneath.
Here's a side view that shows the bottom of the stage and the thrust modules.
Instead of the two standard BE-3U engines planned for the New Glenn upper stage, this design would use two BE-3U power packs - everything except the combustion chamber and nozzle - to feed all of the thruster modules in the ring
At this, you might be asking yourself "Doesn't this look eerily similar to the Stoke design?" And the answer is yes, for a fairly simple reason...
To reuse your second stage, there are two big requirements:
The first is to be able to land the stage. I'm going to assume that you are using engines to land it, which means that you need engines that you will run reliably at sea level. This is why starship has both highly efficient vacuum engines and less efficient sea level engines - the sea level engines are required to perform the landing operations.
Right now the mix is 3 sea level, 3 vacuum, but it's likely that starship will switch to 3 sea level, 6 vacuum in future versions.
The second issue for reusability is being able to make it through reentry. Starship comes in mostly sideways and that allows them to tuck the engines into the rear part of the stage and protect them from reentry heat with the same tiles that protect the rest of the ship.
It adds weight but allows them to use engines with standard bell nozzles.
Stoke and Blue origin are using an approach where their engines work all the way down to sea level, which means they can get away with one type of engine for all the flight and for the landing, which is a useful simplification.
But to make it through reentry they are entering a new ground with a technology nobody has every tried before.
This design is a natural outgrowth of their decision to skip tiles and reenter butt first.
There's a very interesting question about engine efficiency between the SpaceX bell nozzle approach and the stoke/Blue Origin aerospikey approach. For the sake of argument, I'm going to assume that the latter both give you aerospike efficiency.
I've talked briefly about aerospikes in the past - see my video "space - you know rocket nozzles" - so I won't go into the topic too deeply here.
The assertion is that the traditional bell nozzle is optimal only at a specific altitude but the aerospike is optimal at all altitudes, and that the reason rockets don't use aerospike designs is either other concerns having to do with cooling or excess mass or due to a conspiracy, presumably by "big nozzle".
There is a real discussion to be had about the different designs, but the important point here is that we are dealing with second stages and what we care about most is the engine's efficiency when it is operating in vacuum.
So the question is "how will aerospike designs - and presumably these designs - perform as second stages that are only operating in vacuum?"
That's a hard question to answer because the majority of the discussion is about the performance as a first stage engine since we already know how to build vacuum engine nozzles and the air-pressure adaptation feature of aerospikes doesn't help you there.
Which left me with some assertions that aerospikes performed worse than vacuum nozzles but not a lot of real data, and I try to avoid being based on assertions.
Then I happened to come across a post on rocket nozzle types on a blog called, "an ex rocket man's take on it". This blog is written by Gary Johnson, an engineer who used to design rocket nozzles for a living. See the video description for a link.
The blog post is long and involved with a ton of details, but basically he benchmarked a LOX/RP-1 kerosene gas generator engine with both a conventional nozzle and an aerospike.
And then, just for fun, he benchmarked a vacuum nozzle, which obviously would not work for a sea level engine.
The data shows up in this chart, and it's one with a surprising conclusion. Note that we ignore all the low altitude stuff because we only care about vacuum performance.
It's a little hard to read the graph, but what he found is that the vacuum bell nozzle gave a specific impulse of 335, a sea-level bell gave specific impulse of 315, and an aerospike nozzle gave a specific impulse of 290. That the vacuum bell worked better was not a big surprise to me, but that a sea level bell beat the aerospike so easily was a significant surprise.
Plugging those numbers into my Falcon 9 model, I found that the vacuum bell would give a fully-expendable payload of 20.8 tons, the sea level bell would give a payload of 18.2 tons, and the aerospike would only give a payload of 14.9 tons.
That's a reduction of 28% with the aerospike design compared to a full vacuum design. You will be throwing away a lot of payload, and since you are going to reuse the first stage, you will be staging low so your second stage will need to a do a lot of work to get into space.
That is, of course, assuming these designs behave like aerospikes, and they might not. The Blue Origin design has the thrust modules pointing in and really looks like it wants to be an aerospike, but the stoke design looks to have them pointing straight down, so that design may behave more like a whole bunch of normal modules.
My uninformed guess is that the ultimate performance is going to depend on the nozzle size on each of the thrust modules, but bigger nozzles have obvious disadvantages.
Another question is whether the Jarvis design will survive reentry?
Capsules tend to be widest at the base and short, and they are built this way so that the high temperature gas created in front of the heat shield doesn't damage the capsule higher up. Here's a simulation of a SpaceX dragon reentering, and it shows that the sides of the capsule stay in the cool gas.
Even with that, Dragon does not reenter unscathed.
This second stage is straight on the side. It will have the benefit of the thrusters pushing the hot gas out to the side, but it will have to push it out quite a bit to keep it away from the upper part of the stage. That means a lot of propellant use.
This is based on the fairing not coming along for reentry. If the longer fairing is there that makes the problem harder.
The Stoke design uses a shape that is much more capsule shaped and will likely have fewer reentry heating issues.
Another possible issue is the frontal area. Coming in end-wise, the area of the back of the rocket is 63 square meters.
Coming in sideways, the area is 105 square meters. And that's not taking into account of the extra area control fins would add.
Smaller area means faster reentry and higher heating levels. Not good.
Does the cooling scheme works - does active cooling reduce the heat enough to make it practical?
NASA and others have explored cooled structure concepts and transpirational cooling, but there's not any real applications of these approaches as far as I know, so it will require development time and money.
And that's the story of Project Jarvis slash clipper - at least as much of the story I've been able to figure out.
A reusable second stage is a good fit for New Glenn, but it appears that they are trying a new concept that nobody has tried before, which will definitely take a lot of time and effort to develop.
If you enjoyed this video, you might like this research paper...
The NASA budget process is fairly confusing and the way that media reports on it doesn't make things better. This short video is my attempt to reduce the confusion.
This isn't going to be terribly exciting, therefore we're going to start with an Atlas Centaur test from 1962...
Now onto the budget stuff.
We'll start with the NASA fiscal year 2023 budget estimates, a 990 page document that both summarizes all the programs NASA is working on and provides estimates for how much they will cost for 2023.
It would be more correct to replace the word "Estimates" with one of the following:
Dreams
Hopes
Wishes
I'll explain why in a minute...
https://www.nasa.gov/sites/default/files/atoms/files/fy23_nasa_budget_request_full_opt.pdf
Here's how that budget document is created...
NASA is part of the Executive branch, and therefore is managed by the President. In between the president and NASA is the NASA administrator, currently former astronaut and Florida Senator Bill Nelson. You can think of his job as "NASA CEO"; he's in charge of the overall operation of the agency.
The third player in this process is the office of management and budget.
There are discussions between all of these entities; the president has specific ideas of what they want from space, and NASA has ideas on what they want to do, and OMB understands the fiscal realities, and all of those concerns get iterated and refined, and out of the process comes the NASA budget, plus the budgets for the rest of the government.
However...
The constitution says that it isn't the president that decides what money to spend, it's Congress.
That is why the NASA budget is aspirational; it describes what NASA is *hoping* to get.
The senate and house therefore control how much money NASA gets...
To start, both the senate and house create a budget resolution, which is a high-level overview that describes how money will be spent across the government.
These then feed into a joint working group, and the house and senate agree on a Concurrent budget resolution.
But that general resolution has very little detail; it will only say that science, space, and technology will get $43 billion.
Those high level numbers then get sent back to the senate and house, to the committees that own deciding the specifics of how that money will be allocated.
Their decisions then feed back into another shared process to reconcile their different numbers, and that ultimately leads to the NASA appropriations; a detailed budget for how much money NASA has for a fiscal year.
The planetary society nicely provides a summary of the appropriations. Here you can see part of the appropriations for 2022; the table shows the following values:
The first column shows the amount of money that was appropriated in 2021, where "appropriated" means "given to NASA to spend".
The next column shows the president's budget request for 2022.
The next two columns show the house and senate appropriations, which is how much each of the committees thought NASA should spend on individual programs.
The final column shows the amount of money actually appropriated.
Let's look at three programs...
The james webb space telescope is flat across the board; NASA wanted $175.4 million and everybody agreed on that figure.
SLS got more than they requested.
And if you look at the earth science amount, both the senate and house thought the request was reasonable, but when they got together they knocked a few hundred million off of it. That is - as they say - how politics works
https://www.planetary.org/space-policy/nasas-fy-2022-budget
Congressional authorization is a separate process, and it is the way that congress establishes, continues, or modifies federal programs.
NASA cannot start new projects without this authorization.
For example, the NASA authorization act of 2017 said:
(read)
That gave permission for NASA to begin developing what is now known as the Nancy Grace Roman Space Telescope.
Nancy Grace Roman was the first female executive at NASA, and served as NASA's first Chief of Astronomy throughout the 1960s and 1970s, establishing her as one of the "visionary founders of the US civilian space program".
She created NASA's space astronomy program and is known to many as the "Mother of Hubble" for her foundational role in planning the Hubble Space Telescope.
NASA cannot start new programs without authorization, but authorization does not provide any money.
What matters is the money that gets appropriated.
Or as immortalized in "The Right Stuff"...
Here are some examples of weirdness, either in the budgetary process or in the reporting of it.
The A-3 test stand at Stennis started construction in 2007 to test the J-2X engine for NASA's constellation program.
Constellation was cancelled and NASA is no longer has plans to use the J-2X, but the NASA authorization act of 2010 specified that NASA would continue to complete the construction and activation of the A-3 test stand.
The $350 million test stand was completed in 2014 and mothballed; it now costs NASA $700,000 each year to maintain.
https://www.washingtonpost.com/sf/national/2014/12/15/nasas-349-million-monument-to-its-drift/
NASA's original SLS plans were to take the SLS block 1 mobile launcher and modify it to support block to, but Congress instead decided to appropriate money for a second mobile launcher, at an estimated cost of $450 million.
SOFIA, an airborne telescope project, was eliminated from the 2022 budget request, but congress decided to keep it alive by funding it at $85 million
The Nuclear Thermal Propulsion project was also eliminated in the 2022 budget request, but congress decided to fund it at $110 million.
Here's a recent story...
Senate committee approves 2021 NASA authorization, requires second HLS system.
And another one, which says...
... legislation granting more then $10 billion to support NASA's Human Landing System program.
Let's look at the actual text of the authorization bill.
These give the distinct impression that NASA would be making a second award with a significant amount of money. But I'm sure you noticed the specific word used:
Authorization.
This bill is about authorization. The fact that it gives a dollar amount is simply a smokescreen; authorization bills do not appropriate money.
What happened?
Both the house and senate allocated $100 - 150 million extra for HLS in committee, but the final number is the original $1,195 requested by NASA.
Authorization does not equal appropriation
And finally, we come to the NASA authorization act of 2010, which contains authorization for two programs.
First, it authorizes a new follow-on program to the space shuttle known as the Space Launch System.
And second, it authorizes NASA to expand the size of the commercial crew program that had just been started.
But we all know what "authorization" means. How were these two programs appropriated?
Here's a graph showing the SLS budget requests and appropriation amounts. In 2011 and 2012 the amount appropriated was a little bit less than the request, but since then, congress has consistently appropriated more money for SLS than NASA has requested.
Commercial crew was quite different.
In 2011 through 2015, congress appropriated 60%, 45%, 65%, 85%, and 95% of what NASA had requested, and only reached parity in 2016. This lack of money slowed down progress on Commercial Crew significantly.
Here's my quick summary.
NASA budgets are aspirational; it is what NASA is hoping to get.
Congress controls what programs exist and how much money they get. They can decide to fund projects that NASA isn't interested in, not fund projects that NASA wants to do, or underfund projects that they explicitly told NASA to do.
Don't pay attention to stories about authorization bills.
Pay attention to appropriations; that is how congress indicates what is important to them and the actual NASA budget.
If you enjoyed this video, please compose a NASA budget haiku...
As sometimes happens, I found in the midst of working on a video that I needed to do a different video first, about what companies are currently operating at cape canaveral.
While writing that, I realized that talking about the current state of the cape didn't make a lot of sense without some information about the history of the cape.
To get started, there are two different entities that have launch facilities on the cape. There is Kennedy space center, and there is cape Canaveral space force station. They are different organizations but commercial payloads can launch from Space Force locations and military payloads can launch from Kennedy space center locations.
Let's start by clarifying where the cape is. It's on the east coast of Florida, pretty much due east from Orlando, best known for DisneyWorld.
Here's a closer view.
Before we start, we should probably define what a launch complex is.
We're used to big complexes like the Pad 39A complex used for Apollo or Pad 36A and B used for Atlas Centaur, but the term also applies to simple setups like this one used for Redstone or even this one used for Matador cruise missile launches.
While Canaveral is mostly a spaceport now, in the early days it launched a much more diverse set of vehicles.
Note that the launch pad is the physical support for the rocket while the launch complex includes the launch pad and all the support equipment for the pad. A rocket that launches from complex 16 also launches from pad 16, but a launch from complex 36 does not launch from pad 36, for reasons that will become clear.
Also note that in common usage this distinction is very often ignored.
The launch complexes at the cape have a very strange numbering system, one that evolved over time. The following is a short journey through the history of the launch complexes, including the first vehicle that used each launch complex.
We start in 1950 with launch pads 1-4, launching the Bumper 8. It's a German V-2 rocket with a WAC Corporal second stage on top of it.
Moving onto 1951, we have the launch of matador, a surface-to-surface cruise missile with a nuclear payload. It was launched from launch complex A-D.
A few years go by, and in 1955 we see the first launch of a Redstone rocket from launch complex 5 and 6. Redstone was a short-ranged ballistic missile.
Things pick up in 1956, with the launch of the supersonic intercontinental cruise missile, the Navaho from launch complex 9 and 10.
And then vanguard, a scientific launcher, from launch complex 18.
Note that we're already getting a very random pattern of launch site locations and numbers. The air force laid out space as they saw fit. Also note that we're already seeing gaps; there was an allocated space for launch complex 7 and 8, but it was never utilized.
In 1957, things get busy. Thor is an intermediate range missile launched from LC 17. The XSM-73 bull goose is a cruise missile that was designed as a decoy - it was supposed to look like other US bombers on radar. It launched from LC 21.
Atlas - the first intercontinental ballistic missile for the US - is launched from pad 14. And to close out the year, Jupiter - a medium-ranged ballistic missile - launches from pad 26.
In 1958, atlas pads 11-13 go active, and we see a launch of Polaris from LC 25 at the south end of the cape. Polaris was the first submarine launched ballistic missile.
In 1959, LC 15, 16, and 19 go active for titan. Titan is another ballistic missile that was a backup in case the atlas was delayed.
There's also another pad for Polaris - LC 29 - It's not clear why they needed another pad, but they may have been duplicating the way the missile is launched from underwater.
In 1960, things slow down a bit; the final titan pad in lc 20 at the north end comes online.
1961, two more missiles; the Pershing short range missile at LC 30 in the south and the minuteman ICBM at LC 31 and 32 in the middle.
In 1962 the very suborbital Loki research rocket showed up, to be launched from LC 43 near the tip of the cape. The Loki is smaller than many hobby rockets of today.
And the massive Atlas Centaur closed out the year at Launch complex 36, which - just to make things confusing - had two launch pads in one complex. The Centaur stage used with the current Atlas V and upcoming Vulcan rockets from ULA can be directly traced back to this centaur.
1963 and 64 pass by, and we're now we've moved beyond missiles, to the Titan IIIc and launch complex 41. This takes a titan II core and straps two huge solid rocket boosters on the side. The titan III was used as a heavy lifter for department of defense payloads and also for many NASA science missions.
And finally, LC 40 comes online in 1970, also set up to launch titan III rockets.
That takes us to the end of the initial wave of development at Cape Canaveral air force station.
That covered all the air force programs, but of course NASA was getting busy as well in the early 1960s.
This map is from 1962; note that it has been rotated 90 degrees; north is to the left and the Atlantic ocean is at the top.
NASA was getting human spaceflight started with the Mercury program, and they needed rockets for those flights. The obvious rockets to use were the missiles already under development, so they used the launch complexes that were set up for those missiles.
The first three Mercury launches on the Mercury-Redstone rocket launched out of launch complex 5. The Redstone was a small rocket, not big enough to take Mercury into orbit.
The Mercury Orbital launches were on Mercury-Atlas, so they were launched from one of the Atlas sites, launch complex 14.
All of the Gemini launches flew on a Titan II Gemini launch vehicle out of launch complex 19.
In addition, the agena docking target vehicles were launched by Atlas rockets out of one of the Atlas Agena complexes.
The Gemini astronauts would practice the rendezvous and docking maneuvers that were needed for Apollo using the Agena vehicles.
For Apollo, bigger rockets were needed.
At the northern end, complex 34 and 37 were build to support the Saturn I and IB rockets, including the launch of Apollo 7, the only crewed Apollo mission to be launched from Cape Canaveral Air Force Station.
Pad 34 was the site of the Apollo 1 fire that killed astronauts Grissom, White, and Chaffee. The launch platform has been preserved as a memorial to the astronauts.
This picture shows a view looking north from 1964 or 1965
The row of launch pads was known as "ICBM row" or "missile row".
Starting at the south end, we have complex 36 with two pads for the atlas centaur.
Next are the four atlas pads, followed by the four titan pads.
The two Saturn I and IB pads are next.
And then just at the top, you can see pad 40 and 41 for Titan III, along with two assembly buildings. The titan III rockets would be assembled in these buildings and then rolled all the way to the launch pad on train tracks on either side of the access road.
In these pictures, you will often see these strange circular buildings; they are the blockhouses for the launch control centers.
Now that we've done some history, let's move forward to 2022 and see what's going on now...
Here are the active organizations at Cape Canaveral and Kennedy Space Center.
There are no active launch complexes south of the tip of the cape. I'm not sure why; it might be because the air force wishes to preserve those sites - their museum is nearby - and it might be because launches to high inclinations would overfly other launch sites.
We'll start at the south, right near the tip of the cape, where Astra leases space launch complex 46 for their Astra rocket.
Astra's goal is to be able to launch from unimproved sites, so their launch pad equipment is very minimal.
Here's a picture of one of their rockets at the launch site.
Moving north, there is Pad 36 which Blue Origin is modifying to support their New Glenn rocket. Here's what the pad looks like currently. Note that the blockhouse in the picture is the same one we saw in the missile row picture.
There is also the large New Glenn factory in Kennedy Space Center.
Next to the north, we have pad 13 which SpaceX has modified as landing sites 1 and 2 for Falcon 9.
You can see some remnants of the old pad structure, but mostly it's just two concrete landing pads.
The second image is of course a tandem booster landing from a falcon heavy launch.
Pad 16 will host Relativity's Terran 1 launcher.
Yep, it looks like a launch pad. The second picture is from the summer of 2021 showing the strongback that will be used to lift the rocket to a vertical position.
The last of the aspiring small launch companies takes us to pad 20, with the Alpha and beta rockets from Firefly.
Firefly is also planning a factory just across the street from Blue Origin, on New Space Drive (no, I'm not making that up).
Here's the pad, and a rocket sitting on the pad.
The south part of the coast is mostly used for smaller rockets, with New Glenn an obvious exception. We are now moving to bigger rockets as we move north.
Next up the coast is pad 37B, the long-time home for Delta IV rockets, currently launching the last few Delta IV heavy rockets for United Launch Alliance.
In this picture we can see the white assembly building which encloses the rocket.
Delta uses an interesting approach; instead of rolling the rocket out to the launch pad, they assemble the rocket over the launch pad and then roll the assembly building away, as we can see from this shot.
North to space launch complex 40, the first Florida home for SpaceX's Falcon 9 rocket. The space industry loves both creating acronyms and pronouncing them, so space launch complex is abbreviated SLC and pronounced as "slick" If you hear a mention of "slick 40" in a spacex broadcast, that's what it means.
This site was originally used to launch Titan III and IV rockets, and was leased by SpaceX in 2007. The majority of Falcon 9 flights have flown from this site. At the bottom is the integration building where the rocket and payload are assembled, and it is then rolled out to the launch site and lifted vertical by the transporter-erector.
This view shows the characteristic lightning towers to protect the launch vehicle from Florida thunderstorms.
Pad 41 is currently used by United Launch Alliance for their Atlas V rocket, and will also be used for their upcoming Vulcan rocket.
Another old titan site, it looks very similar to the Falcon 9 launch site. ULA assembles the rocket in the integration building at the bottom of the picture and then it is rolled all the way to the launch pad. This is very similar to the approach NASA uses for Apollo, the space shuttle, and the upcoming SLS.
This image shows an Atlas V just emerging from the integration building and heading towards the launch pad.
We have reached the northern boundary of Cape Canaveral Space Force Station and are now moving to Kennedy Space Center, so I'll add some landmarks.
The visitors center is just north of the factory locations for blue origin and firefly, the vertical assembly building - or VAB- is farther north, and the shuttle landing strip is in the northwest.
The first pad is launch complex 48, which isn't actually a working pad yet.
So far, it's just a location where small rockets could launch.
It's roughly halfway between space launch complex 41 and the big pads you can see in the background.
Launch complex 39A is the most storied of NASA's launch pads, hosting Apollo missions to the moon and many space shuttle launches.
It is currently leased by SpaceX and is used for Falcon 9 launches, Dragon crew launches, and Falcon Heavy launches. It is currently the only launch site for astronauts in the United States, though when Starliner launches crew in the near future, it will launch from nearby space launch complex 41.
SpaceX is currently building a launch support tower for Starship next to the Falcon 9 pad.
This picture shows a Crew Dragon on a Falcon 9 rocket. The large arm at the top is used for astronauts to enter the dragon capsule.
And finally, all the way at the north end, we have launch complex 39B, the neglected brother of launch complex 39A.
This pad launched Apollo 10 and the Skylab and Apollo Soyuz missions using Saturn IB rockets.
It launched 53 space shuttle launches.
The current design is a "clean pad" approach designed to support SLS and other commercial launches, but none of the big commercial launchers have been interested, so it's just SLS.
This image show the first SLS rocket and Orion capsule on pad 39B for wet dress rehearsal. This rocket will be used on the upcoming uncrewed Artemis 1 mission around the moon.
If you enjoyed this video and have the ability to do so, please visit Kennedy Space Center. I recommend the bus tours if they are running.
We all know that orbital rockets are hard to build.
It takes a lot of money and technical ability to build rockets that can actually make it to orbit without blowing up.
But once you have that orbital rocket, you'll be able to fly payloads to space and make a lot of money, right?
You will be able to fly payloads to space. Whether anybody will *pay* you to launch those payloads or whether you will make money on those launches is an open question.
That depends upon what kind of payloads customers are wanting to launch and who your competition is.
It's about the launch market.
We're going to start by looking at a market that I think is easier to understand than rockets - the market for tortilla chips. As is usually the case, I'm going to be simplifying a lot of the underlying concepts.
In most cases, there isn't just one market but a bunch of different market segments because consumers want different things.
For example, there is one set of consumers who want plain tortilla chips, and we can call them a separate market segment, or a market all by themselves.
There's a segment of plain chip buyers composed of consumers who are very price conscious. That's another segment.
And then we have endless flavors of chips. There's a segment of consumers who are hooked on doritos nacho cheese flavor, and even on the Taco flavor.
If we want to start our own chip company, it's going to be difficult.
The plain and plain/cheap segments have lots of other companies already competing with each other.
These two doritos flavors have been in production for more than 50 years, and the chance of coming up with a product that can steal away those customers is minimal.
In general, the goal with a new business is not to go where the competition is strong but go where the competition is weak and/or where they are unable to easily go.
So maybe you could come up with something like organic corn chips, or tortilla chips made out of something other then corn. These are better strategies because they are harder for a big company like Frito Lay to follow.
When looking at a company and their chances of success, we need to understand what the markets are and what other players there are in the markets. Is the market very competitive, like the chip market? Or is there a lack of competition in a market that leads to opportunities for new companies.
Business success requires both a good product and a market - or market segment -that gives opportunities for success.
My first question about companies isn't "what are they doing?", it's "what is their target market?", because if there is no market opportunity, they cannot succeed.
Let's take those basic principles of markets and apply them to the rocket world.
And let's look at the market back in 2005, at a time when a certain space company was very young.
We'll start with the commercial satellite market, which was mostly dominated by communication satellites that were launched to geosynchronous transfer orbit, or GTO. This is the most competitive of the launch markets - all you need to do is show up with a rocket that can put these satellites in the desired orbit to join the market.
In 2005, this market was dominated by two launchers.
The European Ariane is the market leader. Ariane 5 is a little weird in that it typically flies GTO missions with two payloads - typically one on the larger side and one on the smaller side - so it gets more than one payload per launch. This dual launch makes the launch cost of $150-$200 million a bit more acceptable. Ariane is launched by ArianeSpace, which is a company owned by 17 European suppliers of Ariane components across 9 countries so it is not the most agile of companies.
From 2000-2005, Ariane flow 43 missions and over 60 payloads.
The second major launcher is the Russian Proton. The Proton is quite a bit cheaper at roughly $65 million per launch, but it unfortunately has a problem. If we look at its launch history, it fails a lot - roughly 10% of the time. In 2000-2005, Proton launched 23 commercial satellites.
There's an obvious question to ask here - why are people willing to risk their business on a launcher that has such poor reliability?
The answer is simple. There is more demand to launch satellites than there is Ariane launch capacity, so from a business standpoint it's better to take the risk of losing the satellite - and the revenue it will generate the next few years - versus the certainty of losing the revenue if you wait to launch on Ariane.
That's one supplier who is great at reliability and service but is expensive and isn't terribly agile, and a second supplier who's a lot cheaper but has reliability problems. And there is a backlog of payloads waiting to be launched.
It appears there is a significant opportunity here if a reliable low-cost launcher were to show up.
The second market is launching general payloads for the air force or launching NASA's science payloads.
Each of these programs have their own requirements, and the requirements to bid on these launches depend on the payload - important payloads require a proven track record, while less important payloads require less of a track record. NASA in particular will fly low-priority payloads on unproven rockets as part of their charter to help commercial launch companies wherever possible.
As the main US commercial launch company in 2005, United Launch Alliance launches the majority of these payloads. Currently there is little competitive pressure on them to launch these missions at low prices.
Another market opportunity for any launcher that can meet the qualification requirements.
The department of defense / air force has a program to launch their important satellites known as the evolved expendable launch vehicle program, or EELV. The program is now known as NSSL, or National security space launch.
Many of these launches are to orbits above low earth orbit and therefore require high performance launch vehicles.
The air force launches many military satellites under this program, and the launchers they used were the Atlas V or the Delta IV.
They are both owned by the United Launch Alliance, a company half-owned by Lockheed Martin and half owned by Boeing that was created in a shotgun wedding brokered by the government when Boeing was found to have committed industrial espionage against Lockheed martin and was therefore barred from competing for government contracts.
For that story, see my video "spies, allies, and enterprise - the strange story of ULA".
The upshot of this arrangement is that ULA has a monopoly on launching payloads and not only are they charging high prices per launch, they are receiving "launch capability payments" from the government for being capable of launching even if they don't have payloads to launch. This is a very lucrative business to be in.
A new launch vehicle could likely take away some of this business, as the government would prefer not to rely on a single company.
There are significant performance and reliability requirements to be able to fly these missions and that sets a high bar for any new company, but a company that can meet those requirements can charge a premium for these launches and still be cheaper than the Atlas V and - especially - the Delta IV Heavy. These launches are rewarded in chunks so it will be very hard to unseat a company launching under this program.
Definitely a market opportunity.
There is one more market, and it's a brand new one.
In 2005, NASA was in the process of spending about $150 billion to complete the international space station.
When completed in 2010-2011, the shuttle would be retired. The shuttle was the only US vehicle that could carry cargo and crew to the space station. Without a new capability, the US would be forced to rely on the Russian progress vehicle to carry cargo and the Russian Soyuz to carry astronauts.
Not a good look for NASA.
In 2005 NASA was spinning up the Constellation program, and the plan was that the Ares I rocket with the Orion capsule on top would take over for the shuttle in carrying crew to ISS.
It wasn't a very good plan - Ares I was going to be very costly to develop and Orion was overengineered for an ISS capsule. The program was just getting underway in 2005, but it was looking to be very costly on a per flight basis, with $1 billion or more per flight showing up in later estimates. It was moving slowly as well.
NASA needed something a *lot* cheaper to get cargo the ISS and it would be really useful if it were ready when shuttle retired.
NASA spun up two programs - a development program named commercial orbital transportation services and an operational program named commercial resupply services.
This was a huge opportunity. An operational contract would likely mean 2 launches per year for many years and they would likely be lucrative. And the program required a capsule to dock with the ISS, so it wasn't purely about launch, and that made it harder for existing launch companies to compete.
There would also likely be a follow-on program to launch crew to ISS to remove the dependency on the Russians, and that could be another lucrative long term project.
To summarize, in 2005 there were significant opportunities in 4 different launch markets. One was a new opportunity, and the other 3 had companies that were not particularly agile.
This was a perfect time to start a rocket company with a medium lift rocket, if only that rocket company could actually build such a rocket.
Obviously SpaceX did get there with the Falcon 9.
In addition to the market opportunities, SpaceX had some other things going for them.
Musk had seed money from his sale of Paypal, and that reduced the amount of outside investment they needed.
Musk had connections to silicon valley investors that could be tapped for development money.
SpaceX had Gwynne Shotwell who had a deep understand of the launch market, and her ability to sell launches on the unproven Falcon 9 to companies that would have launched on Ariane or Proton was critical.
SpaceX was fast moving and very cost sensitive, especially when compared to the existing launch companies.
SpaceX won the COTs contract for ISS cargo, which gave them development money they could definitely use.
And even with very favorable market conditions and development money from NASA, it was a close thing - SpaceX could easily have gone out of business during this period.
SpaceX is a remarkable company but their success relied on market opportunities that they could exploit.
How has the launch market changed since 2005?
Ariane 5 launched 9 times in 2020-2023. Note that this was during the pandemic years.
Proton has pretty much disappeared as a commercial launcher.
Ariane 5 has flown its last flight, and we are now waiting for the first flight of Ariane 6, current scheduled for sometime in 2024.
Because of the delays in Ariane 6 development, there is a year or two of governmental payloads to launch before Ariane 6 can get back to launching commercial payloads.
Proton was obviously replaced by Falcon 9, and from 2020-2023, Falcon 9 has made 19 launches to geosynchronous transfer orbit, the same that Ariane 5 did.
The launch market to GTO is much cooler than in 2005, with many companies trying to understand how starlink might affect their business.
Even with Ariane currently on a break, Falcon 9 could launch more if necessary and it is the price leader in this market, so the market opportunity here is gone.
For the Space Force and NASA payloads, SpaceX is winning pretty much all of the independently bid contracts, so much so that ULA has declined to bid on some of them.
There is still a market opportunity for new launchers to win the launch contract for the NASA payloads that require less reliability.
The current version of National security space launch has contract awards split between United Launch Alliance, who gets 60% of the launches and SpaceX who gets 40% of the launches.
United launch alliance flies both the Atlas V and Delta IV Heavy for these missions, while SpaceX flies both the Falcon 9 and the Falcon Heavy.
Like Ariane, ULA is in the midst of changing rockets, with ULA hoping to fly their Vulcan rocket in 2024.
This has been problematic for ULA; Vulcan needs to fly 3 times before it can fly the NSSL payloads, and it is possible that some of the payloads will be shifted to SpaceX or perhaps launched on one of the few remaining Atlas V rockets if Vulcan is further delayed.
This will get interesting during the next phase of NSSL, as ULA and SpaceX are very likely to win the contracts they have won in the past, but a new "lane" will be added for competitive launches.
That will add new opportunities for companies to bid on, assuming they are cheap enough to beat SpaceX.
For resupply cargo flights to the ISS, SpaceX shares the contract with Northrop Grumman.
For crew flights to the ISS, SpaceX shares the contract with Boeing, though SpaceX is on their 7th operational flight and Boeing has yet to fly a crewed mission.
It is unlikely that any of these providers will be replaced for the remaining lifetime of ISS.
Unlike 2005, in 2023 there are few significant opportunities and you need to compete with SpaceX in most markets.
Is there a way to compete with Falcon 9 with a small rocket.
Falcon 9 is a medium lift rocket that can lift about 15 tons into low earth orbit for $65 million.
There's a region at the left side of the graph that is classified as "small lift launch vehicles. Is this a market opportunity?
Rocket Lab's electron lives way down here at the left edge. It can launch 300 kg payloads for only $7.6 million. It has 36 successful launches. Their payload is small but they have commercial customers who are designing to that payload size.
SpaceX also competes here, with their rideshare flights that are cheaper than Electron if you are okay with their scheduling and their target orbit. SpaceX also supports larger payloads on the rideshare but does not publish their launch prices.
So the answer is that Rocket Lab is already there and competitive and SpaceX is vacuuming up all the companies who care most about price, so there is very little market left for companies that want to compete in the small launch market.
This is apparent to Rocket Lab, which is why they are working on the Falcon 9 - class Neutron rocket.
Atlas V - $109 million to $200 million
Delta IV Heavy - $350 million to $400 million
Vulcan - $110 million?
New Glenn - Who knows? $90 million to $250 million
Ariane 6 - $77 - $126 million
Falcon Heavy - $97 million
Neutron - $50 million
Starship - $15 million to $200 million
It gets worse for new companies.
Let's compare Rocket Lab to a hypothetical new launch company, looking at their cost structures.
Rocket lab has an existing factory for building rockets and multiple launch pads. Our new company has neither of those, so they will need to pay to build them.
There are fixed costs, which are primarily employee salaries plus maintenance costs and the cost of leased equipment and buildings.
Those fixed costs are spread across the number of launches the company manages per year. Rocket lab is on track to launch at least 10 times in 2023, so each launch needs to pay 10% of their fixed costs. Our new company would be lucky to fly twice a year, so at best each launch needs to pay 50% of their fixed costs.
This is a *huge* problem for new companies. You need a high flight rate to get enough business to spread you fixed costs out enough to get a competitive price, but you can't get a high flight rate without a competitive price.
There are also variable costs. Rocket Lab has flown 39 electrons and they are now building a production version of the rocket using an optimized factory process. Our new competitor is building prototype rockets and making changes along the way, and that is inherently a more expensive approach.
Rocket lab is currently developing reuse of their first stage, and hopes to fly a previously-flown stage in the near future. That has the potential to both reduce their per-launch variable costs and also allow them to increase their flight rate. Our new entrant wants reuse - and perhaps is in a better position to design in reuse rather than add it on as Electron has - but when entering the market it's just aspirational.
All of these factors give the established company a significant launch cost advantage over the new entrant.
And there's the additional factor of schedule and reliability. You can call up Rocket Lab to schedule a launch and they will give you a firm launch date, and with 19 successful flights in a row, your launch is likely to be successful. Our new competitor can provide neither a firm date nor a high chance of success.
And that's why starting a launch company is a bad decision...
If you enjoyed this video, build and launch a model rocket.
Pro tip: Never go to launch rockets with only one rocket as you may return home with zero rockets.
I suggest a rocket box...
There is a strategy in business known as "fast follower", where you wait for a company come up with a great idea and develop it and then you swoop in and utilize your advantages to out-compete them in this new area.
I think there may be some examples of that happening in the software industry, but I'm quite sure that it's not going to happen in the launch industry. The majority of the market opportunities that SpaceX took advantage of are gone and to compete in those segments you will need to compete with SpaceX.
But there might be a way...
Amazon's project Kuiper is under development, and it's a competitor to SpaceX's starlink space-based internet project.
Because of that, Amazon has no interest in sending SpaceX money to launch their satellites, so they've created a new market segment which I call "anybody but SpaceX"
Project Kuiper will require a very large number of launches to finish their constellation of 3236 satellites. Amazon has reportedly bought the last 9 Atlas V rockets for launch. They have also bought launches on 3 new rockets, with 18 launches on Ariane 6, 38 launches on ULA's Vulcan, and 12 launches on Blue Origin's New Glenn.
I suspect that there may be undisclosed purchases to fly on Rocket Lab's Neutron rocket, also currently under development.
If a new entrant could create a rocket that would be able to lift Amazon's satellites economically, I'm sure Amazon would buy launches from them, but all of these are big medium-class or larger rockets and that's a hard rocket to develop for new companies.
SpaceX and Rocket Lab are both successful launch companies.
SpaceX is currently dominating the worldwide launch business, but they understand that they can only keep high margins there as long as there isn't significant competition. They are diversifying into satellite internet with Starlink and leveraging their ability to launch cheaply on Falcon 9 - and someday, on starship. And that puts them into the "satellite bus" market as well...
Rocket Lab is perhaps more interesting. They have diversified into hypersonic test, a market where electron is a good fit for the market.
With the money they got when they went public, they diversified into space systems - reaction wheels, star trackers, radios, software, solar panels. That market is one that had not yet been disrupted and they currently make more money off of space systems than launch.
They are also building Neutron, my bet for the first Falcon 9 competitor that will fly. That will allow them to compete for some department of defense launches, and it will also set them up very nicely in the "anybody but spacex" market segment for customers who compete with SpaceX and therefore don't want to give them launch money.
The second company to start up around that time was Rocket Lab, though it took a few years for them to get funding to work on their Electron Rocket and it didn't fly until 2017.
(show electron )
Rocket Lab's target is small launch, with a payload of approximately 300 kg. It's much smaller than the payload of SpaceX's Falcon 9, but it's much cheaper. It would seem that Electron has that market sewn up, but SpaceX has recently been flying rideshare missions where it caries numerous payloads in one launch, and a 300 kg payload on a rideshare is only $1.7 million, much lower than the cost of an electron launch.
This is widely viewed as problematic for Electron, but I'm not sure as markets are weird things. Electron provides flexibility in launch timing and orbit selection that the Falcon 9 rideshare cannot.
It might be that the Falcon 9 rideshare pulls business away from Electron. And it might be that some of the new companies flying on Falcon 9 rideshare for testing their satellites choose Electron when they fly their production satellites.
At this point one of you is saying "what about rocket lab?" That's a great question, thanks for asking.
Electron can carry 300 kg into low earth orbit for about $7.5 million. Which is a lot cheaper than a Falcon 9 launch.
Unfortunately, SpaceX has a rideshare program where you can send a 300 kg into a low earth orbit for $1.65 million.
So Electron's real market is customers who want to fly to LEO in a very specific orbit.
And there is a market there - they did 9 launches in 2022 and will likely hit the same in 2023. And they've come up with a new application for electron - using the first stage as a hypersonic test vehicle. They are currently making money on launch but not a lot.
- customer who have small payloads that they want to put into very specific orbits and are willing to pay a premium for that.
The 2022 launch of Artemis 1 has shown that NASA is on track for being able to send astronauts to lunar orbit using the SLS launcher and Orion Capsule, a key part of the artemis lunar strategy. Next up will be Artemis 2, a trip with astronauts around the moon in 2024, and finally, the first lunar surface mission, Artemis 3 in 2025.
The lander for the lunar surface is a variant of SpaceX's starship vehicle, and obviously it needs to be ready for Artemis 3 or there is no lunar surface mission. There will also be a test flight of starship before putting crew on it that occurs sometime before Artemis 3.
After that, there are future flights for the remainder of the decade, with flights in 2027, 2028, and 2029, using either starship or an alternate lander. It's a slow cadence
Meanwhile, SpaceX has built a starship factory at Boca Chica, Texas that can produce one starship per month, and they have a completed launch site nearby.
If that weren't enough, they are very quickly building a second starship factory at their Roberts Road site at Kennedy Space Center and have already completed a starship launch tower at pad 39A where they currently launch Falcon 9 and Falcon Heavy.
Which brings up a very obvious question. If SpaceX has the capability to build lunar starships more quickly than SLS and Orion can launch, is there a way to use lunar starship without SLS and Orion?
This is commercial moon...
To answer our question, we'll need to dig into the details of what SLS and Orion do and then talk about different options that we might use to replace them.
This talk is about the technical details - I will not, for example, talk about what it might take to make it happen in the current political environment or who might pay for it.
We'll start by looking at the current Artemis architecture. If you want more details, see my video unsurprisingly titled "artemis architecture".
The first step is to launch lunar starship into a low earth orbit, and then fly enough tanker flights to refuel it.
Starship then relights its engines to send it out of low earth orbit and to the earth moon transfer point, where the moons gravity will take over. Starship will then use its engines to get into a near rectilinear halo orbit around the moon. There it will sit and wait.
SLS is then used to launch the Orion capsule with 4 astronauts to the earth moon transfer, and then Orion uses its engines to get into the same near rectilinear halo orbit as starship.
They dock, and two astronauts transfer to starship.
Starship lands on the surface, time passes, they come back up from the surface, get back in Orion, and orion uses its engines to get out of lunar orbit and head for home.
At that point there's an empty lunar starship left in Near rectilinear halo orbit.
I call this approach the artemis reference architecture.
Our options in space are controlled by the energy cost - what we often call the delta v cost. See my solar system road trip video for more information on delta v.
Let's add some labels to our chart to indicate how much delta v we need for each part of the journey. All of these numbers are estimates.
Starting with starship:
It takes 3050 meters per second of delta v to get from low earth orbit to a lunar transfer orbit, and an additional 450 meters per second to get into the near rectilinear halo orbit, for a total of 3500 meters per second of delta v.
Based on Apollo numbers, it will take about 2350 meters per second to get down to the moon and 2300 meters per second to get back up.
SLS generates 3050 meters per second to toss orion to the transfer point, and Orion generates 450 meters per second to brake into the near rectilinear halo orbit, and the same 450 meters per second to get out.
Add all of those up, and we find that starship needs about 8150 meters per second of delta v. This architecture is *really* hard on the landers. Orion requires 900 meters per second.
With that information, we can start exploring architectures that do not rely on SLS and Orion.
To do that, we will need to replace all of the things that SLS and Orion do.
There are three major requirements for an orion replacement.
First, it must have enough delta v to get into lunar orbit and out of lunar orbit, which I'll call 1000 meters per second.
Second, it has to have a robust enough heat shield so that it can reenter the earth's atmosphere at lunar return speeds and keep the interior from getting crispy at energies that are over twice the amount generated returning from low earth orbit.
Third, it must have enough life support to keep the crew alive for the time that they are in the capsule.
We will also need a way to get it to the earth-moon transfer point.
In our first option we will stick to the same architecture but replace both SLS and Orion with other options.
Capsules are very complicated and expensive to develop, so we'll need something that's either in service or close to going into service.
That limits us to two options, either SpaceX's crew dragon, or Boeing's Starliner. Both are designed to carry crew to the international space station, so they weren't really designed for this sort of mission.
Let's see how they stack up against our requirements.
We need 1000 meters per second of delta v. I was unable to find good numbers, but it appears that crew dragon has about 650 meters per second of delta v *if* you are willing to use the SuperDraco abort motors. Starliner looks to be around 500 meters per second, but some of that might be used to put the capsule into orbit.
To reach our 1000 meter per second goal will require modifications to carry more fuel. This would require a larger service module for starliner, and spacex would need to figure out a way to carry fuel in the trunk. Crew dragon masses about 12,000 kilograms at launch, and starliner masses about 13,000 kilograms. If we need to add 1000 m/s of delta v to their thrusters, we end up around 16,500 kg for crew dragon, 17,900 for starliner.
These are both in the "doable, but require some serious work" category.
Looking at heat shields, dragon is supposedly designed to deal with reentry speeds from lunar orbit, and the starliner heat shield is probably only intended to deal with earth orbit reentry, so it would need to be upgraded.
Looking at life support, the inspiration 4 flight of crew dragon lasted 3 days, and starliner is specified for 60 hours.
That's roughly half of what you need to get to the moon and back, but those numbers are with 4 crew members.
If we are okay limiting ourselves to 2 crew members, that bumps those numbers to 6 days and 5 days. That's in the right ballpark for a lunar mission.
If you can solve the delta v issue and add a bit of life support endurance, Crew Dragon looks viable. Starliner requires more work because of the heat shield, but it also looks possible.
Looking at possible launchers, our choices appear to be Vulcan, Falcon Heavy, and New Glenn.
Falcon Heavy can toss around 20,000 kg to the lunar transfer point. ULA's Vulcan rocket can only do about 12,000 kilograms, and Blue Origins upcoming New Glenn is likely pretty close to that.
Given what the modified capsules would mass, Falcon heavy looks like the only practical choice. And it still will probably be tight.
Falcon Heavy is not crew rated but is based on the Falcon 9 which is crew rated.
This option looks viable, you can probably replace SLS and orion with an enhanced version of either capsule, but it would require some reengineering, with the added fuel being a significant obstacle.
Our next option is one I'm calling "A blast from the past"
The constellation program from the early 2000s was targeting low lunar orbit, and the orion design did not have enough delta v to get into that orbit and back out of it.
One of their architectures used an Earth Departure Stage to do the work of getting Orion out of low earth orbit and into low lunar orbit, and that gave orion enough delta v to get out of low lunar orbit and back to earth at the end of the mission.
Can we do something similar?
This approach has a lot of allure, because we already have two crew-rated launchers, and not doing crew rating for Falcon Heavy makes things easier.
But can we launch the heavier versions of the capsules that will be required for Lunar Missions?
The Falcon 9 part is pretty easy; it can launch 16,600 kg in reusable mode, and up to 22,800 kg in expendable mode.
The Atlas V is a bit harder; the highest payload for the Atlas V is 18,850 kg, which would probably work, but that's pretty close the how much the lunar starliner might weigh.
However, there aren't any available Atlas V rockets anyway, so for this mission, Starliner would need to launch on ULA's Vulcan, which can do up to 27,200 kg to low earth orbit.
Or starliner could launch on Falcon 9 as it can fly on different launch vehicles.
Get the heavier capsules into low earth orbit will not be a problem.
Now we need a departure stage. Here's a small list of requirements:
First, it needs to be within the payload capacity of Falcon 9 or Vulcan. That means 22,800 kg for Falcon 9, up to 27,000 kg if we want to use Vulcan. Vulcan Heavy could be an option if we found something heavier.
Second, it needs to be COTS, or commercial off the shelf technology. No sense in spending a ton of time and money to create a new stage.
Third, it needs to be efficient - we don't have much payload and we need a lot of delta v out of it.
The efficiency requirement means we'll need to chose a liquid hydrogen and liquid oxygen, or hydrolox engine. This sort of mission is where hydrolox shines.
There's really only one active US hydrolox stage, and that's Centaur.
Centaur turns out to be a great choice.
It has a dry mass of only 2,250 kilograms, and a propellant mass of 20, 830 kilograms. That gives it a total mass of 23,080 kilograms, just 280 kilograms more than the maximum payload of the Falcon 9 and well within the maximum payload of Vulcan. Cut down the propellant just a bit, and it will launch on either Falcon 9 or the cheaper variant of Vulcan with only four solid rocket motors.
It fits inside of the Falcon 9 extended fairing and the large Vulcan fairing. It might sneak into the short Vulcan fairing, but the clearances look tight.
How much delta v can we get out of centaur pushing one of our heavy capsules?
Here's a chart that shows the delta v for centaur with different capsule masses.
I'll add a line at 3100 meters per second of delta v as that is what it takes to get from low earth orbit to the earth moon transfer point.
We see that we can get 3100 meters per second of delta v even if the capsule has a mass of 18,000 kilograms.
Given the masses that we figured for the capsules, this option looks good for crew dragon, but marginal for starliner because of the new heat shield requirement.
The mission is similar to the reference approach.
Centaur is launched first, then crew dragon. They dock, and then centaur gives crew dragon enough of a push to get it to the earth moon transfer. Dragon gets into lunar orbit and then gets out when it needs to head back to the earth at the end of the mission.
It turns out that Centaur is coincidentally the perfect size for this approach, and it can put crew dragon into the required orbit.
It might also work for starliner.
Option 3 tries to leverage the unique capabilities of starship.
Here's how it works:
We launch our crew on dragon into orbit, and then stash the dragon inside of the lunar starship. Or perhaps docked on the front.
Starship will drop off the dragon in lunar orbit, and after the surface expedition, the crew will get in dragon for the trip home.
This approach has a significant advantage in that dragon only has to supply the 450 meters per second of delta v to get out of lunar orbit and the life support only has to last for the trip home. If the SuperDraco abort motors could be repurposed, the current version of crew dragon might be capable of doing this with minimal modifications. Starship would also need to be modified to either carry dragon internally or to be strong enough to take the stress of a dragon hanging out on the nose.
There is a downside. Lunar starship is nominally specified to carry 50 tons to the lunar surface, but it needs to carry the 12 ton dragon as well, so that reduces the net payload to 38 tons. It would be lower than that if starship cannot carry a full 50 tons.
On the other hand, the NASA requirement for the HLS landers was about 1 ton to the lunar surface, so starship provides many times that.
This option is simple, and uses things SpaceX will already be building.
And finally we get to Option 4, Starship-o-rama
The goal is to do the mission relying only on Starship.
Can we just bring lunar starship back home?
There are two big obstacles.
The first is that we don't have 450 meters per second of delta v to get out of orbit, but even if we did, lunar starship doesn't have the heat shield and fins that would allow it to reenter earth's atmosphere. That would be a non-starter.
But there may be a way to make this work.
We can fuel up a normal starship so that it has 4100 meters per second of fuel, and it will have enough delta-v to get into lunar orbit to pick up the astronauts, get out of lunar orbit, and then come back and land on the earth. That's assuming the starship heat shield can handle lunar return velocities. If so, that's pretty straightforward, and 4100 meters per second only takes about 35% of a full fuel load, so it's fairly straightforward.
It does, however, still require a new lunar starship for every mission.
This approach is easy to do if starship functions the way it is expected to
It does - obviously - require putting crew on starship both for launch and reentry
And finally to option 5, full reuse.
To do this we'll need to get clever about getting lunar starship back to low earth orbit.
If we take the retrieval starship from option 4 and stuff it full of fuel with minimal payload, it will have around 8000 meters per second of delta v. The trip to lunar orbit takes 3500 meters per second leaving 4500 meters per second. That's about 360 tons of fuel.
To get it back to earth will take 450 meters per second, which is surprisingly only 24 tons of fuel. That leaves us with 336 tons of fuel.
Stuff that in the empty lunar starship and that yields 3900 meters per second of delta v, which is enough to get it out of lunar orbit and to brake it gently back into low earth orbit. It's the inverse of how lunar starship got to the moon.
You have multiple choices for crew.
Option 5A puts them in starship for the whole trip. Option 5B uses dragon to get them into low earth orbit and back home.
For this approach, it appears that full reuse is within the capabilities of starship.
Let's summarize the logistics required for each of these options. I'm going to stick with the dragon ones but remember that there may be starliner options.
Note that hardware might be reused, it might be partially expended, or it might be fully expended. See the color code at the bottom.
Option 1 - direct capsule replacement - requires 1 modified crew dragon, 1 falcon heavy, and 1 lunar starship. It also requires 8 tanker flights to refuel lunar starship (that number might be different)
Option 2 - the centaur variant - requires 1 modified crew dragons, 2 falcon 9, 1 centaur, and 1 lunar starship
Option 3 - ride along - requires 1 crew dragon, 1 falcon 9, and 1 lunar starship
Option 4 - starship o rama - requires an normal starship, a lunar starship and 3 extra tanker flights to give it enough fuel.
Option 5A - full starship reuse - requires 1 crew starship and one lunar starship. It also requires 16 tanker flights
And Option 5B - full starship reuse but using capsules - requires a crew dragon, a falcon 9, a starship, and a lunar starship
I came up with some prices for my model.
This is not the way to do this - numbers should be better researched and should factor in launch rate. And they probably should include research and development as well.
But I did it anyway because I know that people want numbers. Just note that these numbers are all wrong, some slightly wrong, some significantly wrong.
And they are intended to be *prices* rather than *costs*
Plug those numbers in and here's what you get - somewhere between $300 and $800 million for all the options.
I'm not going to talk about them in detail because it's really easy to push the numbers around with different assumptions. If you want to play with the numbers, let me know and I'll send you my spreadsheet.
Thanks for your attention.
In summary, there a number of different practical architectures that could be used to get to the lunar surface without relying on SLS and Orion, and they're reasonably affordable.
If I had to choose one, for the short term it would be option 3, ride along. It can probably be done with the least amount of work.
The full reuse of option 5 looks best for the long term
If you enjoyed this video, please draw me a picture of a centaur riding a centaur.
In the 1960s, rocketry was all about missiles - missiles like the titan I, the thor delta, and the atlas.
They were designed for military use and then adapted for scientific use.
But there were others looking at commercial uses for space and an obvious one was communication.
The first true communications satellite was intelsat 1, otherwise known as "Early Bird". It was launched in April of 1965 by a delta d rocket, and it because the first geosynchronous communications satellite.
Satellites in geosynchronous orbit have been the main commercial cargo for many years. And because they are a lucrative market, there has been a lot of competition launching them.
The first few years was all delta, then atlas joined in. During this period, delta launched 30 satellites and atlas launched 16. Delta could launch satellites up to 920 kilograms, and atlas, up to 1515 kilograms.
By the last 3 years of this period, this business had grown to about 20% of the total launch business for both delta and atlas.
It was a good business, and there wasn't any competition. If you were in the west and wanted to launch, you used a US launcher.
But there were two competitors in the wings, one expected, one unexpected.
The expected competitor was the shuttle...
The space shuttle was coming and it was going to revolutionize the launch industry. NASA was projecting very cheap flights and the shuttle could carry multiple satellites, each more than 1000 kilograms.
The second competitor was a bit of a surprise.
It is very well known that the US recruited a large number of german rocket scientists to come to the US, and the USSR had a similar operation.
What is not well known is that the French also had a program, recruiting 90 scientists and engineers and housing them in Vernon in Normandie.
One of the most notable ones was Karl Heinz Bringer, who later changed his name to Henri.
He was the main designer for the Viking engine, a 600 kN engine running on the hypergolic propellants dinitrogen tetroxide and unsymmetrical dimethylhydrazine.
This was important because there was a new player in town.
In 1975, a large group of European countries got together and formed the European space agency. They developed a purely commercial launcher named Ariane based upon the Viking engine. Ariane showed up in three variants, with payloads of 1850, 2180, and 2700 kilograms. These payloads were quite a bit larger than Delta and Atlas, and Ariane supported "dual deployment" where two satellites could ride on a single mission and share the launch cost.
ESA had no launch site locations in Europe, so they chose a site at Kourou on the coast of French Guyana. Their site is only 5 degrees north of the equator, so it takes a small amount of propellant to move geostationary satellites to their final orbit over the equator.
Cape Canaveral is at 28 degrees north and therefore launches from there take considerably more propellant to move over the equator.
The launch site location gives Ariane a built-in payload advantage for geosynchronous satellites
The 1980s was an interesting decade.
Delta and Atlas kept pace, launching 29 comsats for the decade.
Ariane was very strong, with 41 payloads delivered to orbit. This looked to be problematic for Delta and Atlas.
The shuttle started very strong, with 3, 4, and 8 launches in the first 3 years, but in January of 1986, Challenger blew up and there would be no more commercial satellites on shuttle. Shuttle ended up with 16 total launches.
The launch pace was picking up.
The 1990s would bring three new rockets. The delta 2 would adopt the now familiar central rocket with up to 10 solid rocket boosters, it could launch 1820 kilograms to GTO.
The atlas II would keep a traditional design but debuted the centaur 2 upper stage. The base versions could lift 2810-3000 kilograms to GTO, and a version that added 4 solid rockets pushed that up to 3630 kilograms.
Communications satellites were getting bigger.
Ariane 4 came with a confusing 6 different variants. Not only did is support 2 or 4 solid rocket boosters to increase payload, it could alternatively support 2 or 4 liquid rocket boosters.
The simplest variant could launch 2100 kilograms to gto; the beefiest, 4720 kilograms.
These three rockets competed in the 1990s.
Delta II launched 30 payloads, Atlas II launched 27, and Ariane 4 launched a massive 133 payloads.
There was a problem for the US providers.
There were a two incremental attempts at fixing that problem. Lockheed Martin created the Atlas III by tossing the Atlas first stage that dated from the 1960s and creating a new one that relied on the Russian RD-180 engine. It had a payload of 4500 kilograms and flew 6 times with 3 communications satellites launched.
McDonnell Douglas came up with the Delta III. They took the existing first stage and booster configuration and built a new high performance upper stage that used liquid hydrogen and liquid oxygen.
It was rated to carry 3810 kilograms to geosynchronous transfer orbit, but it failed on its first two launches and did not reach the desired orbit on the third, and was quickly cancelled.
The US department of defense had big plans for space launches and no launcher that met their needs and something clearly had to be done.
they came up with the Evolved Expendable Launch Vehicle program in the mid 1990s. The idea is that the government would contract with one company to build a launch vehicle up to the task and that company would then have enough launches to be able to compete with Ariane in the commercial space. That high launch rate would reduce the price per launch.
There were four bidders - McDonnell Douglas, Lockheed Martin, Boeing (this was before the merger with McDonnell Douglas), and alliant Techsystems, or ATK.
At this point the department of defense awarded contracts to McDonnell Douglass and Lockheed Martin, to the surprise of absolutely nobody. They had decided that two launchers was better than one.
Those two companies immediately cried foul, as their bids had been made assuming they would get all the launches. The DoD resolved this by deciding to give money to both companies for being ready to launch even if they didn't technically perform any launches.
You might think that I'm making this up but I assure you that I am not. For the full story, see "spies, allies, and enterprise - the strange story of ULA.
McDonnell Douglass won with the Delta IV Family, which was based on a new common core using a RS-68 engine and was augmented with 2 or 4 solid rocket boosters in the medium + variant, and three common cores with the heavy variant.
The medium variants could lit 5300-7300 kg to geosynchronous transfer orbit, and the beefy heavy variant could lift 14,220 kilograms.
Lockheed martin won with the Atlas V. It came in a thin 4 meter version with up to 2 solid rocket motors and a thick 5 meter version with up to 5 solid rocket motors. Like the Atlas III, it used the Russian RD-180 engine on the booster stage and a new centaur III for the upper stage. It could lift from 4570 kilograms to 8900 kilograms.
These two rockets together would cover all the department of defense requirements.
Ariane was not standing still, and it developed the Ariane 5, which first flew in 1996.
Like both the Atlas V and Delta IV, it debuted a new first stage, this one being powered by a new hydrogen oxygen engine known as Vulcain and adding in two giant solid rocket motors. The first version would launch 6950 kilograms to geosynchronous transfer orbit, and the later enhanced version would launch nearly 10,900 kilograms.
These numbers were quite competitive with the Atlas V and delta IV.
In perhaps weirder news, Lockheed martin in the US partnered with Energiya and Khrunichev in Russia to create international launch services in 1995. Lockheed martin would bring the Atlas V to the table, and the Russians would bring the Proton, which could carry 6300 kilograms to geosynchronous transfer orbit.
In the 2000s we would have 3 new rockets and one very old Russian one.
Delta IV was not a competitive commercial launcher. That one launch in 2002 was the only communications satellite launch for Delta IV, and Delta IV M+ launch.
18 payloads for Atlas, most of which seem to have come from the ILS collaboration.
The idea that EELV would result in commercial sales and cost savings doesn't seem to work out very well.
Ariane not surprisingly doing a ton of payloads. The first few years are a mix of Ariane 4 and 5. 97 total payloads for 10 years, which is pretty nuts.
The real surprise is Proton, which hits 43 payloads, nearly half of what Ariane did.
Proton is selling a *lot* despite having a spotty quality record - in this time period Proton had 77 successful flights and 5 failures, for a 94% success rate. Ariane is way up at about 98%.
The higher failure rate means higher insurance premiums for the payloads and a better chance of your satellite failing. Proton's success is a good sign that there is more launch demand than launch slots.
Falcon 9 started flying in 2010, and flew its first GTO payload in December of 2013.
In the 2010s, Atlas pretty much disappears, and Ariane launches 101 payloads. Proton is about the same as the previous decade at 41 launches, but their flight rate went down significantly in the last 5 years when their success rate dropped to 86%
The lack of additional capacity for Ariane and the quality problems at proton left an opening for SpaceX, and Falcon 9 shows up strong, launching 30 payloads over the last 7 years.
That puts the total pretty high, showing that there was a significant backlog at the beginning of this decade.
And finally we end up in the current decade.
After 2019, Proton was killed by quality issues and geopolitical issues that made ILS an unattractive option, along with SpaceX providing such good service.
ESA had started work on Ariane 6 in 2015 with a targeted first launch in 2020, but it was significantly delayed and has not yet flown. That left Falcon 9 as the only active launcher for communications satellites and they took advantage of it, flying 22 in the decade so far.
It's obviously true that Falcon 9 Block 5 will continue to launch GTO satellites. It currently has 7 launches on the schedule for 2024.
Ariane 6 is still under development. It is scheduled to launch in 2024 but doesn't have any GTO launches scheduled until 2025.
With the advent of Starlink and other planned satellite constellations, it's not clear what the geosynchronous market will be like in future years.
If you enjoyed this video, please send this T shirt commemorating the launch of Intelsat 3-3 on February 2, 1969 to a geosynchronous transfer orbit.
Welcome to the Crazy Nuclear Rocket Engines series.
We will be talking about a number of different engine designs of varying degrees of craziness, but there is some important background information to cover before looking at specific designs. That is the purpose of this video.
I am hugely indebted to the "Atomic Rocketships of the Space Patrol" website by Winchell Chung.
It contains a huge amount of information about all the engines I'll be covering, including links to the source material. It's just a fabulous reference for what is science fact, what is science fiction, and how the two intersect.
There's a link in the video description.
To talk about nuclear rockets we need to understand nuclear fission. There are a few different elements that might be used as nuclear fuel, but uranium is the most common, specifically uranium 235.
The 235 is important as this specific kind - or isotope - of uranium is much less stable than the more common uranium 238.
A Uranium 235 atom has a chance of spontaneously breaking apart. When that happens, we get:
Two new atoms made up from the protons and neutrons that were in the original atom. Here it's xenon 140 and strontium 94, but the elements created vary.
There is also a gamma ray created, a photon of very high-energy radiation.
And there are also a few neutrons created. If these neutrons hit other atoms of uranium 235, they will cause those atoms to break apart, or fission, releasing more neutrons
There are three things that can happen after that initial fission:
In the first case, the first fission produces two neutrons and one of those finds another uranium 235 atom to fission, but after that there is no more fission. The reaction just dies. This is known as "subcritical"; the uranium is mostly just sitting there, as it is in uranium ore.
In the second case, one of the neutrons from each fission causes another atom to fission, so there is a steady level of fission activity. This is known as Critical, and this is where nuclear reactors are operating when they are in a steady state. There might be a little or a lot of fission happening, but the rate is steady.
In the final case, more than one neutron from each fission causes another atom to fission, so the amount of fission going on is increasing. This is known as "supercritical". If this increase is fast enough, we end up with a nuclear explosion.
It is therefore important to control the reaction so it continues at the desired rate.
There are various ways to control the speed of the reaction.
If we have a collection of uranium atoms that are subcritical, we can surround them with water. The neutrons will collide with the hydrogen atoms in the water and slow down, and that increases the chance of each neutron causing another fission. Graphite can also be used for this purpose. These are confusingly called "moderators" because they slow down (moderate) the speed of the neutrons, but they *increase* the rate of reaction.
There are other chemical elements - such as boron or cadmium - that absorb neutrons. They are used in control rods in nuclear reactors; drop them down into the core and they will absorb neutrons and reduce the rate of the reaction.
There are other elements that can reflect or scatter neutrons. If we have a nuclear reactor, the neutrons that come out the side of the reactor core are lost. If we put a neutron reflector like Beryllium next to the core, some of the neutrons will be scattered back to the core and be able to cause further reactions.
To understand what these engine designers are trying to do, we need to understand what make a rocket engine good.
I'll link two videos for background.
What's up with specific impulse talks in depth about what specific impulse is and what affects it.
Planning your solar system road trip explains why you should care about delta v.
Delta v is a measurement of the useful work that we can get out of a rocket. It is defined as the specific impulse times 9.8 times the natural logarithm of the starting mass over the ending mass.
The specific impulse - generally written as Isp - is a measure of the fuel economy of the engine - how much useful work is available for a given amount of propellant.
Starting mass is the mass of the rocket or stage all fueled up and ready to fly.
And ending mass is that same rocket or stage when it has burned up all of it's fuel.
Starting mass divided by ending mass is sometimes known as the mass ratio.
We want more delta v. There are three main ways that we can do it.
We can use an engine that produces a higher specific impulse
We can create a rocket that has a higher starting mass by packing more propellant into it.
Or we can create a rocket that is lighter with a lighter airframe or lighter engine.
Looking at specific impulse in more depth.
Exhaust particles with more energy per unit of mass lead to a better specific impulse.
There are two main rules:
The first is that lighter exhaust particles are better, so using propellants with light atoms is better than heavy atoms.
Nuclear thermal rockets might use different propellants, but they generally use liquid hydrogen, as it gives the lightest exhaust particles. This gives them a big advantage over chemical rockets.
The second rule is that hotter propellant temperatures are better as it means more energy per exhaust particle.
Unfortunately - for reasons we'll explore in later videos - many nuclear engines use relatively low temperatures, which reduces the specific impulse they produce. Since they already use the best propellant choice, the focus of designers is on creating hotter propellant temperatures.
Now looking at mass ratio
Higher mass ratio yields better delta v
The first rule is that lighter engines and structures are better.
Unfortunately, everything in the nuclear rocket is heavy. The nuclear core is heavy, all of the shielding is heavy, and the supporting structure is heavy. All that mass means the ending mass of the rocket is greater.
The second rule is that more propellant mass is better. Unfortunately, hydrogen propellant is not dense, so you can't put much into a given size tank.
What we end up with is an engine with great specific impulse and poor mass ratio. Who wins? Well, it depends on the details - exactly what specific impulse you get and how heavy your engines and tanks are.
This isn't clear without looking at the numbers for a specific scenario.
To overgeneralize, nuclear thermal literature:
Talks about how great the higher specific impulse is.
Ignores or underplays the impact of mass ratio
If you enjoyed this video, please check out the atomic rockets site.
Welcome to Crazy Nuclear Rocket Engines Episode #1.
We will start with the first nuclear rocket engine, NERVA. If you haven't watched it yet, I highly recommend watching the crazy nuclear rocket engines introduction first.
https://www.youtube.com/watch?v=eDNX65d-FBY
Nerva stands for
Nuclear
Engine for
Rocket
Vehicle
Applications
It came out of the "atomic" 1950s, when there were designs for atomic locomotives, atomic planes, and even atomic cars.
To understand how nerva works, we'll start with nuclear power plants.
In a nuclear power plant, we have a reactor that is used to heat high pressure water.
That water travels into a steam generator where it heats up water to make steam, which is fed to a turbine generator to make electricity, which we use to make toast.
The whole point is to use a nuclear reactor as a heat source.
Let's convert that to a nuclear thermal rocket. We'll start with the reactor. We add a water tank to feed water into the reactor, put a rocket nozzle on the outlet, and we get a nice stream of steam coming out the nozzle which gives us thrust.
Unfortunately, it's not a very good rocket, because water molecules are heavy.
We empty the water tank and fill it with liquid hydrogen as propellant, and we get hot hydrogen out the exhaust. Much better from an efficiency perspective.
Here's how we make the reactor.
We mix highly enriched uranium and powdered graphite, put it into a pasta machine and use a die like the one used for rotelle, and then extrude out fuel rods with a bunch of holes in them. The hydrogen will flow through these holes and get heated up.
We then plate the fuel rods with Zirconium Carbide to prevent corrosion, and also create similarly sized "tie tubes". These are arranged in a hexagonal pattern and ultimately into a reactor core.
Around the core are a series of control drums; these contain moderators on one side and neutron absorbers on the other side, and rotating them allows the reaction to be controlled.
The resulting reactor core - which is around 83 centimeters in diameter and 137 centimeters long - is put into a metal tube, which is similar to the combustion chamber on a chemical rocket.
We now start up the reactor and pump liquid hydrogen into the reactor core, which is running at 2400 Kelvin. The heat vaporizes the liquid hydrogen and it expands and leaves the nozzles very quickly.
Because the exhaust is pure hydrogen and hydrogen is very light, the exhaust velocity is very high and that produces a high specific impulse, or fuel economy - roughly double what is possible with the best chemical rockets.
There are some downsides of nuclear thermal rockets.
The first is that 2400 Kelvin core. Hotter is better in terms of specific impulse, and chemical rockets run at around 3500 Kelvin. Unfortunately, the hotter you run nuclear cores the more likely they are to melt, which means there's a tradeoff between temperature and longevity.
The second downside is their weight. The core itself is quite heavy, and heavy shielding is necessary to shield the rest of the rocket from the radioactivity of the core. Thrust to weight ratio is very important for rocket engine performance and the extra weight generally cancels out the advantage of higher specific impulse.
The third downside is the liquid hydrogen propellant. It takes big and heavy tanks and because it is very cold it is difficult to keep liquid.
Oh, yeah. As soon as you start it up the engine it becomes very radioactive and nobody can get near it without a lot of shielding.
In the 1960s there was an extensive development program for Nuclear Thermal Engines run out of Los Alamos, with a test complex at Jackass Flats in New Mexico.
Test configurations were built onto rail cars which were then moved to remote test cells for testing with the engine exhaust pointed up into the air. This shows the Phoebus 1B engine on its way to testing in 1967
These reactors are highly stressed; this one operated at a heat output of 1500 megawatts, which is perhaps half the output of a nuclear power station.
Though the development work was ultimately successful, they did have ongoing issues with both erosion and cracking of the fuel elements.
Here's a video of one of the tests. And yes, they did just send the exhaust up into the air, sometimes with parts of fuel rods.
The last tests were done in March of 1969.
Since NERVA, there has been a lot of further research and design work done on nuclear thermal engines, which, with a few exceptions, has generated lots of paper and very little hardware.
For the last few years, there has been a program at NASA to build an advanced nuclear thermal rocket.
There are small contracts awarded for initial designs for small nuclear thermal engines.
If you want to explore nuclear thermal rockets in more depth, see my video "build your own nuclear space tug!"
Here's a scorecard for nuclear thermal rockets...
On the plus side, they have been built and operated, they are not likely to become explody, and they have a high specific impulse.
On the negative side, the high specific impulse is negated by the heavy engine and heavy shielding required, and they are radioactive as hell both during and after use.
Overall, I'm going to give them a nuclear rocket craziness rating of 3; I don't think they are a great idea but you can probably build a real engine that generally does what you expect it to do.
If you enjoyed this video, please write a tasteful joke about the jackass flats nuclear test site.
This is episode 2 of the series on crazy nuclear rocket engines, where we will examine the centrifugal nuclear thermal rocket design.
If you haven't watched it, please watch the introduction video now.
http://local.ans.org/ne/wp-content/uploads/2021/02/OverviewCNTR-ANS-Winter-2020-summary-paper.pdf
We're going to build a nuclear rocket engine based on a carnival ride. The gravitron.
The gravitron has a large drum that spins around.
The rotational force pushes the occupants to the outside and eventually provides enough force to lift them off their feet.
Here's how we build our engine
We build a drum with an axle at the top so that we can spin it around. At a few thousand RPM
Now we can put liquid uranium on the inside of the drum. The force of the spinning drum holds the molten uranium to the wall, and the little lip at the nozzle end keeps the uranium from just slipping out the nozzle.
We can then inject hydrogen into the middle of the reactor, and the hydrogen will... Well, it won't do that much because it's not in contact with the hot liquid uranium. Worse, our drum wants to melt because it doesn't have any cooling.
Hmm....
We add an outer drum and flood the space between it and the inner drum with hydrogen to cool it. And then we drill a bunch of tiny holes in the drum so that the hydrogen can bubble through the molten uranium and get superheated, to 5000 K. Much hotter than the solid core, and therefore giving a specific impulse of about 1500 to 1800.
That sounds great.
Here's one interpretation of this design. This one use 19 separate rotating cylinders.
https://sci-hub.se/https://arc.aiaa.org/doi/pdf/10.2514/3.28189
https://sci-hub.se/https://arc.aiaa.org/doi/pdf/10.2514/3.2141
http://local.ans.org/ne/wp-content/uploads/2021/02/OverviewCNTR-ANS-Winter-2020-summary-paper.pdf
One of the papers on this reactor has a section labelled "Initial Risk Reduction Efforts", which is engineer-speak for "things we don't know how to do that might violate the laws of physics", and I thought that was an appropriate title.
The first problem is the drum. The uranium will have a lot of force trying to push it through the drum wall and yet the drum has to prevent that and let hydrogen go the other way in large volume.
The rotational mechanism will require tight tolerances and needs to be smooth despite being filled with an inherently sloshy load of molten uranium.
Figuring out how the fuel goes from solid to liquid and ends up in the right place will be challenging.
And finally, there is a lot of hydrogen bubbling through the very hot fuel, and that will inevitably cause some uranium to leave the spinning surface, and some of that will be carried out of the reactor by the fast moving hydrogen.
There's a fun variant of this design called the pebble bed, with the same basic architecture but instead of using molten fuel, you use little uranium pebbles.
They're a bit like peanut M&Ms, with the nuclear fuel acting as the peanut, the carbon layers acting as the chocolate, and the zirconium carbide playing the role of the candy coating.
The advantage of pebble beds is that the pebbles are much stronger than the fuel rods and therefore can be run at a higher temperature. Pebble beds have been used in other reactor designs
In the pebble bed engine, you need to spin the engine fast enough that none of the pebbles escape out the nozzle, which seems complicated with a high hydrogen flow, but probably much easier than it is with liquid uranium.
Here's a scorecard for centrifugal nuclear thermal rockets
On the plus side, the liquid version might yield a specific impulse of 1500-1800, or about double a solid nuclear thermal rocket
On the negatives side...
It might become melty or explody if the molten uranium sloshes around rather than staying in a nice smooth layer.
The mechanical design requires a drum rotating at several thousand RPM
It's unclear how to start up and shut down the reactor
There's a potential for constant uranium loss out the exhaust, which not only loses uranium but exposes the crew to more radiation from fission products outside the shielding.
And like other nuclear reactors, it's heavy and it's radioactive.
And finally, gravitor is more fun
I'll give it a 5 for a nuclear rocket craziness rating.
If you enjoyed this video, please find a gravitron and ride it. We will also accept reasonable substitute rides such as gravitor, rotor, or zero gravity.
Welcome to episode #3 of the series on crazy nuclear rocket engines, where we will examine the colloidal and droplet core designs.
If you haven't watched it, please watch the introduction video now.
https://patents.google.com/patent/US3711370
Here's how the colloidal core works...
You make a big chamber in the shape of a flat wheel, and then you cut a hole in the bottom of it and attach a rocket nozzle.
From a top view, it looks like this.
You then take your fuel particles - uranium particles coated in zirconium carbide for toughness - and mix them with hydrogen.
That mixture is injected through a specially designed set of nozzles, and the nozzles plus vanes in the chamber create a swirling vortex.
The vortex pushes the uranium particles towards the outside of the chamber, where they fission, and the hydrogen is heated and pushed towards the middle, where it exits through the nozzle, producing thrust.
The carefully-tuned vortex keeps the uranium from exiting - at least most of the uranium. You will probably still lose a few kilograms per minute of operation.
All of that gets you a specific impulse of around 1200.
And the droplet core...
Here we take uranium and heat it up to 2000 K (how that is done is left as an exercise to the reader), and then we push it through an atomizer (think perfume mister) so we have a big chamber full of uranium mist. At the same time, we inject hydrogen into the mist.
The top part of the chamber is designed to keep the fission reaction going. We also pump hydrogen in from the side so that the uranium doesn't clump against the walls and melt them. This is where most of the hydrogen heating takes place.
Now we have a problem. Both the hydrogen and the uranium are going to be headed out the nozzle, and we need a way separate the uranium. We use the same approach as in the colloidal reactor, injecting hydrogen so it swirls in the middle of the chamber and pushes the majority of the uranium towards the outside walls.
We now have a lot of extremely hot uranium. We inject liquid lithium in the bottom part of the combustion chamber to cool the uranium down so we can separate it from the lithium for re-use it.
Here's a scorecard for colloidal or droplet cores.
On the plus side, they can generate a decent specific impulse, maybe 1200 to 2000
On the negatives side...
There's a lot of dependence on the power of the vortex to limit the fuel loss, and any uranium or fission fragment loss during operation will increase the radiation on the crew.
Things can easily get melty if the fuel clumps together.
It's not clear how the solid fuel is melted initially
Liquid lithium adds a lot of complexity and we need a lithium/uranium separator
And finally, it's heavy and radioactive.
It gets a solid 5 on the craziness score. I think you can build it, but I also think it will simultaneously melt down and send most of the uranium out the nozzle.
If you enjoyed this video, please eat some jello with blueberries in it.
Welcome to Crazy Nuclear Rocket Engines Episode #4: Pulsed Solid Core
If you haven't watched the introduction video, please do that now.
Expanding on the fission discussion in the introduction, when a uranium 235 atom splits, the are multiple products.
First, there are fission fragments, the new atoms that are made up from most of the mass of the uranium atom.
Second, there are fast moving neutrons that used to be part of the uranium atom and a high-energy gamma ray.
The fission fragments are moving very fast and therefore contain 95% of the energy that is created. That motion - or heat - is what heats up the nuclear core, and - if we get too much of it - what melts the nuclear core.
The neutrons and gamma rays contain 5% of the energy that is created, but it is energy that can travel outside the solid core.
The goal of the pulsed solid core is to harness that 5% to superheat the hydrogen to a higher temperature than the general temperature of the core.
We start with a hydrogen channel that is right next to a thin vertical channel of uranium fuel.
We then use external means to increase the reactivity of the fuel, so it goes from very little fission to a large amount of fission quickly.
This generates a large pulse of neutrons, which travel into the hydrogen fuel and heat it up very quickly. Because we are heating by neutrons, we don't have the temperature limits that we had in the solid core design, so we can end up with much hotter hydrogen and therefore a higher exhaust velocity and specific impulse
Immediately after that pulse, the resulting heat pulse melts the fuel, destroying the engine. Which would seem to be a disadvantage.
However, what if we could put something next to the fuel to suck all the heat away, so that it wouldn't melt. It turns out that to conduct the heat away quickly enough, you need a material that is really good at conducting heat and has a reasonable melting point, so liquid lithium metal is the coolant of choice. The core sits idle until the lithium has pulled enough heat out of the fuel so it can survive another pulse.
But how to create that fission pulse?
Generally speaking, the goal with a reactor is to slowly and smoothly increase the power of the reactor. If we end up with a big pulse of fission, that is generally a bad thing - less like a reactor and more like a bomb.
Interestingly, this is a solved problem.
There is a research reactor design known as TRIGA that is used at many universities. Normally, it's max power is only 250 KW and it typically runs at much lower power levels, but it can also operate in a pulsed mode.
This is a video of TRIGA operating in pulsed mode. Watch for the blue flash.
That blue flash is known as Cherenkov radiation, and it's directly a result of the fission pulse. This pulse was 500 megawatts, or 2000 times the normal maximum output. The loud noise is a control rod being removed or replaced in the core very quickly.
TRIGA accomplishes this by using a very specific Uranium Zirconium Hydride fuel.
When this fuel is cold, it increases the rate of fission
When it gets hot, it decreases the rate of fission.
That automatically creates a pulse about 40 milliseconds, or 1/25th of a second wide.
Triga pulse
https://ansn.iaea.org/Common/documents/Training/TRIGA%20Reactors%20(Safety%20and%20Technology)/pdf/chapter1.pdf
Here's a diagram of the overall propulsion system, which is just like a standard nuclear thermal rocket except for the lithium coolant loop to take the excess heat away.
There's a neat little trick to use some of the heat from the lithium to preheat the hydrogen before it goes into the reactor.
https://sci-hub.se/https://arc.aiaa.org/doi/pdf/10.2514/6.2016-4685
To get sufficient thrust out of this approach the core is arranged as a stack of flat plates of fuel with channels for the hydrogen propellant and the lithium coolant.
Let's talk about heat and power.
A smallish nuclear thermal rocket engine has a thermal power of about 370 megawatts. That suffices to give roughly the thrust of an RL-10 rocket engine.
That energy needs to come from the 5% of the fission energy that goes to neutrons and gamma rays.
The lithium cooling loop therefore needs to deal with 95% of the fission energy, or 7000 megawatts.
It's not quite that bad because we can use the hot lithium to preheat the hydrogen, but any energy we want to use to go above the non-melting core temperature can only come from the 5%.
That waste heat needs to be radiated away. Current satellite radiators - such as these on the international space station - can radiate about 400 watts per square meter.
7000 million divided by 400 is equal to 17 million square meters of radiator.
Here's a scorecard for the pulsed solid core
On the plus side:
The specific impulse is only limited by the ability to remove heat, and might reach 5,000 to 15,000, or perhaps more
It can operate as a standard nuclear thermal rocket.
On the downside:
Liquid metal coolants are very heavy, and you need to be able to liquify all of it at startup. You probably need a dual-loop system.
Heat isn't much of a problem in the standard nuclear thermal reactor because it mostly goes into the hydrogen propellant. Here you need to be able to get rid of 19 times the heat that goes into the propellant. It's simply not feasible.
If the cooling doesn't work exactly right, you end up with molten lithium, hydrogen, and very hot uranium mixing together. That has "core explosion" written all over it, and while it's not a nuclear explosion, it's not going to be kind to the rocket engine.
And, as usual, it's radioactive and heavy, especially heavy because of the lithium cooling loop.
I give this design a Nuclear Rocket Craziness Rating of 7.
If you enjoyed this video, please listen to Amy Lee's haunting vocals on evanescense's 2006 song, Lithium
This is episode 5 of the series on crazy nuclear rocket engines, where we will explore the gas core, open cycle design.
If you haven't watched it, please watch the introduction video now.
Rather than spend engineering effort trying to keep our reactor core cool, let's let it get as hot as it wants.
Conceptually, it's fairly simple; you have something like a combustion chamber that you inject uranium hexafluoride into, and it starts fissioning and creates a uranium plasma, which may reach temperatures of 50,000 K. That is not a misprint.
It heats up the hydrogen that is being pumped into the chamber (along with Tungsten to improve the heat transfer), and the now very hot hydrogen exits the chamber through the rocket nozzle. As does some of the uranium.
One of the problems is that about 5% of the energy is neutrons and gamma rays, and they heat up the combustion chamber. That can be cooled with hydrogen but it limits your power and therefore your specific impulse to 2000-3000. If you want higher specific impulse, you need a radiator system to pull that heat out of the chamber. That will get you to an Isp of 7000 - 10,000.
There is a small matter of radiation...
The core in solid core nuclear rockets emits radiation in all directions, but it is possible to use a shadow shield to absorb most of the radiation, and anything in the shadow - such as the crew compartment - receives much less radiation.
In the open gas core you have still-fissioning nuclear fuel exiting through the nozzle and spreading out, and that fuel is emitting a lot of radiation. It's unfortunately outside the shadow and therefore irradiates the crew location directly.
There are many ideas on how to reduce the amount of uranium lost out the nozzle, pretty much all of which involve vortexes.
Here's a scorecard for the gas core open cycle engine
On the plus side, the core is very hot and can generate a specific impulse of 2000 to even 10,000
On the negatives side...
How do you create the uranium plasma that starts the thing?
There's a lot of dependence on the power of the vortex to keep the uranium from existing through the nozzle.
The chamber needs to deal with a large amount of radiation + neutrons, and it will be challenging to control the meltiness.
It needs a big radiator system to get the high specific impulse
And the crew will likely becomes somewhat glowy.
It gets a solid 8 on the craziness score.
If you enjoyed this video, please invest in global nuclear fuel, LLC.
This is episode 6 of the series on crazy nuclear rocket engines, where we will examine the nuclear light bulb, otherwise known as the gaseous core closed cycle engine.
If you haven't watched it, please watch the introduction video now.
https://ntrs.nasa.gov/citations/19920001892
Here's how we build our light bulb...
Start with a bunch of very thin quartz tubes that wrap around in a circle. That is the outer wall of the "light bulb", and it is cooled by running gas through it.
Inside that is a layer of neon gas, injected so it will swirl around the inside of the chamber in a - you guessed it - in a vortex.
And inside that is our fuel, uranium hexafluoride gas, with more neon inside of that.
That gives you a *very* hot gas nuclear core that puts out a ridiculous amount of radiation, hence the "light bulb" term.
Outside of that goes a jacket of hydrogen propellant. Unfortunately, the hydrogen doesn't readily absorb the radiation that the light bulb puts out, so you need to add some Tungsten to heat up and transfer its heat to the hydrogen.
The goal is to be able to run the core at temperatures up to 20,000 K.
Here are two drawings that make it a bit clearer. The uranium hexafluoride fuel and hydrogen are injected from the left, and the hydrogen exits to the right.
In the vehicle, you put 7 of the bulbs together.
Here's a scorecard for the nuclear light bulb
On the plus side, they can generate a decent specific impulse of around 2000, and the nuclear light bulb is a very bright idea.
On the negatives side...
There's a lot of dependence on the power of the vortex to keep the fluorine contained.
Flourine eats everything. Here's a quote:
Readily capable of detonation or explosive decomposition or explosive reaction at normal temperatures or pressures. Reacts violently with water.
Flourine at 25,000 K is unbelievably reactive.
The quartz containers are fragile and need lots of cooling or they will break.
If you look at the fault tree, every scenario ends with "rocket explodes".
It gets a solid 9 on the craziness score.
If you enjoyed this video, please visit the closest lighthouse.
This is episode 7 of the series on crazy nuclear rocket engines, where we will explore the fizzer.
If you haven't watched the introductory video, please watch it now.
The fizzer is pretty simple.
We start with a long tube like a solid rocket motor.
There is a long rod of uranium 235 running down the middle of the rocket with a cadmium coating on the outside, just enough to keep it from going critical.
Outside the rod is the lithium hydride propellant.
To start the reaction, you drop off a section of the cadmium control sheath at the bottom, and that allows the bottom section of the uranium rod to start fissioning wildly. It heats up the lithium hydroxide so it turns to gas, and an exhaust of lithium, hydrogen, and uranium comes out the bottom of the rocket. The high neutron flux from the burning part continues the process going up the uranium rod.
Here's a scorecard for the fizzer
On the plus side...
There are no moving parts, and the design gives you plenty of thrust.
On the negatives side...
It's excellent and dispersing fission byproducts across a large area
It cannot be throttled down. Or up.
Do not hold in had. Light fuse and get away
It gets a solid 10 on the craziness score.
If you enjoyed this video, please play with a sparkler.
Welcome to crazy nuclear rocket engines #8 - pulse propulsion, otherwise known as "orion".
Which is a great example of the old saying "If all you have is a hammer, everything looks like a nail".
I've included the title pages of two different documents that explore this concept, just to show that it wasn't just a crazy idea somebody came up with, it was a crazy idea that many people researched and discussed seriously.
This would be a great time to watch the introduction if you haven't already.
The concept is pretty simple.
You take whatever vehicle you want to move, and at the bottom you attach a big flat pusher plate, connecting it to the vehicle with some big shock absorbers.
Then you set off a series of explosions near the pusher plate, and the shock wave from the explosion hits the pusher plate and moves it forward, and that moves the whole vehicle forward.
Here's an example, using a test vehicle that holds 5 explosive charges in the central tube.
Let's watch that again. In the second series, you can see the explosives being dropped out of the vehicle before they explode.
That, of course, was a very small prototype test.
What we need is a reasonably-sized vehicle, like this one, one that could carry a healthy amount of payload. Total mass is about 200 tons.
And we will of course need a more powerful pulse unit.
We start with a very small nuclear device, only 150 tons of yield.
We surround the device with a radiation case made of uranium 238, otherwise known as non-fissionable uranium. The purpose of the case is to take the xrays that are created by the nuclear explosion and channel them upwards.
Above that, we have the channel filler made of beryllium oxide. It will absorb the x rays and convert them to heat.
And finally at the top there's a disc of tungsten. The immense heat melts the tungsten and generates a big spray of extremely high-velocity tungsten travelling upwards at speeds in excess of 150,000 meters per second, aimed directly at the pusher plate on the bottom of the vehicle.
Orion is *great* at getting off the ground and into orbit, as the atmosphere enhances the propulsive effects.
This is the US Air force 10 meter Mars mission Orion. It has an ISP of around 3400, a delta-v of 32,000 meters per second, and a payload of 73,000 kg.
It can go from the earth's surface to mars orbit and back to LEO in a single stage. Easily.
You can even do a trip to Jupiter's moon callisto, with a total mission time of about 2.5 years.
Orion did have a slight downside, the issue that launch would involve 200 nuclear weapons exploding in the atmosphere. That pretty much made it a non-starter.
There were some hybrid designs; here's one of a Saturn V first stage and an orion vehicle on top. It doesn't start using the orion drive until it reaches about 90 km in altitude.
And here's another where you would launch the Orion with traditional rockets and put it together in orbit.
Here's a scorecard for Orion...
On the positive side, it's easily scaled to whatever size you want, and mass is not a problem, and the vehicles are quite practical to build. No worries about core temps, radiators, complex fuel designs, etc.
On the downside, 200 bombs to get into orbit is a lot of bombs, and taking off is a bit damaging to the launch site.
Orion turns the craziness score up to 11
If you enjoyed this video, please buy a nice vidalia at the market.
This is episode 9 of the series on crazy nuclear rocket engines, where we will explore the nuclear salt water rocket
If you haven't watched it, please watch the introduction video now.
http://www.npl.washington.edu/AV/altvw56.html
The nuclear salt-water rocket takes a unique approach. Instead of trying to have a controlled reaction inside of the rocket core, it goes for the biggest reaction possible, a "detonation" of the nuclear fuel either inside the combustion chamber or inside the nozzle.
Note that the term "detonation" here is not an editorial comment but the description used by the engine designer
It uses water with 2% uranium tetrabromide in it as a dissolved salt. That is injected at high speed into the combustion chamber and it fissions very rapidly, generating a ton of energy and vaporizing the both the water with the fuel and additional water injected to protect the engine and structure.
The estimated thrust is about 8 Meganewtons - roughly equivalent to the F-1 engine used on the Saturn V first stage - at a specific impulse of 8000.
That is with uranium tetrabromide that is enriched to 20% u-235. If you push the enrichment to 90% uranium 235, that gives a specific impulse of 482,000.
Here's a scorecard for the Nuclear salt water rocket
On the plus side, look at that specific impulse. With highly enriched uranium, it can go anywhere.
On the negatives side...
It's hard to store the fuel without it exploding in the tanks.
It's a continuously exploding atomic weapon inside the engine.
The radiation from the exhaust plume kills the crew even if the rocket survives.
I'm rating it a 25 on the craziness scale, but I'm not sure that is high enough.
If you enjoyed this video, please go swim in some non-nuclear salt water
Dawn Aerospace.
Spaceplanes done right.
Spaceplanes spaceplanes spaceplanes
Aerospace organizations love spaceplanes. What could be better than a vehicle that flies into space and then lands like an airplane?
There are literally hundreds of different designs out there, and quite a few real projects...
There are single-stage-to-orbit projects, like the X-30 national aerospace plane that ran from 1986 to 1993, the X-33 Venture star that ran from 1994 to 2001. There's also the Skylon project that started in the 1990s - perhaps earlier and later - and is still going on, at least for small values of "under development".
Like the other single stage to orbit projects, it's unlikely to ever achieve its goals because single stage to orbit is too darn hard to do. But that's a topic for another video.
There are two stage concepts, an early space shuttle one and a later beta project one. NASA didn't have the budget for two distinct vehicles before shuttle and they certainly didn't have the money during shuttle.
There are private two-stagers, including spacebus from Bristol spaceplanes and space rapid transit from Princeton satellite systems. They also need to do two separate vehicles.
In 2013, the DARPA XS-1 program started, envisioning a two-stage to orbit approach but with an expendable second stage. Boeing's phantom express design was chosen in 2017, but after 3 years of work, Boeing decided it had "fallen out of love" with the concept and withdrew from the program.
Which leaves us with four real spaceplanes, all of which are of the orbital variety.
The Russian Buran shuttle, the US Space shuttle, the boeing X-37, and the dream chaser from Sierra Nevada.
All except the space shuttle are "payload" spaceplanes that are have no boost engines and are carried into orbit on expendable boosters. The Space shuttle used its own engines to provide some of the boost to get into orbit.
And that's the world of spaceplanes. Lots of plans, lots of excitement, four vehicles, two of which are retired, one which is going to fly "real soon now", and one that has flown 6 times.
But... There is another...
No, not this one, though it does technically qualify as a spaceplane.
I'm talking about this prototype, and it's bigger sibling that will likely arrive in a few years...
But I'm getting a bit ahead of myself...
The most-mentioned benefit of spaceplanes is their ability to conduct "airplane-like operations". But the big barrier to launch systems is the development cost, so what matters more is airplane-like *development*.
Let's dig into that.
We know the rocket development cycle pretty well.
We start by building a launch pad. Well, actually, before we build a launch pad, we need to find a site for a launch pad, which is a long and expensive process because you need a specific kind of location plus the ability to build and launch from that location. Such locations are rare.
Now that we have a launch pad, we need to build a rocket. We're typically building our first rocket while we're building our launch pad because the launch pad takes so long.
Once we have the rocket built, we take it to our launch site. If our rocket is small enough and our launch site is accessible by truck, we'll use that. If it's large or we need to cross water, that means a ship.
Then, we need approval from a government licensing agency that our launch can be conducted safely - in the US that means an FAA launch license.
Then, finally, we fuel the rocket, fly the rocket, and collect a lot of data.
We then cycle back to building our next rocket.
The cadence between flights is pretty slow; rocket lab flew electron three times in 18 months, and that is notably faster than most companies. Everything takes a lot of time and money as these are big, handcrafted vehicles that only fly once.
An aircraft-style development approach is very different.
You do need to find a test site, but your needs are simple - you need a runway, hanger buildings, and government approval to fly the sort of test profiles that you plan on flying.
You can't do this from busy commercial airports, but there are countless small airports across the world that might work. You will ultimately want one fairly near to an east coast so that your orbital vehicle can launch to the east over water.
You then need to build your first plane. This is going to take some time - it's a brand new vehicle and planes have a lot of systems.
Then you fly the plane. Over and over and over. You can make small changes to the hardware and software incrementally.
You can fly often if you want - daily or even several time a day. You will work your way up from taxi tests to simple flights and you will slowly expand the envelope - flying higher, flying faster, and performing more stressful maneuvers. If you have a good flight test program, you stand a decent chance of making it a long way through your program with a single vehicle.
At some point you reach the limit of what your current plane can do - you can't learn more from the current design - and you build your next, more capable plane, and fly it a lot.
I would call this approach "Evolutionary". You are taking small steps but you are taking them quickly, and you always know where you are.
I'll label the traditional rocket approach as "revolutionary". In most cases, your first launch is what is known as an "all up test" - you are trying to take a big step, all the way to orbit. Big steps are great when they work, but if they don't work, you may not have the time and money to fly again.
You can still fail with the evolutionary approach, of course.
Until recently, rocket companies were interested in demonstrated technical capability. You wanted to get into the high confidence spot on the right, which was occupied by Atlas V, Delta IV, and Ariane V. Atlas V and Delta IV could command a premium because of their track record.
A rocket like Proton could still garner business without high confidence because they had a lower price.
Then Falcon 9 showed up. It pretty quickly caught up with Proton in terms of reliability, and then very soon moved to join the top 3 performers.
Then SpaceX added reuse into the mixture and changed the rules.
And with a ton of hard work, they have managed to get to the point where they can reuse a Falcon 9 booster on roughly a monthly basis. A huge achievement.
And there are a number of different companies hoping to repeat what SpaceX has done. Nobody has done it yet, though we will likely see some decent attempts in the next few years.
And - of course - SpaceX has aspirations for Starship to be reusable on a daily basis.
The model here is to get your rocket up and running - to demonstrate orbital consistency, if not high confidence - and then work on reusability.
But that's not the only possible approach. What if you start with a spaceplane that is reusable on a daily basis and then push towards the higher capability regime?
I'm not sure which approach is better, but I'm intrigued by organizations that attack a problem in a different way.
Which finally brings us to Dawn Aerospace.
They are building an *airplane* that takes off from a runway using rocket power and flies high enough to get above the karman line and into space. Once in space, it will release an expendable second stage that puts a payload in orbit and the spaceplane will glide back to the launch site.
A first stage with "airplane reusability".
Their current vehicle is named mark 2 aurora.
Dawn Aerospace is located in New Zealand, which means that their pictures and videos have the rugged New Zealand mountains in the background.
Mark 2 is their first vehicle. Like many test vehicles, it's pretty small - 4.8 meters or 16' long, with a maximum takeoff weight of only 350 kilograms, or 770 pounds.
If you look at the rear of the vehicle, you will see two small jet engines. They elected to do phase 1 of their testing with these low thrust jet engines, allowing them to evaluate the flight characteristics of the vehicle and debug their flight software - the vehicle is remotely piloted for testing but can also fly itself.
The point of phase 1 testing with the jet engines is to understand the vehicle well enough to be ready to start testing with rockets.
Here's a video summary of the first five flights.
They fly out of Glentanner Aerodrome on the south island of New Zealand, right on the shores of an alpine lake.
Phase One testing has been completed.
The vehicle made 48 flights, reaching speeds of 200 knots and altitudes of 9000 feet, or 2700 meters. That was what they could achieve with the small jet engines.
Phase 2 removes the jet engines and replaces them with a rocket engine. I've been unable to find a name for the engine other than "main engine", so that's what I'm going with.
Perhaps the best comparison for the main engine is Rocket Lab's Rutherford engine.
Both engines use electric motors to drive their propellant pumps. Both use kerosene as their fuel, though the Rutherford uses the more highly refined RP-1 version. Both are regeneratively cooled.
While the Rutherford uses liquid oxygen as an oxidizer, the main engine uses 90% hydrogen peroxide. This simplifies their world considerably as they do not have to deal with cryogenic liquid oxygen.
The downside of using hydrogen peroxide is a 20% reduction in specific impulse. They have specifically made a tradeoff that values simplified operations and higher reuse over performance.
This engine is - not surprisingly given the size of the vehicle - much smaller in thrust than the Rutherford engine.
Phase 2 is in process, using the same airframe as phase 1 and the rocket engine instead of the jet engines.
Here's the first rocket powered flight...
The first 3 flights of the rocket powered version were done in 3 days in April of 2023. They started by duplicating the tests that they had done with the jet version and will then fly higher and faster.
That is the power of airplane-like development.
This version is known as the Mark II A - it is optimized for easy construction and testing, and can only reach 20 kilometers.
The follow-on Mark II B will have propellant tanks in the wings, a higher thrust engine, a lighter structure, and an RCS system to provide control in space. Their goal is to be able to reach an altitude of 110 kilometers and fly at Mach 3+ with a 5 kilogram payload, with larger payloads possible at lower altitudes. It will provide up to 180 seconds of microgravity.
The payload space is what is known as a "3U" volume, approximately 10 cm x 10 cm x 34 cm , or 4" x 4" x 13". This is a common size for small "cubesat" satellites.
Dawn believes that the mark IIB is commercializable, for applications such as astronomy, earth observation, in-space science, space weather, and technology development.
As a simple reusable system it could easily be better than the alternatives for many applications.
Which takes us to the larger Mark III vehicle.
There isn't a lot of information about this vehicle as it is in early development, but it is an upscaled version of the mark II that is 22 meters long and is much heavier at 23,500 kilograms.
Given what we know, we can make some guesses about how the Mark III and expendable second stage will share the work of getting to orbit.
We'll start with the delta V from the first stage. If you want to learn more about delta v, see my video "planning your solar system road trip".
Note that these are estimates, and not particularly good ones at that.
We'll assume the second stage starts up at 110 kilometers and mach 3.4. We need to figure out the delta v to get the plane to 110 kilometers and to mach 3.4.
To get to 110 kilometers, we'll start with the potential energy, which is simply the mass times the height * the acceleration from gravity. We'll ignore the fact that gravity is reduced at higher altitudes.
Plugging in some numbers, that gives us a little over 1 million kg meters squared over seconds squared, which is conveniently the definition of a joule, so it's 1.08 megajoules.
What we really want to know is that amount of energy expressed as an acceleration, and we can calculate that by treating the potential energy as if it were a kinetic energy.
The kinetic energy is one half the mass times the velocity squared. We refactor to solve for the velocity, and get velocity equal to the square root of twice the kinetic energy divided by the mass, which gives us 1470 meters per second.
The total delta v is that velocity plus the actual velocity at mach 3.4, which ends up being 1003 meters per second, for a total of 2473 meters per second - let's call it 2500 meters per second.
Let's do some comparisons. As we just figured out, the mark III aurora generates 2500 meters per second, which is 27% of the delta-v required to get to orbit.
An airlaunched Pegasus starts out with 720 meters per second of delta v, a bit less than 8% of the requirement to get to orbit. Launcher One from Virgin Orbit is a little less. This is a good illustration what airlaunch isn't really useful - a lot of work to only get 8% of your delta v.
The Falcon 9 on a Starlink launch stages at about 3300 meters per second of delta v, or about 35% of the requirement to get into orbit.
Note that the Mark III and Falcon 9 numbers are considerably wrong, since both had to deal with air resistance and gravity losses to get to their staging point, so the 27% and 35% are lower than the true amounts. Falcon 9 probably spends another 1100 meters per second dealing with that, so it's closer to 45% of the total. I have no idea how to calculate that gravity losses for the Mark III because it's a lifting body and that is a very different world.
So the Mark III looks to be in the right range for a first stage.
Your probably not surprised that I did some delta v estimates for both stages, but they're frankly so bad that I'm embarrassed to show them.
What I can say is that if you make a number of assumptions that I think are reasonable, the mass they give for the Mark III looks reasonable for their target payload.
And that's the dawn aerospace spaceplane, and I hope you understand why I think they are "spaceplanes done right"...
The are using nice developmental approach and they have a decent chance of commercializing their mark II vehicle, providing some capabilities there.
Assuming they don't hit any technical difficulties or money problems, I suspect that they will be successful with Mark III.
Small satellites seem to mostly have decided that the cheap rideshare launches is a more important factor than launching when they want to launch and to a specific orbit. Electron has garnered the remainder of the small launch market and even rocket lab has decided that the market they want to be in is larger launchers.
But if there is an option that allows small payloads to fly when they want to fly and get directly to the orbit they want, that could be a very attractive option.
Like Rocketlab, Dawn Aerospace is not just a launch company - they also make satellite propulsion products.
Their B1 thruster puts out thrust in the 1 Newton range, and they package that into a solution for cubesats named cubedrive.
Their larger B20 thruster has a thrust in the 20 Newton range, and is packaged in their SatDrive integrated solution.
Their thrusters use nitrous oxide as the oxidizer and propylene as the fuel. These are both storable under pressure as liquids, meaning that the rest of the design is simple.
This fuel choice is around 15% less efficient than the typical hypergolic fuels, hydrazine and nitrogen tetroxide, but hydrazine is a really really nasty chemical that requires very specialized equipment, procedures, and licensing to deal with it safely...
Going with a safer but less energetic choice means a tremendous simplification in the whole operations approach, and that is especially important for cubesat builders.
And that's my overview of Dawn Aerospace.
If you enjoyed this video, learn the words to delta dawn, and tell me what, exactly, is that flower she has on?
In the past 5 years we've seen NASA do some amazing things.
They've sent the very capable perseverance rover to Mars, and included the first aerial vehicle to fly on Mars, the ingenuity helicopter.
They finished the long-delayed James Web Space Telescope and successfully deployed it to a challenging location.
Perseverance and ingenuity will cost about $2.7 billion total, and James web is about a $10 billion program. Large amounts of money, but with very meaningful results.
SLS and Orion are inching their way towards their first uncrewed flight around the moon, and so far their cost is something over $40 billion. When operational, they are scheduled to fly up to 1 mission per year.
How can an organization that is so great at science be so terrible at creating a new launch system?
That is the subject of this video
NASA is a large organization, and we're going to start by looking at large organizations and how they operate.
We'll start with the mission of the organization, which is something like this:
Empower every person and every organization on the planet to achieve more
These are typically high, flowery statements. I bet that almost nobody can tell me which organization has this particular mission statement.
Next are the corporate values, which are respect, integrity, and accountability. And that's it.
There's nothing wrong with either the mission or the values, but they are largely motherhood and apple pie - who will stand up to say that empowerment, respect, integrity, and accountability are a bad thing?
Let me add in a few cultural values - growth mindset, customer obsessed, diverse & inclusive, and making a difference. Any guesses?
The company is, in fact, Microsoft. I did leave out the "one Microsoft" value as it was a bit of a giveaway.
This is a public statement of what Microsoft values, how they want to be seen by the world. How does that map to the inside of the company?
What goes on inside any organization is what I've taken to calling "the game". And I'm picking on Microsoft here, but I can confidently tell you that every large organization has their own version of the game.
The goal of the game is simple - it is to move up the ladder in the company, to get more money, more reports, more power, more budget, more respect. The game is about *career*, and very much about career for "the people who matter" - management, and especially executive management.
The rules to the game are a secret - they aren't written down anywhere and they not only differ from one organization to another, they differ in different parts of an organization. You are unlikely to find them in any of the books you read about success.
Here is one set of rules of the game:
Rule 1: Don't be late. It's not good for you or your management if you are the one who is visibly slowing things down. Submit your software even if it's buggy because everybody else is doing the same thing and you won't get attention for buggy code. If you are late, don't speak up; everybody is playing schedule chicken and you just need to be sure you aren't the latest one.
Rule 2: Be a hero. You *will* get kudos for extra effort to stay on schedule even if it's your poor skills that put the schedule at risk.
Rule 3: Be confident.
Rule 4: Manage up. To be promoted, you need the support of your manager, your manager's peers, and your second-level manager. They need to know your name and be comfortable that you are the kind of person they would be comfortable working with.
Rule 5: Don't make waves. The game is predictable when everybody is following the rules and that makes it easier for everybody.
Let's look at a little example of how the game works.
Here's part of an organization, with three first level lead managers and a second-level manager they report two. The reports are small because they are generally not participants in the game, which they would typically refer to as "politics".
Let's suppose the first lead implements an idea that results in a 10% improvement in productivity. That's good for the lead; they are helping the company and it's good for the manager as it shows that their reports are doing a good job.
Now let's suppose the third lead implements an idea that results in a 200% improvement in productivity. That sounds great.
This idea makes the other two leads look bad, and if given the choice between implementing the same idea or making the third lead look bad, they will choose the easier choice. The third lead has not made friends with his peers.
The third lead has also caused problems for his manager. If anybody finds out about this improvement, they will expect the manager to implement it across all three leads, and the increased performance in the manager's group will not make the manager's peers or manager happy.
Finally, the third lead had three reports to do the work they were doing, but they are now 200% more productive and only need one. But one report is too few, so the lead can't be a lead any more.
That's what happens if your idea is very successful. If you try something and it isn't successful - or it isn't obviously successful - be assured that your peers will not forget about it going forward.
There is generally one high-level value that the game rewards the most, and that is conformance. Follow the rules and stay within the norms of your peers, and you'll do great.
You do need to know what the mission is but in most cases you aren't evaluated in your contribution to the mission, because it's simply not feasible to do so.
The values are useful inside the game; you can refer to them in a positive way to move people ahead, and in a negative way to hold people back. "I like Joe but I don't think he has enough of a growth mindset to take on larger challenges".
That's the way the game is often played in large organizations with professional management. But it's not the only way it is played...
Let's say you are a small launch company still run by a founder. You likely have a very different kind of mission and values.
Note that the missions and values are very specific and are *not* motherhood and apple pie. You can actually use them to decide how you should act as an employee and you can use them for employee evaluations.
The game in this company is likely very simple - know the mission and follow the values.
And this can work because a) the founder is in control and not willing to employ anybody who is trying to play a different game and b) the only path to success in the game goes through mission and values because otherwise the company fails.
It's this sort of alignment between the mission, the values, and the game that can make small startups so nimble.
If you were wondering, the mission and values here are from Rocket Lab. That's true even if you weren't wondering...
Another way of stating what we just went over is the following
Read:
The incentives are what drive the game.
Thankfully, we can finally move to NASA.
NASA is a big and complex organization. The two notable parts are the Mission directorates and the various centers. Roughly, mission directorates define work and centers perform work.
We'll focus on the 5 mission directorates.
The first is aeronautics research, which spends a little less than a billion per year
The second is space operations, which includes running the international space station and crew flights to and from the station. It spends about $4 billion each year.
Science is about $8 billion.
Exploration systems development includes everything Artemis - SLS, Orion, gateway, and the human landing system. It's about $7 billion.
And finally, there's Space Technology, which covers - not surprisingly - development of new technology for future NASA missions. Things like advanced thrusters, nuclear thermal engines, nuclear reactors for space power, etc. It's about a billion and a half.
We're going to focus on the science and exploration systems development directorates because they show the best and worst of NASA.
How does science at NASA work?
Science starts with figuring out we know and what is interesting to research.
At NASA, science is guided by a series of decadal - ten year - surveys created by the national academies of science. The commissions for each area of study pull together the top researchers in the field to decide what is known in a specific area, what questions are interesting, and how to answer those questions. This information is summarized in the decadal survey documents, each running several hundred pages in length. The NASA ones are:
Planetary science, earth, solar, biological and physical, and astronomy
In some cases these are about doing things in space, in some cases they aren't - the astronomy one covers both space-based observatories and ground based observatories.
The actual research happens outside of NASA, in a large number of independent projects - some small, some medium, and a few large.
To figure out what research NASA should fund, NASA maintains a series of science directorates that align with the decadal surveys. These directorates are responsible for looking at the goals in the decadal surveys, looking at the research proposals that have been made in that area, and choosing the ones that best accomplish those goals.
This sort of model is not unique to NASA; it's pretty much how all government funded science is handled.
Let's look at a few examples.
List separate parts.
Decadal surveys
Competitive proposal program
Psyche & book that talks about it.
List of ongoing programs.
Notable failures, some that turned into successes (hubble, JWST).
NASA science succeeds through a competitive grant environment and farming out the majority of the actual work to academia, including the operation of JPL, which is a federal research laboratory.
They do what they need to do - to organize and choose what is important - and leave the rest to academia. And they do a large number of programs.
Look at history of their biggest programs?
The recent DART mission was built by the applied physics laboratory at Johns Hopkins university.
It also flew the lee-cheeya cube satellite, built by argotec for the Italian space agency.
The currently postponed psyche mission spacecraft was built for Arizona state university by maxar technologies, and the instruments are created and operated by johns Hopkins applied physics laboratory, UCLA, MIT, and JPL.
If you want to know more about what it takes to put a planetary mission together, I highly recommend the book "A Portrait of the Scientist as a Young Woman" in which Lindy Elkins-Tanton describes her experience as a researcher, including being the primary investigator - or group leader - for psyche.
Another good example is the Perseverance Rover. The rover was created by the jet propulsion laboratory, but the 7 instruments that ride on it were chosen from 60 proposals and come from research institutions around the world.
You've noticed that I put up two different JPL logos, one with NASA and one without.
JPL is a federally funded research laboratory that is run by the California institute of technology, so it's more like a research arm of a university. Funding for JPL comes through NASA but NASA does not run JPL in the way it runs the NASA centers, though JPL and NASA do work very closely together.
NASA has had some issues on the science side.
In 1990 the hubble telescope was launched, with optics of unprecedented quality. Unfortunately, in an attempt to save money, NASA had chosen a bid that did not subject the telescope to full-scale testing on the ground, and an error in a mirror testing device resulting in a mirror that was perfectly made to the wrong specifications and therefore yielded poor results.
Three years later NASA installed a set of corrective lenses - and other upgrades - as part of a NASA servicing mission using the space shuttle, which fixed the problem.
The James Web Space Telescope program was started in 1998, and at that time the plan was that it would launch in 2007 at a cost of $1 billion.
As the years went by, the project slipped and got more expensive, finally launching 15 years late and at nearly 10 times the original budget. That increase in cost meant that it took $9 billion away from other research projects.
But these concerns have largely gone away due to the early success of JWST, not only in terms of science but in producing such phenomenal images.
Here's the mission and values for the NASA science directorate. Their game is mostly just "do science" - the things that make you successful in science will likely make you successful here. It is a government bureaucracy, but that's what big federally-funded science is, and being a bureaucracy has not been a significant obstacle.
They have a number of success factors.
First, they choose projects based on the decadal surveys, which is the best scientific opinion about how to use limited funds.
Second, they choose a mix of big and small missions.
Third, they leave the details to the mission teams - NASA provides the grant money and the teams go off and execute using those funds.
Fourth, they provide oversight, help, and coordination for those mission teams, both in figuring out how to create their proposal, who to work with, and how to work well.
Fifth, they leverage expertise around the world.
The NASA science directorate has had some issues, but they are generally science done well.
We'll now switch from the science side to the human spaceflight and exploration side.
During Apollo, things were very simple.
The mission was beat the Russians
The values were beat the Russians
And the game was beat the Russians.
This unity of purpose is one of the main reasons Apollo was successful; it was a *national* goal, and while all the contractors were happy to get Apollo money, nobody wanted to be the one who kept the country from winning the race.
On July 20th, 1969, Neal Armstrong and Buzz Aldrin became the first humans to land on the moon, and the US won the race to the moon.
This was a wonderful day for NASA and all the contractors that worked on Apollo.
It was also a terrible day.
The whole point of Apollo was to achieve what Apollo 11 achieved - to land humans on the moon and return them to earth. The mission was accomplished, the race was done.
NASA's budget had already been going down since the peak in 1966 and by 1970 it would be half the peak. NASA had already made painful cuts in their workforce, and with the race won, it was possible that congress would eliminate all human spaceflight funding.
That would be bad for all the NASA centers that worked on Apollo, and bad for the managers employed at NASA Headquarters in Washington, DC.
And bad for the major contractors that worked on apollo and their numerous subcontractors
And bad for the congresspeople with NASA centers or major contractors in the areas they represented.
The situation defined the game; it was:
Keep NASA spaceflight centers open
Keep money flowing to NASA contractors
Keep votes flowing to congresspeople
Preserve NASA management jobs
The mission was literally "whatever we can sell to keep the game going"...
And the values - I'm sure there were values, but against an existential threat to NASA's existence, there wasn't much time for values.
NASA had some ideas for the post Apollo era - space stations, nuclear tugs, moon bases, and even Mars bases - but they would be very expensive and none of them was politically possible as both congress and president Nixon were uninterested in further exploration.
There was another option:
The current government and DoD launcher was the Titan III, which was $20-30 million per launch. That's about $150-230 million in 2022 dollars, so roughly what an Atlas V would cost.
The government wanted a cheaper launcher than the Titan III, and that provided an opportunity for NASA.
NASA believed they could build a "space shuttle" for $5 billion and it would cost $5 million per flight.
With a bit of math, you can come up with this chart, and it shows that if you fly 28 times per year for 15 years, you will save money with the shuttle approach. NASA projected that they could fly up to 60 missions per year, at which point it would be much cheaper than Titan III.
NASA horse traded with the Air Force to get DoD support for the shuttle, and based on that support and this economic argument, the shuttle program was approved and NASA was saved.
The mission became "Build and Fly Shuttle".
You've probably noted that "fly the same vehicle to the same place over and over" really isn't an exploration goal, but remember that it's the game the matters.
Shuttle was an unbelievable success; it put NASA in a stable funding environment for 30 years and there was no need to go back to congress to justify a new program. Money for NASA, money for contractors, votes for congresspeople. Stable careers all around. For those that matter, it was great, and that is why NASA flew it for so long.
It's true that it was very expensive, it killed 14 astronauts and destroyed two $ billion orbiters, and it never went beyond low earth orbit, but none of those factors are important to those that matter.
It is widely believed that the Columbia disaster was what led to the cancellation of the shuttle program, but that's not really true. The investigators did not recommend cancellation; they merely recommended a recertification process similar to the ones that are common for military aircraft. NASA had plans to operate shuttle to 2020 or even beyond.
But in 2004, the Bush White house released a new "Vision for Space Exploration".
The mission in the previous years had been merely "fly shuttle" and "build ISS".
There was now a new mission. The first two items are similar - fly shuttle to finish ISS and then retire it, and then operate the ISS.
And then there are things that look like real missions - robotic missions to the moon followed by human ones, and then the same thing for Mars.
It would be really great for NASA to do something beyond low earth orbit with a defined mission, but there's a big problem.
These first two items were already consuming 100% of NASA's exploration budget. The plan was to continue them until ISS was done - probably 5 years - while at the same time start up a program with a similar scope to Apollo. It's pretty obvious that there was a budget issue here; there was no way NASA could afford working on the things they said they would be working on.
But remember that it's the game that matters, and there was certainly a lot of opportunity here for everybody involved, so NASA started down this road with the constellation program, including the Ares I crew launcher, the Ares V cargo launcher, and a new deep space capsule known as Orion. The two Ares rockets were shuttle-derived because that is what the rules of the game required.
In 2010, there was a fight between the Obama administration - who wanted to cancel Constellation because of the budget issues - and congress, who wanted to keep the game going.
Not surprisingly, congress won, and the mission was the big orange rocket, Orion, and maybe Mars at some sufficiently distant point in the future.
The sole bright spot was the commercial crew program.
And then finally, in 2017 we got a rebranding of the previous program into "Artemis", and shift towards the moon instead of maybe mars at some point.
And that's the state where we currently are. The Game is still mostly driving the NASA exploration programs, though the success of commercial crew and the award of the lunar lander to SpaceX has changed things a bit.
The main problem is with the exploration systems development mission directorate.
If we look at their mission description, it's all about system development. This is why, until 2017, there was no real mission for SLS or Orion; the goal of the directorate is simply to develop them.
What we need is an independent decadal survey for *exploration* and a NASA mission directorate in charge of implementing that survey.
But, of course, that hasn't happened, because it would interfere with the game.
Is it possible to change the rules of the game? I believe the answer is yes, but that's a topic for another video.
If you enjoyed this video, please join my kickstarter for the NASA exploration game.
So you want to rent an astronaut.
How much will it cost?
I'd like to make clear that I'm not talking about getting an astronaut to do something on earth. If you want that, you need to go through NASA's astronaut appearances office. There are not surprisingly quite a few conditions and you aren't going to get an official astronaut visit for Suzie's birthday party, no matter how much she loves space.
You could book a former astronaut as a speaker, which will cost you somewhere between $10,000 and $100,000.
Or you could go with a backup plan
https://www.nasa.gov/wp-content/uploads/2015/02/astronaut_appearance_guidelines_2020_updates.pdf
What I'm talking about is wanting to hire an astronaut to do something in space. Maybe you have an experiment on the international space station that needs some human attention. Or maybe you want to do some research on the surface of the moon.
For the international space station, the question is easy to answer - you can get astronauts to work on your experiment for $130,000 per hour. That is the price for astronaut time, but it's not the cost - it's well-known that NASA subsidizes research on ISS.
What is the real cost for an hour of time by an astronaut in low-earth orbit? That's our first question.
Here's the NASA price on a graph...
We'll start with the Inspiration 4 flight that was conducted in 2021 on Crew Dragon.
This flight was 2 days and 23 hours, but some of that was getting up into orbit and getting ready to return, so we'll drop down to 2 days.
Now we have to define how much of their time was useful time. There's no standard way to do that for this flight, so I'm going to arbitrarily say that they worked 8 hours of time per day.
8 hours per day for each of 2 days for four astronauts means 64 hours of useful time during the mission.
Jared Isaacman has said that the cost was less than $200 million, so I'm going to use that as an upper bound, and a bit of math gives us a cost of $3.1 million per hour.
Adding in the Inspiration 4 data, it is about 24 times more costly per hour than what NASA charges.
Next up is the Axiom AX-2 mission to the space station.
On this mission, the participants spend 8 days on ISS.
I'll assume they only spend 6 hours doing useful work each day.
That gives each participant 48 hours of useful work time.
We don't know the actual price or cost for that participant, but I'm going to put it at $60 million, which is in the right ballpark.
Math says that's $1.25 million dollars per hour.
Axiom is quite a bit cheaper - it has similar launch costs, but it has an 8-day mission so there is more time.
On the space station, we have the advantage that astronauts are there for a long term, so their per hour cost should be cheaper. I thought about doing this on a per-expedition cost, but it's really complicated so I'm going to use a shortcut and look at a year on the station.
NASA astronauts work 5.5 days per week, which means 286 days per year.
They work 6.5 hours per day, for a total of 1859 hours/year per astronaut.
Multiply that by 4 astronauts, and that's 7436 hours/year.
For 2022, NASA requested $1.32 billion to operate the space station and a further $1.77 billion for cargo and crew services, for a total of $3.1 billion.
Take 3.1 billion and divide it by 7436 hours, and we get $417,000 per hour.
ISS astronauts have long missions, so their cost per hour is fairly low - about 3.6 times the NASA price.
But it's more complicated than that.
Back in 2003 a National Academies report noted that NASA estimates that it takes 2.5 astronauts to keep the space station operational, which means that with the current crew of 4 astronauts, only 1.5 is available to do research.
That means we need to redo the calculations using 2789 hours, and we find that the per hour cost goes up to $1.1 million.
Add in the overhead of running the ISS, and the numbers look quite a bit worse - 8.5 times the amount that NASA is charging.
To come up with a "full cost" number, we would want to include the cost to construct the ISS.
We can simply add up the design cost and the launch cost to get that number. For the design cost, there was an ongoing fight about what ISS construction costs were while it was being built and nobody could agree on the cost to design and build the ISS modules. The cost to launch depends on what you think a shuttle flight costs and there is no agreed-upon number there.
When faced with this situation, I typically do a low/mid/high approach. I think that ISS cost the US more than $50 billion to build and less than $150 billion, so those are the low and high and $100 billion is conveniently in the middle.
We can allocate that across the 20 year lifetime to get 2.5, 5, and 7.5 billion dollars a year of construction cost.
Add in the yearly $3.1 billion cost, do some math, and our answers are $2 million, $2.9 million, and $3.8 million per hour of useful time.
When we add in the ISS construction costs, the cost per hour ends up being roughly what Inspiration 4 cost.
ISS is really cool but wow it's expensive.
One interesting number from this price chart is that NASA only allocates 90 hours of astronaut time a year for commercial activities, only about 3% of the total amount of astronaut time.
That's not a lot of time devoted to commercial research.
Since we've looked at LEO we definitely want to estimate how much lunar surface time costs.
We'll start with Apollo.
Apollo cost $257 billion in 2020 dollars. If we add up the EVA time for each of the missions, we end up with 80.4 hours total.
Do the math - $257 billion divided by 80.4 hours * two astronauts and we get $1.6 billion per hour of EVA time.
We'll start our lunar chart with that amount.
But perhaps that is unfair.
NASA ran all of Mercury and Gemini and built a lot of costly infrastructure for Apollo. If we look at Apollo 17 and only consider the incremental cost of that flight, we get cost of $450 million in 1972, or about $2.8 billion in 2020 dollars.
With 22 hours of EVA, that brings us down to $63 million per hour of EVA.
This is obviously far less than the whole cost.
How pricey will Artemis III be?
There are a lot of unknowns but we can still come up with some numbers.
SLS and Orion can fly about once a year and their overall budget is about $4 billion, so we'll use that as their cost. We also need a Starship lander, and for that I'm going to use the price of the second Starship lander, $1.15 billion. That gives us $5.15 billion in total costs.
We known Artemis III will spend approximately 7 days on the surface because that will be the first launch opportunity to dock with Orion to head home.
Assuming 6 hours per day of EVA for 6 of those days gives us 36 hours.
Do the math, and that gives us $5.15 billion / (36 * 2) = $71.5 million per hour
Interestingly, pretty close to Apollo 17.
This of course ignores the development costs for Artemis.
I'm going to arbitrarily assume that there are 6 lunar missions and half are one week and half are two weeks. That gives us 3 missions at 36 hours EVA time for 108 hours and 3 missions at 72 hours for 216 hours, or 324 hours total for all 6 missions.
Costs are complicated...
SLS cost about $27 billion to develop in 2022 dollars, and it spends $2.5 billion a year for 1 flight, so $15 billion in flight costs and $42 billion total.
Orion cost $26 billion to develop and $1.5 billion per flight, so 6 flights gives us $35 billion total
Starship costs $1.8 billion to develop and $1.2 billion per flight for 3 flights, so $5.4 billion total
Blue moon costs $1.9 billion to develop and $1.5 billion per flight, so 6.4 billion total.
Add that all up, and we get $89 billion. We're missing some costs - the cost of the spacesuits from Axiom, the cost of the CLPS exploratory flights, and any costs for gateway.
We take that 89 billion and divide it by 324 hours for each of the 2 astronauts, and we get $137 million per hour.
There might be more astronauts on some missions which would reduce the price.
What did we learn?
First, astronaut time in LEO is really expensive, and astronaut time on the moon is ridiculously expensive.
Second, NASA heavily subsidizes commercial experiments on ISS, at least when it comes to astronaut time. That means that applying NASA numbers towards commercial space stations probably doesn't work very well. But that's a topic for another video.
Third, Artemis looks cheap, but it's built on existing shuttle parts, uses apollo era infrastructure, the landers are provided to NASA at a huge discount, and the on-moon mission times are going to be a lot longer than the Apollo ones.
Finally, Suzy is going to be disappointed at her birthday party.
If you enjoyed this video, please send me an Artemis decal for a buck ninety five.
In the early 2010s, Russian commercial space was flying high. They were doing a lot of business and bringing in a lot of money.
Then something happened that would break their once vibrant program.
We'll start by looking at the state of the Russian space program from 2010 through 2014
The Proton rocket was their workhorse launcher for commercial satellite launches.
From 2010 to 2014, they had 43 successful launches. A rough guesstimate for the cost is around $90 million per launch, or $3.9 billion total for this 5 year period. Note that actual launch prices are usually confidential, so these are just estimates.
International proton launches are sold by international launch services, a joint venture of Lockheed Martin (LM) in the United states and Khrunichev and Energia in Russia, and there is likely some overhead to that arrangement, so the money going to Russia will be less than these numbers
Russia sells commercial launches using the Soyuz launcher out of the Baikonur launch site in Kazackstan.
Once again, there is a joint venture at work; Starsem is a joint venture with Arianespace, Roscosmos, and Progress Rocket Space Center.
In 2010 through 2014, they performed 5 commercial launches, at a very rough estimate of $60 million per launch.
In 2011 Arianespace added a program to launch Soyuz rockets from Guiana Space Center in French Guiana.
This launch site is nearly at the equator and that allows them to launch to orbits that would not be reachable from Baikonur
They did 7 launches from 2010 through 2014 at around $110 million per launch, or $770 million total
With the retirement of the Space shuttle, NASA had no way to way to get astronauts to the International Space Station, so they starting buying seats in the Russian Soyuz capsule.
During this time period, NASA bought 6 seats every year for prices that went up over time, costing a total of $1.5 billion dollars.
https://oig.nasa.gov/docs/IG-20-005.pdf
Finally, the Atlas V rocket launched by ULA uses the excellent Russian RD-180 engine.
The Atlas V flew 32 times during that period, and at an estimated cost of $15 million per engine, that's $500 million dollars total.
If we add all of those up, we get around $1.4 billion per year.
It's important to note that the money collected is what is known as hard currency, as these costs are paid for in strong currencies such as US dollars, Japanese Yen, or Euros.
The Russian Ruble is not considered to be a hard currency; it is less desirable.
This means that the money from these services makes it easy for Russia to buy products on the world market.
Now we'll roll forward to the present day.
This graphs shows the Proton launches in purple and the SpaceX commercial launches in green. As you can see, as soon as Falcon 9 was flying regularly, most of the commercial business for Proton dried up, with only 1 commercial launch in the last 4 years.
It's not just Falcon 9, however; if we look at failures we can see that proton had a string of failures these years.
Including this notable one
Here's the history of commercial Soyuz launches from Baikonur. The program was largely dead until a rebirth in 2020, as OneWeb bought a number of launches for their constellation. Presumably OneWeb chose Soyuz over Falcon 9 because SpaceX's Starlink competes with OneWeb's constellation.
The Russians have recently chosen not to honor the contract for future OneWeb launches, and those launches will be instead be done by Falcon 9.
The Soyuz launches by ESA in Guiana have been steady since the start of the program in 2011. The majority of these launches were for European governments payloads.
There is a huge question mark going forward, as Russian has suspended these launches.
NASA fairly consistently bought 6 seats per year until crew dragon was ready, and then those seats and the revenue disappeared.
This chart looks at NASA and department of defense launches for Atlas V and Falcon 9. Each of the Falcon 9 launches is one less RD-180 engine that Russian sells to ULA.
The reliance on Russian engines has been the subject of much debate in the US government and a lot of weaseling around trying to come up with a policy. Currently there is a requirement that no bids after 2022 can use the RD-180, but in early 2022 Russia announced an end to sales and support of the RD-180 (fade out RD-180).
ULA will be replacing the Atlas V with the Vulcan, which is currently waiting on the delivery of BE-4 engines from Blue Origin.
Here's a fiscal update.
We'll start with the information from 2010 through 2014, where the Russian space industry was bringing in about $1.4 billion dollars a year in hard currency.
Roll forward to 2021, and we see that the revenue from Proton has disappeared, and that's primarily because of Falcon 9; the competitive price and high reliability of Falcon 9 simply pushed proton out of the market.
The revenue from flying astronauts to ISS has also disappeared, also because of Falcon 9 and crew dragon.
The sole bright spot is the revitalization of the Baikonur Soyuz launches with the one web deal.
Overall, in 2021 the amount of hard currency the program brought in was half of what it was in earlier years, and it's possible that the OneWeb launches cost less than $500 million.
Fast forward to 2022, Roscosmos has decided to cancel their launch deal with OneWeb, cancel any future sales of the RD-180, and suspend the Soyuz operations in Guiana, effectively killing off their commercial space program.
If you enjoyed this video, please write an ode to it
It's pretty clear that SpaceX is the most capable aerospace organization right now. More capable than NASA, the Russian ROSCOSMOS, and India's ISRO. Maybe there might be a discussion compared to the Chinese.
Many companies want to emulate SpaceX, and there is a lot of discussion about whether they will or not. Many of those discussions ignore the fact that SpaceX benefited from a rather unique set of circumstances when building Falcon 9.
In this video we're going to take a journey into the world of alternate universes and look at a number of situations that, if different, would likely have led either a drastically different SpaceX or no SpaceX at all.
I'm going to skip the early days of SpaceX, the days when they were working on Falcon 1. If you want to understand those days, read Eric Berger's excellent book "liftoff". We are going to focus on Falcon 9.
We know that the retirement of the space shuttle in 2011 meant NASA needed a way to carry supplies to the Space Station, and for that they created the commercial Orbital Transportation Services program, or COTs.
That led to SpaceX getting help to build Falcon 9 and Cargo Dragon.
That takes us to our first scenario:
Scenario 1: What if shuttle kept flying?
Columbia disintegrated on reentry in February of 2003, killing all 7 members of the STS 107 crew.
If the Columbia accident hadn't occurred - and if there no other losses of shuttle - shuttle could have flown to 2015 or even 2020, and the COTs program wouldn't have been available when SpaceX needed it.
After the accident, the Columbia Accident Review Board created a report that detailed the causes of the accident and made recommendations going forward. Not only did they detail the changes to make before shuttle returned to flight, they detailed what should be done to continue flying after 2010.
If that plan was followed - and if there were no further accidents - the shuttle could have flown to 2015 or beyond, eliminating the chance for SpaceX to get a COTs contract.
But president George W. Bush decided that he wanted a different space program, including a plan to return to the moon by 2020 in preparation for the human exploration of Mars. That new plan required the retirement of shuttle in 2010 after the completion of the space station so that the shuttle money could be shifted to the new exploration programs.
The new plan is what led to shuttle being retired, and would set the stage for SpaceX many years later.
Scenario 2: What if NASA had a replacement vehicle to service ISS, so that COTs would not be needed?
In the early 2000s, NASA had an orbital space plane program that would have launched small crew and cargo vehicles on commercial launchers to augment the space shuttle. These space planes would have worked well to serve ISS.
And yes, one of them is a capsule. NASA was okay calling that a space plane, and so am I.
The orbital space plane program morphed into the crew exploration vehicle for the constellation program. There would be two vehicles built by two separate contractors and flown to see which one was the best deal for NASA to purchase. This Lockheed Martin design would have launched on top of a commercial launcher and had both earth orbit and lunar variants.
When the constellation program started, new NASA administrator Michael Griffin restructured the program to be a capsule-based design that was too heavy to be launched on the commercial launchers. This would eventually be renamed to Orion.
If that rescope hadn't happened, the crew exploration vehicles might have been a valid solution to carry crew and cargo to ISS. COTs would not have been needed.
Administrator Griffin mandated that the rockets used in the constellation program would be shuttle derived.
That gave us the crewed Ares I that used a solid rocket booster derived from shuttle, a new second stage, and the orion capsule on top. It also gave us the Ares V cargo vehicle to perform the heavy lifting required for the moon program.
While Ares I was considered a solution to get astronauts to ISS, using a moon capsule for that mission was a poor choice in terms of complexity and expense. There was no solution to fly cargo to the ISS in constellation.
There was a competing shuttle-based design known as "direct" or "Jupiter" that would use shuttle components with as little modification as possible. NASA also could have developed a "clean sheet" solution for ISS cargo.
Any cargo solution available early might have meant that COTs was not required.
Scenario 3: What if SpaceX did not succeed at COTs?
This one is going to be a bit complicated. I said that I wasn't going to talk about Falcon 1, but I've included the flights here to give a little context.
To start up the commercial space mandate, NASA decided to give $227 million to Kistler Aerospace in 2004. Kistler was headed by former NASA associate administrator Dr. George Mueller.
SpaceX protested to the general accounting office that this was an illegal non-competitive award, and the GAO quickly told NASA that there was no way they would win this suit. NASA pulled the award in June of 2004, and started the more formal process used for COTs.
The initial program got spun up in late 2005, with 6 semifinalists selected in May of 2006 and SpaceX selected as a phase 1 winner in august of 2006, with an award of $278 million. I'm going to ignore the other participants in COTs.
Note that between the startup of the program and the selection, SpaceX launched their first Falcon 1 flight. If NASA had started COTs 6 months earlier, NASA would have been making their decisions before Falcon 1 had gotten off of the pad. It's not clear that SpaceX would have been a winner without that technical achievement, and this is a way that SpaceX might not have gotten the COTs contract.
There was one more problem with COTs. NASA had arbitrarily decided that $500 million was enough money for two participants, and it was apparent by 2009 that the program would not be a success without additional money. This was problematic because Congress hadn't been very excited about COTs during early appropriations.
The program was in trouble.
Then something unexpected happened. President Obama had set up the Augustine commission to review the current state of NASA's human spaceflight program, and the commission released their report, "Seeking a human spaceflight program worthy of a great nation", and in the congressional hearings that followed the release, commission members advocated giving an additional $200 million to the program.
Even more unexpectedly, when the NASA authorization act of 2010 came out of congress, it gave an additional $300 million to the program, and $118 million of that went to SpaceX.
Without that extra money it is unlikely that SpaceX's COTs contract would have been successful.
There was another issue...
The original concept was that the contractors would finish the development of their vehicles and then NASA would order flight vehicles to be used for actual resupply missions.
But shuttle was retiring in 2010, and NASA was concerned that the COTs contractors would not be nimble enough to start flying soon enough after shuttle retired, so they made a decision to push things forward and get the program up and running on much earlier.
The released a request for proposal in April of 2008, got the proposals and then made awards in December of 2008. Note that at the time when SpaceX needed to be figuring out what their firm-fixed price bid was for flying falcon 9 and cargo dragon, they had been only working on Falcon 9 for two years and had yet to reach orbit successfully with Falcon 1.
They were supposed to somehow figure out the operational costs for a vehicle that would not fly for two more years, in mid 2010.
Not surprisingly, they did not estimate as well as they had hoped and the operational CRS flights did not provide the profits they had planned on, leaving the company cash-strapped. This is why SpaceX raised their prices significantly for the CRS-2 contract.
This timing was an unfortunate choice by NASA that was unfair to both contractors - firm fixed price contracts are supposed to be used when the products are well understood, and that clearly could not be the case 4 years before the first flights to the space station.
Scenario 4: What if SpaceX did not get commercial launches?
(Add satellites next to Ariane..., show on next slide as well.
In 2010, Ariane 5 was flying consistently.
One the problems of Arianespace is that their production is spread across many countries and companies, and that means production planning is very complex and has a long lead time, so it's difficult for them to increase or decrease the flight rate.
This means they cannot easily respond to increased demand.
That left an opening for a competitor, and that competitor was the Russian Proton rocket.
It was priced around $75 million, which was an attractive discount on the Ariane price. The downside of proton was that it had a long history of poor quality, and if you flew on Proton there was a chance that your payload would not survive.
But there was enough backlog that customers would prefer to take their chances rather than wait years to get their satellite into orbit.
If Ariane 5 was able to fly more or Proton had better quality, the opportunity to launch these commercial satellites would not have existed.
This provided a market opportunity for SpaceX.
I said I was only going to talk about external factors, but I need to break that here.
If SpaceX did not have the skills of Gwynn Shotwell to sell launches on this new and unproven rocket, it is likely that they would not have been able to exploit this market.
Scenario 5: What if SpaceX could not launch for NASA and the department of defense?
In 2010, if you were in NASA or the department of defense and you wanted to launch a payload, you had three rocket choices.
You could launch on the Delta IV Medium plus, the atlas V, or if you really needed a big payload to a hard destination like geostationary orbit, a delta IV Heavy. For these launches, you would pay $164 million, around $160 million, and a mind-boggling $420 million.
All of these rockets were manufactured by United Launch Alliance, a company that had a virtual launch monopoly on US government payloads since it was formed in 2006. You might think the high prices were due to a monopoly, but Lockheed Martin and McDonnel Douglass (before the Boeing merger) had been the two providers for the air force's evolved expendable launch vehicle program since 1994 and had found that having assured business with the government was a great opportunity to make huge profits.
Falcon 9 was able to enter this market because it was so much cheaper than the ULA rockets, costing perhaps $90 million for a government launch.
Entering this market took some time and some lawsuits, but it turned into a steady source of high profits for SpaceX.
If there had been a competitive US launch provider, this opportunity would not have been available.
What have we learned?
Space benefited from some unique circumstances that helped pay for falcon 9 and made it relatively easy for them to get launch contracts.
SpaceX saw these circumstances took advantage of them.
If you are evaluating other launch companies, note that NASA is not currently helping companies to build their rockets and instead of competing with Proton and ULA, new launch companies have to compete with SpaceX.
If you enjoyed this video, please send some lucky space rockets. I have calculated that for my salute to Falcon 9, I will 69 of these packages.
https://www.nasa.gov/wp-content/uploads/2015/01/617036main_396093main_hsf_cmte_finalreport.pdf?emrc=e76114
In the past 20 years, two presidents tried to fix NASA human spaceflight. Neither of their plans really worked, but both led to situations that would require NASA to fundamentally change, both in the past and in the future.
It's a weird story.
The story starts on February 1st, 2003, when the space shuttle Columbia broke up on reentry because of damage to the leading edge of the wing caused by foam that came off the external tank attachment to the shuttle.
The conventional view is that this was the reason that the shuttle program was cancelled, but that's not actually what happened.
Cancelling shuttle was never the plan. The plan was to bet on the status quo and keep flying shuttle for a long time.
And that's not surprising.
Everybody involved with shuttle at NASA, all the shuttle contractors spread across the country, and all the congresspeople in areas that have NASA centers or NASA contractors are used to the status quo, and many of the managers have built their careers around the shuttle program.
They would have to be idiots to want to give that up - the status quo approach is generally the low drama approach.
And I'm not picking on NASA; this is true for pretty much all successful big organizations.
And there was another big factor - Congress and NASA had decided to invest many billions of dollars to build the international space station and shuttle was the ticket to both finish the station and to carry both cargo and crew to the station.
I've come across a number of assertions that the Columbia Accident Report said that shuttle should be retired. Here's what the board actually said...
The board both believed that the shuttle was safe enough and could be recertified to fly into the 2020s.
Everything was aligned behind returning to flight and continuing to fly shuttle.
But not everybody...
The status was not quo.
NASA administrator Sean O'Keefe sensed that a big change in direction might be possible, but he would need support from the executive branch to figure out what that change should be and to implement it.
After discussion with the top people in the Bush administration, he convinced Stephen Hadley, deputy director of the highly influential National Security Council to work across agencies to figure out what the space plan would be going forward. That group eventually decided that going back to the moon was probably the best option.
President George W. Bush was signed up for the plan; he wanted to see if he could be successful where his father's space exploration initiative was not.
This new initiative was announced by president bush on January 14th, 2004.
It had 4 major goals:
The first was to return the shuttle to flight as soon as possible to complete the international space station, and then retire it. This seems like a strange approach, but the new space initiative required money and it was not possible from a budgetary perspective unless the very expensive shuttle program stopped.
The second was to develop a new spacecraft, the crew exploration vehicle and conduct a crewed mission by 2014. That would leave a small gap between shuttle retiring and new missions to the ISS, but it wouldn't be terrible.
Return to the moon by 2020
And do human missions to Mars and beyond
This would turn out to be a watershed moment, a moment that changed that path of NASA forever. But I don't want to give away the ending...
NASA responded soon after with their fleshing out of the plan, called The Vision for Space Exploration, which is pretty much everything that NASA had been talking about in exploration for the past couple decades...
There were two big problems...
The first is that the program NASA outlined was very ambitious and it would be very hard to accomplish under any budget NASA could expect.
The second problem was that the vision did not align with all the people who were happy with the shuttle status quo.
The status quo group bought a lucky break when Administrator O'Keefe decided to leave his position in early 2005. His official reason was that becoming Chancellor of Louisiana State University would pay better and allow him to better support his family, but it was clear that he was very tired from Columbia and the strain of trying to reform NASA.
He was replaced by Michael Griffin, who was much more in line with the traditional NASA approach. He was comfortable with the approach building new rockets from shuttle parts and argued that it would be quicker than starting from scratch. That was key to the approach outlined in NASA's exploration systems architecture study
But Griffin had a problem, and that problem was the commercial space act.
The commercial space act says the following:
(read)
There are 7 exceptions, and you need to fly commercial unless you can hit one of those exceptions.
Griffin decided to lean heavily on exception #2. Obviously big missions like going to the moon were things that only NASA could do.
To meet that exception the mission had to be modified.
The crew exploration contract was modified in process to change from a "bring what you want" design that could fly on commercial launchers to a "big capsule with a service module" design.
Orion was just big enough that it was too heavy to fly Atlas V or Delta IV, and any way, those launchers would need new second stages and they use small solid rocket boosters that aren't safe, not like the big solid rocket boosters that NASA flies.
That may sound like it violates the spirit of the commercial space act, but it's only a violation if somebody complains.
It was therefore decided that NASA would build two new rockets.
The new rockets are Ares I and Ares V, both mostly made out of shuttle parts, which was what Griffin had decided. That mostly made the status quo crowd happy; the NASA centers would be able to work on the same sort of projects and most of the manufacturers would still be making shuttle parts.
NASA had another need, and that was supplying the Space Station with cargo. It was hard to make an argument that commercial providers could not launch payloads into low earth orbit since pretty much every commercial rocket was designed to do exactly that, so we got
The commercial orbital transportation services program in 2006. This was the first time that NASA human spaceflight had agreed to buy a service from a commercial provider.
After a bit of drama, NASA settled on two providers. The first was a sure bet, with Orbital Sciences flying their Antares rocket and Cygnus resupply spacecraft. It took them 6 years and they were a couple years after the shuttle retired, but Congress slow-rolled the initial funding to slow things down.
Yes, it is strange that congress underfunds a program directly authorized by congress and in line with a space act that congress had also passed, but it's important to remember that legislation and funding are two separate things run by different people.
The status quo folks liked Orbital Sciences because they had a proven track record and had a goal of selling their services to the government and making a nice profit and nothing beyond that.
The second award went to an upstart company know as Space Exploration Technologies who had just flown their first rocket to orbit. They built a medium launch lifter plus a cargo capsule named dragon to go with it and their first flight to ISS was in May of 2012, about 16 months earlier than Orbital Sciences.
This made NASA happy - they still relied on the Russians to carry astronauts to ISS but at least they could get cargo.
And the Ares I rocket would soon be ready to fly astronauts.
Time to look back again.
NASA Spaceflight launched the first US astronaut into space in 1961 in project Mercury, and followed with Gemini in 1965, and finally, Apollo 11 reached the moon in 1969. Three unique systems in 10 years, though Mercury and Gemini did fly on existing missiles.
In 1974, NASA started working on the space shuttle, and in 1981 it finally took flight. During the 20 years from 1961 to 1981, NASA was a development organization; designing new things and figuring out how to make them work. That took a special kind of organizational culture and a special kind of management to make it work.
In 1981, NASA stopped doing development; their goal now was to just fly shuttle over and over with minimal modifications. That creates a different kind of organization and NASA was run by managers instead of by engineers; many of the Apollo and Shuttle engineers left or retired.
That shift was certainly a factor in the loss of both Challenger and Columbia, but that's not the issue I want to talk about here. The problem at hand is that NASA spaceflight has signed up to do development work on this huge new project that is arguably harder than Apollo in architecture, and it's been 30 years since they developed shuttle.
NASA started executing on Constellation.
During this time, congress was getting a little concerned about depending on Russia for both cargo - commercial cargo hadn't flow yet - and crew access to the international space station.
The 2008 NASA authorization act directed NASA to use commercial provided crew services to the space station as much as possible and limit the use of the crew exploration vehicle - soon to be named Orion - to missions beyond low earth orbit.
That seems like a pretty clear directive, but remember that authorizations tell NASA that it can do something but don't actually provide money to do it.
The appropriations act included this, which basically says that if NASA has any leftover money from the COTs cargo program they should spend it on commercial crew. Not exactly a ringing endorsement of the concept, and NASA is still officially the organization that flies humans in the US.
As is common with large NASA programs that spend a lot of money, the government accountability office had been watching the Constellation Program, and in August of 2009, they issued this report, entitled:
Constellation program would go a lot better if NASA would decide what they are building and why.
That's a bit of a paraphrase; the actual title is "constellation program cost and schedule will remain uncertain until a sound business case is established".
In October 2009, the Ares I-X was launched to gather performance data on the first stage solid rocket booster.
It flew a standard shuttle 4 segment solid rocket booster, with a fake 5th segment on top and then parachutes for recovery. On top of that was a fake second stage, a fake orion, and a fake launch escape system. The fake sections were instrumented to gather data about the test.
The test was successful, with the booster performing great, separating, and then parachuting back into the sea.
Presumably this test was intended to show the progress NASA had made, but what it amply demonstrated is that NASA had no new hardware ready to fly for Ares I with the exception of the larger parachutes designed for the 5 segment booster.
What did we learn from constellation?
The first is that NASA doesn't know how to run a development program any more. This really wasn't that much of a surprise; President Bush had put together a commission in 2004 and one of their conclusions was (read).
The second is that NASA had no solution for ISS crew and cargo. It was 2010, shuttle was going to retire very soon, and NASA was not on a path to flying a replacement. The only solution was to fly astronauts and cargo on Russian rockets and that makes NASA look very stupid.
In 2009, President Obama took office and immediately commissioned a review of the current space program, and that committee released their report in October of 2009. This is one of the best reviews I've seen done and is still worth a read now; there's a link in the video description.
This graph is eye opening. The projected budget for constellation was $10 billion a year. The FY10 budget would likely set that level to $7 billion a year, and this chart does not include the cost for the ground systems - launch pads etc. - which would move to constellation when shuttle retired. That meant that Constellation was $3-4 billion short per year even if it met the NASA estimates, which it was already exceeding.
The report listed a number of different options on how to proceed, noted that constellation had already spent $10 billion over 5 years and only produced one test launch. Obama looked at the different options and decided to cancel constellation in 2010, excluding it from the proposed budget request.
https://www.nasa.gov/wp-content/uploads/2015/01/617036main_396093main_hsf_cmte_finalreport.pdf?emrc=e76114
The reaction of the status quo was swift. Florida Senator Bill Nelson - yes, the same Bill Nelson that current runs NASA - claimed that "the president made a mistake", playing the "you aren't the boss of me!" card. He was correct; it is Congress that has the power to decide what NASA does, and they have been happy to defund programs that the president chose and fund programs that the president did not choose.
Nelson went on to tell Obama that he will see him in the parking lot after school, or, to be more specific, at the President's Space summit on April 15th.
Like President Bush, Obama was hoping to transform how NASA worked, and he laid out his thoughts in a speech at the summit.
Not surprisingly, his first point was to cancel constellation because it was not fulfilling its promise in many ways.
His second point was to buy space transportation from commercial operations. The idea of flying astronauts on non-NASA vehicles really chafes with those who believe in NASA uniqueness, but it was really clear that NASA had no current plan to carry astronauts to ISS, which means that it was either commercial space or the Russians.
The next point was to skip the moon and go directly to Mars. Here I think Obama was really underinformed about the current state of NASA's capabilities; NASA couldn't execute on constellation to get to the moon and mars was oh so much harder.
Then the two big ones - develop orion into a crew rescue vehicle for ISS, and spend $3 billion to research a new heavy lift rocket, focusing on new designs, new materials, and new technologies.
The status quo looked at this and said, "I think we can work with this"...
Obama's plan morphed in congress into the NASA authorization act of 2010.
Congress gave a big check mark to cancelling constellation, so there was no mention of constellation in the act.
Buying commercial transportation also got a check mark, very grudgingly, and Congress decided to underfund it for the first 4 years. Congress apparently prefers a world in which the Russians fly astronauts over a world where US companies fly astronauts.
The moon or mars question got punted down the road. The act talks extensively about missions but typically refers to them as "missions beyond low earth orbit"
The Orion section got a check but only once it was rescoped to continue the Orion program from Constellation, with a requirement to be able to deliver crew and cargo to the ISS if necessary. Remember that congress hates commercial crew and cargo.
And finally, the status quo loved the idea of a new heavy lift rocket, which they would call the "space launch system", or SLS.
They didn't like all this talk about new designs, new materials and new technologies, so they got rid of all that.
To make it very clear, they spelled things out in detail.
The act required extending or modifying existing vehicle development contracts.
It required the retention, modification, and development on critical skills and capabilities that just happened to be those that were used by shuttle.
It had to use anything that was shuttle-derived or had been started on Ares I.
It was basically a very clear requirement that SLS had to be shuttle derived, which was exactly what had been going on with constellation. So the real change was to rebrand constellation as SLS/Orion and get rid of the Ares I option. The status quo made sure there was no chance NASA could make a different choice.
So, once again, a President had failed to redirect NASA. The status quo still reigned supreme...
But maybe not...
I have said in the past the George W. Bush was merely trying to create a legacy for himself, but the thoughtful way in which his administration worked with administrator O'Keefe has impressed me. They wanted a legacy, but they were also tired of astronauts being killed on flights carrying cargo into low earth orbit.
Their plan did not work the way they had hoped, and the status quo and Administrator Griffin gave us a shuttle-derived plan.
But their plan did uncover NASA's inability to build a new vehicle to service ISS, and that gave us the unicorn that is SpaceX.
Without their plan and the cancellation of shuttle, there's a decent chance that SpaceX does not survive.
The plan from the Obama administration also did not work as intended, but like the Bush plan, it set a couple things in motion.
With the decision to cancel Constellation is was very clear that NASA had no solution to carrying crew to ISS, and that meant either commercial crew had to go forward - in line with what congress required - or NASA had to keep sending large sum of money to Russia and accept a space station that was chronically undercrewed.
We therefore got SpaceX's crew Dragon, and Boeing's Starliner capsule.
Since shuttle retired, crew dragon has flown 50 humans into space, and starliner is in the middle of its first crewed flight, flying 2 astronauts.
Vehicles developed by NASA's exploration division have flown a total of zero humans during that time period.
NASA has not just lost their monopoly on launching humans from the US, they have lost their ability. That may come back with Artemis II, but in an area where NASA was the only game in town they are lagging far behind.
And this has also spawned the Polaris program, which is Jared Isaacman and SpaceX's very own Gemini program, bypassing NASA to develop those capabilities.
And, because SLS and Orion have never been part of a coherent moon architecture, the heavy lift rocket and capsule that they built could only get humans into lunar orbit.
The constellation project had a big lander known as Altair, but it didn't make it into to SLS & Orion world, so the Artemis moon story is a story without a NASA LANDER.
The huge SLS and Orion can get into an easy-to-get-to lunar orbit, known as Near Rectilinear Halo Orbit, and Orion can get back home, but all they provide is a sightseeing tour.
NASA needed a lander, but SLS couldn't carry one that was big enough.
Stealing a slide from my video on the Artemis architecture - go watch that for the details -
We see that SLS launches Orion and Orion brakes into near a near rectilinear halo orbit.
A lander is launched into low earth orbit, refueled, and flies out and meets up with Orion. Astronauts transfer over, and the lander lands then. On the return trip, it goes back to orion, transfers the astronauts back, and Orion heads back for home, leaving the lander in lunar orbit.
NASA has contracted to get two different landers. SpaceX will build a version of Starship optimized for the lunar mission, and Blue Origin will build a lander known as blue moon.
What may not be clear is that is technically quite a bit harder to get a lander to the moon, land, and get back into orbit than merely to get a capsule in to orbit, so NASA is helping pay for two projects that are much more capable than SLS and Orion, and there are obvious ways that those architectures could be modified to do the entire moon mission without SLS and Orion. And with SLS & Orion costing on the order of $4 billion and only capable of flying once a year, it seems likely that the commercial moon architectures will be developed.
This happened because the status quo was more focused on the money and jobs side when they set of SLS and Orion and didn't consider that there might be commercial companies capable of outdoing what NASA could do.
And that's where we are.
Two presidents made attempts to reform and reimagine how NASA does things, and both of those attempts failed - they did not achieve what the presidents wanted to achieve.
But they both forced NASA - and more importantly, congress - to adapt, and those adaptations change the US spaceflight world in ways that pushed NASA into a world where NASA relies on commercial providers far more than anybody had thought possible, and that trend is only going to continue.
If you enjoyed this video, please buy yourself this book...
A while back Elon said the following...
(read)
That's a very interesting statement and I'm curious about whether it is correct.
To figure that out, we'll need to answer two questions:
The first is "how would increased gravity effect rockets?", and the second is whether that effect is bigger for reusable rockets.
To answer these questions - and to see if Elon is correct - we need to dig into the physics. This will get complicated for a minute or two and then it will get simpler. This part will not be on the test.
The energy required to get into a 400 kilometer orbit - about where the space station orbits - is determined by a number of factors.
We need to get up from the earth's surface to orbital altitude, and we can express that using this equation.
We then need to get up to orbital velocity, and we can figure out the energy cost using this equation.
Then we need to deal with the losses. We waste energy fighting gravity to get into orbit, and we waste energy fighting against drag, and there are two equations to estimate that.
We add all these up, and we get a very precise answer that is wrong. It's wrong because treating the altitude and velocity as separate quantities assumes that rockets are launched to the target altitude and then turn sideways. They don't do this because it's less efficient to do so.
The robust way to come up with the true energy cost is with a simulation that calculates all of these values on the fly. That's a lot of work and different rocket architectures fly different trajectories to minimize losses.
I'm not that ambitious, so I'm going to make what is often known as a simplifying assumption.
The drag loss is fairly small, so we will just ignore it.
All three of the remaining terms depend on gravity or the mass of the earth but the velocity energy is the largest one when , so I'm going to ignore the other two and just look at the velocity energy.
It won't give us the right answer, but it will give us an answer that is close enough.
A little time with Excel yielded this graph, which summarizes the change in energy required to get into orbit in different gravity environments.
In the middle we see that at earth standard gravity the ratio is 1.0.
If we move out, a 10% reduction in earth gravity results in an energy ratio of 0.95 and a 10% increase results in a ratio of 1.05.
We now have some numbers that we can plug into a real example...
Which will of course be the Falcon 9, which I chose because it's very familiar, it has both expendable and partially reusable modes, I already have a model for it, and it's the cutest rocket around.
My model is written in Excel and is based on the rocket equation. I can feed in a payload amount and it will tell me how much delta v - how much energy - the Falcon 9 can give a payload of that mass. If you want to know more about the rocket equation, see my video "the care and feeding of the rocket equation".
We can plug in the maximum payload of the Falcon 9 to low earth orbit - 22,800 kilograms - and the model will say that the Falcon 9 can generate 9262 meters per second of delta v when flown in fully expendable mode.
That number is pretty much in the middle of the 9100-9400 meter per second range generally quoted to get into low earth orbit, so it's a validation that the model is decent.
I would also like to estimate the delta v required to fly in the other two modes. The maximum payload for a drone ship recovery is about 16,600 kilograms which yields a delta v of 10,090 meters per second, and the maximum payload for a return to launch site is about 13,000 kilograms and about 10,700 meters per second of delta v.
Those are the values that we see with standard earth gravity.
We can now combine the gravity data with the data from the model.
Our gravity data gave us these estimated energy multipliers. Multiplying our expendable delta V of 9262 meters per second by those multipliers, we get the required delta v for those three gravity scenarios. Not surprisingly, it takes less energy if gravity is lower and more energy if gravity is higher.
And then finally, we can use the Falcon 9 model to translate to the payload amounts Falcon 9 can carry while hitting those delta v values.
If gravity was 10% less, the Falcon 9 expendable payload would be about 27,000 kilograms, and if gravity was 10% greater, it would be about 19,000 kilograms.
Do that for all three modes of the Falcon 9 and for a set of gravity ratios, and we get this chart.
There's definitely an effect, but it's not a huge one - 10% heavier gravity leads to 15-20% loss of payload. I would say that Falcon 9 isn't terribly sensitive to differences in gravity. Not close to the "impossible" situation that Elon talked about.
Let's look at that quote again...
He very specifically says full reusability, and since the Falcon 9 is only partially reusable, it isn't the scenario that he is talking about.
He's talking about Starship.
I took a similar approach with Starship and plugged its numbers into my model. Since it's very much under development, Starship is a moving - and sometimes exploding - target, so I used what I know about the OFT-4 flight and its supposed payload of about 50 tons.
Add the starship data, and we get this graph.
Starship is hugely sensitive to changes in gravity, note that a 10% increase in gravity drops the payload from 50 tons down to 23 tons, and a 20% increase drops it all the way down to 1 ton.
At this point I hope you are asking "why, oh why is starship so bad?"
And for that we need to do a quick comparison between the second stages of falcon 9 and starship.
The Falcon 9 second stage is very light, with an empty mass of a little less than 4 tons. It carries 112 tons of propellant and, in droneship mode, 17 tons of payload. The total mass is 139 tons. We can look at this as percentages, with 3% going to empty mass, 84% to propellant, and 13% to payload.
Those are fairly typical numbers for a second stage.
Let's look at starship.
130 ton empty mass, 1200 tons propellant and 50 tons of payload , for a total of 1380 tons, or pretty close to 10 times the mass of the Falcon 9 second stage.
For percentages, the empty mass is 9% of the overall mass. That is the penalty of reuse; the stage requires coverage for the engines in the rear, heat shield tiles to survive reentry, and fins to guide it. It is also made of stainless steel, which is heavier than the aluminum lithium alloy used on the Falcon 9.
The propellant mass is a bit higher at 87%, and that leaves only 4 % for payload.
This explains why higher gravity is so much worse for starship; increased gravity reduces the mass a rocket can lift into orbit and the only place that reduction can come is in the payload. Higher gravity might require a 2% reduction in overall mass to compensate, and that only reduces the Falcon 9 payload from 17 tons to 14 tons, but 2% is half the payload for starship.
That's the story about starship and gravity, and I would consider the quote to be confirmed.
My model didn't quite show it was impossible at 10%, but I know that the SpaceX models are much better than mine is and we did see a very big effect.
If you enjoyed this video, please send me "Can't explain", a special edition DVD and book set.
The space shuttle first launched in April of 1981. (video) It was the world's first reusable launch system, with both the orbiter and the solid rocket boosters recovered and reflown and only the large external fuel tank thrown away.
Shuttle flew for 30 years, and during that time it remained the only operational system that recovered and reflew parts.
It's now 42 years later, and SpaceX is getting ready for the first flight of starship, which, if successful, will be the first fully reusable orbital launcher.
Is starship the follow-on to shuttle, the version 2 that NASA never built?
The space shuttle first launched in April of 1981. (video) It was the world's first reusable launch system, with both the orbiter and the solid rocket boosters recovered and reflown and only the large external fuel tank thrown away.
Shuttle flew for 30 years, and during that time it remained the only operational system that recovered and reflew parts. It was, however, never the cheap and reliable system that NASA had hoped for.
It's now 42 years later, and SpaceX is getting ready for the first flight of starship, which, if successful, will be the first fully reusable orbital launcher.
Is starship the follow-on to shuttle, the version 2 that NASA never built?
To answer this question, we'll need to figure out what NASA might want from shuttle 2.0.
Based on these NASA proposals from before and after shuttle was developed, I'm going to assert that shuttle version two should be a fully-reusable, two staged vehicle.
We're going to focus on the second stage because that is the more challenging part of the design and that makes comparisons with shuttle easier.
Here's a nice video of the shuttle second stage in action.
The second stage is an orbiter that's about 37 meters long with 4.6 meter fuselage, and a much larger external fuel tank that measures 49 meters by 8.4 meters.
Here's the orbiter plus the external fuel tank.
To build a fully reusable version, we need to solve a significant packing problem.
The expended external tank contains roughly 2000 cubic meters of propellant, and we will need to find a way to fit that propellant into the orbiter body.
The orbiter payload bay is about 300 cubic meters in volume, so this is a significant challenge. We will definitely need a bigger orbiter.
We can reduce that 2000 cubic meter requirement if we change our architecture.
The current shuttle is a "stage and a half" design - the orbiter engines ignite on the ground and run all the way to orbit. That requires quite a bit more fuel.
It also uses solid rocket boosters that aren't as beefy as a separate booster stage, so the upper stage needs to do more work.
If we get rid of the solid rockets and put a booster underneath the shuttle, we can cut down the requirements quite a bit, down to perhaps 1250 cubic meters.
The v1 orbiter is 37 meters long and has a wingspan of 24 meters.
If we scale it up to hold all the propellant for our reusable version, we end up with truly huge 66 meter orbiter with a 43 meter wingspan.
That assumes that upsizing the orbiter only adds 26 tons of mass - the mass of the external tank. If the bigger orbiter gains more mass than that, it will need more fuel capacity and will end up even larger.
This bigger orbiter still has the same performance as the small one - it can lift about 27 tons to low earth orbit - but it would get rid of the expendable external tank.
NASA had wrested with the size issue during the initial design of the orbiter. This is a drawing from North American Rockwell that shows that the shuttle would be 62 meters long if it carried all of its propellant internally.
It would be 37 meters if it carried the liquid hydrogen in an external tank.
And it would be 34 meters if both liquid hydrogen and liquid oxygen were carried in an external tank. Bigger means more expensive and more volume that you need to protect with heat shields.
NASA had a very limited budget, so they chose the external tank, giving up their dream of a fully reusable system.
For shuttle 2.0, it would be great if we could build a smaller orbiter than the mega shuttle
The problem is really one of propellant density.
The space shuttle uses the RS-25, or space shuttle main engine. Its propellant is liquid hydrogen and oxygen, or hydrolox.
Our second stage variant carries 460 tons of propellant, and each ton of that propellant requires 2.8 cubic meters of tank volume, for a total of 1290 cubic meters of tankage. It takes a lot of space because liquid hydrogen is very low density.
What if we replaced the RS-25 with an engine that uses a different fuel?
SpaceX's Raptor engine burns liquid methane and oxygen, or methalox. This propellant combination is not as efficient as hydrolox, so it takes about 50% more fuel, or 700 tons total.
However, the liquid methane is much denser than the liquid hydrogen, so it only takes 1.2 cubic meters of tank to hold a ton of fuel, and a total of 840 cubic meters to hold 700 tons.
The result is that the required tank volume for the Raptor version is 38% less than the RS-25 version. That means a significantly smaller orbiter or, carrying more propellant for a larger payload.
We will need to decide whether shuttle 2.0 should be a plane or something else?
The 1960s were a time of space planes, some real like the X-15, some proposed like the Dyna Soar, and many others as design concepts.
The language that started the shuttle program was pretty clear, specifying a chemically fueled shuttle operating between the surface of the earth and low-earth orbit in an airline-type mode.
It doesn't actually say that it would be a space *plane* - a craft that flies like a plane - but it was certainly implied, and from what I can tell all of the early shuttle designs were space planes and NASA never looked seriously at other options.
It turns out that there are a number of issues with the spaceplane design.
The shuttle required a very strong airframe that makes up the bottom half of the fuselage because the payload doors cannot carry any structural loads. That strong airframe is heavy.
It also required large and heavy wings and landing gear.
I'm going to make a few comparisons that are a bit unfair.
The shuttle has an empty of 71,000 kilograms, and its maximum payload to low earth orbit is 27,500 kilograms. That means that the total mass to orbit is 98,500 kilograms, and only 27% of the mass to orbit is payload.
If you want add in the external tank, the numbers go up by the mass of the external tank, and only 22% of the mass hauled to orbit is payload.
Looking at the Atlas V with its Centaur second stage, we see that the empty second stage is a very light 2300 kilograms and carries a payload of 18,850 kilograms, giving a payload percentage of 89%.
The Falcon 9 second stage has a mass of 3900 kilograms and a payload of 22,800 kilograms, for a payload percentage of 85%.
As I noted, this is an unfair comparison because neither the Atlas V nor Falcon 9 have reusable second stages, but it does illustrate one of the reasons shuttle v1 was so big and so expensive - it carried nearly 125,000 kilograms of mass to orbit and only about 27,000 kilograms was payload.
The point is that the shuttle architecture was reusable but it wasn't very efficient, and one of the reasons is the big and heavy orbiter.
Cost is also an important factor.
When space shuttle Endeavour was built in 1987 to replace Challenger, it cost $1.8 billion, and it was only that cheap because there were a large number of structural spares leftover from the construction of the original orbiters. Our bigger fully reusable orbiter would obviously cost more.
This is one answer to "why isn't there a shuttle 2.0?" If you are looking at vehicles that are a few billion dollars each, your development project is going to be very expensive, and with NASA spending their money flying shuttle there was no money to create a replacement.
Is there an alternative?
A way to build a vehicle that is fully reusable, more efficient, and cheaper to build and operate?
If we look back at vintage science fiction, we find a large number of spacecraft that are tail-landers, for the very simple reason that spaceplanes require both atmosphere and runways.
And, of course, one of NASA's most famous spacecraft was a tail lander...
Those second stages I was talking about earlier look interesting. We know how to make large rocket stages cheaply.
Can we get one from orbit back to the earth's surface?
Without adding the disadvantages of turning it into a plane.
The first challenge with any reentry vehicle is thermal protection - a way to deal with the heat of reentry.
There are two traditional approaches to deal with this.
The first is the ablative heat shield used on capsules like Dragon and Orion - it has a material that slowly chars and vaporizes to deal with the heat.
And the second is a tiled approach used on the shuttle and vehicles like the X-37. This thermal protection can be reused but may need maintenance between flights.
Since we want to be easily reusable, some sort of tile-based approach makes more sense.
We will therefore need to cover the bottom of our second stage with heat shield tiles.
Our second challenge is to control our vehicle orientation and flight path. If the heat shield ends up facing upwards, this might happen.
Capsules do this with a neat trick...
Their center of mass is designed so that the capsule is stable with the heat shield pointing in the right direction, and it is also offset to one side by placing heavier equipment on that side. That tilts the heat shield to the side and generates lift.
That lift can then be used to control the path of the capsule by rotating the capsule around with its thrusters.
Unfortunately, we have two problems. The first is that our vehicle is mostly symmetrical in mass distribution, so there's not a lot of natural forces keeping the heat shield on the bottom.
The second problem is that we have a lot of heavy engines and structure at the bottom of our second stage, and no engines at the top, so our center of gravity is doing to be mostly near one end.
Which means we can't use the low tech "capsule" approach, and instead we need an approach with active control surfaces, like the shuttle, but not a plane.
We can do this fairly simply.
Our vehicle is a long cylinder with heat shield on the bottom, and the air is rushing past it.
We need a way to keep the heat shield on the bottom.
If we add fins to the sides, we can move them from one side to the other.
Push them one direction, and it will cause a rotation in one direction. Push then the other way, it will cause a rotation in the opposite direction.
That's what we need to keep the heat shield facing our direction of travel.
Looking from the side, if the fins at the left are up and the ones at the right are down, there is more drag on the right, and that rotates one direction.
Flip them the other way, and we get the opposite rotation.
That's two degrees of control.
And finally, looking from the top, with this arrangement, we rotate to the right
And this arrangement rotates to the left.
The fins give us the ability to control the vehicle during reentry.
These fins do, of course, make our heat shield harder - we need to deal with more complex curves, but that's a solvable problem. We'll just need to get more complex with our tile shape.
And now we need to move onto landing. The first question to ask is "can we use a parachute?"
The solid rocket boosters from the shuttle program weighed about 90 tons, so roughly what our second stage might weigh, and they were parachuted down and landed in the water. They were towed to port, disassembled, and then shipped off to the factory to be reused.
The SRBs used for shuttle had 20 millimeter thick steel walls. Liquid fueled stages are much thinner, with walls around 4 millimeters thick, or even thinner. That makes it harder to survive the force of landing.
And remember that our stage has heavy and expensive engines that don't react well to being dipped in salt water, while solid rocket boosters are mostly just big empty tubes.
Regardless, even if this did work, parachutes aren't very controllable.
At this point, somebody is probably saying "what about electron?"
Electron is working towards recovering their booster with a helicopter, as shown in this test flight.
Unfortunately for this approach, the electron booster weighs about 1 ton, and the stage we want to do this with might be up to 100 tons. There are no helicopters that can lift that much weight.
So we're going to use the fins to guide us and then, as we near the ground, use thrusters and rockets to spin back to vertical and land.
Which leads us to a vehicle like this...
Which brings us back to our original question. Is starship shuttle version 2?
The answer is "Yes", in the sense that it will, if successful, do the things that shuttle could do, but do them better.
The answer is "No", as it uses a different architecture than shuttle and is wildly more ambitious in its goals than shuttle
If you enjoyed this video, please listen to track one of "Freedom at point zero".
SpaceX has been focusing on getting ready to launch starship at their Boca Chica facility in Texas, but launching Starship from Cape Canaveral in Florida has always been part of the plan.
The details of those plans seem to have gotten a bit more firm recently, and that makes it a good time to talk about this topic.
To launch quickly, it probably needs to be a site that SpaceX Controls. SpaceX has three sites;
launch complex 39A where they launch Falcon 9, Crew Dragon, and Falcon Heavy.
Space launch complex 40, where they launch Falcon 9.
And space launch complex 13, which they use for landing falcon 9 boosters on some missions.
The landing zone doesn't really help, so the candidates are space launch complex 40 or complex 39A.
How about Space Launch Complex 40? You can call it "slick 40" if you want to sound cool.
Changes here wouldn't affect Falcon heavy or crew dragon launches.
Unfortunately, it's close to other users.
To the north we have slick 41, which ULA uses to launch sensitive DoD payloads as part of National security space launch and also Starliner astronauts to the ISS. This makes it a very valuable launch site, and unfortunately their launch integration building is quite a bit to the south of the launch pad, placing it only 1700 meters from slick 40. To the southwest, there is an electrical substation at 1180 meters. And finally, the very popular astronaut beach house is 1450 meters to the northeast.
More problematic, the pad is fairly small - only about 330 meters across. That means that all the pad facilities are close to the launch pad.
Everything on the pad is designed for medium rockets and it doesn't have facilities for cryogenic fuels like liquid methane.
Another factor in pad 39A's favor is that SpaceX already has an environmental assessment done to launch starship and super heavy from launch complex 39A.
The plan is to build a launch tower like the one at Boca Chica in the indicated location.
https://netspublic.grc.nasa.gov/main/20190801_Final_DRAFT_EA_SpaceX_Starship.pdf
As of June 2022, SpaceX has a number of launch tower sections constructed and has started assembling them at pad 39A. The curved structure on the right is probably a flame diverter.
If you look closely you can see two humans fitting a crossbeam in place.
Initially, starship and super heavy will be shipped from boca chica to florida.
I discuss how that might happen at length in the linked video.
In the longer term, SpaceX has two needs.
SpaceX needs a place to build starship in florida so they don't have to ship it.
And spacex needs an additional launch site for starship, as one site isn't going to be enough.
In 2018 SpaceX leased a plot of land near the Kennedy Space Center Visitor's Center, known as the Roberts Road site.
Here is the site on a map
Is this big enough for a starship factory?
I did a quick estimate of the area that Starship is using at Boca Chica, and it came out to about 2.8 million square feet, or 263,000 square meters.
Interestingly, the Roberts road site is pretty much exactly the same size, about 2.8 million square feet.
It's not clear exactly how SpaceX plans on using the whole site, but there are some initial site plans to deal with water management that can be referenced to make some guesses.
The hanger on the right - since completed - was started as a building to do Falcon 9 refurbishment, and it is probably devoted to that operation rather than Starship.
This space here is currently being used to construct the tower sections for the pad 39A launch tower. You can see the concrete bases for the legs of the tower.
The space on the far left is being used for surface water management. The elevation of Kennedy space center is only about 7' about sea level, and these big ponds to manage surface water are very common.
Near the water is a large factory building, roughly as big as the 4 large tents at Boca Chica. Next to it is a possible extension to the factory building.
And then finally, two buildings that appear to be high bays; they look to be the same size as the mega high bay that Spacex is Building at Boca Chica.
For a little sense of scale, I put a Falcon 9 with dragon in the lagoon. This is a big site.
This site is about 5.7 miles to Pad 39A.
That's more the double the distance from the Boca Chica factory to the launch site there, but less than half the distance from Blue Origin's factory to launch complex 36.
That covers the factory side. What about an additional launch site?
If you transplanted the Boca Chica starship launch site to LC49, here's what you would get.
Plenty of room for multiple towers if SpaceX is willing to accept close separation, and given their choices on LC39A, that appears to be true.
There's another possible option that *might* happen.
Back in Cape Canaveral Space Force Station, ULA launches Delta IV rockets out of LC 37, but there are only 3 remaining delta IV launches remaining; one in 2022, one in 2023, and one in 2024, and after that it is done flying.
Would ULA give up their lease to SpaceX?
They might want to keep it for future use, but one of the focuses of Tory Bruno as CEO of ULA is to reduce the duplication of facilities. They have shut down the Delta IV factory to save money, and it makes little sense to continue to pay for a launch complex you aren't using. That assumes that the cost of keeping it around is meaningful and/or there's a requirement to use it if you want to keep it.
How would this site work for starship?
The pad was built for Saturn I and IB which were considerably smaller than Saturn V and therefore smaller than Starship, but LC37 was planned to have two launch towers so the actual pad area is pretty big - about 600 meters in size.
It currently has liquid hydrogen and LOX infrastructure, which is easily adapted to liquid methane and LOX.
It has good separation. The closest pad is launch complex 34, but that has been abandoned for years and is the location of the Apollo I Memorial, which would be far down on the list of sites to redevelop.
To the south, there's LC 20 which is currently leased by Relatively, but it's 2200 meters away.
To the north, the situation is even better. The nearest launch pad is 3600 meters away, and that launchpad is slick 40, which SpaceX already leases.
Here's an image of the Starship launch complex at Boca Chica. It's about 300 meters by 200 meters.
And here's what it looks like on top of space launch complex 37. It fits very easily.
In fact, you could likely built a whole series of pads along this section of land if you could get environmental approval to do so.
We can call it "starship row"...
If you enjoyed this video, please answer this question:
2 hours to use the skylab shower.
Danger of inhaled water.
2 hours to use the skylab shower.
Danger of inhaled water.
Mia and the martians:
https://www.indiegogo.com/projects/mia-and-the-martians-publishing-campaign#/
Off Nominal Podcase episode
https://offnom.com/episodes/166
Gateway white paper, 2022:
https://www.nasa.gov/wp-content/uploads/2023/10/acr22-wp-gateway-the-cislunar-springboard.pdf?emrc=5a0843
GAO-24-106878 NASA should document and communicate plans to address gateway's mass risk
https://www.gao.gov/assets/gao-24-106878.pdf
NASA crewed mars mission architecture (2023)
https://x.com/KenKirtland17/status/1745931455199338512?lang=en
Artemis lunar architecture video
https://www.youtube.com/watch?v=O2IBV_XSu60
Asteroid retrieval crewed mission:
https://ntrs.nasa.gov/api/citations/20140008285/downloads/20140008285.pdf
Overview of mission design for NASA asteroid redirect robotic mission concept.
https://www.trylam.com/files/IEPC_2013_Overview_of_Mission_Design_for_NASA_ARRM.pdf
Lunar Gateway - or just gateway - is the name of the space station that NASA is planning on putting in a near rectilinear halo orbit around the moon, the same orbit that the orion capsule will be using for Artemis lunar landing missions.
You can think of it as a space station like ISS but smaller and not continuously occupied.
How did this program come about? Like many NASA programs, there's a bit of a story.
After the columbia accident in 2003, president George W. Bush and NASA administrator Sean O'Keefe worked to put NASA on a new path, and that resulted in the Vision for Space Exploration in February 2004.
I've sometimes described it as the "full meal deal" for NASA; they put pretty much everything they'd wanted to do into the vision.
With the shuttle scheduled to be retired, that meant NASA needed a replacement rocket, and the constellation program came up with two, the Ares V, which was an absolute beast and the Ares I which would carry a new astronaut capsule named Orion.
Both were heavily based on shuttle technology for a number of reasons, some of them sound technical reasons, some of them purely political. The first goal of the program was to return astronauts to the surface of the moon.
The constellation program made slow progress; during the first 4 years it only produced a bad test version of the Ares I.
When President Obama came into office in 2009, he commissioned a review of NASA and the resulting report said NASA had too much on their plate; they could not afford finishing the space station, operating the space station, and building the hardware that constellation would require.
Obama ended up "cancelling" constellation, and there are air quotes around the word "cancelled" because although NASA is under the control of the executive branch, their money all comes from congress.
Congress did not take kindly to what they viewed as the interference of the president. After some negotiation, we ended up with the following:
The president got a commitment for NASA to pursue commercial cargo and crew programs to the ISS.
Congress got a program to develop a "space launch system" that could get out of low earth orbit to access cis-lunar space, and that rocket would be designed to carry at least 130 tons to low earth orbit.
The fine print required SLS to be made out of shuttle parts and utilize constellation contracts wherever possible. Congress was not going to run the risk that NASA would choose a new architecture this time.
Congress also told NASA to keep working on the Orion capsule that was part of constellation so it could be launched by SLS.
We can tell the priorities of Congress by noting that the act that authorized SLS gave no specifics of the missions that it would fly, just that they would be beyond low earth orbit.
That left NASA with a problem. What would they do with this new launch system and new capsule?
Obama had directed NASA to skip going to the moon in favor of visiting an asteroid and then going to Mars.
The asteroid visit quickly got shot down; getting humans to a near-earth asteroid to collect samples and returns is really, really hard - surely harder than a moon mission and probably harder than a Mars mission.
So NASA flexed and came up with a new mission...
NASA would launch a small probe to rendezvous with a near-earth asteroid to either capture a small asteroid or pick up a boulder from the surface of a larger asteroid. It would then return that material into a storage orbit around the moon. It would be a very long mission; about 3 years to get to the asteroid and another 4 years to get the asteroid back into the storage orbit. The long timespans are because the probe would need to use very efficient but low thrust ion thrusters to be able to accomplish the mission.
That illustrates pretty well why a crewed mission wasn't possible...
Once the asteroid was in a distant retrograde lunar orbit, NASA would launch an Orion to dock with the probe, and astronauts would do EVAs to grab pieces of the asteroid and bring them inside the probe/capsule for study.
NASA called it the "Asteroid Redirect Mission" which appears to be an attempt to make people think this was a planetary protection mission rather than a science one.
If you are asking why NASA didn't bring the asteroid back to the ISS where they could perform the analysis and easily return samples back to earth, shame on you. The point is coming up with a use for SLS and Orion, not to do things efficiently.
NASA did quite a bit of work figuring out what it would take to modify Orion for the mission, but the main problem was that it didn't provide much use for SLS and Orion - only one or two launches.
During this period NASA began working on designs for a full space station. Some of these were planned to be in the same distant retrograde orbit that the asteroid mission would have used, but they were also studying the near rectilinear halo orbit as that gave access both to deep space and the lunar surface.
During this time the station went from being named the deep space gateway to the lunar orbital platform - gateway, or "LOP-G", and finally to the Lunar Gateway.
Why does NASA want a new space station? There are two big factors
For years the main US mission control center has been at Johnson Space Center in Houston Texas. It has handled NASA flights since Gemini 4 in 1965, covering all the Apollo flights, all the shuttle flights, and since shuttle retired, it has become the ISS mission control center.
ISS is scheduled to go away in 2030. Mission control will still need to handle the Artemis missions, but those are planned to run once a year for a few weeks. If there are commercial space stations that NASA uses, those companies will have their own mission control centers.
Johnson stands to lose a lot of work and therefore a lot of jobs when this happens and that isn't something that the management at Johnson and the politicians in Texas want to happen.
The solution is gateway. It's a space station and while it doesn't have people in it all the time, it still needs to be managed all the time.
The second reason is related to SLS.
To get SLS launched quickly, NASA adapted an upper stage from the Delta IV rocket and named it the ICPS. It works fine but because it was designed for Delta IV, it's comically small for SLS and can only put Orion into lunar orbit.
NASA is working on a more appropriately sized upper stage - the exploration upper stage - so that it can carry both Orion and extra cargo to lunar orbit. That is the SLS Block 1B version.
Currently, the only cargo it might carry is gateway modules, so Gateway provides a justification for spending a lot of money - at least $5B - on SLS block 1B.
NASA was set. They would get SLS block 1B finished and use it to launch gateway and that makes gateway a part of the moon program.
But then - as often happens - another president has different ideas...
In 2017, President Donald Trump issued a memorandum known as "space policy directive 1" and essentially flipped from the Obama policy of going to mars back to the George W. Bush policy of going back to the moon first. President Trump wanted to get back to the moon "real quick like" (not an actual quote).
That presented a real problem.
The development of the exploration upper stage and the second mobile launcher has been continually delayed and would not be ready quickly enough. NASA needed a plan to do a lunar mission without the lunar gateway.
The solution was to restructure the Artemis 3 flight so that it could complete the first mission back to the lunar surface without gateway, with a target of 2024, since moved out to 2026.
The obvious problem here is that if you can complete a lunar landing without gateway, NASA's assertion that gateway is critical to lunar surface missions is a little harder to support.
So while all this artemis fun was happening, the gateway folks realized they might have a way around the Artemis 3 issue. In their original plan they couldn't launch any modules until Artemis IV and you need two modules to make gateway minimally useful, so that's after Artemis V.
But if you put the first two modules together and launched them on a Falcon Heavy, they could launch them in 2027 and get there in 2028, approximately when Artemis IV would launch. Assuming Artemis IV is a block 1b SLS, that would give NASA 3 gateway modules for the second lunar landing flight.
And that's the story of gateway and why NASA is spending money on it.
At this point it's helpful to learn a little bit more about gateway. It's a bit like a mini version of the international space station in terms of overall size.
It's also smaller when it comes to the modules themselves.
This is the ISS module Harmony nestled inside the payload bay of Discovery. At 4.4 meters or 14 feet in diameter the module pretty much fills up the entire payload bay of shuttle, and it's 7.2 meters or 24 feet long. It has a mass of 14.3 tons.
The gateway Halo habitation and logistics module is only 3 meters in diameter and 7.5 meters long and is expected to mass 8 to 9 tons.
This small diameter has proven to be challenging.
This picture shows astronaut Suni Williams inside the harmony module on ISS, and notice how much space is taken up by machinery, equipment and storage.
Now take away about 1.4 meters, or 4 feet of the diameter. Even with significant work to save space, gateway will be pretty cramped.
The other issue is mass...
Falcon 9 could easily launch a module like harmony to a low earth orbit and land the first stage downrange on a drone ship, but gateway is in a much tougher-to-reach destination.
Launching the gateway power and propulsion module and the halo habitation module together can be done in a single Falcon Heavy launch as long as the modules hit their mass targets, which so far has been problematic for NASA. That doesn't actually get them to lunar orbit but to an orbit where gateway can get the rest of the way there by itself, as long as you're okay with it taking a year.
That single launch - currently planned for 2027 - gives us the two core modules for the gateway station.
If you look at the modules, you will find a lot of international partners, and the plan is to carry those other modules to gateway starting with the Artemis IV mission.
The exceptions are the SpaceX or Blue Origin lunar landing systems, which get to gateway by themselves, and the Dragon XL and HTV-XG logistics modules.
The challenge and cost to get to a lunar orbit - and the current limitation that the only way for astronauts to get there is on an Orion on top of an SLS rocket - means that gateway will not be a continuously inhabited station like ISS. It will be inhabited once a year when lunar missions stop by, and the claim is that there might be crew there for up to 3 months.
How much will gateway cost?
I'm going to give my usual less-than-useful answer - we don't know.
And that's as unsatisfying as usual, so I'll tell you what I do know.
A general account office report on gateway notes that NASA committed to launching the gateway initial capacity - that's the PPE and HALO modules - by December 2027 at a cost of $5.3 billion.
We also find that between fiscal years 2018 and 2029, NASA anticipates spending over $7 billion to build and operate the gateway.
You can put whatever level of credence you think is appropriate on those cost estimates; the first one is only about 9 months old and it seems unlikely that it will survive until the launch in 2027.
I've already covered why I think NASA wants to build gateway, but to be fair we should look at what NASA says.
In 2022 NASA wrote a white paper explaining gateway, and it starts this way.
Okay. That's not a terrible justification, but when I read this, a few of the words jump out at me.
Critical - why is it critical?
Enables - how does it enable?
Steppingstone - how does it make getting to mars easier?
Permanent presence - I do not think those words mean what you think they mean.
You should really go and read the white paper yourself - it's linked in the description - but I do have a few thoughts to share:
NASA talks about science and lists a few of the experiments that will fly on it.
The advantages of hosting your experiment on gateway is that you probably don't have to pay to get it into space and you will get services from the station that you would need to build into your spacecraft otherwise. And it's likely a lot cheaper and easier to get into the gateway program than it is to get NASA to fund you as an independent science mission. On the other hand, you can only fly once a year at best and your experiment has to be remotely operable.
I think there's a good argument to doing experiments on Gateway if you have strong reasons to build it, but I don't think there's a scientific justification to spend the money just to do science there.
NASA plans to use gateway as a waystation between the earth and the lunar surface. How will that work?
Artemis 3 is designed to land on the moon without gateway. Orion will launch with 4 astronauts, but only two of them will move to the starship lunar lander to spend six and a half days on the moon, despite starship having a huge amount of crew space compared to Orion.
The other two astronauts are stuck in Orion for a week. Or, if gateway happens to be done, they will spend that time camped out in Gateway.
This is the worst space consolation prize.
You've worked for years to do something you dreamed about, you get to go to the initial party, but for the main event you are stuck watching it on TV like everybody else.
It's not clear what the plan is for Artemis 4 - the second starship lander is supposedly an enhanced version that could probably carry all 4 astronauts to the surface, so there may be no need for a place to chill while the surface astronauts do their stuff.
Blue Origin's blue moon lander is specified to carry 4 astronauts starting with artemis V.
Gateway ends up being that cheap airport hotel you stay in because you can't get a flight that works for your schedule.
There's another problem that recently came to light.
When NASA designed the power and propulsion module that keeps the station properly oriented, they were thinking of small lunar landers like this Lockheed Martin proposal.
What NASA bought was this, an absolute giant of a lunar lander, one that has a mass approximately 18 times the mass used when designing the PPE module.
In simple terms, it is likely that the PPE cannot control the station when starship is docked.
This problem occurs to a lesser degree with smaller landers like Blue Moon and even with the cargo vehicles that dock with the station. It's not clear yet what the solution might be - it might involve revised software for PPE or it might involve coordinating the PPE thrusters with those of the visiting ship. Or worse, a redesigned PPE.
Another use that NASA mentions for Gateway is as a stop on the way to Mars.
To understand this, we'll need to contrast the SpaceX approach with the one that involves gateway.
In the SpaceX one, a Mars starship refuels in earth orbit - likely from a starship propellant depot - and then lands on Mars. It refuels there - either by making its own propellant or possibly from starship tankers - and then flies back to earth, reenters, and lands.
In this approach, you build your spacecraft on the ground, launch into orbit, and if everything is okay, head to Mars. If not, it's easy to get back to earth and fix things for another try.
Somewhat surprisingly, it's not that hard to get to Mars - it takes about 3600 meters per second of delta v from low earth orbit if you aerobrake on your way to landing on Mars, the same way capsules use the earth's atmosphere to slow down on reentry. It's actually a lot easier than getting to the surface of the moon.
Getting from Mars back to Earth is a lot harder than getting there, requiring 5700 meters per second of delta v.
This wonderful graphic from Ken Kirtland IV shows the NASA Mars Mission Architecture. I don't want to get into the details, but in this design the Mars spacecraft is assembled at gateway.
In our mars direct approach, it took 3600 meters per second of delta v to get to mars and 5700 meters per second to get back.
In this approach, we need to get all of our spacecraft parts into the near rectilinear halo orbit that gateway uses. It's not particularly expensive, but it still takes 450 meters per second to get in and 450 meters per second to get out, bumping up our total to get to from low earth orbit to Mars from 3600 to 4500 meters per second, a 25% increase.
Using gateway means you are assembling your spacecraft in a remote location that increases your fuel costs.
This is like me wanting to fly from seattle to Syracuse New York, but deciding to do it by flying to southern Oregon, staying there for a week, then flying on a different plane back to seattle and then turning to fly directly to Syracuse without landing.
All this approach does is make the architecture more difficult and riskier.
NASA likes to call gateway a springboard to Mars, but from an energy standpoint it's space's largest marshmallow soaking up energy that we could use to carry more payload.
There is a caveat.
If you can build enough lunar industrial capacity to create rocket propellant and launch it into orbit - and be able to do it cheaper than propellant lifted from the earth - then stopping off at the moon on the way to Mars may make sense.
We're a long way from that capability at this time.
NASA also argues that gateway builds on the international cooperation on ISS and aligns with the goals of the Artemis Accords.
Gateway involves the usual suspects - JAXA, NASA, ESA, CSA, and adds in one new entrant, the united arab emirates.
All of those groups will be expecting trips to the lunar surface for their astronauts as part of participating in Gateway. The rest of the Artemis Accords countries can expect no trips to the lunar surface as part of Artemis because Orion is too expensive and flies so rarely, and it's not like there is any other current solution to take them to Gateway to spend time there.
NASA persists in implying that the Artemis moon program and the Artemis accords are tightly linked. SLS and Orion are mostly NASA programs with an assist from the ESA for the Orion service module.
And that's the story.
If NASA wants to build an international partnership around the moon, it should be spending its effort on the lunar surface, because that's where everybody wants to go, and there's a lot more variety of things to be done on the moon than in orbit around it.
If you liked this video, I'd love some tickets to see the gateway...
But actually, what I'd really like you to do is go to the Mia and the Martians project on indiegogo and pledge your support.
What this world needs is more cool space books for kids.
Link is in the description, along with a link to the Off Nominal podcast discussion of the book.
The last nuclear rocket engines were tested about 50 years ago. Since then, there has been a lot of talk - and a lot of claims - about nuclear thermal engines, and it's not clear what they can really do.
NASA has recently started a nuclear propulsion program.
What is the program about, and what can we expect from it?
I'm sure some of you are saying, *Another* nuclear rocket video?
I've done a couple; the most recent on is linked in the corner if you care.
When I started on this one I figured that I could just put up a slide that says nucular thermal rockets are stoopid and be done with it. Not much of a video, really, and not a lot to motivate me to cover the same ground for a third time.
So I did some writing, got bored, and went on to other topics.
Then a few weeks later I came back to it, took another look.
And I... Am... Persuaded... that this program is a good idea...
Which is a stunning reversal. Don't get me wrong; I still think this is a fairly transparent effort to channel some of that sweet NASA money into the bank accounts of companies that have words like "nuclear" or "atomic" in their names.
But it's more than that.
Here are my reasons...
The first reason is that this program is not about building a nuclear rocket engine. Take away all of the rockety parts like turbo pumps and hydrogen tanks and nozzles; this is a program about building a nuclear rocket engine reactor.
The second reason is that this isn't actually a NASA program, though NASA is paying for it.
The program is being run by the US Department of Energy, which is the central point for nuclear activities in the US Government. DoE has very specific ideas on how nuclear projects should be done, and that means that this is about building a real reactor - and by real I mean practical to make, robust, safe, and all the other things you would want in any nuclear reactor. This is categorically different than the "here's a paper about a nuclear rocket engine" approach I've seen so often.
The DoE has chosen Idaho National Laboratory as the location where this work will be managed, and the operation of this laboratory is contracted to Bechtel.
If you read the documents, "contractor" means Bechtel, as they are coordinating the program and "subcontractor" means a company that will be participating in the program. It's a little confusing.
The third reason came after reading Lori Garver's "Escaping Gravity" book, where she talks in depth about NASA existing to do things for the country.
This is NASA's mission statement.
This is exactly the kind of research that NASA should be doing; whether nuclear thermal engines are practical is a significant unknown, and if it can be answered it will affect future planning.
So, yes, I am persuaded.
The details of the process are in this document that I'll link to in the description.
If you understand what NASA did with commercial resupply and commercial crew for ISS, this multi-phase approach will look familiar.
The project uses the standard approach used by Department of Energy projects.
The first part is conceptual design, also known as "phase 1" for this project. The goal is to complete the conceptual design to a 30% level - an "is this feasible" level of detail - and then complete a review.
Any subcontractors that complete that 30% review successfully will be eligible to move to the next phase.
https://www.energy.gov/sites/prod/files/2018/10/f56/EM_CNS%20_Engineering%20and%20Design_SRP_Final%20Version_EM%20Review_%20SEPTEMBER%2020....pdf
Phase 2 is known as the preliminary design phase, and will be awarded with a separate contract, very likely for much more money.
In it, the subcontractor needs to mature their 30% design to a 90% design to feed into the 90% design review.
Any subcontractor who successfully passes this phase will have the opportunity to participate in phase 3.
https://www.energy.gov/sites/prod/files/2018/10/f56/EM_CNS%20_Engineering%20and%20Design_SRP_Final%20Version_EM%20Review_%20SEPTEMBER%2020....pdf
Phase 3 involves creating two prototype reactors - one unfueled and one fully fueled and ready for testing.
https://www.energy.gov/sites/prod/files/2018/10/f56/EM_CNS%20_Engineering%20and%20Design_SRP_Final%20Version_EM%20Review_%20SEPTEMBER%2020....pdf
Successful completion of phase 3 presumably means NASA would be looking at flight hardware to be tested in orbit.
NASA has not specified a reactor design, though there are a few obvious possibilities. I expect a NERVA-like design, but I'm really hoping one of the contractors will try a pebble bed design.
There are a number of requirements the design must meet.
The first requirement is that the engine will use uranium that is 5-20% uranium 235. This precludes some of the early designs that used uranium that was highly enriched, to more than 90% uranium 235. You can be sure that everybody is going to specify 20% as going lower than that just makes their lives harder.
Interestingly, the congressional authorization that set up this program said, "use low enriched uranium if feasible", and NASA decided to make that a requirement. I view that as a good thing.
The second requirement is to use hydrogen as a propellant, though "reaction mass" is perhaps a better term than propellant as it is heated, not burned.
No other comment here as hydrogen is the only propellant that is going to meet the other requirements.
One of the possible advantages of nuclear thermal engines is their higher specific impulse - which gives them better fuel economy.
The goal is to reach a specific impulse of 900 seconds - roughly double what the best chemical rockets can do - and that requires a reactor outlet temperature of 2700 Kelvin to achieve the required exhaust velocity.
900 seconds Isp translates to 8829 meters per second in exhaust velocity, and that requires roughly 2500 kelvin, so 2700 should get them to that level of performance.
The thrust requirement is 55.6 kilonewtons. They express it as a hydrogen flow rate to ensure that the engine preserves the target specific impulse even at the maximum thrust.
This is almost precisely half of the thrust available from an RL-10 engine.
That's pretty small as engines go.
There is a second thrust requirement, of 111.2 kN
Read this as "you should be able to scale your engine up to RL-10 size".
I'm not excited about this second requirement.
I can see an argument for starting small - a program to develop a smaller reactor is likely cheaper. But the way that you scale up to a bigger size is by making the core longer or fatter, and that affects both the reactor design and the rest of the rocket engine.
The big problem here is that I don't understand what is required to meet this part of the requirement. A contractor could just do both designs, and that would be great, but what if they choose not to do that? Do they just promise that it will work in a bigger version with little or no additional technology development.
If NASA wants an RL-10 sized engine, they should just specify that as the size. We all know what happens with a "block 1, block 2" approach.
NASA will accept a reactor that masses 3,500 kg but hope that it will be as light as 2,500 kg.
The mass numbers here and those quoted for other nuclear thermal designs tend to be pretty weasely as nobody has built real systems with them. We have actual numbers from NERVA but they were never designed to be light so that doesn't matter.
For this design the minimum thrust/mass is 1.6, but they hope for 2.3.
We can compare to a couple of proposed engines; the SNRE that was commonly referenced at NASA comes in at 2500 kg for 72.9 kilonewtons of thrust for a 2.9
and the enhanced SNRE at 3250 kg for 111.7 kilonewtons of thrust for a 3.5
This requirement really disappoints me; many nuclear thermal advocates talk about light reactors with advanced materials, and this is quite a bit worse than the ones that were proposed 20+ years ago. Note that does not include the rockety parts - the turbopumps, the nozzle, etc. - though those are fairly light.
The majority of nuclear thermal reactor designs use an internal shield inside of the reactor vessel to reduce the amount of radiation exiting the pressure vessel and a larger external shield outside the reactor. Those are likely to be heavy structures. There is a list of reactor components that should be part of the mass estimate and shielding is not listed.
I'd be happier if there was a requirement for the maximum radiation coming out of the reactor, so that the shielding mass would count and the designs would be on equal footings.
Just to compare, the RL-10 has a thrust/mass of about 47.
The reactor must maintain full power thrust for 2 hours minimum, preferably for 5 hours. This is total operational time across multiple starts.
It should also be capable of 5 starts and shutdowns.
This is unfortunately not well defined. A start is defined as going to "operational power levels", but the word "operational" is not defined. A shutdown is defined as occurring from any power level.
Why is this important?
During NERVA, there were ongoing problems with erosion of the fuel elements due to the hot hydrogen. There were also issues with differing expansion rates causing cracking in the core.
These requirements should therefore be designed to verify that the reactor doesn't have these issues.
I would expect it to be something like:
A reactor cycle is:
The reactor starts at ambient temperature (earth ambient or space ambient?)
The reactor is started and power is increased to full power
Full power is held for at least 15 minutes
The reactor is shut down
Hydrogen flow is continued until the reactor returns to the original temperature.
That would much better simulate the use of the reactor in real-world conditions.
As for reaching these requirements, I suspect that this will be the area where going beyond NERVA will be most important.
The longest that a NERVA ran at full power was just over an hour, and that was at about 2500 Kelvin.
There have been many start/shutdown tests - XE prime did 28 tests - but they only ran up to about 1300 Kelvin for many of them.
https://www.osti.gov/servlets/purl/4257093
https://www.osti.gov/servlets/purl/822135
https://inis.iaea.org/collection/NCLCollectionStore/_Public/05/105/5105500.pdf?r=1
There are many deliverables that go along with the requirements. The deliverables show that the Department of Energy clearly has a very clear idea of how to build well-behaved reactors
First, there is the actual 30% design of the reactor.
Then there is a performance analysis of the reactor - does it meet the requirements, what are the stresses during startup, operation, and shutdown, and what will keep it from blowing up if your launch vehicle drops it in the ocean.
There is a mass estimate - a very detailed mass estimate - of all the parts of the reactor.
And then there are cost and schedule estimates. How much will it cost to build the reactor, and how long will it take?
They must define the interfaces between the reactor and the rest of the engine, describe the hydrogen flow, how the proper balance between nozzle cooling and turbopump power is achieved, and which aspects of the system are most likely to cause issues.
There is a description of instrumentation and control.
The purpose of the control analysis will be to provide information that supports development of a system that safely controls the reactor and engine systems when they are linked together. The fission rate of the reactor is dependent upon the core design, the control drums that are used to control reactivity, the flow of the hydrogen - which also affects reactivity - and other factors.
There is a reactor manufacturing plan that describes how to build the reactor, how it is assembled into an actual engine, what facilities are needed.
There is a test plan that describes how both the components and overall reactor will be tested, and what facilities might be used.
There is a quality assurance plan, which will describe what requirements there should be for testing nuclear thermal reactors for spacecraft propulsion.
And finally, there is a technology readiness assessment, which details what technology development is required for this design to work and how the subcontractor will get there.
This is a large number of deliverables and goes into much more depth than any of the designs that I've seen.
Phase 2 takes the 30% design and pushes it to a 90% design, ready for another big review process.
In addition, it requires the construction of a unit cell of the reactor core, basically enough of the core to be representative of the overall core design, and at least half scale.
In a nerva-style core, it would include the green fuel rods with hydrogen channels in them and the red and blue tie rods that structurally hold the core together. This is a test of "can you build this thing using the approach you say".
These unit cells will be tested in operational reactors such as the Transient Test Reactor at Idaho National Laboratory or the High Flux Isotope Reactor at Oak Ridge National Laboratory.
The point of this testing is to heat up the unit cells in a real reactor, run hydrogen through them, and verify that they are well-behaved under actual reactor conditions
They will very likely also be tested in NASA's NTREES simulator, which heats hydrogen up to operating temperatures - 2700 kelvin - and verifies that the core designs can tolerate it.
This is very useful as the cores do not become radioactive and can be carefully examined.
Finally, phase 3 gets to real reactors...
First a reactor that is mostly identical to the final one but with the uranium replaced with either depleted uranium - the leftover from the enrichment process - or natural, unenriched uranium.
It's a reactor you could plug into the actual engine and everything would fit, except that it just wouldn't work.
And then there's the final product, a reactor that is complete including the appropriate enriched uranium.
The fueled reactor prototype shall be suitable for use in nuclear testing of the reactor designed for demonstration of the ability to control the reactor under a variety of normal operation and accident conditions.
It's designed to be integrated into a rocket engine for real testing.
Beyond phase 3 probably means an on-orbit test.
What companies are involved?
One of the requirements is that you need to include a company that - in the department of energy's opinion - knows how to deal with enriched uranium, so we will see companies that we aren't used to seeing the space world.
NASA awarded contracts to three groups of companies.
The first partnership is BWX technologies who are suppliers of fuel components and fuel, and Lockheed martin would obviously do the spacey part.
Next up we have general atomics, X energy, and Aerojet Rocketdyne. A traditional nuclear company, a new nuclear company, and an experienced rocket company.
Finally, the last partnership is a big group. A new nuclear company, blue origin presumably for engine skills, GE Hitachi nuclear, GE research, Framatone (a French reactor company) and Materion, a high-performance materials company.
I frankly don't see a good way to handicap these companies until we see what comes out of phase 1, but I'm happy there are three groups.
Perhaps the biggest surprise to me is that the phase 1 awards are only about $5 million each. There's a lot of detailed engineering work to be done for a small amount of money, so presumably these groups hope to see a lot more later.
If you enjoyed this video, please help me rhyme "framatome" for a commemorative poem I am writing...
https://www.ulalaunch.com/docs/default-source/supporting-technologies/launch-vehicle-recovery-and-reuse-(aiaa-space-2015).pdf
https://forum.nasaspaceflight.com/index.php?topic=37390.0
Ever since SpaceX has been flying Falcon 9, United Launch Alliance has been producing information that compares what ULA does to what SpaceX does.
Some people have considered that information to be PR, the kind of public relations material that any good corporation routinely produces to advocate for its products.
Others have chosen a less kind label, claiming the information is full of untruths or outright lies.
Which is it?
We're going to walk through some of those informational exhibits, examine the details, and the you can decide how you would label them.
Here's our first exhibit, which I hadn't seen or didn't remember, but luckily Eric Berger posted it a while back.
Let's start with the support to NASA section. SpaceX has gotten $2.5 billion from NASA, have only launched 3 times, and that means NASA is paying $840 million per flight. As they say in the PR biz, the optics aren't very good - SpaceX has gotten a ton of money and hasn't given much value back.
My initial goals was to validate all the ULA numbers, but because of the way the contracts were written it was hard to find the data I needed and I frankly got lazy, so I'm going to take them at their word.
At least mostly at their word.
The data says that ULA did 11 launches for NASA from 2007 to 5/7/14. I went looking for data, and I found that they launched 17 missions during that time period.
This is just the kind of attention to detail that makes you confident in a set of data.
Let's see if we can validate this SpaceX number, the $2.5 billion number.
SpaceX had two contracts during this time period.
The first is the commercial orbital transportation services, or COTs money that came to SpaceX through a Space Act Agreement (link to funding video here?). This money was spent to develop Falcon 9 and Cargo Dragon. SpaceX kicked in $454 million of their own money for this project
The second is the commercial resupply services, or CRS contract, which was for $1.6 billion dollars for 12 flights
Add those up, and we get a number that I'm going to call $2 billion. Can we find the extra half billion?
NASA did make two launch contracts with spacex in 2016 for Jason-3 and TESS, and that added another $169 million.
The thing that isn't readily apparent is that this scorecard is talking about contracts. Contracts are an agreement to do something at some point in the future in exchange for a given amount of money when that thing is done. In particular, the CRS contract is a "pay as you go" contract; NASA will pay SpaceX some money when they order a mission the balance when the mission is completed.
So the contract number mostly represents possible future payments, not money that has already been paid.
Note that the line in the scorecard very carefully says "NASA contracted $/launch" is does not say "cost per launch". Though you can argue that cost per lb does not do that.
Let's look at the actual SpaceX numbers...
We'll see if there are any other NASA contracts during that time period.
There was the deep space climate observer, otherwise known as DISCOVR. SpaceX got a $97 million contract to launch it, but that contract was made by the Air Force, not NASA, so it doesn't belong here.
Jason-3 was contracted on Jul 2012 for $82 million and TESS was contracted in November of 2016 for $87.
But these two missions illustrate the problem with the approach ULA is using. Companies do not get paid the money for the launch when the contract is signed. They may be paid a small amount along the way, but the bulk of their payment comes when the payload is launched.
These missions were launched in 2016 and 2018, so they don't fall in this period in terms of revenue.
ULA is playing a cute little game here. Note that the label does not say Cost per launch, it says NASA contracted $ per launch. They're taking the amount of contracts awarded in a period and dividing it by the number of actual flights in that same period. And they're implying that it is cost per launch.
A few lines down, they calculate a "cost per pound" metric by taking the total contract value and dividing it by the number of pounds. That's not a true value because they cost they are using is not associated with the flights they are using.
Let's see if we can come up with a better value for SpaceX...
SpaceX flew two demo flights under COTS. The second demo flight successfully made it to the space station.
The cost of both of these flights was covered under the COTS contract, which at this point has been fully paid - SpaceX has received all that money.
There were then three operational CRS flights during this time period out of the 12 contracted. Under the contract, those cost $133 million each for a total of $399 million and an overall total of $ 795 million. If we look at those three flights, that's a per-flight cost of $265 million. If we add in the successful demo 2 flight, that's a per flight cost of $199 million.
The $265 million is pretty close to the per-mission cost of ULA, and the $199 million is less, though as I noted they actually flew 17 flights and the $2.6B they listed is not what they charged for those flights.
Let's go onto the mass numbers. The number they're using for SpaceX is not mass delivered to orbit as a dragon is about 13,000 pounds by itself, but I'm fine with using "mass delivered to the space station".
Here are the mass numbers of cargo delivered by these missions. CRS-3 sees a big jump because it's the first flight of the much larger and more capable Falcon 9 V1.1. The total is 9594 lbs.
I now challenge you to find a reasonable combination of these masses that gives you the 6,671 lbs that ULA claims. I have had no success. These numbers are the ones NASA gives out so they are readily available and I find it confusing the ULA number is wrong.
That 9594 lbs delivered to the ISS comes to $83,000 per lb. That is pretty high compared to the ULA figure, but we need to consider that each of these launches includes a capsule that navigates to the ISS and keeps most of the payload pressurized. It also returns payload back to earth. The ULA numbers are purely the mass of the payload on top of the third stage.
If you include the dragon mass, the number is about $13,000 per lb.
NASA was very pleased with COTs and CRS - the up front cost was $953 million, which was about 10 months worth of the space shuttle's yearly upgrade budget. And the cost per kilogram of cargo delivered was 1/3 to ½ of what shuttle cost.
Moving down to the industrial base impact, I have two things to say.
First, the label on this is "cost scorecard" and none of these are about cost.
Second, if you don't know how many suppliers your competitor has you might as well give up doing comparisons that require knowing that data. I do not doubt that ULA has far more "rocket job" suppliers than SpaceX does, but the question marks here do not inspire confidence.
Finally, on technical and financial transparency...
There is no way that ULA can evaluate how much information SpaceX gives to customers because they are not a SpaceX customer. I do think the ULA payload guides have more information than the Falcon 9 guide, but a) there is a lot of information there and b) it's pretty obvious that customers can ask SpaceX for more information.
The cost reduction part caught my eye. First, they talk about costs - which are generally the internal costs - rather than price for ULA and then talk about price for SpaceX. I'm going to assume that they mean price for both because customers do not care about internal costs, they care about price.
ULA is claiming a 5% reduction year-over-year.
Let's look at a couple of missions.
In 2009, ULA launched the LRO/LCROSS mission for NASA on an Atlas V 401 rocket, and received $132 million from NASA for a contract signed in 2007.
In 2014, ULA launched the TDRS-L satellite, also on an Atlas V 401 rocket. The contract for that launch was signed in 2013, and in 6 years inflation would have pushed the price from $132 million up to $148 million. With a 5% reduction in cost every year, we would expect that launch to be 36% cheaper, or approximately $110 million.
The TDRS launch was part of a contract for four Atlas V launches that according to NASA cost "approximately $600 million". So the cost of this one launch was around $150 million, or pretty much what we would expect it to cost.
ULA *did* reduce their Atlas V prices considerably a few years later to better compete with SpaceX.
For financial transparency I don't think it's worth the effort to go into the details, but there are two points to make.
The first is that SpaceX is fully compliant with the requirements for working with NASA.
The second is that I'm not sure what point ULA is trying to make on actual prices; the prices on the website are for satellite launches, and the contracts with NASA are flying a cargo capsule to ISS. It's pretty obvious that the prices will be different as the second requires a capsule. It's also very common for government launches to cost extra because they require extra services. ULA knows this as they launched two comsats during this period.
So what do you think about "know the facts, understand the truth"?
Next, we'll talk about a paper that ULA wrote about launch vehicle recovery and reuse and published in September of 2015. It's mostly about ULA's approach for reuse that they call "SMART reuse", but there's one short section titled that "Economics of Reuse" that has generated a lot of discussion.
I've linked the paper in the description for the video, and I've also included a link to a very interesting discussion about the model on NasaSpaceFlight.com, with links to the spreadsheet.
The model used is fairly complex - overly complex in my opinion - but the part that generated the most heat was this graph which asserted that ULA's SMART reuse approach would reach break-even at 2 flights but it would take the Falcon 9 reuse approach 10 flights to break even.
The math used is fairly ugly so I'm going to try an alternate approach.
But first, we need to talk about differences between rockets...
I'm going to use Atlas V because that was the launcher flying at the time the paper was written, but the results apply to Vulcan as well. And I'm going to talk about payload to a 400 km low earth orbit.
Atlas 5 is a "choose your performance" rocket. The base "501" configuration can lift about 8 tons. Add a single solid rocket booster, and that gives you 10.7 tons of performance. Keep doing this all the way up to 5 boosters, which gives you 18.5 tons of performance.
The important point here is that once you get above the 501, there is never much extra performance with an Atlas V.
The Falcon 9 is also a "choose your performance" rocket, but in a very different way. If you need the full performance of 22.8 tons, you can fly expended (you can't actually use this performance for LEO but you can for higher energy orbits).
If you need less than 17.5 tons, you can land on the drone ship.
If you need less than 13.5 tons, you can land back at the launch site.
The ULA approach uses $ per kilogram to orbit as their measure.
If you dig into the NASASpaceFlight.com thread on this topic, you will find some discussion with Dr. Sowers about why that metric was chosen.
Let's explore an alternate metric - looking at reuse based on an analysis of likely payload sizes.
7400 (3185kg) - 7420 (8) + 7425 (2) = 10
7300 (2800 kg) - 7326 (2) + 7320 (7) = 9
7900 (6000 kg) - 7925 (31) + 7920 (11) = 42
We can do a simplified analysis based on market data, using the ULA launch data from 2000-2009. I've normalized all these launches as if they were LEO launches to simplify things.
We'll start with the atlas V. It flies with 0 to 5 solid rocket boosters with the configurations 501 for no boosters to 551 for 5 boosters. I've put them on the chart here to show what the LEO capability of each configuration is. And I converted the 4 meter variants to 5 meters. The no booster variant flew 9 times, 1 booster flew twice, two boosters flew 5 times, 3 boosters flew twice, and 5 boosters flew once.
The Delta II was still flying a lot during this period. The 7300 variant flew 9 times, the 7400 flew 4 times, and the 7900 flew a staggering 42 times. Like the atlas V, the vertical location of the number shows the max LEO payload for each of these variants.
The Delta IV flew some. The M variant flew 3 times, the M+(5,4) variant flew once, and the M+(4,2) variant flew 4 times. In addition the Delta Heavy flew 3 times, putting it way at the top of the graph.
ULA flew 91 flights , and this provides a reasonable approximation of what the government commercial market was during this decade.
. Looking at the Atlas V launches from 2000-2009, there were 9 launches with no boosters, 2 with 1, 5 with 2, 2 with 3, zero with 4, and 1 with 5.
This data is skewed from a market perspective because it's just US government launches, but it's a starting point.
Let's push those over to the SpaceX side, nothing that this is a little unfair because these are Falcon 9 2024 capabilities.
Starting at the top, we end up with one payload that needs an expendable ride, seven that can use drone ship recovery, and 11 that can use return to launch site
Delta IV Medium had 8 launches during this time, in three different variants. I'm going to charitably say half of those are drone ship and half of those are return to launch site.
Of 31 launches, 1 would be expendable class, 11 would land on the drone ship, and 19 would land at the launch site, for a total of 30 out of 31 reusable launches with no loss of payload.
The big point here is that reuse depends on how your rocket size relates to the size of payloads that customers want to launch. Falcon 9 version 1.0 could only lift 10.4 tons to low earth orbit, so it would only have been able to lift 12 of these payloads and only in expendable mode. Reuse on Falcon 9 Version 1 would not have been practical.
What would that market allocation mean for a launcher like the Falcon 9.
The falcon 9 can carry 22.8 tons into LEO in expendable mode - though it would need changes to the payload adapter and second stage to carry that much. It can carry about 17 tons with recovery on the drone ship, and about 13 tons with recovery back at the launch site. Let's overlay our ULA market data on our scale.
Before we evaluate these, we need to do a bit of a mental adjustment. Most of these launches were not launches to LEO and the ULA launchers have a bit of a payload advantage on higher energy orbits, so they payload numbers for all of these launches need to be pushed up a bit. What do we notice?
All 61 of the delta II launches, 11 of the Atlas V launches, and 4 of the delta IV launches can all be handled by Falcon 9 in the return to launch site scenario.
11 of the launches could be handled by a first stage landing on the drone ship.
Only one launch required expending the booster.
And three launches - the three Delta IV Heavy launches - could not be handle at all by the Falcon 9.
If we do some statistics, 82% could be handled by landing at the launch side, 12% by landing on a drone ship, and 1% by expending the booster. 3% would require a different rocket.
The key is to correctly size the rocket to the market. If we look at the payload capacity of the Falcon 9 version 1.0, we see that it *maybe* could have been reusable with the delta II payloads but it could not have carried most of the Atlas V or Delta IV payloads.
But there is a case where the dollars per kilogram to orbit may be more applicable. And that is constellations...
I don't want to try to explain the model from the paper, so we're going to go with one that I believe is simpler.
We are going to be comparing Falcon 9 in expendable mode to Falcon 9 with the first stage landing on the drone ship.
And let's just throw some numbers at the overall cost. As usual, these numbers are for entertainment purposes only and do not constitute real Falcon 9 costs.
We will say the first stage costs 20 million, the second stage costs 10 million, and there's another 5 million someplace. And we'll also say that recovery and refurbishment costs 5 million dollars.
We'll start with boosters that are flown 5 times. The reusable vehicle carries 17 tons 5 times, for a total of 85 tons. The expendable vehicle caries 22.8 tons 4 times, for a total of 91 tons. It's about 7% more than the reusable scenario.
For the expendable scenario, we buy four first stages at $20 million each, four second stages at 10 million, four others at 5 million each, for a total of $140 million. For the drone ship, we buy one first stage, four recovery and refurbishings, 5 second stages, 5 others, for a total of $115 million. 82% of the expendable cost
You can push these numbers around pretty easily. Push the cost of the first stage down, the cost of refurbishings up, or the payload penalty up, and you can easily make the expendable version look better.
If we bump that up to 10 launches per booster, we get down to 77%.
At 20 reuses, it goes to 59%.
Note that the savings is likely underestimated, especially in this scenario. The factory space and workforce required to build 20 first stages and 20 second stages will be significantly more than building 1 first stage and 20 second stages, and the increase in fixed costs would be make the expendable scenario more expensive.
The ULA analysis is less broken in this scenario, but I think the real problem is that they chose some values that happened to make their point look good and didn't bother to show other cases where they didn't look so good.
We start with a fun graph, a modification of one used previously.
It lists 8 different destinations and then groups them into one group that is "low energy" and another that is high energy.
If we read the associated text, it says the following:
The missions on the right require more energy and significantly more complexity. In fact, all of the missions seem to fall into two broad categories: those on the left we'll call "Low Energy" and those on the right that we'll call "High Energy."
In rocketry, we talk about energy in terms of "delta v", which is the amount of velocity change we need to be able to make to our spacecraft to get into a specific orbit. If you want the details, see my "planning your solar system road trip" video.
This image is a delta-v map of the terrestrial planets in the solar system.
Starting on the earth's surface, it takes about 9300 meters per second of delta v to get into a low earth orbit. If you want to go to a harder destination, it will take more energy - if you want to send your payload to the moon it will take about 3100 meters/second of additional delta v. Even though we express it in terms of velocity, it's really a measure of energy.
To get a payload to the moon, the rocket has to have a lot of propellant left over when it gets to low earth orbit, and therefore it can't as much payload. The moon is a higher energy destination.
But ULA has decided that they are going to redefine a commonly understood concept in a different way.
They do include delta V as one of the things they show, but note that the delta v in the "high energy" GTO enhanced mission is lower than the "low energy" interplanetary mission.
The other factors they chart are the number of burns that the second stage needs to make, the longest coast time the second stage makes, and the time to separation, presumably after the last burn.
What they are trying to convey is the kind of second stage that you need to complete these missions, which is a fine concept. Maybe they could convey it in a different way:
Perhaps like this, from a 2022 article by "Tory Bruno", which says the missions to the right have increasing difficulty and complexity. Apparently the 2020 Tory Bruno felt differently than the 2024 Tory Bruno.
As somebody who cares about effective presentation of data, the color coding on the bars is a bit confusing and you really need to start your Y axis at zero if you want to make it understandable.
Switching back, ULA has decided that their new brand is "high energy".
Let's move onto rocket architectures. This slide is comparing two rockets...
The rocket on the right is very obviously Vulcan Centaur. But what is this rocket? It's a weird generic rocket with a single first stage engine and a second stage inside of some kind of fairing. Since starship isn't aiming at this sort of market, we're going to assume that the rockets here are falcon 9 and falcon heavy.
I'm not going to go over all the individual items shown here, but I have a few thoughts to share.
The first is that most of the missions that that ULA is talking about are National security space launch missions for the department of defense, and SpaceX has flown 9 Falcon 9 missions and 3 Falcon Heavy missions as part of this program. Vulcan is a new rocket and has currently not certified to fly these missions, so it has flown zero missions of this category. That means that the things ULA is claiming are unique to ULA are already flying for SpaceX.
There's one more thing I want to cover here. ULA claims that the upper stage of the rocket climbs to low earth orbit, and that's certainly true; Falcon 9 stages low and early to make recovery and reuse possible.
It's this other claim that Vulcan's upper stage transits from low earth orbit.
This is covered in the text as well...
Bruno writes, (read)
So, the assertion is that the Vulcan booster first stage will carry Centaur V nearly to LEO.
Interesting claim...
If you want the full slide that explains where I got these numbers, you can find it at the end of the video.
We have pretty good numbers on the new Centaur V stage and that means we can do some decent calculations. We are going to be looking at a mission all the way to geostationary orbit because that is the one that is the highest energy. Looking at the delta-v map of the solar system, we find that it takes 13,160 meters per second of delta v to get there. This is an approximate amount but it's good enough for our purposes.
Vulcan supports three different configurations for sending payloads go geostationary orbit. The two booster VC2S version can deliver 2.5 tons, the four booster VC4S version can deliver 4.8 tons, and the six booster VC6S version. There's also an "extended" version that uses a slightly more efficient engine so that Vulcan can deliver 7 tons to meet the NSSL requirement.
From the information we know about Vulcan centaur, we can figure out the amount of useful work it can do, measured in terms of delta v. We get 9050, 8094, and 7695 meters per second. This is very good performance - the Centaur V is a light stage and the engine it uses is very efficient.
We can now subtract these numbers from the 13,160 meters per second required to estimate the contribution of the first stage. We get 4110, 5066, and 5464 meters per second of delta v. Suffice it to say that none of these numbers are remotely near the amount of delta v it takes to get to LEO; the last one is only 59%.
I will of course compare these numbers to those of the Falcon 9. SpaceX does not publish payload numbers to geostationary orbit, so I'm using an estimate of 3.9 tons. With that payload, the second stage can generate 9694 meters per second of delta v and the booster contributes only 3466 meters per second. Much less than the Vulcan, which is not surprising because the Falcon 9 stages early.
We can also look at the thrust to mass of the second stages. The Centaur V thrust to mass ratio is very low - only about 0.35 - while the Falcon 9 thrust to mass ratio is about 0.76.
Why are these two designs so different?
The big factor is that ULA buys their engines. For a long long time, the only engine you could buy for an upper stage is the RL-10, and ULA has used them for years. They are very efficient engines, burning liquid hydrogen and liquid oxygen but they are low thrust and they are expensive. Atlas II used two of them, Atlas III and Atlas V used a single one - perhaps to reduce costs - and the bigger Vulcan Centaur uses two of them. If the engines you can buy are low thrust the only thing you can do is build a less capable second stage and then a beefy first stage to get it where it needs to be.
Here's a example of an Atlas V launch; you can see that the first stage is great at accelerating but when the second stage kicks in the low thrust/weight ratio really penalizes the stage.
It's actually worse than it looks; if we look at the altitude graph we can see that the first stage puts a lot of energy into vertical velocity - using a lofted trajectory - so that the second stage can concentrate on horizontal speed.
That's one of the downsides of using the RL-10 in your second stage. But if you are ULA, your only other choice is to use the BE-3U from Blue Origin but since you already depend on their BE-4 for the first stage that's not very palatable.
Using this engine means that you need a big beefy first stage and you need to stage late because the thrust to weight ratio is so low that if you stage early, your second stage will fall back into the ocean.
Why are these two designs so different?
The big factor is that ULA buys their engines. For a long long time, the only engine you can buy for an upper stage is the RL-10. ULA has used them for years. They are very efficient engines, burning liquid hydrogen and liquid oxygen but they are low thrust and they are expensive. Atlas II used two of them, Atlas III and V used a single one - perhaps to reduce costs - and the bigger Vulcan Centaur uses two of them.
Here's a example of an Atlas V launch; you can see that the first stage is great at accelerating but when the second stage kicks in the low thrust/weight ratio really penalizes the stage.
It's actually worse than it looks; if we look at the altitude graph we can see that the first stage puts a lot of energy into vertical velocity - using a lofted trajectory - so that the second stage can concentrate on horizontal speed.
That's one of the downsides of using the RL-10 in your second stage. But if you are ULA, your only other choice is to use the BE-3U from Blue Origin but since you already depend on their BE-4 for the first stage that's not very palatable.
Using this engine means that you need a big beefy first stage and you need to stage late because the thrust to weight ratio is so low that if you stage early, your second stage will fall back into the ocean.
If you enjoyed this video, please send me this original painting, "Love in the time of Polaris", available for the modest sum of $5400.
Okay. Let's do some fact checking. The rocket in the picture is clearly a Vulcan Centaur and we're going to be choosing a mission to geostationary orbit because that's the hardest when it comes to energy.
What we're going to do is calculate how much delta v that centaur v second stage will generate with the posted payloads, and we know that any delta v that the second stage can't generate must come from the current stage.
I'm using the best numbers that I could find but some of them will be wrong and delta-v estimation ignores a number of real-world effects like gravity losses, so the numbers are approximate.
The VC0S version without boosters can't get any payload to geostationary orbit, so we'll only look at the three other versions. We have a set of measures and calculations that we'll use.
There are some that are fixed - the empty mass of the stage is always 5500 kilograms, or 5.5 tons. The amount of propellant is always 54 tons, and the thrust of the two RL-10 engines is always 21 kilograms.
The payload to GEO for the VC2S version is only 2.5 tons, which puts the total mass at 62 tons and the final mass - the empty mass plus the payload mass - at 8 tons. That gives a mass ratio of 7.75 - the ratio of the starting mass full of propellant and the final mass empty of propellant.
With the total mass and thrust, we can calculate the thrust/mass ratio, and it's a disappointing 0.35. It's not uncommon for second stages to have ratios less than 1, but this is quite low. That is in line with "let the booster do all the work".
We can plug our numbers into the rocket equation and we find that this combination gives us 9050 meters per second of delta v. Which is a lot. If you want to know more about how that was calculated, see my video "the care and feeding of the rocket equation".
We know that our total delta v is around 13,160 meters per second, so that means the first stage furnished 4110 meters per second of delta v. Not an amount that I would call "nearly to LEO", which takes over 9000.
Hmm. Let's look at more boosters....
The only changes to the second stages are the slightly higher payloads which don't push the numbers around a lot.
We see that with 4 boosters the Centaur V second stage generates 8094 meters per second of delta v for a first stage contribution of 5066 meters per second.
With 6 boosters its 7695 meters per second of delta v, and a first stage contribution of 5464 meters per second.
Suffice it to say that none of these numbers are remotely near LEO; the last one is only 59%.
On November 16th 2022, SLS and Orion finally took flight on the first flight in NASA's Artemis moon mission. It was a long mission, running 25 days and landing on December 11th.
NASA management was very pleased.
*read quote*
NASA immediately started their post-flight review process.
NASA would spend roughly 3 months reviewing the data, as part of the Artemis 1 post-flight assessment review.
On March 7th of 2023, they held a press conference where they declared the mission a success but noted there were still reviews underway.
The press release described some of the issues that were still being worked on, but did not release the details of the post flight assessment review.
NASA released this picture of the flown Orion capsule heat shield, and provided the following status:
(read)
(read)
This sort of thing happens all the time, especially on test flights, and it's not much of a story. On May 6th, 2024, ULA scrubbed the first crewed launch of Boeing's starliner capsule because of a misbehaving valve on the centaur upper stage.
(read)
"Out of an abundance of caution..." is a fairly standard phrasing for "it's probably okay, but we're going to do something anyway..."
The NASA information about the heat shield had the same tone.
What we didn't know at the time is that there were pictures of the bottom of the heat shield. I don't think anybody had this sort of damage in mind when they read the NASA phrase "... wore away differently than predicted".
What has happened is that NASA management decided to do their best to minimize the issue with the heat shield and not release these photos.
Which was their first mistake, and honestly it's a pretty stupid move...
Many organizations will at some point need to deal with a situation that involves negative PR, and there are many guides and PR professionals who would be happy to help.
Here's a guide from Forbes. I've picked out 5 of their 13 rules...
Take responsibility. Do not try to cover up the PR crisis, it will only worsen the damage.
Be proactive, be transparent, be accountable.
Get ahead of the story. You want information to be coming from you rather than from other sources.
Turn off the fan. Don't fuel the fire.
Be prepared. Anticipate potential scenarios and establish internal protocols for handling them.
This is how you keep a bad situation from getting worse. Before we get to the "getting worse" part, we need to talk about heat shields...
The heat shields for Apollo were made of a fiberglass hexagonal structure that was filled with resin. This is a fragment of the heat shield that was flown on the Apollo AS-202 uncrewed test flight in 1968. You can see the cell structure quite clearly.
The downside of the apollo approach is that it is very labor intensive; every cell in the heat shield is filled by hand.
The material they use is known as avcoat.
For Orion, Lockheed Martin used the same approach. This is the heat shield that flew on the Orion EFT-1 test flight on top of a Delta IV Heavy in 2014.
If we look at a close up, we see the same hexagonal structure used for Apollo.
Boeing's starliner also uses the honeycomb cell structure filled with ablator material.
In the 1980s, NASA invented a new heat shield material known as the Phenolic impregnated carbon ablator - or PICA (pee cah) - for use on planetary missions, and it performed well on the Mars Science Laboratory mission in 2012 which delivered the Curiosity rover to the surface.
That led to a collaboration between SpaceX and NASA to create a new material known as PICA-X which is used on the Crew Dragon. That material is created in tiles that are then applied to the base of the capsule.
After the EFT-1 test flight, NASA and Lockheed Martin decided to go with a different heat shield for Orion. Instead of the hexagonal cell structure that had been used on EFT-1 and Apollo, this approach uses blocks of AVCOAT.
NASA said this about the change.
(read)
The important thing to note is that Artemis 1 flew with a brand new heat shield design that has never been flown before with a material - AVCOAT blocks - that also had not been flown before.
This is very much a non-NASA thing to do - SLS and Orion have leaned on old technology as much as possible because Artemis 1 is the only uncrewed test flight and therefore things need to work the first time.
In the picture at the right, it appears that most of the erosion issues are next to joints between blocks, but the left pictures shows there can be issues inside the blocks.
Now that we understand the heat shields a bit better, we can return to our story...
Soon after the review of Artemis 1, the NASA office of the inspector general started an audit to review NASA's readiness for the Artemis II crewed mission to lunar orbit. The provided the following justification for why such an audit was important.
(read)
These audits are part of NASA's normal procedure; this is one of 14 audits conducted on different facets of the Artemis program since 2020 so it's something that NASA organizations will expect.
The weird part here is that the Artemis folks - the Exploration Systems Development mission directorate, or ESDMD - did not provide the auditors with the materials from the post-flight assessment review that was completed in February.
In the OIG report, they say that they did not review the actual post flight assessment review because some information was not finalized or reviewed by NASA management.
It's pretty clear they asked for it and were told no. Which is the second stupid move by management.
For the next year, OIG works on their review in the absence of the review information, and, apparently, without status updates on issues from the Artemis folks.
Eventually, they finish their audit and write their report.
At that point, it is customary to send the report to the audited organization for comments on the conclusions and recommendations of the audit. Their official response will be added to the report as an appendix.
At this point, ESDMD management realizes that the a) they don't have a solution for the heat shield issue yet and b) there will probably be a reaction to the heat shield pictures, so they wisely decide to get out in front of the report and release the information themselves so they can control the narrative.
Just kidding - management has never read the Forbes article and is apparently ignoring what their PR people are advising.
There are 6 recommendations in the report.
(read)
ESDMD provides status on all 6 of them in their response. Three have been completed.
Two heat shield issues are still active with an expected completion data of June 30th, and there is a launch imagery issue that is expected to be completed on September 30th.
The imagery one is a bit perplexing. There were some issues with misalignment of cameras and camera fogginess, but most of the issues came because the camera exposures was not adjusted for the night launch. It is 15 months since Artemis 1 flew but it is going to take another 6 months to figure out how to set up the cameras properly.
ESDMD concurs with all the recommendations, which means they agree they are the right things to do going forward.
Then they decide to turn up the fan.
In the beginning of their response, they write the following:
(read)
Maybe if ESDMD was more forthcoming when working with OIG, they would have known the details of what was being worked on rather than having to discover it on their own.
Then they add
(read)
ESDMD has apparently forgotten that OIG gets the last word:
(read)
(read)
ESDMD apparently made little effort to keep OIG auditors in the loop with the original flight review content or updates along the way, and complaining about situation that they are responsible for looks like whining.
It's amusing that ESDMD has no inkling of what is going to happen when the report is released; their biggest concern is that the report talks about things that are already fixed and that, for some reason, somebody might think that OIG is in charge of technical solutions in ESDMD.
Apparently nobody who mattered was paying attention enough to notice that disturbing heat shield picture or note that their hope that they would have a solution before it was disclosed did not pan out.
May 1st and the report is released. The response is what is expected and not at all what NASA wants. The OIG report is driving the story, and ESDMD's goal of shifting blame onto OIG is ignored, which is good, as it makes them look not only non-transparent, but petty.
However, veteran space reporter Eric Berger writes a more insightful story, commenting on ESDMD's response to the report.
Berger notes that (read). This is spot on, except that he uses "downplayed" rather than the phrase I would use, which is "covered up"...
He discusses the final comment by Catherine Koerner, the associate administrator for ESDMD.
Berger comments (read)
I agree fully with Berger's analysis, and I would go farther and say that this demonstrates that NASA management is more concerned about appearance and politics between internal groups than with their actual mission.
Was that the right interpretation of the quote?
Berger asked for a response from NASA, and they provided this. (read)
First, this response says that the interpretation is wrong without explaining what the sentence actually meant. It's a non-response, and the items they mention are a) already covered and b) pretty clearly their own fault.
Second, it's pretty clear that a group that is not transparent with their internal auditors is even worse with the public.
Third, telling the press that they are wrong is a stupid thing to do even if they are wrong. All you are doing is further damaging the relationship.
Take responsibility, and turn off the fan....
Roughly a week after the release of the OIG report, we learned that NASA had commissioned an independent review team in late April to have "another set of eyes" and plan to either be done at the end of June or this summer, depending on who you talk to.
NASA has not revealed who is on the independent review team and what their exact charter is.
And that's the current state with the Orion Heat Shield.
A few parting thoughts.
The independent review is a good change in plan since a year of analysis did not resolve the problem, but it's not clear what "independent" means in this context or who the reviewers are. Finishing by the end of June seems unlikely since that was the target date before the independent review was spun up.
ESDMD management is not trustworthy or transparent. They can't even work well with their auditors, much less provide information to the public as they are required to by the Space Act of 1958. Given that,
ESDMD is going to have to deal with a lot of freedom of information act - or FOIA - requests in the coming years because that is the only way to get the full story.
If you enjoyed this video, please send me this tile that supposedly had something to do with the Soviet Buran space shuttle.
What will it take to fly people on starship?
One of my recent videos provides information on the history of reliability analysis and risk tolerance at NASA, and you might want to go watch it first.
If you follow online discussions, there have been strong opinions about flying crew on starship. Here are a few examples...
Are these opinions right?
I certainly have an opinion, but if you've watched any of my videos you're probably expecting it's going to take a while to get there.
Just to keep things from getting boring, let's do something different.
(read)
Now that we've gotten that out of the way, feel free to tell me why I'm wrong in the comment sections. Or... maybe you could watch the rest of the video first, and then decide.
In reference to the space shuttle, NASA administrator James C. Fletcher said:
(read)
When I read that, another quote sprang to mind.
(read)
The problem I'm having is with this word "safe". Because "safe" is not a binary thing.
What we really care about is "safe enough".
Or, to put it more simply, is the benefit we get from the activity worth the risk of engaging in the activity.
Formula 1 drivers at one point accepted a risk of dying of 1 in 100 per year, and that met the bar for "safe enough"
What about a different scenario?
What about a 1 in 10,000 risk per year? Is that safe enough?
I don't think so.
The current number is probably about 1 in 5,000,000 per year, and there are certainly people who would argue that that isn't safe enough.
The main point here is that risk is a relative thing based on the scenario.
***
1/100 is historical data from Formula 1; newer races are likely much safer.
The average child rides 5 miles on a school bus and does this 200 (ish) days a year, so about 1000 miles per year.
NHTSA says the fatality rate is 0.2 fatalities per 100 million vehicle miles travelled, or 1 fatality per 500 million miles. Assume 10 children per bus, and that puts the per child per year risk at
500 million * 10 / 1000 = 1 in 5,000,000
What is safe enough to put humans on starship?
Here are five vehicles that NASA has either flown astronauts on or plans to fly astronauts on.
We have the mighty Saturn V rocket used on the Apollo missions.
We have the Russian Soyuz that carried 77 astronauts to the space station and back for NASA
We have the space shuttle, which NASA flew for 30 years.
We have the Falcon 9 and Crew Dragon.
We have the SLS rocket and Orion capsule, which NASA will be flying on upcoming Artemis missions
Time for a pop quiz!
Your task is to rank these launch vehicles according to safety
Spend a few seconds thinking about that. I'll wait.
Here are the results...
Coming in at #5, we have the Apollo Saturn V. We don't actually have a true estimate for this vehicle because NASA chose not to do one, but it's probably around a 1 in 10 chance of losing the crew per flight on a full lunar mission.
Coming in at #4, we have the space shuttle. NASA was happy to fly shuttle for about 20 years without good risk estimates, but when they finally did them, early shuttle flights were estimated at about 1 in 10. By 2010, shuttle got to 1 in 90, at which point it was retired because it wasn't safe enough. I think I might be making that up...
Coming in at #3, we have the SLS Orion, where NASA has chosen a 1 in 75 target for a mission around the moon and back.
Coming in at #2, we have the ubiquitous Russian Soyuz at about 1 in 100. This has been a very reliable launcher since the early 1970s, though there have been quality concerns recently.
And finally at #1, we have the Falcon 9 and Crew Dragon, at a NASA-verified 1 in 276.
How did you do? Any surprises?
The point of this quiz is that NASA has been comfortable flying astronauts on vehicles with relatively poor or even unknown safety profiles, and their newest launcher is any safer than shuttle was. So it's a little strange to consider them the keepers of safety for human launch.
If you care about this subject at all, you should go buy a wonderful book by Rand Simberg titled "Safe is not an option - How a futile obsession with getting everyone back alive is killing our expansion in space".
His main point is that we know that death is a possibility in many human endeavors - driving, flying a private plane, skydiving - and also within many jobs - commercial diving, fishing, research on antarctica, and the military. And we, as a society, have learned to deal with it. That's not to say we don't try to make those activities safer, merely that risk is inherent in life and many of these things are worth doing.
The ebook is only $5, and it's well worth your time.
Now that the preliminaries are out of the way, we can talk about Starship.
Just one more clarification...
Crewed missions are composed of three parts; there is the ascent into orbit, the actual mission, and then the return from orbit. I'm going to skip the mission risks.
We'll start with the ascent phase:
Broadly speaking, there are four big risk areas.
The first is what is generally called underperformance; there is some sort of problem with the propulsion system and the vehicle therefore can't get into its desired orbit.
The second is a vehicle breakup, which might come from aerodynamic forces or it could come from catastrophic failure of the propulsion system.
The third is a failure of the electrical system or some other mission-critical system
And the fourth is some sort of environmental condition that puts the crew at risk - a fluid leak or a fire.
We'll start by talking about underperformance, and that takes us to talking about engines.
How reliable can we expect Raptor to be? Let's look at two engines.
The Merlin engine used on the Falcon 9 has had 3 failures in 1500 uses, giving a failure rate of 1 in 500.
The RS-25 engine used on the space shuttle had 1 failure in 405 uses, for 1 in 405.
1 in 500 looks like a good starting point.
We can get some context by looking at Falcon 9 first.
We'll start by assuming each engine has a 1 in 500 chance of failing on any flight.
Falcon 9 has 9 engines on the booster and 1 on the second stage.
The booster will abort if 2 engines fail, and the second stage will abort if its only engine fails.
Time for some math. For 9 engines, the change of no failures is 0.998 raised to the 9th power, or 0.982. That puts the chance of a single engine failure at 0.018. We need two engines to fail for an abort, so that gives us an abort probability of 0.00032, or 1 in 3000. That is the power of redundancy; having more engines makes it more likely 1 engine will fail but redundancy makes it less likely that enough engines will fail to cause a problem.
Looking at the second stage, it's chance of an engine failure is 1 in 500 or 0.002, and since there is only a single engine, that's the full probability as well.
If we look at the aggregate across both stages, we get 1 in 430 as the overall chance of requiring an abort due to engine shutdown, with most of the risk coming from the second stage.
Now let's look at starship, keeping the same 1 in 500 chance of a failure.
Current information suggests the booster will have 33 engines and the second stage will have 9. What effect do you think that will have on reliability?
The booster will be okay with 3 failures but will abort with 4, and I'm asserting that the second stage will be okay with one failure and will abort with 2.
For the first stage, the chance of one engine failing is 0.064. To abort, this needs to happen 4 times, for a probability of 0.000017, or 1 in 60,000. That's a really big number for a rocket.
The second stage has a single engine failure probability of 0.018, but it needs two to fail to trigger an abort, with a probability of 0.0003, or 1 in 3000.
In the aggregate, the risk is dominated by the second stage so the overall value is about 1 in 3000.
That is a very low chance of underperformance, and that's for a fairly low expected level of performance from Raptor.
If Raptor manages 1 in 1000, the aggregate probability goes to 1 in 12,000
Onto vehicle breakup
I looked at vehicle breakup due to engine failure to see if there are other examples like this Antares one from 2015.
Atlas had a few engine failures that might have been that bad, but none since the 1960s. Delta had 1 in the 1960s, and zero since then. Ariane had zero. Soyuz had zero.
And the space shuttle and falcon 9 both had zero.
This suggests that the Antares failure is probably an outlier - but if you watched the NASA crew safety video I did, you know the dangers of trying to generalize from events that are rare. The best we could ever say is "the historical data suggests that failures that are more impactful than engine shutdown are very rare", with some weasel words about how rare they might be.
The are other risks during launch; the vehicle might breakup due to aerodynamic forces, there might be an electrical failure, or there might be environmental issues.
Despite the regular mention of Max Q as if it is a significant issue in launch, failures due to aerodynamics appear to be exceedingly rare.
I don't have figures about electrical failure or environmental issues or any ideas on how to estimate those and they aren't unique to starship, so I'm going to skip them.
I think most of you have been waiting for the return phase discussion.
The big risks I see are burning up on reentry, engine failure on landing, failure of the flip maneuver, or failure of the catch
The obvious question to ask about reentry is whether starship is like shuttle. What are the chances that it will have the sort of problem that doomed Columbia?
There are two important differences that make this less likely.
Shuttle tiles were attached to the orbiter using adhesive, in a painstaking process that took NASA a long time to develop.
Starship tiles are attached to metal pins that are welded to the vehicle structure, and are therefore *likely* to be more durable. This will, of course, need to be validated during orbital flight testing.
More important, however, is two design choices made for shuttle.
The first is the architecture of the shuttle; mounting the orbiter on the side of the fuel tank exposes the fragile thermal protection system to damage from foam debris that comes from the external tank, and this was an ongoing problem in the program - 80% of the flights where imagery was available showed foam shedding.
The second is the fuel choice for the shuttle; the shuttle's choice of hydrogen required that the external tank be insulated to keep the extremely cold liquid hydrogen from boiling away, and that meant foam sprayed on the outside of the tank. In the picture, the upper circle shows the bipod ramp where the shuttle attaches to the external tank; on Columbia, a chunk of foam came from this location and impacted on the wing where the second lower red circle is drawn.
If you want to know more about why NASA chose this design, I've linked my video "why does the space shuttle look so weird?" in the upper corner. There were other shuttle concepts that did not have this issue.
On starship, there is nothing next to the thermal protection system, and therefore it will not be damaged from debris impact on launch.
Finally, we've come to the scenario that most of you are thinking of.
(video)
And those were certainly three exciting failures.
But there were also two successes:
(video)
We can do some math and look at the odds of failure due to Raptor failure, assuming once again that raptor fails once in 500 uses.
There are 3 landing engines and to crash all three need to fail.
The chance of a 3 engine failures is the chance of one engine failing - 0.002 - cubed.
Or a whole lot of zeros followed by an 8, or 1 in 125 million.
If raptor is a 1 in 500 engine, the chance of all 3 engines failing due to engine issues is miniscule.
We are much more likely to see system issues; there might be fuel system issues, fuel contamination, weather issues. Issues like these are more likely to occur than a triple engine failure.
Can the Falcon 9 booster landing record tell us anything?
The falcon 9 has had direct engine problems in 2 of 105 landings, or about 1 in 50
Falcon 9 does not have the landing engine redundancy starship will have, nor was the landing hardware designed to be crew-rated
If it did, we'd expect it would be 1 in 125,000 or so.
SpaceX is planning on catching both super heavy and starship at the launch tower.
I have no idea how to estimate the risk of that.
I can, however, surmise that SpaceX does not think it is significantly riskier than a landing-leg approach, and if it is more risk, they could use landing legs as a fallback.
How about a hybrid system? Put a crew dragon capsule inside of starship, and use that as an abort option.
This seems like a great idea, but has some problems...
The super draco escape engines use toxic and explosive propellants, and therefore there is risk in just having those systems in your spacecraft. Remember that a Crew Dragon capsule exploded during tests in 2019.
The real question is whether the increase in safety in the abort scenario outweighs the decrease of safety in the normal scenario?
The answer isn't clear; if Raptor is really as reliable as we suspect, having an abort system with super dracos could increase the risk substantially.
I'd also like to touch on SpaceX's confidence in powered landing
SpaceX flew SN15 successfully on May 5th of 2021
9 months later, they have flown zero additional landing tests
Further, the recently-announced Polaris program features two crew dragon flights and the first crew starship flight.
Both of these show that SpaceX is very confident in their powered landing approach.
Let's go back to my original statement...
After looking at the ascent and reentry risks, I see no reason to suspect that Starship will be less risky than Crew Dragon on Falcon 9 and every reason to suspect that it will be considerably safer.
I think the common belief that it is high is because parachutes are commonplace and their reliability is not well appreciated and we've all seen starship prototypes blow up on landing.
One more point about safety, then I promise we'll start talking about Starship.
Let's say you are going on a space holiday for 7 days. You need to ascend into orbit, stay for a week, and then come back and land.
And let's just say that the ascent and descent both have a 1 in 500 chance of killing the passengers, and the orbital stay has a 1 in 5000 chance.
We can convert those probabilities to success rates, multiply them together, convert it back, and get 1 in 238.
Over time, we improve our landing so it's 1 in 1000. That will push our overall risk down to 1 in 412, a big improvement
Now let's change the scenario; instead of staying in orbit we are going to spend 7 days on the surface of the moon, and that part of the mission has a 1 in 50 chance of death.
That gives us a 1 in 42 overall risk.
Now posit the same increase in landing reliability, to 1 in 1000. All that gives us is 1 in 44.
Which brings up another conclusion. On risky missions, the less risky parts don't matter - they don't contribute much to the overall risk. You can spend a huge amount of effort there and make minimal gains.
Okay, now that we've covered that, time to talk about Starship.
Wait, one more topic
We need to talk about abort systems and their impact on reliability. We'll use SLS as an example.
NASA's target goal for SLS is 1 in 300 on ascent, or 99.66% reliable. Let's say that the base reliability of the launcher is 1 in 100. That only gives us 99%.
Take the Orion capsule and add a launch escape system to it. Let's assume that system will save the crew 66% of the time.
So, we can take the 1% chance of needing the launch escape system and the 66 % success rate when we need it, and figure out that we get an increase in 0.66% in the survival rate, pushing us up to 99.66% total, or the 1 in 300 we are hoping for.
Abort systems are great.
Now lets look at the nominal, or non-abort scenario. For reentry and landing to succeed, the abort system needs to be jettissoned from the capsule. If that doesn't work, the crew cannot reenter.
Let's say that works 98% of the time and fails 2% of the time.
We can take that two percent chance of failure times the 99% of the time an abort isn't needed, and that will lead to a loss of the crew 1.96% of the time, reducing our overall survival rate to 97.94%, or 1 in 43.
Abort systems are terrible.
Let's look at another example.
Finally, let's talk about the details of starship.
Like shuttle, there are two basic kinds of problems that can happen on ascent.
The first class is underperformance; something has gone wrong with one or more of the engines and we therefore don't have the thrust we expect.
The second class are major issues. The booster explodes, there is a toxic gas leak, the electrical system fails, that sort of thing.
We'll star
Our options depend upon what the performance of Super Heavy was and what the performance of Starship is; failures with either stage might cause us to explore our abort options.
The shuttle had 5 abort options.
If the energy deficit was small, they could abort to a lower-than-expected orbit.
If there is enough energy to get near to orbit, they can travel around the earth once and then land.
The next option is to land at an airport in europe.
If the ground track is convenient, they could land on the east cost or the Bahamas
And final, they could return to land at the launch site.
Assume Super heavy has a significant issue. Super heavy can either keep flying and stage when it runs out of fuel, or it can stage immediately.
If the choice is to stage immediately, the options depend upon which starship is flying. If it's a 6 engine starship, the thrust/weight ratio is less than 1 and that means it's not possible to abort while sitting on the pad.
If it's the 9 engine starship, the thrust to weight is probably greater than 1, and it's possible to abort while sitting on the pad, though SpaceX has talked about stretching Starship to carry more fuel and that might change that.
Pratt & Whitney JT9D - 747-100
General Electric GE90 - 777
If you look around, you will find a lot of explanations why airlines don't have parachutes for passengers that explain how impractical it would be and how most accidents wouldn't provide time to use the parachutes.
All those points are true.
But they largely view parachutes as pointless, and they are wrong in that.
Pratt & Whitney JT9D - 747-100
General Electric GE90 - 777
Before we dive into things, I have two links to share with you.
The first is my video on Space Shuttle abort modes, as it's very useful to understand what the options were for the shuttle.
The point here is that safety changes can be good and they can be bad. We can express this in numerical terms.
We can look at how much better our mitigation is, multiply it by the chance of that scenario coming up, and get an estimate of the safety improvement
We can look at the problems our mitigation might cause, multiply it by the chance of that scenario, and get an estimate of the safety loss
Applying this to our parachute scenario with some made-up numbers, let's assume that the parachute can save 50% of the people that would otherwise die, and the chance of that scenario is one in one thousand. That gives us an improvement of half of one in one thousand, or one in two thousand.
Let's assume that the scenario where parachutes slow down evacuation only results in 10% extra deaths and the chance of that is one in one hundred, which would results in a reduction in safety of one in one thousand.
The point being that safety losses in other scenarios can outweigh the safety gains in the scenario you are trying to address.
Here's a video of how airplane evacuations are supposed to happen, during an evacuation test of the Airbus A380.
Now ask yourself, what would happen if 10% of those people were wearing bulky parachutes and tried to make their way off the plane?
Aviation disasters are very well studied, and we know that the time it takes to get off the plane can be the difference between life or death. We also know that passengers do not follow instructions, there are documented cases where people have died because they inflated their life vests inside the plane and could therefore not get out the exits that were slightly under water.
Parachutes would kill passengers who otherwise would have lived.
Pogo is a longitudinal - forward and backwards - vibration that occurs in rockets that use liquid fueled engines.
It was an issue with the Titan II missile, shown here in the Gemini V launch, and the Saturn V rocket, shown here in the Apollo 8 launch.
Pogo can range from a minor annoyance to a mission-ending problem.
Pogo is a complex phenomena.
If you want to dig deeper, I recommend "NASA Experience with Pogo in Human Spaceflight Vehicles", which I have linked in the description of the video.
If you want more depth including all the math, I've linked to some other resources that should make you happy.
Add link:
https://ntrs.nasa.gov/api/citations/20080018689/downloads/20080018689.pdf
https://ntrs.nasa.gov/citations/19730007197
https://sci-hub.se/https://www.sciencedirect.com/science/article/pii/B9781483198217500144
https://ntrs.nasa.gov/citations/19710016604
https://www.researchgate.net/publication/268472794_Overview_of_Combustion_Instabilities_in_Liquid_Rocket_Engines_-_Coupling_Mechanisms_Control_Techniques
We're going to start with a very simple model that most of you should be familiar with, a common playground swing.
We know that to make a swing go higher, we need to push it at the right moment, ideally when it is just starting to go forward. Push it the wrong time - say when it's coming back - and it will swing lower, not higher.
The optimal frequency - the frequency at which pushing it produces the biggest effect - is known as the resonant frequency.
We're talking about liquid fueled rockets, and all liquid fueled rockets have propellant tanks, and they all have feed pipes at the bottom to convey the propellants to the engines. Here we see the first stage of a Saturn V, with a small RP-1 fuel tank at the bottom with short feed lines and a large liquid oxygen tank at the top with long fuel lines.
We'll be looking at the liquid oxygen tank in this example.
We know that if we have a tank filled with liquid, the pressure at the top of the tank is much lower than the pressure at the bottom of the tank.
If we look at a propellant tank, at the very top there will be no pressure, near the bottom of the tank there will be more pressure - I've arbitrary chosen 1 atmosphere - and then at the bottom of the feedline, there will be even more pressure, say, two atmospheres.
This pressure is from the gravity pushing down on the propellant, or 1 g of acceleration
If the rocket is accelerating upwards at 1 g, the total acceleration would be 2g, and that will double the pressure.
Or, to generalize, more acceleration in the rocket will lead to more pressure.
How would we get more acceleration? We would get it if the engine thrust increased.
Let's assume there's a little increase in thrust, which produces a little increase of acceleration. That will produce an increase in pressure, but it won't happen instantly.
What we get is a pressure wave that starts at the top of the tank and travels towards the bottom. It will travel at the speed of sound in whatever liquid is in the tank. If it's liquid oxygen, it's about 1050 meters per second.
The important point is that there is a bit of a delay between the increased thrust of the engine and the increased pressure at the bottom of the feed line.
Now we'll add the rest of the rocket. There is a pump connected to the tank, and that is connected to the combustion chamber of the rocket. The engine is running, and that produces thrust that is pushing the rocket forward.
Rocket engines do not produce power that is totally smooth - there are variations - or instabilities - in the combustion. Some are smoother, some are rougher, but they all have pressure variations, and therefore they have thrust variations. The pressure spikes up and the thrust spikes up.
That leads to a pressure wave in the propellant that travels down to the combustion chamber, which enters the combustion chamber and causes a second pressure and thrust spike.
This situation is very much like the swing we looked at. If the pressure pulse arrives at just the right time - as the pressure is already going up - it's like the push on the swing that sends it higher. We end up with an oscillation that continues as long as the proper conditions persist.
This is formally known as "coupled structure-propulsion instability" - the structure of the rocket and the behavior of the engine are working together to produce a problem.
It's more commonly known as "Pogo", because the up and down thrust variations are like the motion of somebody on a pogo stick.
Pogo is affected by a number of different factors.
The pressure in the empty part of the tank, known as "ullage pressure" affects the pressure to the pump
Fuller tanks produce a stronger pressure wave.
Denser propellants produce stronger pressure waves.
Longer pipes produce stronger pressure waves
The rocket structure itself can flex and change the length of the pipe.
The size and design of the engine effects how smooth the combustion is and the effect of pressure waves on combustion pressure.
More engines will reduce the thrust variation due to random fluctuations in an individual engine
And finally, the vehicle mass matters. When a stage is full of fuel and heavy, a given amount of thrust variation produces a small acceleration variation. As the stage uses propellant and gets lighter, that same thrust variation produces a much greater variation in acceleration, and therefore a stronger pressure wave.
Let's look at an example.
https://www.researchgate.net/publication/268472794_Overview_of_Combustion_Instabilities_in_Liquid_Rocket_Engines_-_Coupling_Mechanisms_Control_Techniques/link/5a2f42920f7e9bfe81702dbc/download
Titan II was initially developed as a missile by the air force.
This chart shows some of the vibration levels measured during early flights, and it has a Pogo problem - significantly exceeding the air force acceptance threshold of 1 g on many of the flights.
The Titan II used storable hypergolic propellants, with Nitrogen Tetroxide as the oxidizer and a stable hydrazine blend known as Aerozine 50 as the fuel. These propellants are relatively similar in density and therefore either could be the culprit.
Based on a mathematical model of the propellant system and the resonant frequency of the entire Titan II, engineers zeroed in on the oxidizer feed lines. What they needed was a way to reduce the pressure wave.
There is a similar problem in household water systems. If water is flowing through pipes and is shut off rapidly - say in a washing machine - the result is a big pressure spike called "water hammer" that can damage the pipes. The fix is to add air chambers on the pipes.
For the Titan II, they installed a standpipe near the oxidizer pump - a short piece of vertical pipe with a closed end. During fueling, the standpipe would be filled with oxidizer except for a small space filled with inert gas, nitrogen in this case.
When a pressure wave comes down the feedpipe, it normally went straight into the pump. With the standpipe on the side, some of the pressure wave was diverted into the standpipe where it compressed the inert gas.
This reduced the size of the pressure wave.
They flew with this fix, and it had a big impact, but not the way they had been hoping - the G oscillation went up to 9 Gs.
The NASA MSC Director remarked to his Manned Space Flight Management Council that he saw one hope: "the fact that the addition of the surge chamber affected the oscillation problem may indicate that the work is being done in the right place."
Pogo was not well understood in the early 1960s and the analysis tools available were quite primitive.
Further analysis pointed the finger at cavitation in the turbopumps. Cavitation occurs when the pressure exceeds the vapor pressure of the liquid, and bubbles form. These changed the frequencies in the calculations, and new calculations showed that it was the fuel feedlines that were more problematic.
The next series of flights flew without the oxidizer standpipe and replaced steel feedlines with aluminum. That produced this series of results, which was good enough for the air force.
But not quite low enough for NASA - their goal was ¼ G.
NASA had done considerable testing in the centrifuge to determine the effects of G loads on astronauts, and they had chosen the ¼ G threshold based on those tests.
More work needed to be done.
NASA made a few phone calls and called in some favors, and the air force agreed to fund further development.
They decided to install a fuel accumulator to reduce the pressure waves on the fuel feedlines and a redesigned oxidizer standpipe based on the updated analysis.
Tests with that new configuration validated the theory, with the lowest Titan II vibration levels seen, and good enough for NASA's requirements.
Performance during Gemini flights was mostly within the ¼ g specification, with Gemini V as one outlier.
Gemini V had a long hold before launch after the nitrogen gas was introduced into the oxidizer standpipe. The nitrogen readily dissolved in the nitrogen tetroxide oxidizer and most of the space at the top of the standpipe was instead filled with oxidizer.
This same effect happens in residential plumbing air chambers.
Onto Apollo...
On the second test flight of the Saturn V, there was significant Pogo during the first stage segment. This is a good illustration of how a vehicle may be stable during part of the flight but as vehicle mass goes down, pogo may show up.
This is a close up view of the pogo portion of that flight . At the start of Pogo, the acceleration is changing between roughly 2.5 and 3 Gs, and at the worst, it's between roughly 2.5 and 3.5 Gs.
That's far above the NASA requirement for Gemini, and it was measured all the way at the lunar module simulator, so it's representative of what the astronauts would feel on launch.
Perplexingly, the first launch of the Saturn V did not have this issue, and NASA and the contractors had done considerable work to ensure the Saturn V would be Pogo free.
Deeper analysis using more data showed that the first stage was marginally stable; on the first launch, the conditions were just on the stable side, and on the second launch, on the unstable side.
The problem was compounded by the huge F-1 engines used on the Saturn V; they used very big combustion chambers and were inherently not as stable as smaller engines. This is one of the reasons that nobody builds engines this big.
NASA addressed the pogo problem with an accumulator on the liquid oxygen line; there was a big cavity pressurized with helium, and that served to damp down the pressure wave, and it fixed the first stage pogo.
Interestingly, the crewed flight of Apollo 8 exhibited an instance of Pogo on the second stage with a very different cause.
Most engines need to be able to rotate - or gimbal - to control the path of the rocket. That presents a problem - the engine needs to move without breaking the propellant lines that are attached to the stage.
That is accomplished with a flexible duct with bellows that can extend, contract, and bend, and that accommodates the movement.
The center J-2 engine on the Saturn V second stage was mounted to a cross structure that was attached to the stage at the outer edges but not in the middle. This design set the stage for a problem.
Here we have the engine mounted to the cross beam supported by the tank, with a flexible bellows in the middle. There would actually be two bellows, one for each propellant.
Start the engine...
and the thrust of the engine bends the crossbar upwards a bit, and that shortens the bellows. The reduced volume of the bellows increases the pressure of the propellants, that results in more thrust in the engine
And that bends things up a bit more.
What happens next depends on a number of factors. Early in the flight of the stage, the engine just finds a stable spot and stays there.
But as the fuel is burned off, conditions change.
Let's say we get a little random spike in thrust that bends the beam up even more. This beam really wants to unbend, so when the increased thrust goes away, it does that...
That lengthens the bellows, the engine pressure drops, and the thrust drops. We have now set up a pattern - each time the thrust drops, the bellows lengthen, and that results in more thrust drop...
This continues as the beam unbends and the thrust keeps dropping, perhaps with the beam bending in the opposite direction.
We are now at the bottom position. The bellows have refilled, the engine is producing normal thrust, and the beam wants to bend forward again. The cycle repeats.
This data is from apollo 8.
The top trace shows the effect on the Apollo command module of the pogo. It was at 8-10 hertz, so think of watching a movie in a theater with a very good sound system that you feel. That's what the astronauts felt, but it was only about 1/10th of a g, so it was noticeable but not significant.
But remember that the engine was flexing the support beams up and down. The bottom trace shows the movement of the engine, and it was at 9 gs, which was getting close to the structural limits at around 15 gs.
Here's a chart showing the acceleration of the center engine during 4 Apollo flights - 9, 10, 12, and 13. For the first 3, it was fine - within structural limits and therefore deemed not to be a flight hazard.
On Apollo 13 at about 160 seconds, the center engine experienced an acceleration of 34 gs. The top of this chart is only 12 gs, so it's really an unprecedented level of acceleration.
The reduction in pressure when the engine flexed down was so significant that the engine thought it had run out of propellant and shut itself down, and that saved the vehicle from structural failure on the next cycle.
https://ntrs.nasa.gov/citations/19860001748
https://ntrs.nasa.gov/api/citations/20080018689/downloads/20080018689.pdf
The fix was pretty simple - add an accumulator as was done on the first stage - but there was very little room. NASA ended up taking an existing part of the piping, drilling holes in it, and encasing it in a drum that was pressurized with helium. That was enough to fix the problem.
Onto the space shuttle
The space shuttle presented NASA with a significant problem with respect to pogo.
The traditional approach was to analyze the rocket structurally, but the shuttle is a very strange launcher. Not only does it have 4 separate parts - the orbiter, two boosters, and a huge external tank - both the payload mass and location differed from mission to mission. NASA created models but it made the analysis very difficult.
They also planned on flying crew on the first mission and losing the crew or the orbiter would be very bad.
They therefore decided to take a more proactive approach. Instead of the "fly and fix" approach used on Gemini and Apollo, they would do up-front analysis, design a motor that would be as "Pogo proof" as possible, and validate it with testing that that subjected the engine to pressure pulses.
The engine used in the shuttle is the space shuttle main engine, or RS-25. This engine uses liquid hydrogen as a fuel and liquid oxygen as the oxidizer. These fuels are very different in density, with liquid oxygen 16 times denser than liquid hydrogen. That means that any pressure waves in the liquid hydrogen feed will be much smaller, and it turns out that you can just ignore the hydrogen side.
Like many engines, the RS-25 has high pressure turbopumps to feed the fuel and oxidizer into the combustion chamber, but under testing, they found that these pumps would cavitate because of insufficient input pressure, which reduces pump efficiency and can damage the pump.
To fix the problem, they added two low pressure turbopumps to take the propellant from the tanks and raise its pressure high enough so the high pressure pumps would not cavitate. This is one reason the RS-25 is so complex and so expensive; it has the systems of a normal rocket engine with two extra turbopumps grafted to the outside.
The low pressure turbopumps are fixed rather than attached to gimbals, so any pipes or ducts between them are connected with flexible bellows. The ones shown here connect the outputs of the low pressure turbopumps to the inputs of the high pressure turbopumps. These are just some of the big ones; there are flexible connections on every pipe that goes between the low pressure side and the high pressure side.
The initial plan was to put the pogo suppression in the traditional place, between the oxidizer feedlines and the pumps, but analysis showed that there were still scenarios where pogo might show up in the plumbing between the two oxidizer turbopumps, so the pogo suppression is placed at the input to the high pressure oxidizer turbopump.
The system is fairly complex. The accumulator is pre-charged with helium at a specific pressure before launch but after launch it is switched over to gaseous oxygen.
Note the location of the accumulator in relation to the flexible bellows that connect the low pressure and high pressure turbopumps. NASA found that for the system to work as desired, the accumulator needed to be on the high pressure side of the bellows. This is likely due to possible volume changes in the bellows part of the line causing issues, similar to the ones seen on the J-2 engine on the Saturn V.
http://large.stanford.edu/courses/2011/ph240/nguyen1/docs/SSME_PRESENTATION.pdf
Here we have a wonderful picture of the system.
At the top we have the Low Pressure liquid oxygen coming from the low pressure turbopump. It goes through two bellows that allow the engine to gimbal, heads down, and turns to head towards the inlet of the high pressure LOX turbopump.
Directly before the input to the turbopump is the pogo accumulator, as indicated by a helpful Aeroject Rocketdyne employee. Note that the accumulator is as close as possible to the high pressure turbopump.
With this system, shuttle did not have pogo issues.
What about other rockets? There aren't a lot of details available, other than mentions that Ariane, Atlas V, Delta IV, and Falcon 9 all use anti-pogo measures.
What about starship and super-heavy?
I haven't seen any evidence one way or the other.
Super heavy has the advantage of having a *lot* of engines - so many that instabilities might not be a problem on ascent, though it will use a small number of engines to land. Starship is likely in a similar situation.
My guess is that SpaceX is probably using pogo control techniques - either traditional ones or active controls.
If you enjoyed this video, please listen to Just Another Day by Oingo Boingo - or buy yourself a slinky.
Add link to: https://www.youtube.com/watch?v=FFU_5Ng-4DM
In 2021, we saw the Inspiration 4 crew spend nearly 3 days in space in a Crew Dragon capsule.
On the surface, it looked like a billionaire taking 3 people with him to have some fun in orbit as part of a fundraising event.
A very successful fundraising event, raising $243 million for St. Jude Children's Research Hospital.
But there were a few strange things about it...
The first was the amount of training they did. They flew in high performance jets owned by a company that Isaacman formed in 2012 and sold in 2019, Draken International. Draken has had training contracts for the US air force, the US Marine Corp, and the US Navy.
They did zero-g training in an airplane that flies parabolic paths, one of the so-called "vomit comets"
They went on arduous treks in the mountains.
Thet did centrifuge training.
And they did long hours of simulator training for their flight.
There's a name for this sort of training. It's called *astronaut training*, and Isaacman's goal was to create a crew of civilian astronauts for this mission.
But why? That was a lot of extra work and expense for everybody who was involved, and it wasn't really aligned with their fundraising mission. They also did quite a bit of scientific research on their flight, a cool thing to do, but also not really aligned with their mission.
There were some strange things on the SpaceX side as well. They took one of their 4 crew dragon capsules offline to make special modifications for the Inspiration 4 flight, including this large glass cupola instead of the normal hatch.
This cupola gave the crew a very unique view of the earth during the mission.
There was also some strangeness about the amount that it cost.
The total cost of a crew dragon mission for NASA is about $360 million
Axiom space missions to spend 8 days on ISS supposedly cost $55 million per seat, or $220 million total.
Isaacman did not disclose the exact cost of inspiration 4, but he did say that it was less than $200 million.
This is confusing. SpaceX did a lot of extra work on the capsule for this mission and the crew spent more time training than the Axiom crew and yet SpaceX charged less for it.
It was a bit perplexing, but after a while it became clear.
Inspiration 4 was not just about fundraising or inspiring the world, though it did that quite well...
Inspiration 4 was a commercial project Mercury, a first step towards a larger program.
Why would you want to build a commercial program?
NASA has spent many years and many dollars on Mercury, Gemini, Apollo, and the space shuttle and come up with impressive solutions.
Why not just use what NASA has done.
The first problem is one of practicality. The designs in the historic capsules and space shuttle are at least 50 years old, and the technology that they used no longer exists. Those designs were based on the limitations of technology at that time, and there are many better solutions now.
Maybe you could build off of the Orion capsule used in the Artemis program. It started development in about 2005, so it's only about 20 years old.
Unfortunately, the Orion capsules cost $663 million each, and that's after spending $21 billion developing them.
Space suits are another issue. The extravehicular mobility units used on the ISS cost about 12 to 15 million per suit.
The NASA solutions are simply too costly for commercial providers.
What is needed for commercial human spaceflight is affordable solutions and to get there will require new development work.
If inspiration 4 equals project mercury, then we should expect something like project Gemini - or gemiknee if you want to use NASA's official pronunciation.
What would we expect that program to look like?
Gemini had a number of goals.
Rendezvous, docking, and maneuvering. Gemini used both Agena targets and other Gemini flights.
Long duration missions. Here we see Frank Borman and James Lovell after spending 14 days in Gemini 7.
Precision re-entry, landing, and recovery.
Extravehicular Activity, accomplished by Ed White in Gemini 4
Scientific Investigations across all the flights.
Of these, SpaceX has already figured out how to do two of them - the rendezvous and reentry ones. That leaves three other goals...
The overall program is called Polaris, and the gemini analog is named Polaris dawn.
It will feature the first commercial spacewalk.
It will do health impact research.
They will attempt a high altitude orbit at 700 km to better understand the radiation environment in that orbit
They will test a laser-based communication system using Starlink
They will spend 5 days in orbit. Quite a bit shorter than Borman and Lovell, but those astronauts were very good sports.
This flight hits all of the Gemini goals and they are attempting to do them in a single flight.
They will once again be using "Resilience" for this mission, this time modified to support tethered EVA and depressurization and repressurization of the capsule.
They will also take the existing launch and entry suit and perform significant modifications to allow the crew to have mobility while on EVA. These new suits should be roughly the equivalent of the ISS space suits without the backpack that allows spacewalks without an umbilical connected to the capsule.
This mission is very much a collaboration between Isaacman's Polaris and SpaceX.
Jared Isaacman and Scott Poteet come from the Polaris side and Sarah Gillis and Anna Menon are both Lead Space Operations Engineers at SpaceX working on Crew Dragon.
This will very much be a "when we are ready" mission, but we do know that summer of 2024 is the earliest possible date.
After Polaris Dawn is over, there is the excitingly named Mission II. There is no available information except that it will be another crew dragon mission and we can be sure that it will once again fly on Resilience.
At some unspecified time in the future, we get mission III, which is the first starship flight carrying crew.
SpaceX and Polaris are developing human spaceflight capability without ties to NASA.
It is a natural fit for SpaceX's requirements for moon and mars missions.
It is the same business plan Isaacman used to build Draken International, but applied to space. If there are going to be a lot of people working on commercial space missions, they will need to be properly trained.
And that's how you create your very own commercial human space program.
If you enjoyed this video, please send me this original painting, "Love in the time of Polaris", available for the modest sum of $5400.
I was answering this question on Reddit for probably the tenth time a few days ago, and I remembered that I was planning on doing a video the topic, so here it is.
Why are some rockets able to do propulsive landings of the first stage while others are not?
There are a number of factors to consider that control whether propulsive landing is possible and practical.
The first factor is the payload margin that you have. Can you spare the fuel to land?
Looking at the Falcon 9, we see that in expended mode, it can launch 22,800 kilograms to low earth orbit and 8300 kilograms to geosynchronous transfer orbit.
However, if the first stage is landed, that reduces the payloads to 16,250 kilograms to LEO and 5500 kilograms to GTO, for a payload penalty of 29% and 34%
Is that okay? It depends upon the market you are targeting.
If the majority of customers are here in reusable land, then you are in good shape; you can afford the performance penalty for most of the payloads that you launch.
If your market is over here in expendable land, then you have issues; you are not able to use your reusability.
For Falcon 9, their market is dominated by payloads that allow them to fly reusable, so they are in good shape.
This was *not* true for the first version of Falcon 9; its expendable payloads were less than the current Falcon 9's reusable payloads, and that meant it could not serve the target market in reusable mode.
Another factor is whether you use solid rocket boosters on your launch vehicle, as the Vulcan and Ariane 6 will.
These vehicles tune their performance by adding just enough solid rocket boosters to support a specific payload mass. For example, they might use the base rocket plus 2 solid rocket boosters.
To land the first stage, they need more payload margin, which will mean the base rocket plus 4 solid rocket boosters.
The two additional solid rocket boosters might cost $10-15 million, which makes reuse much less attractive.
The second factor is the difficulty of reentry, and this is controlled by when the rocket stages.
Falcon 9 has a beefy second stage and stages quite early - 120 seconds into the flight, at an altitude of 65 kilometers and a speed of 2200 meters per second.
Atlas V from ULA - and the upcoming Vulcan launcher - have a wimpy upper stage, and do not stage until much later - at 260 seconds into the flight at an altitude of 170 km and a speed of 4800 meters per second.
If we remember that kinetic energy is ½ the mass times velocity squared, we can arbitrarily set the amount of energy Falcon 9 has to deal with to 1, and then calculate that the amount of energy Atlas V would have to deal with. The ratio is 4.8, and that means it would be much harder to protect an Atlas V first stage than a Falcon 9 first stage.
The third factor is the engine count, which determines if we can use the engines to land.
The Falcon 9 uses 9 Merlin 1D engines which produce a thrust of 845 kN each or 7605 kN total.
A single engine can be throttled down to about 482 kN, and with an empty first stage, that produces a thrust/weight ratio of about 1.9. That is low enough that it is relatively easy to control the velocity and land, but it does require the so-called "hoverslam" maneuver to do so.
The atlas v uses a single RD-180 engine that produces 3830 kn of thrust, and it can be throttled down to 1800 kN. With the atlas V first stage, that produces a thrust/weight ratio of 8.7, which makes propulsive landing unfeasible.
The fourth factor is the availability of engines and their cost.
For rocketlab's neutron, they will be using 7 Archimedes engines. They are building their own engines, and their goal is to create a cheap and simple engine - very similar to SpaceX's Merlin.
ULA, however, does not build their own engines so they are limited to commercially available engines. They have chosen the Blue Origin BE-4 for Vulcan, but the goal of blue origin is not to make those engines as cheap as possible or to make them a convenient size; their goal is to make as much profit as practical on an engine they are already making for New Glenn.
Their other issue is that they are inherently limited to engines that are available. They also considered the AR-1 from Aerojet Rocketdyne, and then could maybe consider the SpaceX Raptor or Merlin, but all of these have the same problem; that other company wants to make a profit on each engine.
The last factor is the cost to benefit ratio. Is it worth the investment to develop reuse.
Neutron is a clean-sheet design, so they are doing all the engineering from scratch. Vulcan is an Atlas V derived launcher that uses the same second stage as Atlas V, and a redesign to support propulsive landing would be more costly.
Neutron is designed to compete in the commercial market and therefore needs to be an advanced design, but Vulcan is designed to launch National Security Space Launch payloads for the department of defense.
Neutron is designed to have a high flight rate, while Vulcan is designed to only launch a few times a year.
And finally, neutron needs to be competitive in price while the NSSL contracts value reliability and performance over price.
It simply may not be worth the money to develop reusability for Vulcan, though Amazon has recently ordered a lot of launches for their project Kuiper constellation and that will change the cost/benefit analysis for Vulcan.
Let's apply the factors to specific launchers.
Falcon 9 has sufficient payload margin to launch for most of their customers and land the first stage, their reentry is relatively easy, and they have engines figured out. Their cost/benefit is clearly pretty good.
Electron is like Falcon 9 in most ways, but unfortunately they are only a 200 kg to orbit launcher, and reuse is harder for lighter rockets, so doing so would reduce their payload to under 100 kg. That alone makes propulsive landing non-viable for them.
They are, however, experimenting with catching their first stage with a helicopter for reuse.
Starship is pretty obvious; it has all the characteristics to support both first stage and second stage reuse.
Neutron is like an evolved version of Falcon 9, and therefore has the same benefits.
New Glenn looks like a nice reusable design based on the factors I listed, though it's not yet clear how reuse affects their payload margin and the markets they wish to serve.
Ariane 6 does not score well. They don't have the payload margin and they use solid rocket boosters so the economics of reuse are less compelling, and they stage high and fast. They do have the possibility of new engines but they do not plan to do that in this launcher. It's not clear that they launch often enough to see a benefit from reuse.
They are, however, starting work on an "Ariane Next" launcher that looks much more like the existing reusable launchers, including either reusable liquid boosters or triple core like Falcon Heavy.
Finally, we come to Vulcan.
Like Ariane 6, it doesn't have the margin and depends on solid rocket boosters, and it stages high and fast. ULA doesn't build engines.
They have, however, talked about what they call "SMART Reuse", where the engine pod separates from the booster, deploys a reentry shield, and then is caught in mid air by a helicopter, using the same approach as Electron. That would save the expensive engines, and it would have a smaller payload impact than propulsive landing, but would have larger costs to refly.
SMART reuse hadn't been talked about in the last few years, presumably because the cost/benefit didn't make it worthwhile, but with the recent Project Kuiper contracts, ULA has said they will be developing it.
And that's the story on propulsive landing, and what is required to make it work.
If you enjoyed this video, please send me $1 for every Falcon 9 landing.
The air force help fund Raptor development for possible use on a falcon 9 second stage. This would be a smaller raptor to better fit the size needed for the Falcon 9. The air force put in a chunk of money - up to $61 million initially, and then more than $40 million a year later.
Why was the air force interested in a different engine? That requires a bit of an explanation...
This goes back to the Air Force - now Space Force - launch requirements. The air force was looking for new options to launch their satellites, since they were fully dependent on ULA's Atlas V and Delta IV launchers. These launchers were very reliable but they were also very expensive.
To win part of the NSSL contract, you need to be able to hit all of these orbits. The top 7 are relatively easy, but there are two at the bottom that are hard to get to. They are geostationary orbits that require the launcher to do a lot of work. Let's explore why.
Here's a diagram that should help.
The goal for a geostationary payload is to get out to about 36,000 km and establish a circular orbit. This first diagram shows the launch, with the red part showing powered launch and the yellow part showing a long coast phase. This initial boost puts the payload what is known as a geostationary transfer (GTO) orbit. But there's a problem - only the high part of the orbit - the apogee - is at the proper altitude; the lower part of the orbit (the perigee) is still close to the earth. To get to the proper orbit we need to do this by circularizing the orbit; by thrusting at the apogee, we can change the orbit progressively #press #press #press until we hit the circular orbit we want. There also may be an adjustment to the orbital inclination at the same time.
There are three basic ways to do this.
The first is with hypergolic thrusters built into the payload. These thrusters put quite a bit of thrust, so they can circularize the orbit in one to three burns. This was the most common way of doing circularization for a long time, though it does have a relatively low Isp and therefore a lower payload capacity.
The second way is to use ion thrusters - sometimes known as electric propulsion - on the payload. These thrusters put out very low thrust and therefore need to run continuously for months to circularize the orbit, but they have an Isp around 2000 or maybe higher, so that means smaller fuel tanks and therefore larger payload capacity and/or longer lifespan. They do require large solar panels to supply the electrical power, but big communications satellites already have large solar panels, so that works out well. Pretty much all the new GEO communications satellites use ion thrusters.
The third way is to use the engines built into the second stage. Those are generally pretty big, so they can do everything in a single burn, but there are two downsides.
The first is that both the payload and the second stage need to be circularized, and that reduces the payload significantly because you are circularizing both the payload and the second stage.
The second is that it takes a long time to get out there - 5 hours or more - and that means the second stage needs to last for that long so that it can perform the burn. That's especially hard for a Falcon 9, because it needs to keep its RP-1 fuel warm for 5 hours so it can perform the burn. That second problem is why the air force was interested in a different second stage engine.
That's not why you came here, however, so let's move on to talking about putting a raptor on the Falcon 9 second stage.
The results of putting the raptor on the first stage were less exciting than many had hoped. Will it be different this time? Let's see...
Let's talk about the competition...
The Merlin Vacuum engine is a Gas Generator engine, using RP-1 (refined kerosene) and liquid oxygen, or kerolox.
Going up against the Merlin is the Raptor, a full-flow staged combustion engine using liquid methane and liquid oxygen, or methalox.
And, just to have some fun, I decided to add in the venerable RS-25 or space shuttle main engine. It's a fuel-rich staged combustion engine using liquid hydrogen and liquid oxygen, or hydrolox
Going through some parameters, the Merlin puts out 981 kilonewtons of thrust, the Raptor puts out around 2000 kilonewtons - all raptor numbers are estimates - and the RS-25 puts out 2280 kN.
For Isp, the Merlin has a healthy 348, the Raptor 380, and the RS-25 wins with an Isp of 452.
I couldn't find an exact weight for the Merlin, so I added about 50% to account for the added size and complexity and got 750 kilograms. The raptor is probably around 2000 kilograms, and the RS-25 is a hefty 3177 kilograms.
Finally, the nozzle size of the Merlin is 3.03 meters, which fits inside the 3.7 meter diameter of the Falcon 9. The Raptor is smaller at around 2.5 meters, and the RS-25 is still smaller at 2.4 meters.
There's an interesting question here - why do these big vacuum engines have such a smaller nozzle than the Merlin. Both the Raptor and RS-25 have regenerative nozzles that are cooled by passing one of the propellants through them, but if you want to test them before launch - and you do - they have to have nozzles that are small enough to remain stable at sea level. And of course the RS-25 had to function at liftoff for the shuttle. Most of the Merlin nozzle has no cooling channels, so it can be tested without most of the nozzle. The raptor also has to fit in the space provided in starship. The larger nozzle is one reason the Merlin is so close in Isp to the challengers.
We'll start by seeing how our engine choices affect our propellant load.
The Merlin burns RP-1 which has a density of 840 kilograms per cubic meter, and runs at a mixture ratio of 2.36 parts oxygen to 1 part RP-1. If we know the second stage holds about 100 cubic meters of propellant, that amounts to 63% liquid oxygen and 37% RP-1 by volume. The total amount of propellant is around 111,000 kilograms.
Moving onto the raptor, it burns liquid methane which has a density of 445 kilograms per cubic meter, which is about half the density of RP-1. That would lead us to expect it to require about double the space for the liquid methane, but the mixture ratio is higher - more liquid oxygen - so when we do the math it turns out that the stage would be 56% lox, 44% liquid methane. Removing some lox and replacing the RP-1 with liquid methane reduces the mass of the propellant we can carry, down to around 90,000 kg, or about 20% less.
Finally, the RS-25 burns liquid hydrogen with a density of only 71 kilograms per cubic meter. It makes up a bit my having a high mixture ratio, but even with that we end up with a LOX tank that's only 25% of the total and the remaining 75% is needed for liquid hydrogen. That drastically reduces the amount of propellant we can carry, down to about 37,000 kg. The low density of liquid hydrogen is a big issue for designs that use it.
Now we can look at the results we get from those stages. We'll be talking about delta-v, so I put a link up in the corner to another video.
The Merlin has a low Isp - only 348 but is crams a bunch of propellant into the stage, so that gives is a high mass ratio - the ratio to the mass fully fueled to the mass empty. That gives is a delta-v of about 6500 meters per second.
The Raptor has a higher Isp but a worse mass ratio because we could only put 80 % as much propellent in the stage. It comes out at 6260 meters/second, which is in the same ballpark as the Merlin.
The RS-25 has a great Isp, but it falls down on the mass ratio, and that essentially kills its performance, only netting about 4380 meters per second of delta v, a distant third.
One of the other factors that affects performance is gravity losses, so I wanted to see if I could calculate the gravity losses for each option, but I ran into another issue.
The merlin vacuum launching a starlink payload will burn for somewhere around 380 seconds. The more powerful raptor with less propellant will burn for about 155 seconds. And the much lighter RS-25 version will only burn for about 70 seconds.
The short burn times coupled with the high thrust create a problem. The Merlin starts at a bit less than a G and then will end up at a bit over 5 Gs if it's not throttled down.
The Raptor starts and 2 Gs and rapidly increases, hitting nearly 11 Gs.
And the RS-25 starts at 4 Gs and also trends up above 10 Gs.
The engines can be throttled down a bit, but not really enough to fix the issue, and at lower thrust their Isp will be reduced.
It turns out there was a reason the air force wanted a smaller raptor...
Let's see what we can do.
The Raptor is roughly twice the power we need, so I'm going to create a new engine.
It's known as the Halftor, with a thrust of 1000 kN, and it's only going to mass 1300 kg because it's smaller.
The RS-25 is even more overpowered than the Raptor because the stage is so light, so we're going to need to go to one quarter scale
That will be the RS-6 point two five, with 570 kN of thrust and only 1300 kg of mass
How will these work?
We can now rerun the delta V calculations as the engines are lighter.
The halftor moves up 100 meters per second to 6365 meters per second.
The RS-6.25 moves up 250 meters per second to 4630 meters per second.
That's a pretty good gain for both of them, but still less than the Merlin.
Now we can flip back and look at our graph of remaining propellant and we can see that now both of the candidate motors run for a whole lot longer. The halftor runs for 330 seconds and the RS 6.25 runs for about 285 seconds, which is much more reasonable..
With the acceleration graph we see the Halftor reaches about the same acceleration as the Merlin. It finishes a bit faster, so it's going to gain a bit in gravity losses, which will put the Merlin and the Halftor closer together.
The RS 6.25 follows a similar curve to the other engines, but it kindof burns out a bit too early because of the amount of propellant. It's not really possible to give it both the higher thrust and the longer burn time because of the lack of propellant mass.
Now that we've done our analysis we can reach our conclusion
The halftor upper stage doesn't have any real advantages over the merlin, and the RS-6.25 lags far behind as it simply does not pack enough propellant to be competitive.
Why would you want to do the halftor?
Not so fast!
As we noted earlier, both the halftor and the Rs 6.25 end up being lighter than the Merlin simply because their propellant weighs less.
This means the first stage can produce more delta-v...
Skipping the math to get new numbers, we'll start with the second stage delta-v for each of the options.
Then we can add in the first stage.
For a star link launch, with the merlin, the first stage gives 3770 meters per second and the full stack nets a total of 10,270 meters per second.
Going to the Halftor, the first stage goes up to 4060 meters per second and the full stack nets a total of 10, 425 meters per second. That's 1.5% more, and there's probably a little more than that with reduced gravity losses in the second stage.
The RS-6.25 puts up a strong showing with a first stage of 5190 meter per second and a toal of 9820 meter per second, only 4.4% less than the Merlin. The lightness of the second stage redeems itself a bit, but it's still not competitive.
What did we learn?
First, the Air Force was interested in the Raptor upper stage engine primarily to handle long duration flights that go directly to geostationary orbit. In 2020 the F9 was certified to do those flights, so this is not longer an issue from the air force or I guess the space force perspective.
Second, to use raptor technology on the Falcon 9 would require a smaller engine and it would only produce a few percentage points of performance gain. It's unlikely to be worth it financially and that's why SpaceX is no longer working on it.
Third, the RS-6.25 is my favorite made-up-engine name
Launch Vehicle Recover and Reuse Paper
https://www.ulalaunch.com/docs/default-source/supporting-technologies/launch-vehicle-recovery-and-reuse-(aiaa-space-2015).pdf
Business case and spreadsheet post
https://forum.nasaspaceflight.com/index.php?topic=37390.0
Dr. Sowers' comments on a longer reuse thread.
https://forum.nasaspaceflight.com/index.php?topic=37390.msg1470446#msg1470446
For a long time, ULA has been producing PR material that refers to SpaceX and the Falcon 9. This is a second part of a series where we look at the details of one piece of PR material to decide whether it's PR, or whether it's lies.
For this part, we're going to be looking at the famous - or perhaps notorious - ULA paper on launch vehicle recovery and reuse.
And here's the paper, released in 2015. There is some interesting stuff here; there's a map of ULA's SMART (Sensible Modular Autonomous Return Technology) reuse approach, and more interestingly, this chart that tells us how much the various parts of their rockets cost. If you want to read the whole paper, I've linked to it in the description.
But that's not why you are here. You are here for the graph...
This is the graph, and this is the text that goes along with it (read).
And note that when ULA says "booster fly back", they are talking about drone ship landing, not flying back to the launch site.
They determine the reuse index based upon the cost per kilogram of payload.
The model is based on a fairly complicated equation and contains a few terms that I frankly don't think are very important.
Before we spend more time with it, I want to talk about models in general and then we will look at a simpler reuse model.
We build models to answer questions.
The question the paper is trying to answer is whether ULA reusability is better than SpaceX reuse. That's a very hard question to answer because SMART reuse was theoretical at the time the paper was published and still remains theoretical in early 2024.
So I'm going to put that question aside and talk about the broader question, are reusable rockets cheaper?
That question is pretty easy to answer, and the answer is "it depends". We can envision cases where reusable rockets would be cheaper and we can envision cases where they would be more expensive. We need a better question.
Is Falcon 9 in reusable mode cheaper?
That's a better question as it allows us to deal with specifics, but its still too simplistic as there's not going to be a simple yes/no answer.
What factors affect the cost of Falcon 9 reuse?
This is an interesting question. There are a obviously a lot of factors, so it would be good to limit them down.
How do recovery/refurbishment cost and first stage lifetime affect the cost of Falcon 9 reuse?
This is good - it's specific enough that we can probably come up with a simple model.
We aren't spacex so we won't have real data, but we can come up with some estimates that will give us a deeper understanding of reuse for the Falcon 9.
The model I'm introducing is wrong in specifics but it's right enough.
We are going to be comparing Falcon 9 in expendable mode to Falcon 9 with the first stage landing on the drone ship. Space says that Falcon 9 can carry 22.8 tons to low earth orbit in expendable mode, and we've seen it launch 17 tons and land on the drone ship for starlink launches, so those numbers will be decently accurate.
We will assume for the model that the first stage cost is $20 million the second stage cost is $10 million.
We've already identified the two variables we want to look at - the first stage lifetime and the recovery and refurbishment cost. You could use the same model to explore how second stage cost affects things if you wanted.
Our equations are pretty simple - the expendable cost is the cost of the first stage plus the cost of the second stage.
The reusable cost is the cost of the first stage divided by the number of times we expect it to fly plus the recovery/refurb cost plus the second stage cost.
That gives the cost per flight, but we haven't accounted for the difference in payload, so we divide those costs by the payloads to figure out the cost per kilogram.
All of this feeds into a pretty simple Excel spreadsheet which will generate a graph with our two variables.
Here's the graph. The horizontal axis is the number of times a first stage flies before it is scrapped. The vertical axis is the cost in dollars per kilogram. The yellow line shows the expendable cost, which of course doesn't vary. Note that I'm talking about cost using a very simple model, so the numbers will be much lower than the price that SpaceX charges.
The different lines are for different recovery and refurbishment costs - the blue line assumes it only costs $2 million, and the green line assumes it costs $14 million.
The green line tells us something interesting right off the bat. We learn that if it costs $14 million for recovery and refurbishment, reuse is not practical - the payload loss is greater than the savings. That's true at $12 million as well.
Looking at the other lines, we can see that recovery cost is the critical factor in whether reuse is practical. If you can get to $6 million it's a lot more interesting than $10 million.
Another interesting thing is that first stage lifetime isn't super important. Once we get to 10 flights, the cost per flight of the stage itself is down to $2 million, and pushing that to 20 flights only reduces that down to $1 million. After ten flights the overall flight cost is dominated by the cost of the second stage and the cost of recovery and refurbishment.
I think this model is a success. The big thing we learned is that reuse depends heavily on the recovery and refurbishment costs. If those are high enough, it doesn't matter how many times you reuse the first stage. And if they are low enough, reuse starts saving you money very quickly.
Let's return the graph from the paper.
(rework)
The problem with this graph is that it doesn't tell us anything useful. Like my graph, there's a mathematical model behind it but it only shows us one graph line for each reuse scenario and it doesn't tell us what the underlying assumptions are. As we saw with our model, we can play with the inputs to come up with whatever result we want.
I would therefore rate the utility of this graph as minimal - I can't evaluate what ULA has done to achieve these results and I can't validate whether their assumptions make sense.
ULA makes a number of points in their text.
Smart is profitable in 2 flights versus 10 for booster reuse
The reason is the much higher payload reduction in booster reuse
Return to launch site is never profitable because the payload reduction is too high.
It concludes with the following (read).
This is honestly hilarious. The graph shows the results for a single set of values and the only variable is the number of uses, and therefore it *cannot* show the relative importance of the different terms of the equation since different values of those terms are not explored. I guess it can maybe show that lower performance loss is better than higher performance loss since the SMART curve assumes less and the Falcon 9 assumes more, but that is honestly probably the simplest thing to figure out about reuse.
I feel like this graph was pulled out of a more in-depth research paper, one that addressed my concerns, but this paper is all that was available.
Then this happened.
The creator of the model posted on a NASASpaceFlight.com forum thread - I've linked it in the description of the video - and included both an explanation of the business case and an excel model, presumably the one they used to create the data.
This gives us a lot more detail on the model behind the graph.
The business case explanation paper gives a lot more detail about the equation they are using in the model, and that is useful.
We also get information about the actual parameter values they used.
All they have for any of the parameters is a single value. I really want to see some sensitivity analysis, where you define a parameter range for each parameter and see how differences in that value change the result, as we did with our model. This is especially important because of the fuzziness of the numbers - I especially like that one is quote "based on internet chatter".
Two parameters seem especially interesting...
This model's p value assumes a 30% penalty with drone ship reusability, but the actual numbers I used in my model suggest it's only 25%.
K in the model is the fraction of the overall hardware cost that is reused. That could easily be a higher fraction.
Maybe the spreadsheet can help...
Considering the complexity of the model, I was surprised at how simple the spreadsheet was.
I modified it so we can do sensitivity analysis on the k and p variables.
Here's the resulting graph for the Falcon 9 case.
The blue line is the one used in the chart that says that breakeven for reuse is 10 flights. Let's play a bit.
The green line shows the result if you save 45% of the cost of hardware rather that 40%. The pushes the breakeven down to 6 flights.
The red line shows what happens if the payload loss is 25% rather than the 30% used for the blue line. That puts breakeven at 5 flights.
If you make both of those changes, breakeven is at 4 flights.
Not surprisingly, we've found that the model is very sensitive to those values. The 25% payload loss value seems to be a better value than the 30% used, though it's fair to note that the 25% number is on the fully refined block 5 version of Falcon 9 and it took SpaceX years to get there. The recovery fraction change may not have made sense when the paper was written but probably does when fairing recovery is taken into consideration.
It's pretty clearly both the model and the graph were created to tell a story rather that do an analysis.
There's one more thing in the business case paper that I want to cover...
The business justification document says the following at the top.
Read.
Hmm...
Here's a wonderful chart from Zapata talks nasa that shows cost per kilogram of a bunch of different rockets.
Way over here on the right side we have the Falcon Heavy flying the USSF-44 mission. The cost is a little over $2000 per kilogram, making it the current leader, and since that parameter is the single most important one, that is why everybody flies Falcon Heavy and those Electron flights that you have been watching are just figments of your imagination.
The ULA statement is so obviously not true that I'm wondering who they have proofreading their material.
Consider two payloads
The Gazelle satellite is 110 kilograms in mass and it is destined for a low earth orbit.
The Jupiter 3 communications satellite is 9200 kilograms in mass and needs to go to the higher energy geosynchronous transfer orbit.
It's not surprising that gazelle was launched by the small Rocket Lab Electron and Jupiter 3 required the might of the Falcon Heavy.
It's not cost per kilogram that matters. What matters is the answer to a few simple questions...
What payload market can my rocket cover?
How big is that market?
Those together define the business opportunity. If the market for my rocket is only 5 flights a year, it's probably not enough to support a company.
Who are my competitors in that market?
What do they charge?
This is about the competition in the market. The market for Rocket Lab's electron isn't that big because they payload is small, but they mostly have the small satellite launch market to themselves.
And then, for the sake of this discussion, how does reuse affect the answer to these questions?
Before we look at a market analysis, I think there are two markets where cost per kilogram is very important.
The first is starlink. For constellation launches, you want to design your satellites so you can fit as many as possible in the fairing and hit the maximum payload to a given orbit.
The second is station resupply. You will want to pack as many supplies as possible into the capsule. The Dragon 2 shown here is volume limited - NASA runs out of space to carry supplies before they run out of payload - but other resupply scenarios might use a bigger vehicle.
Propellant tankers would probably be another example.
The forum thread yields more comment from Doctor Sowers.
This is getting *so close* to the right way to do analysis. What Dr. Sowers is missing is that you don't need to pick one mass for your analysis. You can pick the spread of payload masses that exist in the market you want to cover.
What matters is how your reusability approach affects the market that you want to cover.
To do an analysis, you need to figure out the market, and that is a lot of work. I'm going to use a subset of market, which is all payloads launched by ULA from 2000-2009.
This is complicated because ULA flew 5 different rockets during that decade. There was the atlas II, atlas III, and atlas V, with 10 different variants. The was the Delta II, with 3 basic variants and the delta IV, with another 4 variants. And that's ignoring fairing size differences.
Initially I was going to exclude the Atlas II and III and the Delta II because they weren't actually competitive in payload with Falcon 9, but it turns out that a) they flew a lot of missions and b) they were pricier than I expected.
All numbers are in 2006 dollars adjusted for inflation.
We'll classify all the ULA flights based the maximum payload to low earth orbit of the specific rocket variant. This is an oversimplification but the only good way I came up with to compare the flights. I'll do a bit of an adjustment later to correct for an error that this may introduce.
The number of flights for each variant is put in a vertical position that shows what that variant's payload to low earth orbit is.
For the atlas 2, there were 7 of the IIA variant and 12 of the IIAS variant.
For atlas 3, there were two of the IIIA and 4 of the beefier IIIB variant
For the atlas v, there were 9 launches of the 501 variant with no boosters, 2 with one booster, 5 with two boosters, 2 with three boosters, and 1 with 5 boosters.
For delta II, there were 9 of the 7300 series, 10 of the 7400 series, and an impressive 42 of the 7900 series.
And finally, for the delta IV, there were three flights of the M variant, one flight of the M+(5,4) variant, and four flights of the M+(4,2) version. In addition, there were three delta IV heavy flights.
That is a lot of flights, 116 by my count. This a decent analysis of the payloads that ULA flew in that decade; it ignores the missions flown by other launchers.
What sort of reuse would be possible if Falcon 9 launched these payloads?
Here's the current Falcon 9 with its payload capabilities in expended, drone ship, and return to launch site modes. It's a little unfair to use these numbers because they are the mature Falcon 9, but the Falcon 9 in 2017 - when reuse got moving - was relatively close.
We'll fly the ULA launches over to the Falcon 9 side
Because I'm looking at launches to low earth orbit and the ULA rockets do a bit better on higher energy launches, I'm going to draw some boundaries that are lower than the actual Falcon 9 payloads. This also adjusts for the early Falcon 9s being less performant than the current one.
All of the Atlas II, atlas III, and delta II payloads plus 15 from Atlas V and Delta IV can be handled by falcon 9 in return to launch site mode, for 101 launches total. 11 further flights require a drone ship landing. One flight cannot be flown reusably and requires an expendable launch. And three missions require another rocket.
For 112 out of 116 launches or 96% the Falcon 9 has sufficient margin to put the payload in the proper orbit in one of the reusable modes. On these flights there is no penalty for reusability. The evaluation is not $ per kilogram, it is whether you can put a customer's payload into orbit.
Note that this is only possible because Falcon 9 is well-sized for these launches and has the margin it needs. If we look at Falcon 9 Version 1.0, we see that it can only launch 78% in fully expended mode and would be able to do less in reusable mode.
We could do the same analysis for communications satellite launches, looking at the actual Falcon 9 flights, and we'd find that many of them can be launched in drone ship mode but some need an expendable launch.
What have we discovered?
That's the end of part 2.
What do you think of this paper and the famous graph? Leave your opinion in the comments.
If you enjoyed this video, buy this print
https://www.universetoday.com/17040/nasa-to-install-shock-absorbers-to-mitigate-thrust-oscillation/
It's complicated.
(rework. How do you maximize delta-v for a given stage?
Let's use Falcon 9 as an example.
The first decision is the engine, and SpaceX already had the Merlin and thought they could build a vacuum version. They knew what their NASA ISS resupply contract required, and they had initial ideas on how much Dragon would mass.
So, you start there. You have a second stage and it's going to have a big engine for a second stage because it's going to be based on Merlin. Choose an arbitrary size for your propellant tanks, and that controls how much delta-v you can get. You now have a wet weight for the payload of the second stage and a rough idea of how much delta v you need for the first stage. Pick a size for your propellant tanks, choose the appropriate number of engines so you have enough thrust to get off the ground. Does that give you enough
Rockets that aren't clean sheet designs have more constraints. ULA's Vulcan, for example, was going to fly centaur as the second stage, and that meant
Start with Falcon 9 approach...
You need to stage slow enough so that the first stage can reenter and land without too much reentry heating. It's also helpful to do it early so you aren't very far from your launch site. And you need a first stage that doesn't use solid rockets as they will a) push your staging later and b) be expendable (yes shuttle reused solids, but it didn't really save any money). And obviously you need enough extra propellant to come back and land successfully or wings if you want to do with wings (don't, it's probably a bad decision).
Then you need a big beefy second stage that can do most of the work to get into orbit.
That's falcon 9.
If you want to make your second stage reusable, it's a lot harder. You need to design a second stage that is also a reentry vehicle, which means some way of protecting it against the heat, some way to control it during reentry and while in the atmosphere, and some way to land it. All of these capabilities add mass, and every kilogram of mass you add is one less kilogram of payload that you get from the system. SpaceX looked at making the second stage of falcon 9 reusable, but it turns out that the added mass removes enough payload mass to make it uneconomical. For example, Falcon 9 can *barely* launch most geosync satellites on the geosynchronous transfer orbit that they need, and full reuse would mean losing those customers.
That's why starship is so big; it masses more than 8 times what Falcon 9 does at liftoff.
Yes.
Solar is great because it's low tech and has no moving parts; you just put the panels out on the surface, hook them up, and you're making power. If you want more power, deploy more panels. And there are numerous companies throughout the world continuously working to make solar panels cheaper and lighter.
But solar obviously doesn't work through the night. You can deal with that by batteries, but they are heavy, and anybody on earth with an off-grid solar system knows that they are also very expensive.
Nuclear would be a great solution for the initial power and longer-term baseline power. There a research project with one of my favorite names - The Kilowatt Reactor Using Stirling TechnologyY - or Krusty - that built and testing a prototype for a 1000 watt reactor that only weighs 400 kg. Just the thing to take on an initial mission. Put it up a sufficient distance from your base, run a cable to carry power, turn it on, and it will pretty much run itself for 10 years. You can use the same technology up to 10 kw, and NASA says that 4 of those would be enough power for 4-6 astronauts.
https://ntrs.nasa.gov/api/citations/20205009350/downloads/03-KRUSTY%20Reactor%20Design.pdf
There are lots of launch assist methods out there; balloons (rockoons), rockets dropped from planes, all the various gun designs (see link), spinlaunch (link),
Wikipedia non-rocket spacelaunch
Then there are more esoteric approaches, like launch loops, lightcraft, microwave propulsion, space elevators, Orbital rings, sky hooks, etc.
I'm going to do a video on at least some of these in the future, but the short answer is that I'm not very excited about them, as they require one or more of the following:
Development of advanced materials that will, at best, be very expensive to create.
Deployment of large and expensive structures on the ground
Deployment of large and expensive structures in orbit
Very large power supplies (the energy needs to come from somewhere)
Require complex operations, such as connecting an ascending vehicle with a rotating tether travelling at many times the speed of sound.
The problem is that these are pretty much all multi-billion $ projects, and those sort of projects tend to take longer and cost more than you would think, and the cost/kg of the resulting solution isn't easy to predict. And these are all big structures that provide no utility until they are totally done and if your engineering isn't good enough, you could blow up the whole investment in a single day.
We already have a very good launch assist method, known as a "first stage". There is literally tons of prior art so you can easily predict how much it will cost and what it will do, it's cheap to build the first one, and if it blows up on the pad, you just rebuild the pad, figure out what went wrong, fix it, and go back to launching.
And we now know how to reuse this launch assist method.
Tanks on earth - just put a pump between them or use gravity to move the fluid.
Tanks in orbit; everything is floating around so you can't easily separate the fluid from the liquid. If you are using hypergolic propellants, you can use containers with pressurize membranes - like residential water tanks used by people who have wells - but those membranes do not play well with cryogenic liquids.
So, to do with cryogenic propellants, you need a way to separate the liquids. You can do this with a little bit of thrust - known as ullage thrust - to settle the tanks. Or you could hook the rockets together with a tether and spin them to settle the propellant.
My guess is that the most practical method is to use some very light thrusters, build some good liquid/gas separators, and then small pumps to do the pumping. You can probably power the pumps with "ullage gas" from the tanks and then exhaust that both to provide thrust and to get the power to move the liquids.
Disclosure: I have an amusement-sized investment in rocket lab.
There are numerous small companies trying to break into the launch market. I get it - rockets are cool - but it makes very little sense as a business strategy.
Add in link to markets and moats talk.
The basic problem is the cost of rockets is:
Fixed cost (factory, engineers, launch pad, etc) / launch rate + hardware/propellant cost per launch
SpaceX is going to launch 50 times in 2022, so their fixed costs are going to be spread over 50 launches. How are you possibly going to compete with that when you are trying to launch a new rocket and get up to 5 launches a year?
To put it in market terms, launch is largely a *commodity*, and commodity markets are dominated by the cheapest provider. As evidenced by SpaceX walking over the bulk of its competition.
What you want is something that makes you different, what we call a "differentiator". ULA has one in that it is one of two companies certified to launch NSSL payloads. Good luck trying to take that away, as the bar is very high.
Rocket Lab is able to undercut spacex for launches that want a specific orbit and since they're the only real smallsat launcher right now, that is working for them. But there are other launchers out there that are more capable waiting in the wings.
Photon is their attempt to differentiate - to provide end-to-end services that make it much easier for companies who want to get into space to get what they want. There is a market there, though it's not clear how big it is yet, but the satellite bus market has largely been dominated by boeing and (others), who make big busses. Rocketlab has the chance to move into the lower mass market and, through vertical integration, do the same thing to that market that SpaceX did to launch.
I think it's a great move strategically.
Maybe a 5?
I think it can be built and will probably work and I don't think it's likely to blow up your ship.
But thrust seems problematic. Only some of the fission produces free fragments and only some of those can be aimed in the right direction. The designs claim that there is some magnetic approach to steer the fragments in the right direction, but the exhaust velocity is something like 2% the speed of light, so good luck with that.
The more complex designs are probably a 7.
I think the jury is still out on nuclear propulsion making sense *at all*. With any luck the current NASA program (link) will produce a real reactor and then maybe we can end up with some real numbers to compare with chemical engines.
The NASA design, however, is pretty conservative when it comes to mass, and that doesn't bode well for overall performance. If they only do as well as NASA hopes, they will end up with a thrust/mass ratio of about 1/16th of what you get from an RL-10.
But I could be wrong; advocates have for years been saying "advanced lightweight materials", and we'll see what they come up with.
I decline the contract.
$100 million isn't much for design, and now I need a nuclear design - see last question - plus everything else.
It's a $1 billion task IMO.
In 1979, the Russians carried out a space experiment on rat reproduction but were unable to complete it because of weightlessness.
Japanese researchers have fertilized mouse eggs in an artificial micro-g device called a clinostat.
But that's not what you are really asking.
In 1992, astronauts Mark Lee and Jan Davis met during training for STS-47 and secretly got married without telling NASA until it was too late to change the crew, and they flew together in shuttle. If the rest of the crew was cooperative, experimentation would have been possible.
But there have been a lot of humans on both shuttle and ISS, and there's been a lot of speculation.
To my knowledge, nobody is talking.
If you enjoyed this video, please check out the atomic rockets site.
Is there a way to help rockets get off the ground using other technologies?
It turns out that there's one industry that has a lot of ways that could help.
This is rockets and roller coasters
We're going to use the small launcher electron from rocket lab. I chose it because of its size and because Electron would be a great name for a roller coaster.
Electron is 18 meters tall, weighs 12.5 tons, and has 9 engines on the first stage that put out 224 kilonewtons of thrust.
Here's a launch...
Watching this launch, we can figure out a few things.
In the first 4 seconds, it reaches an altitude of 18 meters and a speed of 37 kilometers per hour or 10.3 meters per second. That gives an approximate acceleration of 2.6 meters per second squared.
That's about ¼ g of acceleration, and since it also fights against gravity, its total acceleration is about 1.25 g. That is what we need from a launch assist system.
Why roller coasters?
The traditional roller coaster used a long lift hill to provide the coaster train with its initial velocity.
But the current crop of roller coasters use another method to generate a lot of velocity in just a few seconds - the same thing we are hoping to do to help the Electron....
There are a number of different technologies...
A weight drop uses the energy in a large counterweight.
Some coasters use compressed air.
Some use hydraulics.
And many of the newer ones use linear motors.
They can achieve accelerations in the 1.5 - 3 g range, and that's exactly what we're looking for.
I'm going to concentrate on the linear synchronous motor, as it's easy to build and low maintenance. I found two case studies for coasters using this launch technology.
The first is from the Italian company Zamperla.
In this design, the train weight is 16 tons, about 25% heavier than Electron. It launches at a speed of 20 meters per second, more than we need. It takes about double the distance that electron does, so we would perhaps want a version with a little more oomph. And it takes 4 seconds to do that.
It does take quite a bit of power to launch - about 2.7 megawatts. But this is only for 4 seconds, so it uses a trivial amount of energy, about 3 kilowatt hours of electricity worth perhaps 30 cents.
The second case study is from Piak design in the Netherlands. It's similar to the first one; the launch speed is a little faster and the duration is shorter, which could mean that it's more powerful, or it could mean that the train is lighter.
Either way, it's in the right class when it comes to power.
I was a bit surprised to find that building a launch assist system is practical, at least for a small rocket like electron. Larger rockets like a Falcon 9 are much heavier and would require much larger systems.
But are economical?
To do that, we need to know how much these launch systems are going to cost.
I would expect this kind of information to be proprietary - talking about detailed costs just makes it easier for your competitors - and I wasn't surprised there weren't any specific numbers.
Based on what I did read, I'm going to estimate that the sort of systems we are looking for are going to cost a couple of million dollars. It's a conservative estimate.
How much propellant would we save with such a system?
I couldn't find any figures on propellant costs for Electron, but I did find some rough ones for Falcon 9, which uses the same propellants.
Falcon 9's has a total propellant mass of 489 tons, and the propellant cost is about $300,000 per mission.
That means the propellant costs approximately $600 per ton.
It's surprisingly easy to calculate how much propellant a rocket burns.
The specific impulse of a rocket engine is equal to the thrust in newtons divided by the mass flow rate times 9.8.
You can see my video "what's up with specific impulse?" to learn more, but for this discussion all you need to know is that specific impulse is a measure of a rocket engine's efficiency at converting propellant mass into thrust.
We can rearrange this to solve for the mass flow rate, and we find that it is equal to the thrust divided by the specific impulse multiplied by 9.8. That 9.8 is there to convert between the specific impulse and the exhaust velocity.
Plugging in numbers for the Rutherford engine on Electron, it has a thrust of 25000 newtons and a specific impulse of 311, and we find that it has a mass flow rate of 8.2 kilograms per second.
The electron has 9 engines, so it's total mass flow rate is 74 kilograms per second.
The launch assist will operate for 4 seconds, and the propellant savings during that period will be 296 kilograms.
We now have enough information to determine the breakeven point where the launch system makes economic sense.
The cost per launch of the launch assist system is equal to the total cost divided by the number of launches, and the savings is equal to the propellant cost per ton multiplied by the number of tons saved by the system.
Plugging in numbers, we're using $2 million as the cost of the launch assist system, $600/ton for the price of propellant, and 296 kilograms - or 0.296 tons - as the amount of propellant saved, we find that our launch assist saves $180
Solving for number of launches, we get about 5500 launches to break even.
Which is the answer to the question of why we don't see these systems, often summarized in the saying "propellant is cheap".
If you enjoyed this video, please go ride a roller coaster, and remember to shout "wheeeeeee!"
Space people sometimes talk about satellite buses.
We know what a satellite is, but what does that have to do with a bus?
For all satellites, there are some common things we need...
a structure that everything can mount to
electrical power
communications
command and data handling to tell the satellite what to do and to get data back from it
propulsion to move it around and station keeping to keep it where it should be
attitude control to keep it pointing in the right direction
guidance, navigation, and control to make sure it gets to the right place
Deployment to deploy spacecraft parts or other satellites
Launch vehicle integration so we can launch it.
All of those things are built into the main body of the satellite, and that is called the satellite bus. It does all the spacey things so that you can focus on what you care about. You might want to do earth imaging, or high power communications, or experiments, or a lunar rover.
What you want to do will drive what the satellite bus needs to do.
Do you need a big structure or a small one? Do you have tiny power requirements or huge ones? Are you hanging around in low earth orbit or going to mars?
There are many different players in the satellite bus market.
Since the James Webb telescope is a spacecraft rather than a satellite, it has a spacecraft bus.
The spacecraft bus is a fully custom creation to support the requirements of the instrumentation and telescope.
There's no exact cost for the spacecraft bus, but the overall project cost $10 billion and a few billion is a fair guess for the spacecraft bus.
For the Mars Curiousity Rover, the spacecraft bus is the body of the rover, known as the warm electronics box.
NASA reused the basic rover design in the later Perseverance rover.
Geostationary communications satellites are very common, and there a number of suppliers that produce satellite buses for this usage. For example Boeing produces the BSS-702P, Thales Aleia Space produces the spacebus 4000, and Lockheed martin produces the A2100. There are a number of other manufacturers.
These are spacecraft buses that are very specialized to the requirements of these satellites - they have enough propulsion to move the satellites from the geosynchronous transfer orbit they are launched to the final orbit, high power systems to operate the satellite, and communications electronics for the satellite to use.
These are very complex satellite buses and are therefore expensive with some of them costing more than $400 million.
The NASA spacecraft and communications satellite buses are traditional designs that have been costly.
There are, however, some new approaches being used elsewhere in the satellite market. An obvious example is cubesats, and there are many many companies working to provide very small satellite - or cubesat - solutions.
To pick one of those companies, endurosat makes cubesat platforms in many different sizes, and you can go directly on their website and buy one of those platforms. A very different approach than traditional satellite bus companies.
There are other companies working on bigger satellite buses.
Firefly aerospace calls their bus elytra "el itra", and it comes in three versions.
Elytra dawn is the simple one that works in Low Earth Orbit to deliver or host satellites.
Elytra dusk is targeted from low earth orbit to geostationary orbit, with bigger payloads across a wider range of orbits
Elytra dark is designed to support missions to lunar orbit and beyond.
We see the same pattern with Rocket Lab, with four different platforms based on a common heritage of hardware and each platform targeting a specific scenario.
We have also seen SpaceX take their Starlink communications satellite and adapt that satellite bus to Starshield, a satellite for the department of defense.
And that's satellite buses. Doing all the hard spacey stuff so you don't have to.
If you enjoyed this video, please watch the best space bus
The creators of the TV Series For All Mankind had a problem...
Their story line required a large moon base that would grow larger over the years.
The Apollo lunar module could only carry a small amount of mass to the lunar surface - a ton or so. A modified "LM truck" would replace the ascent module with a cargo module, and that would carry perhaps 5 tons of cargo. That's enough for extended stays on the lunar surface, but too small to build and support a big base.
What was needed was a cheap rocket that could fly regularly.
Here's the solution to the problem - a rocket with enough oomph to support a lunar base - and other projects.
But interestingly, that rocket did not come from the minds of the shows writers...
It came from the mind of rocket designer Robert Truax while working at the rocket engine manufacturer Aerojet in the early 1960s.
This is sea dragon.
Sea Dragon was not just an idea, it was a very well defined proposal in three volumes, with the summary occupying 284 pages and the operational plan an additional 418 pages. Volume 2 is unfortunately not available online.
There is far, far more detail available in these documents than I can cover in a short video. I've linked to them in the video description.
The Sea Dragon approach puts heavy emphasis on the "ship" part of the term "spaceship", with it being based more on shipbuilding approaches and technologies than what we think of as aerospace.
Here we see rockets being assembled in a dry-dock environment, and then towed out through a canal to be launched at sea.
Here's another view of the rocket leaving the manufacturing area.
Once at the launch site, the rocket propellants would be loaded..
In preparation for launch the rocket needed to be flipped around so that it was vertical in the water. This was done using a set of 6 ballast tanks attached to the nozzle of the first stage engine.
The ballast tanks would be filled with a material such as drilling mud that is heavier than water. The weight will cause the rocket to switch to a vertical orientation.
The for all mankind team thought it would be cooler if the rocket started out fully submerged, but in fact it floated with most of the front section out of the water.
Now it's time to launch...
When the launch checkout is complete, the four auxiliary engines are ignited, and then the large main engine. When performance is verified, the ballast tanks are released and the rocket climbs out of the water and into the atmosphere.
The ballast assembly is recovered for reuse.
We can now examine the components of this very unique rocket.
The engine is a pressure-fed design - the propellant is stored in tanks under pressure and that pressure feeds the propellant into the combustion chamber. Pressure fed rockets are simple and cheap, but to get decent performance they require high pressure tanks, and those tanks are thick and heavy.
The first stage engine uses the very common mixture of liquid oxygen and RP-1 kerosene, known as "kerolox". The LOX tank is pressurized to 16 times atmospheric pressure, or 16 bar, and uses an aluminum tank 53 mm thick, or 2.1 inches. The RP-1 tank is pressurized to 30 bar, and uses an aluminum tank 106 mm thick, or 4.1 inches. Because of the difficulty of dealing with aluminum sheets of that thickness, they also considered a steel tank for the RP-1, which would only be 30mm or 1.2 inches thick.
These are very thick tanks - the Saturn V first stage tanks were made from aluminum and were 4 to 6.5 millimeters in thickness, though they did have internal reinforcement.
The LOX tank is pressurized by heating the liquid oxygen in the engine and returning the oxygen gas to the tank. This doesn't work for RP-1, so there is an addition tank holding methane gas under high pressure that is used to keep the RP-1 tank pressurized.
It's time to compare the sea dragon first stage engine to another engine, and the obvious choice is the Rocketdyne F-1 used in the Saturn V first stage.
The engine cycle of the F-1 is gas generator, burning propellants to drive turbopumps. The sea dragon is pressure fed.
The chamber pressure in the F-1 is 70 bar, while Sea Dragon is a much lower 21 bar. This is primarily because of the much lower pressure in the pressure fed design.
The specific impulse - or efficiency - of the F-1 is 263, while the sea dragon is only 242 despite running the same propellant mix.
The F-1 is a very large rocket engine, with a nozzle that is 3.7 meters or about 12 feet across. The Sea Dragon engine is 22.8 meters, or about 75 feet across. Which is huge, and it's smaller in proportion to the F-1 engine - if it was an equally efficient nozzle it would be nearly 50 meters across.
And finally, the F-1 thrust is 6.7 meganewtons, the most powerful single-nozzle engine ever made. The sea dragon is 350 mega newtons. It would take 52 F-1 engines to equal that thrust.
The F-1 engine was, of course, very difficult to develop because the large combustion chamber led to combustion instabilities - variations in thrust that can break engines. Not good.
The Sea Dragon summary has this to say about it.
(read)
Is that correct? I'm not a rocket engine designer, so my opinion shouldn't carry much weight, but I am skeptical.
81 seconds after launch, the vehicle reaches 38 kilometers in altitude and a velocity of 1770 meters per second, and the first stage is done. This is a very early staging by rocket standards - the Saturn V and Falcon 9 are early staging rockets, and they are much higher and much faster. Sea Dragon has only about 60% of their kinetic energy at staging...
Can you guess why?
At staging, the first stage detaches from the rest of the vehicle, and then deploys a large drag flute - a ring that is inflated from the methane tank. This ring provides drag to keep the first stage in the proper orientation and slow during reentry, and it will ultimately impact the surface of the sea at 90 meters/second or 300 feet per second. This results in a g-load of only 22 g - it's slow enough and the stage is robust enough due to those very thick tank skins - that there is no damage to the vehicle.
The drag flute is composed of two layers - the inner structural layer and an outer ablative layer that is replaced after each use.
Note that the front of the first stage has a cone shape. This provides an aerodynamic shape for reentry, and the nose tucks inside the nozzle of the second stage engine to reduce vehicle mass and length.
Now we move onto the second stage. You may have noticed a weird texture on the outside of the first stage. That is an expandable nozzle skirt for the second stage engine made out of stainless steel that is approximately 4mm, or 1/6" thick. The thin stainless steel is corrugated to give it strength, and attached to the second stage engine nozzle. During launch, the nozzle is held next to the skin with retention bands.
During staging, the bands are removed, and the corrugations allow the end of the nozzle that is not attached to the engine to expand. This larger nozzle significantly increases the efficiency of the engine.
Here's a cutaway of the second stage. The second stage engine burns liquid hydrogen and liquid oxygen, or hydrolox. It has an immense hydrogen tank because liquid hydrogen is not dense. In this variant, there's a payload tank at the front to carry propellant, but it could also be a conventional payload section.
The four auxiliary motors each have their own tiny liquid oxygen tanks - this means that there is no requirement for feed lines from the forward liquid oxygen tank back to those engines.
Like the first stage, it is a pressure fed engine design but both the pressures and stresses are low, so the tanks are made of aluminum that's only 18mm or about 0.7" thick.
There is a second stage recovery plan as well, using a similar drag flute but one that is protected against reentry heat.
Once again we need an engine to compare to, and the obvious choice is the Rocketdyne J-2 used on the Saturn V second and third stages.
The engine cycle of the J-2 is gas generator, burning propellants to drive turbopumps. The sea dragon is pressure fed.
The chamber pressure in the J-2 is 54 bar, while Sea Dragon is a much lower 5 bar. This is a low pressure design.
The specific impulse - or efficiency - of the J-2 is 421, while the sea dragon is 409. Sea Dragon does well here because of that giant nozzle, a full 42 meters across.
And - finally - the thrust is about 62 times the thrust of the J-2.
How big was Sea dragon?
These existing super heavy launchers have takeoff thrusts of 23 to 45 mega newtons, and can lift payloads from 57 tons up to 140 tons into low earth orbit.
Starship has a takeoff thrust of 74 mN, and targets a low earth payload of 150 tons in reusable mode up to 250 tons in expendable mode.
Sea dragon is 150 meters tall and has a 23 meter diameter. It has a takeoff thrust of 350 meganewtons. It is a monster that can lift 500 tons to low earth orbit.
What will all this cost?
The development was estimated at $2.86 billion, or $28 billion in 2023 dollars. The Saturn V cost about $37 B in current dollars for research and development and all the flight hardware.
For the vehicle, the second stage engine is $2.1 million to build and the rest of the stage is $8.4 million.
The first stage is $29 million and the engine is $8.3 million.
That gives a total of $47.8 million. Here are the numbers in current dollars.
There is a complicated model for estimating payload costs to orbit which gives $10-$30 per pound, $22 to $66 per kilogram. That's $215 to $645 per kilogram in 2023 dollars. For comparison, Falcon 9 is about $3600 per kilogram in reusable mode.
Was sea dragon practical?
Will the big single engines work?
As I noted earlier, the US had big problems getting the F-1 to be reliable, and the Soviets went to multiple chambers for engines like the RD-170 because they couldn't solve the instability problems. And that was with engines that were a fraction of the size of the sea dragon engines.
I think it's unlikely that they could be made to work.
Was there an alternative?
Perhaps
If you built a lot of pressure-fed small engines, you could probably get them to work. It would change your packaging a lot, and you might get less benefit from the second-stage nozzle - though there's no reason you couldn't use this giant nozzle with a bunch of small engines inside of it.
Would recovery work?
For the first stage, probably. It just isn't going very fast so it has much less kinetic energy and more drag than a Falcon 9. And you aren't going to do much to that hugely strong tank section, though you might mess up the engine or engines.
For the second stage, even with the big drag flute that I didn't bother to model, I don't think so. The front of the structure is all aluminum and that is going to get far too melty during the early part of reentry. There's a reason that shuttle had heat shield tiles on top of its aluminum skin.
Would the seawater eat up everything?
Seawater and aluminum are not a happy couple. Perhaps it would be possible to coat and seal the aluminum structure, but it would likely be problematic, especially with the engines.
Was sea dragon practical?
Not really. There's a reason NASA built the Saturn V and didn't build the Sea Dragon.
But it's a wild and fun design, and was perfect for a TV series that needed a really big rocket.
If you enjoyed this video, watch "for all mankind". Even though it plays a bit loose with history at times, it's well-written and entertaining.
Here's a hypothetical for you.
Let's say that you are wanting to build a space station like ISS, which lives at an altitude of 407 kilometers and an inclination of 51 degrees. If you want to understand what those number mean, please watch my "you know orbits" video.
You have two launch systems you can use to launch all of your space station modules.
You have the classic space shuttle. And you have the modern Falcon 9.
Which option will get your space station built in the fewest number of flights? It's going to depend on which option can carry more payload.
I did not look at every space shuttle flight to the space station, but I did find that STS-126 was a little more than 17,000 kilograms, and I'm going to use that as the maximum payload. It's close enough.
SpaceX has never flown a Falcon 9 mission to the space station, so we'll need to look for something similar. Starlink Group 4-37 carried 16,500 kilograms to 570 kilometers at 53 degrees inclination, so I'm going to assert that it can also get 17,000 kilograms to ISS. That's with first stage booster reuse. It's probably over 20,000 kilograms without reuse.
This result is a bit surprising - the shuttle is a massive launch system, but it's not any more capable than the much smaller Falcon 9.
What is going on here?
The falcon 9 second stage brings the payload to the space station. It's fairly small and light, massing only 4,000 kilograms.
That means the total mass to the destination is 21,000 kilograms, and 81% of that is useful payload.
I think you've probably figured out what's coming next.
The space shuttle brings along a 78,000 kilogram orbiter with the payload, which means the total mass is 95,000 kilograms and only 18% of that is payload. Shuttle can bring far more mass to the space station, but unfortunately the vast majority of that mass is orbiter.
Which brings up an interesting question.
If we want to build a space station quickly, is there a way to create something shuttle-like that can carry much more payload.
We can chop off the whole front crew section as we don't need astronauts, clip off the tail, and remove the heavy wings, landing gear, and frame.
Do that, and you end up with something like this. A long tubular vehicle with shuttle engines.
NASA called this version "shuttle cargo", or "shuttle C"
It reuses the external tank, the shuttle rocket boosters, and the same main engines. It has a payload bay roughly the size of orbiter.
In the late 1980s, NASA was doing planning studies for Space Station "Freedom", the predecessor of the international space station, and they wanted to explore different ways of building it, including shuttle c.
Freedom would be placed at 400 kilometers of altitude and at an inclination of 28.5 degrees, the natural inclination for launches from Cape Canaveral.
The orbiter can carry 17,000 kilograms to that orbit.
A version of shuttle C powered by three space shuttle main engines would be able to lift 66,000 kilograms. That is a very significant difference in payload.
Working that into their assembly plan, they found that it would take 17 shuttle flights to build the station.
If the shuttle C was available late in the assembly process, it would take 9 shuttle flights plus 3 shuttle c flights, for a total of 12 flights.
If it was available early in the assembly process, it would take 6 shuttle flights plus 4 shuttle c flights, for a total of 10 flights.
These same ratios would apply for building the international space station.
So why didn't NASA build Shuttle C?
To understand why we need to look at the stakeholders in the shuttle and space station programs.
We'll start with the astronauts...
(read)
Astronauts want to fly missions. They don't want this.
For the space shuttle program overall, shuttle C would mean new development work in a program that last did significant work in the 1980s, and that would be good.
But overall, shuttle would fly less, and flying less means fewer employees, smaller budgets, and less prestige.
And shuttle C might go over budget or fail.
Staying with the existing plan is much more comfortable.
For the supporters of the Space Station, shuttle C could conceivably result in earlier completion of the space shuttle. That would be nice but it's not really a big career builder.
But current manufacturing and assembly plans are all based on using shuttle. Changing those would add a lot of complexity and more work.
And most importantly, the space station shuttle launches are not paid for by the space station program - those come from the shuttle program - so there's no monetary or budget incentive to do shuttle C.
The main shuttle contractors were united space alliance who did shuttle processing, ATK Thiokol who did the solid rocket moters, Lockheed martin who built the external tanks, and Aerojet Rocketdyne who maintained the shuttle engines.
Fewer flights means fewer solid rocket boosters, fewer external tanks, and less time spent on shuttle processing.
It would means more shuttle engines, but NASA would probably want a cheaper version.
Shuttle C generally makes things worse for the contractors.
The last set of stakeholders are the politicians.
Fewer flights means fewer jobs at NASA centers and contractor locations, and therefore less influence and smaller campaign donations for congresspeople.
Looking across all the stakeholders, it's pretty clear that shuttle C is a solution to a problem that they do not have - it's at odds with what they are trying to accomplish.
And that's why Shuttle C was never built. It was a solution to a technical problem but the objectives of all the program stakeholders was better served by flying more shuttle flights.
If you enjoyed this video, here's a different take on shuttle sea...
In August of 1985, the space shuttle Discovery touched down to mark the end of STS-51-I, the 20th shuttle mission.
It was the 6th shuttle mission for the year, and it seemed that NASA had hit a good cadence for shuttle flights.
On May 15th of 1986, challenger would launch the Ulysses solar probe, and 5 days later, Atlantis would launch the Galileo probe to Jupiter.
These were important science missions, and they would fly with the shuttle centaur upper stage in their payload bay to send the probes on their journey.
Some were excited about the upcoming missions. Some feared that a malfunction in the stage would turn the shuttle into a large fireball.
This is the story of shuttle centaur.
The need for shuttle upper stages like shuttle centaur is a result of shuttle architecture.
To understand the problem, we'll need a little bit of rocket physics.
When we put a payload into orbit with a two stage rocket, it's not just the payload that is put into that orbit, it's the second stage as well. For this Atlas V, the second stage weighs 2300 kilograms.
We're going to be looking at the payload that this rocket can lift to different orbits. If you want to know more about the orbits I'll be comparing, see the "You know orbits" video
The Atlas V in this configuration can put about 20,500 kilograms into low earth orbit. Of that, 2300 kilograms is the mass of the centaur second stage. That is a payload tax that we have to pay, and for this orbit, it's about 11% of the total.
A higher orbit - in this case a GTO-1800 transfer orbit - takes more energy to get into, so the total mass is about half the LEO mass. The mass of the second stage didn't change, and that means the second stage is 21% of the total, leaving only 79% for the payload.
Go up to GTO-1500, and the second stage is up to 25% of the total, leaving only 75% for the payload.
Finally with a geostationary orbit - the hardest one in this graph - the second stage is up to 37% of the total with only 63% for payload.
If we kept going to higher energy orbits, at some point the second stage would consume all of the mass and the payload would be zero.
It's therefore important to keep the second stage as light as possible, to reduce the payload tax.
The centaur III that flies on the Atlas V is a great second stage because it is so light.
The propellant tanks are the thickness of a US dime, 1.35 millimeters or 1/20th of an inch. They are an example of a balloon tank, a tank with walls so thin that if it loses pressurization it will collapse.
The space shuttle orbiter is not a light second stage - it is a very heavy second stage at approximately 110,000 kilograms at launch.
The external tank is also carried to orbit, and it masses a further 58,500 kilograms, for a total mass to orbit of 168,500 kilograms.
Let's recreate our graph, this time for shuttle.
The shuttle graph looks very different.
Even at a very low orbit of 204 kilometers, only 14 % of the mass to orbit goes to payload.
On an ISS trip, that drops down to only 9 % of the mass to orbit.
The highest altitude shuttle flew to was 614 kilometers to launch hubble, and that was only 7% of the total payload.
Fly to 960 kilometers and there is no useful payload.
All of these orbits are low earth orbits. Because of the mass of the orbiter and external tank, the shuttle cannot get out of low earth orbit.
This was a problem as shuttle was intended to launch the majority of US satellites including geosynchronous satellites like NASA's tracking and data relay satellites and commercial satellites like Optus A1 or Leasat 1.
What was needed was a way to get these satellites from a low earth orbit to their destination orbit.
This was a solved problem. The Titan III launcher has the same issue to a lesser degree, and it had been flying the transtage, a small upper stage using hypergolic propellants. The transtage functions as a third stage for the rocket and it's light so the payload tax is low.
The air force was already working on the problem - they were planning on a simpler and cheaper upper stage that relied on two solid rocket motors - to fly on the Titan 34D and later the Titan IV. NASA had been thinking about a "space tug" upper stage that would work with shuttle, but that was a long term project.
NASA and the air force decided to collaborate on the solid rocket solution until the space tug was ready, and the solid rocket solution was therefore named the Interim Upper Stage.
In late 1977 NASA abandoned the space tug work, leaving the interim upper stage the only upper stage.
Nobody wanted to change the acronym, so they tried a bunch of different words starting with "I" until they settled on "Inertial"
That's not actually true - they chose "inertial" because they needed an I word and the stage used Inertial Guidance.
There would be a two stage IUS that would be used by the Titan and the shuttle for satellites and a three stage version that NASA would use for planetary missions.
https://web.archive.org/web/20140417071751/http://www.aerospace.org/wp-content/uploads/crosslink/V4N1.pdf
As IUS development was underway, NASA was having second thoughts.
The Inertial upper stage worked fine for launching NASA satellites, but it fell short for planetary missions. NASA wanted to fly the Galileo mission with a simple flyby of mars to gain velocity and then on to Jupiter, a trajectory that would involve 3 years of travel time.
But IUS wasn't up to the task of putting Galileo on that trajectory - it couldn't generate enough velocity. The only way to get Galileo to Jupiter with IUS was a trajectory called VEEGA, for Venus Earth Earth Gravity Assist.
IUS would launch the probe but not towards Jupiter, but inward towards Venus. About 4 months later an encounter with Venus would give it a velocity boost and sent it back to earth, where ten months later it would get a boost from earth that would fling it out towards Jupiter, only to come back to earth two years later to get another boost, finally reaching a velocity that was high enough to send it to Jupiter. The trip would take more than 6 years.
NASA was not excited by the complexity of the mission and extending the travel time as it cost them $50 million per year to keep the program running.
NASA's solution was the Centaur, a hydrogen oxygen upper stage that first flew in 1962 and has been flying ever since. It existed, it worked reliably, it was much more efficient, and it carried more propellant than IUS. It could do Galileo the way NASA wanted to do it and would enable other ambitious science missions.
There was a bit of a problem. While it was possible to fit centaur into the payload bay, there wasn't much room left for a satellite in front of it. Worse, it put a lot of mass forward in the payload bay, too far forward for the shuttle to be stable during reentry if it had to come back with centaur and the payload inside.
This shuttle picture is of Endeavour on STS-123, and the large module at the back of the payload bay is the Japanese Kibo (keybow) lab module. It needs to be that far back in the payload bay because of its mass.
What NASA needed was a modified version of centaur, one that was both shorter in length and wider around to concentrate the mass of the fuel - and the payload on top of it - as far towards the back of the payload bay as possible.
NASA decided to build two versions - the shorter one is known as Centaur G, optimized for department of defense payloads up to 40' long, and the longer Centaur G Prime is designed for planetary missions.
The air force would fund Centaur G, and NASA would fund Centaur G Prime.
Congress agreed to pay for it, and the Shuttle Centaur program was born. You can think of shuttle centaur as the overall program and centaur g and centaur g prime as the specific stages under that program.
Shuttle centaur would be a major program and things quickly got political.
The obvious choice for the program was Lewis Research Center in Ohio - since renamed to Glenn Research Center. They had done the feasibility study for Shuttle Centaur, they had reconfigured the Atlas-Centaur stage to work on Titan, and they had had worked on earth orbital missions, lunar missions, and planetary missions like Viking, Voyager, and Pioneer. And the center had lost much of its work when the nuclear propulsion program was cancelled, so they had room and bandwidth to do the work. The were the centaur experts.
They weren't, however, part of the "shuttle club" of Johnson Space Center in Texas, Marshal Space Flight Center in Alabama, and Kennedy Space Center. The directors of those three centers wrote an "eyes only" letter to NASA Administrator Alan Lovelace where they asserted
Therefore, we believe, when consideration is given to all relevant factors, it is in the best interest of the program and the Agency that the existing JSC/MSFC/KSC team be responsible for the Wide Body Centaur program
They specifically thought that Marshal should manage centaur development.
As often happens in large organizations, their secret letter quickly made it back to Lewis, and to say the were displeased would be an understatement.
Administrator Lovelace ultimately chose Lewis, but this initial interaction would set the stage for a contentious relationship.
There were a lot of moving parts in this program.
With Lewis in charge of design and management they had to work closely with Johnson and Kennedy. But it was much more complicated - they had to collaborate with the air force for the version of Centaur that would launch department of defense satellites, with JPL on the design of Gallileo and - just to make things interesting - with the European space agency who was building the Ulysses probe.
And there were the contractors doing the actual work, most prominently Martin Marietta who was building the actual centaur g stages.
One of the managers at Lewis later said,
"It was a very substantial development program. When we sold it, I don't think we emphasized
how much substance there was to the development. I think we said, it's a Centaur, we have
them all over the place, we will just make one that will fit in the cargo bay. It was a relatively
straightforward technical job, but managerially it was extraordinarily demanding."
There were also some significant differences in goals.
Lewis' goal was purely to fly shuttle centaur for both NASA and Air Force missions.
Johnson had a more complex job. They needed to keep shuttle safe and keep the shuttle flying missions on an increasingly faster cadence. Somewhere down on their list of goals was to fly shuttle centaur, and while they might think shuttle centaur was a good thing for NASA, it was just a lot of extra work for them. It's not surprising that they would spend more time on the downsides of shuttle centaur than the advantages.
And Johnson saw lots of disadvantages.
They were deeply acquainted with the difficulties of working with liquid hydrogen, and while centaur G only held about 4% of the hydrogen in the shuttle external tank, it held it inside the shuttle payload bay in a very thin tank. There needed to be a way to vent that hydrogen out of the tank in some abort scenarios, which added complexity. And centaur G also had a big tank of liquid oxygen.
The NASA astronaut corp was also headquartered at Johnson space center, and they made their opinion of shuttle centaur very clear when they labelled it "the death star". Some astronauts said they would never fly with it.
Getting everybody aligned enough to get shuttle centaur approved to fly on shuttle was a big job, but it was finally achieved. NASA would fly galileo and Ulysses on centaur G.
But progress was made. A support structure known as the centaur integrated support structure was designed.
The first centaur G was under construction at Martin Marietta's factory.
In August of 1985 the first centaur g was rolled out.
At the start of shuttle centaur, Lewis called up Johnson and asked, "How much payload will shuttle be able to lift? We need to know for shuttle centaur". And Johnson replied "29,450 kilograms."
Well, actually, Johnson put together a team that looked at the question and generated a report, and the number they gave was 65,000 pounds, but it's essentially the same.
And Lewis said, "thanks, that what we needed to know".
They used this information to figure out their mass budget for launching a planetary mission with Centaur G.
There are three components of the system. The first is the probe itself, and they used Galileo in their planning because that was the mission that took the most energy.
The second component is the centaur stage. It has to carry enough propellant to be able to give Galileo the push that it needs.
The third component is the payload bay structure that supports the launch, along with electronics and other systems to do the launch. Think of it as an orbital launch pad.
Lewis worked with JPL on the design of the probe and Martin Marietta on the centaur and support structure, and came up with these numbers:
The Galileo probe would have a mass of 2560 kilograms.
Centaur would have a mass of 22,800 kilograms, and that would be enough to give Galileo the push that it needed.
The payload bay support equipment has a mass of 3700 kilograms.
Add all those up, and the total payload mass is 29,060 kilograms, which means there was a margin of 390 kilograms.
It's a little more complex than this - because the earth and Jupiter are always moving the amount of fuel to get there changes from one launch day to another, so what you want is to have enough margin so that you can have a comfortable amount of launch opportunities.
During development, the shuttle gained weight and every kilogram of extra weight meant one less kilogram of payload. If the shuttle gained 2000 kilograms - only a couple percent - the payload would go down to 27,450 kilograms, which would leave a margin of negative 1610 kilograms.
Shuttle could no longer lift the centaur g and enough propellant to carry Galileo to Jupiter.
But perhaps there was an option...
The space shuttle main engine had a rated thrust of 2.09 meganewtons.
But the engine team continued development, made a host of changes, and certified the engine as safe at 104% thrust, or 2.18 meganewtons. That engine power was used on STS-6 and later, and while it helped, it limited the number of galileo launch days to 6, which was judged to be too few.
But there was a full power setting at 109% which produced 2.28 meganewtons of thrust, and that was enough to make the Galileo mission practical. NASA associate administrator Jesse Moore loved the idea and immediately approved it, despite the objection of the propulsion engineers at Marshall who asserted that the full power level was only intended for abort scenarios.
In the Fall of 1985, Centaur G, Ulysses, and Galileo were all ready and there were firm launch dates in May of 1986. The shuttle astronauts were still not happy but they had decided to fly the mission.
In a few short months, Shuttle Centaur would fly.
Then Challenger was destroyed during ascent on a cold January day, and the world within NASA shifted.
Three things were very clear in the time after Challenger...
The first is that there would be a new review of the overall safety of launching shuttle centaur as a payload, and it might be decided that the concept was inherently too risky.
The second was that there were going to be safety changes that would make shuttle heavier and that would further decrease the payload capability. It eventually ended up at 25,000 kilograms.
The third was that the 109% full power setting would not be allowed in the future.
Put those together, and it was pretty clear that the planned Galileo mission could not take place as shuttle could not lift the required payload.
On June 19th of 1986, NASA administrator James Fletcher cancelled the program, saying that Shuttle Centaur
NASA and the air force had spent $1 billion on the project, $2.8 billion in 2024 dollars.
Galileo was finally launched on October 18th, 1989 on the STS-34 shuttle mission.
It used the standard two-stage IUS that the air force developed and therefore required the complex VEEGA trajectory.
Was Centaur G headed for disaster?
We don't know.
It does seem that the choice to launch Galileo at a 109% engine setting - one that was never used during the shuttle program - was exactly the kind of decision making that led to the Challenger disaster. Perhaps the engines would have worked fine, but it's important to remember that shuttle was very intolerant to main engine failure during many parts of a flight, and even a safe shutdown of an engine could lead to loss of the shuttle and crew.
See my shuttle abort modes video for more details.
My opinion is that there was a good chance that the Galileo launch would have led to disaster and the chief reason it didn't happen is that Challenger happened first.
If you want more details, there are three important references. The first is "Taming Hydrogen - the Centaur Upper Stage Rocket".
https://forum.nasaspaceflight.com/index.php?action=dlattach;topic=2398.0;attach=620740
https://books.google.com/books?id=5MVGAQAAMAAJ&pg=PA579&lpg=PA579&dq=%22Study+of+Upper+Stage+Alternatives%22&source=bl&ots=lcolWNSe1s&sig=ACfU3U2tgUNsM5QH-mfYmYF9I9I8hevgUA&hl=en&sa=X&ved=2ahUKEwjXs6LFiPWDAxV7CjQIHa1EDGEQ6AF6BAgJEAM#v=onepage&q=%22Study%20of%20Upper%20Stage%20Alternatives%22&f=false
Soon after that, in 1980, NASA presented their plans to congress.
https://play.google.com/books/reader?id=5MVGAQAAMAAJ&pg=GBS.PA578&hl=en
And there was a later
https://ntrs.nasa.gov/citations/19820011429
If you enjoyed this video, please send me a framed picture of your favorite centaur image...
https://earth.google.com/web/@29.56026109,-95.08503954,27.65235519a,0d,60.00000007y,300.54006047h,68.90341192t,0r/data=CiQSIhIgN2Y3ZTA1ZTg2Y2E1MTFlNzk5YzI1YjJmNTFhNjA3NTIiGgoWZGxfcDZ5M1NDRllBQUFRdnhnYnlQURAC
SLS has obviously had a number of scrubs...
They scrubbed in the green run test.
They scrubbed in the wet dress rehearsal
And they've scrubbed twice when trying to launch.
What's going on?
There are 4 factors worth talking about...
The fact that SLS uses hydrogen has been discussed a lot.
I won't go into the details of why SLS uses hydrogen, other than to say it uses hydrogen because shuttle used hydrogen.
Hydrogen is indeed a pain to deal with; it wants to leak out of everything and if it catches fire, you can't see it burn. Then there's this neat thing known as hydrogen embrittlement, where the small hydrogen atoms worm their way into cracks in metals and make them worse. Neat.
However, hydrogen is well understood; it has been used successfully and routinely for years in rockets like the Atlas V Centaur, the Delta IV, and the Ariane 5.
And in fact it was not a large source of launch day scrubs on shuttle.
This is a ULA pathfinder from their upcoming Vulcan rocket, a core stage complete with development BE-4 engines from Blue Origin.
A pathfinder is a preliminary version of part of a rocket that is used to practice operations and find issues.
This particular one is a launch pad fueling pathfinder.
ULA will perform a series of fueling tests to
Validate launch pad infrastructure with the rocket
Evaluate countdown procedures
Train the launch team in advance of the first mission
They will perform these tests repeatedly to validate that the equipment is reliable and the procedures work as expected.
This is the SLS fueling pathfinder. You are probably thinking that it looks very much like the Artemis 1 rocket, and that is because it is the Artemis 1 rocket - NASA chose not to build a pathfinder that could be used for fueling tests in order to save money.
That means that the wet dress rehearsal was the first time that NASA had tried to fuel an SLS rocket. It failed the first test, failed the second test, and that is all the testing that NASA did. They have not proven that they can fuel the rocket reliably - if anything they have proven that they cannot.
This is a view from ULA's integration building at pad 41 on cape Canaveral space force station, watching an Atlas V rocket journey out to the launch pad.
The time to get from the building to the launch pad?
40 minutes...
This is SLS on the mobile transporter
SLS takes 6-12 hours to get from the vehicle assembly building to the launch pad.
Which means that it's a day to roll one way, so the minimum round trip time is about 3 days, which means rolling back to the VAB is a big deal.
For Apollo, NASA built this large mobile service structure that provided access to various parts of the Saturn V rocket and totally enclosed the Apollo spacecraft. This allowed them to do some work on the launch pad.
For shuttle, NASA went further, and created a rotating service structure that would cover most of the orbiter when closed. Not only could NASA access many of the orbiter systems, the structure included a full clean room that could be used to access the shuttle payload bay.
Lacking an easy way to work on the launch pad SLS harder to work with.
It's currently September 5th, 2022. How long has it been since a team has launched a specific rocket?
Falcon 9 flew last night, so it's been a day.
Atlas V flew a month ago.
Delta IV heavy doesn't fly very often. It flew 16 months ago
And Ariane 5 flew 3 months ago.
All of these teams are experienced in launching rockets and have done it relatively recently.
When did NASA last fly a rocket? It's been 12 years since the last shuttle flight.
When did NASA last fly a new rocket? It was the first flight of shuttle, 41 years ago. NASA simply does not have the institutional knowledge of what it takes to launch.
NASA has a number of structural issues that make it hard to make progress with SLS, and they've made choices that make it even harder.
I unfortunately don't see any sign that they are trying to build a system that is reliable, so we can expect that there will continue to be ongoing issues as the program progresses.
If you enjoyed this video, please conduct your own fueling tests and report your results.
Welcome to Stupid Idea #4, solar power satellites.
Here's the basic idea...
We build a big satellite out in geosynchronous orbit - 36,000 kilometers - so that it stays in the same place over the equator. That satellite has large arrays of solar cells that can convert the sun's light into electricity.
This is a good place for solar cells; in earth orbit there is 1360 watts of energy falling on each square meter of solar cell, and the sun shines 24 hours per day. Multiply those together, and that means we get 32.6 kilowatt hours of power for every square meter of solar cell each day.
Compare that to solar cells on the surface of the earth, in this case a solar farm in Utah. I chose Utah because it's quite sunny but not as sunny as New Mexico, so it's a reasonably average place in the American Southwest. We know that sunlight is blocked by clouds, we know that the atmosphere absorbs some of it, and obviously the sun doesn't shine at night. There is information from the National Laboratories that tells us the average amount of solar radiation per day in a given location. For Utah, it's about 5.5 kilowatt hours per square meter per day.
If we divide 32.6 by 5.5, we find that, over a year, a solar cell in geosynchronous orbit produces 5.9 times the power of one on the ground in Utah. That is the exciting part about solar power satellites, you get much more power from a given area of solar cells.
There is one problem - getting that electricity to the ground. That is done by converting the electricity to microwaves, and beaming - or aiming - those microwaves at a small spot on the earth, where we have placed a large antenna known as a rectenna. That converts the microwaves back into electricity.
Not surprisingly, there are a ton of factors involved in comparing space-based and terrestrial solar projects, and I didn't want to make a thesis project out of it, so I came up with a simple game plan.
First, I'm going to choose a midsize solar project on the earth
Second, I'm going to find a solar power satellite that produces a comparable amount of power
Third, I'm going to compare the two options.
For the terrestrial option, I chose Red Hills renewable energy park in Utah.
It was completed in 2015 at a cost of $188 million, and covers 2.11 square kilometers of land.
It will generate 104 megawatts of power at full output and about 206 million kWh / year
If we divide that by the number of hours in a year, we will find that the average output is 23.5 megawatts. That is the size of solar satellite we are looking for.
Nicely, there's a study that aligns pretty well with the Red Hills renewable energy park. The SPS-ALPHA study, completed in 2012, has a pilot plant option that delivers 18 MW to earth, which will generate about 80% of the power that Red Hills will generate in a year.
This project looks at the full cost of such a project, and it estimates it will cost about $4.5 billion to complete, and over the 10 year lifetime it will be able to generate power for $3.26 per kilowatt hour.
https://ntrs.nasa.gov/api/citations/20190002466/downloads/20190002466.pdf
Red Hills makes 206 million kwh per year, and the construction cost allocated over 10 years gives a cost of $19 million per year. Do the math, and that comes out to 9 cents per kilowatt hour generated.
That's compared to $3.26 for SPS-ALPHA
But it's actually worse than that.
SPS-ALPHA estimates a launch cost of $1500/kg to get to LEO and $1500/kg to get from LEO to GEO, for $3000 per kilogram total.
The mass of this option is a little over a million kilograms, for a total launch cost of $3.2 billion
I have no idea what transportation system they were thinking of in 2012.
Falcon 9 to LEO on starlink missions is likely around $3200 / kg, but that's only to LEO.
Falcon 9 to GTO-1800 is probably around $9100 / kg, but you need to get from GTO to GEO
The only real option is Falcon Heavy, which will run around $13,000 / kg.
1 million kilograms at $13,000 per kilogram is a bit more than $13.6 billion, so the total cost is not $4.4 billion, it's $14.3 billion.
That pushes the per-kilowatt hour cost to $10.40.
But what about starship?
Let's assume starship can take 100 tons to LEO and can be refueled in 6 tanker flights with enough propellant to get to GEO.
Assume $10 million per launch.
That means $70 million for 100 tons to GEO, or $700 / kg
That puts the launch cost at $700 million, the total cost at $2 billion, and the cost per kilowatt hour at $1.55, or 17 times.
Which actually is decent for a pilot project; the launch cost is significant but not killing us.
There is a production version of SPS-ALPHA that assumes 100 times more power, or about 2 gigawatts. It also assume much cheaper equipment and transportation costs of $1000/kg to GEO - around what starship might be able to do. And higher efficiencies in all of the electronics, and much cheaper prices for all the equipment because you are making a lot of them
Put all that together, and you get a cost per kilowatt hour of $0.09, the same as Red Hills.
That looks competitive.
Well, not quite. The $0.09 per kilowatt hour cost of red hills assumed a 10 year lifetime, but a 30 year lifetime is more appropriate, which means we're more likely to see $0.03 per kilowatt hour.
I am, of course, glossing over some important details that impact both cost and practicality.
We'll start by looking at issues with the terrestrial (land based) solar projects.
The power generated varies from month to month.
Here's a chart that shows the monthly variation of sunlight in Utah, with the red line showing the 5.5 kilowatt hours per square meter per day value that we were basing our comparison on. We can see in December that we are falling short of that amount of power.
While there are longer-term power storage approaches that might allow us to save power from one season to use it the next season, they aren't very well developed yet and they are expensive.
The simplest thing to do is to simply increase the size of the system by about 75%. The green line shows that at our lowest-power month, we are now at the amount of power we get from the satellite solution.
We now have a lot of excess power in the sunny summer months.
https://maps.nrel.gov/nsrdb-viewer/
https://www.solarenergylocal.com/states/utah/salt-lake-city/
https://books.google.com/books?id=ixJSAQAAMAAJ&pg=PA328&lpg=PA328&dq=salt+lake+city+solar+radiation+by+hour&source=bl&ots=RX7VpcQqv3&sig=ACfU3U3McZe0-kTMmvHvuxTfgrZF7OZLSg&hl=en&sa=X&ved=2ahUKEwjwhLmBwaX3AhXpKzQIHUo8DPkQ6AF6BAg5EAM#v=onepage&q=salt%20lake%20city%20solar%20radiation%20by%20hour&f=false
Excess power is an ongoing problem with renewable power.
One approach might be for the industries that use a lot of power - the chemical, metals, and cement industries, for example - to target their production at the times of the year when power was abundant and therefore much cheaper.
In particularly, jet fuel can be synthesized using water, carbon dioxide from the air, and power from the sun, and fuels could be stockpiled during the summer months for use during the winter months.
There are also hourly variations; here's some data from Utah that shows how the power varies by hour.
What we need is the ability to spread that out so that we can supply some of the power when the sun isn't shining, perhaps at the level of the green line. There are different technologies for this, but the most versatile one is battery storage.
12 hours worth of power at 18 megawatts is 216 megawatt hours.
The cost in 2020 of battery storage systems is about $350 per kilowatt hour, so adding batteries to support that 12 hour requirement is about $75 million.
At an installation like Red Hills, that increases the cost by about 40%.
https://www.pnnl.gov/sites/default/files/media/file/Final%20-%20ESGC%20Cost%20Performance%20Report%2012-11-2020.pdf
Onto the space based solar issues.
The effects of high-power microwave beams are not fully known.
They might interfere with satellites, either geosync satellites nearby or satellites in lower orbit.
They might impact aircraft
They might harm birds
And they might interfere with radio astronomy.
All of these would need more study.
"Beaming microwaves to the ground" sounds a lot like "orbital death ray" to me...
In reality, microwaves can't be focused tightly enough to be a useful weapon.
Let's assume a satellite that puts out 1 gigawatt of microwaves. The rectenna for that will be at least 1 kilometer across, or 785,000 square meters. That gives is 1270 watts per square meter.
Our old friend the sun is around 1000 watts per square meter.
We can reduce the power density simply by increasing the size of the rectenna.
This picture shows a relatively current map of the satellites in gosynchronous orbit.
It looks quite crowded, but the geosynch space is quite big. Under international agreement, you can launch to a location if you can guarantee that you won't interfere with the current satellites; this is where the beam effects become very important. There also may be additional regulations for a specific country.
If you want to deploy solar power satellites, you will need to develop a workable international approach to do so.
In summary, even with the rosiest assumptions, power from solar power satellites is around 3 times the cost of terrestrial solar power.
And it gets worse...
The solar industry is very hot, and there are thousands of companies competing in the production of everything that goes into a large-scale solar project - the panels, the electronics, the tracking machinery, the physical hardware, the construction process, etc.
Large scale terrestrial projects will continue to benefit from all of that innovation, while there aren't thousands of companies working on solar power satellites. They would benefit from the cheaper prices of solar cells, but most of the project cost comes from other things.
Here's a fun little chart from the US National Renewable Energy Laboratory that looks at project costs.
Red hills was completed in 2015, so we would expect that the cost would be about $1.93 per watt, which is pretty close to the $1.81 per watt that it cost.
Note what happens if you move ahead 5 years. The total cost goes down to $0.94 per watt, or less than half. So that $0.03 / kwh of power cost on a project like red hills will be cut roughly in half for the next project only 5 years later. This process will continue.
https://www.nrel.gov/docs/fy21osti/77324.pdf
To complete the summary, it's only going to get worse as terrestrial projects get cheaper.
If you enjoyed this video, please carve a fish out of a carrot
There are a number of common questions about rocket launch sites...
Why aren't there more launch sites?
Why don't we launch rockets from mountains?
Why are the launch sites in areas that flood so easily?
Why does everybody launch from Cape Canaveral?
Our subject is rocket launch sites, and specifically, what makes a good one.
The whole point of a rocket is to put a payload into a specific orbit, so we'll need to talk a bit about how orbits affect launch site selection.
If you want a lot more information on orbits, go and watch my "Space: You know orbits" video.
The first principle is that closer to the equator is better. There are two reasons for this.
The first is that the earth is rotating towards the east. At the equator the speed is 1038 miles per hour, or 1670 kilometers per hour. If we launch to the east, that rotational velocity is added to the velocity of the rocket and that means more payload to orbit with a given rocket.
As we move away from the equator, the velocity decreases. 20 degrees of latitude north or south, and the velocity is about 6% less, and it is 25% less at 40 degrees and, and 50% less at 60 degrees.
The second effect depends on the kind of orbit you are targeting.
Orbits that go directly around the equator are called equatorial. Other orbits are at an angle to the equator, and they spend some time north of the equator and some time south. The angle to the equator is known as the inclination of the orbit.
Because of the way the physics works, the inclination of the orbit you can easily reach is equal to the latitude of the launch site. Cape Canaveral is at 28 degrees north latitude, so launching directly east gives an orbit that is inclined at 28 degrees.
A higher inclination orbit can be reached from cape canaveral by aiming the rocket to the northeast, and it can also be reached by aiming the rocket to the southeast. Higher inclination orbits result in slightly reduced payloads.
For launches to Low earth orbit, it is technically possible to reduce the inclination during or after the launch, but it costs so much energy that that we generally just say it's impossible. A rocket that could lift 22,000 kilograms to a 28 degree orbit could only lift 4,000 kilograms to a 0 degree orbit.
This generally isn't an issue as low earth orbits that are low inclination aren't very useful for anything - higher inclinations are far more useful.
For orbits that are on the way to geosynchronous orbit - where zero inclination is very useful - it is possible to get there from higher latitude launch sites, but it has a energy cost and therefore a payload cost. Most of these payloads launch to an intermediate geosynchronous transfer orbit and the satellite does the rest of the work of getting into the final orbit.
If we are heading to a typical geosynchronous transfer orbit like GTO-1800, we'll say our hypothetical rocket launched from the equator will have a payload of 6500 kilograms.
Launch that same rocket from a launch site at 28 degrees, and it only lifts a payload of 5500 kilograms. It's about a 15% reduction in payload, significant but not so bad that you can't launch a geosynchronous satellite from that launch site.
Move to 51 degrees, and the payload is down to 4000 kilograms, about a 40% reduction in payload, and that's pretty bad.
Those two reasons are why launch sites near the equator are generally preferred. There are exceptions that I'll cover later.
Many launch sites are on the east coast, so that the expended first stages can land in the water.
Cape Canaveral in the US, Rocket lab LC 1 in new Zealand, and Satish Dhawan in India are all on the east coast. All three areas are lucky enough to have good east coast locations.
Europe, on the other hand, is less lucky - they don't have any convenient east coasts for spaceports that wouldn't drop used stages on other countries.
The European Space Agency therefore launches from Guiana space center in French Guiana on the north coast of south America. This location is also only 5 degrees from the equator, which makes it a great site for low inclination launches.
Russian does have an eastern shore and they do in fact have a spaceport near that point, va-stoich-ne Vostochny Cosmodrome. It's at 51 degrees north, so it has very limited access to lower inclination orbits. The Soyuz 2 rockets it launches are manufactured in Samara, more than 6400 kilometers away by rail. The first launch at Vostochny was in 2016 and it has only launched a few rockets.
The main Russian launch site is at Baikonur in Kazakhstan. It's at 46 degrees north, but launching directly to the east means that the spent rocket stages would drop on China or Mongolia, so launches from Baikonur aim a little bit to the north and that means their minimum orbital inclination is 51 degrees, the same as their new launch site at Vostochny.
The limited inclinations accessible from Baikonur is one of the reasons that the international space station orbit is inclined at 51 degrees - it's a little less efficient for the US to launch to that inclination but it's pretty impossible for Russian to launch to any lower inclination.
This does mean that Russia drops used rocket stages over their own territory, but given how hard it is to get to their eastern coast, it's an understandable decision.
China does have an eastern coast and in fact the Wenchang launch site is nicely situated at only 19 degrees of latitude.
That is, however, one of their new sites.
Their most active launch site and that one that supports crewed launches is gee-o-quan Jiuquan at 41 degrees of latitude, and that's why their space station is at 41 degrees of inclination. The spent stages from that launch site - and from other inland sites - do fall on Chinese territory, sometimes hitting towns.
If you are launching to polar orbits, you don't need an eastern coast - you can launch either to the north or to the south - and the latitude of your launch site is largely unimportant because you don't get a boost from the earth's rotation.
The Russians launch from plea-setsk Plesetsk, all the way up at 62 degrees, and the Swedes are planning on launching from keeruna Kiruna at 68 degrees. Most of the US polar launches come from Vandenberg space force base in California, where launches are made over the water to the south.
There are often some constraints on launching. Guiana space centre is nice because it is largely free of constraints, and launches can point directly east to head to low inclination orbits or north or past north to get to polar orbits.
Cape Canaveral is more restricted. It can't get higher than 57 degrees because it would fly over the eastern seaboard of the US, and if it tries to aim too far south it will fly over islands with people on them.
It's possible to fly what is known as a "dog leg" trajectory - one that is deliberately curved to avoid flying over land. This is one that SpaceX has used to launch from Cape Canaveral into a polar orbit. That path keeps them offshore Florida and puts the drone ship in a reasonable position for the first stage to land.
India has similar problems with their launch site, and they will also fly a dog leg when going to a polar trajectory.
Dog leg trajectories not surprisingly are sub optimal and reduce payload slightly.
There are many other factors that make a good launch site. It's very useful if it's not close to anything.
Cape Canaveral is well separated from the mainland, with NASA's launch complex 39 to the north, cape Canaveral space force station to the east, and Kennedy space center in the middle. The closest non-space areas are Titusville to the west - more than 10 miles from the big NASA pads - and Port Canaveral on the south.
It also helps that Cape Canaveral has been launching rockets for more than 50 years, so nobody around the area is surprised by them.
This pattern is repeated at India's Satish Dhawan; there are pads near the coast, solid rocket manufacturing in the same area, and pretty much nothing nearby.
SpaceX managed to find a pretty good site at Boca Chica on the east coast of Texas just north of Mexico.
The launch pad is near the coast, and the closest built up areas are the starship factory location and what is left of a small retirement village. There is really nothing else around.
But if we zoom out a bit, we see South padre island to the north at about 5 miles of distance and port Isabel just a little bit farther.
This is an area that is not used to rocket launches and starship is a very big rocket, so currently spacex is limited to 6 launches per year from this site.
The launch site has to be a site where you are wanted.
The left picture shows the distribution of population on the east coast, and you'll note that the east coast of Florida is population dense, which means launch sites are limited. Cape Canaveral is a set of barrier islands that were low-lying and swampy, and therefore they were available when the government was looking for land to test missiles.
Far to the north in Virginia, there is Wallops Flight Facility, also a piece of land that is away from population centers.
Camden county in Georgia attempted to create a spaceport near Woodbine, Georgia over a number of years, trying to attract NASA and more recently SpaceX, but in 2022 voters rejected the project.
The site needs to be accessible enough to get your rockets and equipment to the site.
A Falcon 9 is pretty much the maximum size rocket you can put on a truck and ship across the country. It's about 3.7 meters, or 12 feet wide. It's thin as rockets go.
Other rockets are much bigger. The Saturn V was 10 meters, or 33 feet, and the SLS is slight smaller, at 8.4 meters or 28 feet.
Both of these rockets are shipped to the launch site on a barge, and having water access is therefore a very useful thing.
Or you can decide to build the rocket in a factory right next to the launch site, as SpaceX is doing with Starship, Blue Origin is doing with New Glenn, and Rocket Lab is doing with neutron.
You do obviously need a site for the factory and a practical way to transport the rocket to the launch site if you are going to go with this approach
It needs to be permissible to build whatever launch infrastructure you need.
This picture shows the construction of Pad 39A at kennedy space center in 1965.
In those days, there were no issues with taking 160 acres or 65 hectares of wetland and filling it all in to create a launch site.
In current days, there is a huge review process to do anything in sensitive areas. SpaceX went through it with their Boca Chica launch site. Rocket Lab went through it for their launch site in New Zealand. It generally takes many years of work and a lot of money to undertake such a project, and approval is not guaranteed.
If you want more detail on what is actually involved in this sort of environmental review, I've linked to the Boca Chica review documents in the video description:
https://www.faa.gov/space/environmental/nepa_docs/spacex_texas_eis
In most cases, it's far more straightforward to find a spot at an existing launch location, where the previous construction has paved the way for new use and the surrounding population is used to the impact of rocket launch.
These locations are of course desirable for many launch companies, with sites at Cape Canaveral difficult to get and sites at Vandenberg air force base in California even harder to get.
If you want to understand more about the history of cape Canaveral and what companies are currently operating there, see my video "Launches at Cape Canaveral - then and now"
Your rocket is going to need consumables - liquid oxygen and a fuel such as liquid hydrogen, liquid methane, or the refined kerosene known as RP-1.
You need suppliers for all of these products in the quantities that you require. Being near a site that already has that infrastructure makes it cheaper - for example, Air Liquide already has a liquid oxygen production facility near Kennedy Space Center.
And now, you know launch sites...
If you enjoyed this video, you might enjoy the atlas of space rocket launch sites.
There are numerous orbits for satellites - enough that the topic can get confusing.
That's the point of this video - to get rid of that confusion.
First we need to understand what it means to be in orbit. There's a classic thought experiment that works well.
If we have a body floating in space and give it a kick to the side, it will move to the side but the earth's gravity will soon pull it down so it crashes into the surface.
Give it a harder kick, and it goes farther around the earth before it hits the surface.
Give it just the right kick, and the satellite will go fast enough so that it falls towards the earth at the same rate the earth curves away from it. That is what it means to be in orbit
Here's a bit of math for those who like that sort of stuff.
The required velocity to stay in orbit is given by the following equation, where G is the gravitational constant, mass is the mass of the earth, and R is the radius of the orbit measured from the center of the earth.
The approach is a bit inconvenient because we measure orbits by how far they are above the surface, so there's an alternate version that uses the orbital height and adds in the radius of the earth, about 6.3 million meters.
We can chart the speed at various altitudes.
Way down at 400 kilometers, the orbital speed is around 17,200 miles per hour, or 26,800 kilometers per hour.
Out at 35,000 kilometers, it's 6900 miles per hour or 11,200 kilometers per hour - less than half of the velocity than it was at 400 kilometers.
The velocity goes down because the earth's gravity is reduced as we move away from it.
This graph demonstrates how difficult it is to get into various orbits...
Getting into a 400 kilometer orbit takes about 9400 meters per second of what is called delta v - the amount of work the rocket has to do. The Falcon 9 can carry around 17,000 kilograms in reusable mode to 400 kilometers.
Getting to higher orbits is much, much harder. Getting to 35000 kilometers takes over 13,000 meters per second of delta V, and that limits Falcon 9 to perhaps 4000 kilograms for that orbit.
Now we can dive into the details of an orbit.
There are three main factors that define an orbit - the altitude of an orbit, the inclination of an orbit, and the eccentricity of an orbit.
Altitude is the first factor
We'll start with low earth orbit. The earth is about 12,700 kilometers in diameter, so a low earth orbit of less than 2000 kilometers is pretty close to the earth's surface.
Low earth orbit is above most of the atmosphere but there is still enough atmosphere to slow satellites down, so the space station needs to be reboosted periodically to keep it from reentering the atmosphere.
This is also the orbital range where the hubble space telescope lives, at 560 kilometers, and the SpaceX starlink satellites, at 550 kilometers, 1150 kilometers, and 340 kilometers.
Medium earth orbit is a bit weird. It runs from the top of 2000 km all the way out to a very specific distance of 35,786 kilometers. More about that number in a minute...
Probably the most well-known satellites in this orbit are navigation satellites, with the US GPS system, the Russian Glonass system, and the European galileo system. All orbit at around 20,000 kilometers.
Geosynchronous orbit is at 35,786 kilometers of altitude, and has an orbital speed of 11,067 kilometers per hour or 6876 miles per hour.
At this altitude, the speed at which the satellite orbits is exactly the speed that the earth turns, which gives an orbital period of 24 hours, which is the definition of a geosynchronous orbit.
This is very convenient for communications satellites, and it turns out that there are over 500 of them in this orbit.
High earth orbit is geosynchronous orbit and beyond.
Two satellites in distant earth orbits are the interstellar boundary explorer, which looks at the interaction between the solar system and interstellar space, and the transiting exoplanet survey satellite - or TESS - searches for exoplanets.
Being in a distant earth orbit keeps the satellite close enough that it is easily interacted with but far enough away so the earth will not degrade the science.
The second orbital factor is inclination.
Inclination describes how tilted the orbit is compared to the earth's axis.
An orbit that goes directly around the equator of the earth is not surprisingly known as an equatorial orbit, and an orbit that goes over the earth's poles is known a polar orbit.
We would say that an equatorial orbit has an inclination of zero degrees, and a polar orbit has an inclination of 90 degrees.
There are inclinations between those two orbits - you might choose to launch to an inclination of 51.6 degrees. There are two big factors that drive the choice of inclination.
The first is ground track
Ground track simply defines the path that the satellite travels over the earth's surface.
Let's say that we want to provide TV service over North America. We need a fixed position so that our customers know where to point their satellite dishes, so we choose a geosynchronous orbit that has zero inclination - otherwise known as a geostationary orbit. The satellite orbits at the same rate that the earth turns, so from our ground perspective, it looks like it's just parked over a spot on the equator.
A company like Direct TV uses multiple geostationary satellites to serve different parts of north America.
The geostationary orbit is a long way away from the earth, so the signal that it generates can reach most latitudes on the earth except for the poles, though the farther from the equator on earth, the more the dish needs to be tilted and the greater the likelihood there is an obstruction in the way. This prevents them from providing effective service at high latitudes.
The location of these satellites makes them poorly suited for earth observation - they only see one side of the earth and only from a long distance.
For earth observation, we want an orbit that covers most of the earth's surface, and that is a polar orbit.
The earth slowly spins underneath the satellite, and each orbit covers a different patch of the earth's surface.
That leads to a ground track that looks like this.
Both the animation and this ground track were created by Jake Low's excellent ground track visualizer.
I said earlier that polar orbits were 90 degrees, but technically any orbit from 60 degrees to 120 degrees is a polar orbit.
The majority of earth observation satellites fly a very specific orbit known as a
The majority of earth observation satellites fly a very specific polar orbit known as a "sun synchronous orbit".
Wikipedia will be happy to share with you all the complex details and math, but the basic concept is that the sun synchronous orbit is set up so that the orbit rotates - the technical term is "precesses" - about 1 degree eastward every day. That keeps the orbit synchronized with the sun so the satellite passes over any given point of the surface at the same local mean solar time.
This gives consistent lighting and that is hugely useful for any satellite that does imaging.
The most popular sun synchronous orbit is at 567 kilometers and an inclination of 98 degrees. It covers most of the earth's surface, provides 15 orbits per day, and high enough that the satellite will last in orbit for a long time.
SpaceX's starlink uses a variety of inclinations, with satellites at 42 degrees, 48 degrees, 53 degrees, 70 degrees, and 97.6 degrees. They use this mix of inclinations to provide more satellite coverage where needed.
The second important factor for inclination has to do with the accessibility of orbits from a specific launch site.
Cape Canaveral in Florida has a latitude of 28.5 degrees north, and because of the way the physics works, that means the "natural" inclination for the launch site is 28.5 degrees of inclination.
It *is* possible to get to lower inclinations, but it takes quite a bit of energy to do so and that means less payload to orbit. This is especially true for low earth orbit because the orbital speeds are so much higher.
It's costly enough that it's common to say that it's not possible to launch to an inclination lower than the launch site's.
It is possible, however, to reach higher inclinations fairly easily simply by aiming to the north or south during launch. There is a reduction in payload to those orbits but it is much cheaper than launching to a lower inclination.
The accessibility of orbits has some significant impacts on international cooperation.
NASA launches from Cape Canaveral at 28.5 degrees, while the Russians launch from Baikonur in Kazakhstan at 51.6 degrees.
To make the international space station possible, it was placed at 51.6 degrees so that it was possible for both Russia and the United States to reach it. That inclination also has the advantage of covering more of the earth's surface.
China launches the taikonauts from Juiquan satellite center to an orbit at 41.5 degrees. There have been suggestions that Russia might stop collaborating with the ISS partners and start working with the Chinese, but because of the difference in inclinations, the Russians cannot reach the Chinese space station with Soyuz from Baikonur.
The European space agency launches from Guiana Space Center in French Guiana at an inclination of only 5 degrees. This is a very desirable location because it is so close to the equator, though dealing with the jungle conditions is challenging.
Because I know people will ask, SpaceX's boca chica launch site in Texas is at 26 degrees north. Vandenberg space force base in California is at 35 degrees north, but it is generally used for polar launches to the south where the natural latitude is much less important.
The final orbital factor is eccentricity
Let's look at launching a satellite to geostationary orbit.
We start by launching to a very elliptical orbit, where the high point or apogee is all the way out at 36,000 kilometers and the lowest point or perigee is all the way down at 180 kilometers. This is known as a geosynchronous transfer orbit. The orbit is inclined at the latitude of our launch site.
We need to do two things to get to our final orbit.
First, we need to circularize our orbit - to raise the perigee so that it is the same as the apogee.
Second, we need to get rid of whatever inclination we got from our launch site.
We do this by thrusting in the direction of our orbit at the high point of the orbit, with a little correction to reduce the inclination. We do this at the high point of the orbit because that is the point where we are travelling the slowest and therefore it is the point where the cost of the inclination change is the smallest.
That puts us into a less inclined orbit with a higher perigee.
We will repeat that several times until we get to a circular orbit with no inclination.
This was the typical approach used by satellites.
Another technique is to use a really big rocket motor - typically the second stage motor itself - to make the required changes using a single burn.
The Department of Defense launches that go directly to geosynchronous orbit use this technique.
Most modern satellites use electric thrusters that are very efficient but have very low thrust.
Because of the low thrust, the orbit cannot be adjusted at the apogee; instead there is a complex approach where the thrust direction is changed throughout the orbit to achieve the desired orbit efficiently. It takes many orbits and months of time to complete this maneuver.
There are some satellites that are deliberately placed into a highly eccentric orbit. The Soviets and Russians have used an orbit known as Molniya for communications satellites since the mid 1960s as it provides good coverage at the high latitudes near the poles where geostationary satellites do not work well.
The orbit is inclined at 63.4 degrees. The height varies between missions, but a typical molniya orbit is very eccentric, with a perigee of 600 kilometers and an apogee all the way out at 39,700 kilometers.
The satellite travels much slower during the high part of its orbit and therefore spends most of its time over one hemisphere. Three satellites in this orbit can therefore provide continuous coverage.
The ground track of the molniya orbit is a bit weird. The satellite orbital period is 12 hours, so the ground track just alternates between two locations. Looking at the two locations, you can understand why both the soviets/Russians and the US are fond of this orbit for reconnaissance satellites.
One of the downsides of the molniya orbit is that the satellite will need to be designed to deal with the high radiation of the Van Allen belts every orbit.
The near rectilinear halo orbit used on NASA's Artemis missions to the moon uses an highly eccentric orbit that is similar in concept to the Molniya orbit.
It has a closest point - known as perilune rather than perigee because the orbit is around the moon - of 1600 km, and a apolune of 68,260 km. This is a big orbit and the moon doesn't have a lot of gravity, so the orbital period is 6.5 days. That gives good coverage of the Artemis landing sites at the south lunar pole, though for continuous coverage multiple satellites would be needed.
Probably the strangest orbit is the Tundra orbit.
It is much less elliptical than the molniya orbit, with a perigee of 25,000 kilometers and an apogee of 39,700 kilometers. This keeps it out of most of the radiation from the Van Allen Belts. It shares the 63.4 degree inclination that molniya uses.
You'll notice that apogee is a little bit higher than the apogee of a circular geosynchronous orbit, and the perigee is a bit lower.
You therefore shouldn't be surprised to find out that Tundra is a geosynchronous orbit - the orbital period is 24 hours. But it's not geostationary as the orbit is both highly inclined and eccentric.
That gives it a ground track that is frankly bizarre - a teardrop shape with a little loop in the high latitudes. Sirius Satellite radio used three satellites in this orbit to provide service to North America - at any time, two satellites were above the and they could therefore provide continuous service.
Tundra is generally a better choice than Molniya but it takes more energy to get there.
And that's all I wanted to cover, and now, you know orbits...
If you enjoyed this video, please explain the relevance of this image
I was talking about space capsules with a friend of mine recently, and we got onto the topic of parachutes.
He asked me an interesting question.
Isn't that a solved problem?
Parachutes have been used in human spaceflight for more than 50 years. Haven't we figured out how to make them work?
It's a fair question. After all, Jacques Garnerin built and tested this design back in 1797, so one would think that it would be figured out by now.
But it turns out that parachutes are still cranky beasts that are not well understood. Especially when used in spacecraft.
We'll start by looking at the parachute system used by the Orion capsule, first designed during the Constellation program in 2005 and recently flown on the Artemis 1 mission.
The parachute's role starts at about 26000 feet of altitude with the capsule travelling at 325 miles per hour. That's 8 kilometers and 523 kilometers per hour.
The first problem is that none of the parachute hardware is exposed - it's protected during flight by the forward bay cover. We need to get rid of the cover. Unfortunately, the airflow behind the capsule can be quite chaotic, as you can see from this simulation. If we just release the cover, it could get stuck or come back and collide with the capsule.
We need something that will actively pull the cover away from the capsule.
We can do it with parachutes, but we'll need to get the parachutes away from the capsule enough so that they will work reliably.
We need something like a party popper, where we use a small pyrotechnic device to generate a lot of gas and use that gas to propel our payload - in this case a parachute - away from the capsule.
The big version we need is known as a parachute mortar
On the left we have an electrically triggered pyrotechnic device that can produce a lot of hot gas.
That gas travels through a plenum - or open chamber.
It exerts pressure against the sabot. That's a term from military mortars - in this case, I think "piston" is probably a better word.
The pressure pushes against the parachute pack. Because space is very limited in capsules, the parachute is very tightly compressed in the mortar, to roughly the density of wood.
Once the pressure on the parachute pack is high enough, it will blow the lid off the end and the parachute will deploy.
Parachute mortars are sophisticated devices - they need to use the right amount of pressure but not too much and the parachutes have to be carefully packed so they deploy correctly.
All of this needs to be tested. Like this:
Mortars deploy the parachute far away from the capsule. They significantly increase the chance that a parachute will deploy and won't get tangled with other parachutes.
Three mortars deploy small chutes connected to the front bay cover of the capsule and the pull of the chutes plus small thrusters pull the cover away from the capsule.
This design was tested on the ground and ultimately tested in airdrops.
Next, we will deploy two 23' parachutes known as drogues. They stabilize the capsule and begin the process of slowing it down. Like the front bay cover parachutes, they are also deployed by mortars, shown here in green and yellow.
There's a problem, however. If we toss these chutes out of a capsule that is going very fast, they will quickly inflate and tear themselves apart.
Which would be bad.
This is solved by using a technique known as reefing. A reefing line is wrapped around the bottom part of the parachute and it keeps the parachute from opening fully and therefore limits the amount of drag produced and the shock on the system.
You can think of reefing as a way to make a parachute temporarily function as a smaller parachute. After a period of time, the reefing line is released and the parachute will open fully.
This is a diagram from the Gemini program showing the reefed and disreefed - or open - appearance of the parachute.
The Orion capsule drogues use two-stage reefing. They start in a fully reefed condition, then progress to a partially open state, and then finally to a fully open state.
Reefing of course adds complexity to parachute design and testing.
Each drogue parachute is attached to the capsule at three points. The lines between each parachute and those points need to be routed in channels designed in the capsule so that they are protected from damage and can deploy successfully.
The left drogue harness is shown in yellow, and the right drogue harness is shown in red.
The drogue chutes have done their job, but we need to be sure they don't get tangled up with the main chutes, so we need to get rid of them. This is done with a line cutter. The line goes through this end of the cutter.
The line will ultimately be cut with the blade, but we really don't want that to happen ahead of time, so the piston is held in place by two shear pins that keep the blade away from the line.
When we want to cut the line, an electrical signal is sent to the pyrotechnic cartridges - two of them in case one fails - and they generate a large amount of gas, exerting enough pressure on the piston to break the shear pins and fire the blade into the anvil, cutting the line.
Once that happens, the capsule is free falling again.
We are finally ready to deploy the main parachutes.
The mains are really big - the opened parachute is 116' in diameter and weighs 310 lbs. The mortars that we used for the much smaller drogues were about as big as we could fit in the capsule, so we need a different approach...
The mains are packed in parachute bags that are specially shaped to fit in the available space. This is not a new concept - military cargo parachutes are packed into deployment bags for the same purpose - but the capsule ones need to be packed much more tightly.
The bags fit into the capsule, but we need a way to deploy them quickly.
We do this with *another* set of parachutes known as pilot chutes. At 11' across and 11 lbs, they are tiny, but when they are deployed, they pull the main chute bags away from the vehicle and the chutes stream out of the bags.
The main chutes also go through two stages of reefing before they open fully. After landing, a line cutter detaches the parachutes from the capsule so it will stay upright even in high winds or seas.
Here's a nice video showing main parachute deployment using a pilot parachute, this time from a Starliner test.
This is a diagram of the parachute components used in Orion, looking down from the top of the capsule. At the bottom you can see the two drogue mortars with the chutes inside of them. You can see the three pilot chute mortars next to the main chutes they attach to, and finally the large main chutes packed in their bags.
The width of the parachute section is about 3 meters at the bottom and gets smaller toward the top of the capsule.
Let's count up what it takes to get Orion safely back to the earth's surface.
That's three chutes and mortars to remove the cover, two drogues and mortars, and three pilots and mortars to deploy the three main chutes. 8 mortars and 11 chutes in total.
How does that compare to other capsules?
Apollo used the same scheme, except that it only used one chute and mortar to remove the cover. 6 mortars and 9 chutes.
Boeing's starliner also uses this approach, and uses two chutes and mortars to remove the cover. 7 mortars and 10 chutes total.
SpaceX has - predictably - done something different with crew dragon.
Crew Dragon stores its parachutes in compartments shown in green and yellow rather than under a removable cover, so there is no need for cover removal chutes.
They do have two drogue parachutes deployed with mortars. Rather than cutting the drogues free and then using pilot parachutes to deploy the mains, Crew Dragon uses the drogues to deploy the mains, so it does not need pilot chutes. Initially, crew dragon had 3 main chutes but a fourth was added during testing. That means SpaceX uses 6 parachutes instead of the 9-11 used with the other capsules and only two mortars.
This video shows the crew dragon under the drogue chutes, then extracting the main chutes. We can clearly see the main chutes start fully reefed and then open in two steps.
The Russian Soyuz is also different. It uses two small pilot parachutes to deploy a single braking parachute, which I suspect is similar to a drogue. That drogue then deploys a single large main parachute.
This system looks to be lacking in redundancy, but there is a full backup system that can take over if the main system fails. The backup main parachute is smaller and results in a slightly higher landing velocity than the main system.
These are complex systems, and while it is possible to do some analysis on them, they all rely on testing.
Lots of testing - the Orion parachute test program ran for 10 years.
The tests are used to find issues that need to be addressed, but just as importantly, they give empirical - or real-world - data on the reliability of the parachute systems.
To explore this I'd like to introduce you to my high power rocket, "Spot", which uses a parachute for recovery.
Let's say that I launched it 10,000 times, and the parachute worked 9675 times and failed 325 times. With those numbers, we'd feel pretty confident that it was 97% reliable because we have a lot of empirical data.
Let's reduce the number of launches to 100. We get 97 successes and 3 failures. Looks like 97% reliable, but we are less confident in the results, because we have less data.
Now we drop it down to only 10 launches. We get 10 successes and no failures, which means it is 100% reliable by calculation.
We know that number is wrong because nothing is 100% reliable. But we would like to be able to say something useful about the reliability of our parachute system other than "it hasn't failed yet".
There is an engineering approach that lets us do that, and its basically about quantifying the chance that we are just getting lucky in our testing.
We fly Spot 4 times and the system works every time. We can figure out mathematically that there is a 90% chance that the reliability of the parachute is 50% or greater, and a 10% chance that we got lucky and the reliability is less than 50%.
We aren't super excited by that - 50% reliability doesn't sound very good - but we know more than just saying "it worked 4 times", and - importantly - it helps remind us that we might have just gotten lucky.
As we do more testing, our estimate of reliability goes up. With 11 successes, we are 90% confident that our reliability is at least 80%, and if we go all the way to 45 successes, we are 90% confident that our reliability is at least 95%.
But there's still a 10% chance we could be wrong, and the actual reliability could be lower.
What if we want to be more confident?
We can calculate another curve, a curve that shows what we can determine if we want to be 99% sure of our estimate.
On this curve, we need 24 successes to be 99% sure we have 80 percent reliability, 47 for 90 percent reliability, and 89 tests for 95 percent reliability.
In this sort of situation, you need to test a lot to get even modest amounts of information on reliability. And the whole thing seems to be counter-intuitive - a system that works 89 times in a row seems extremely reliable, but we actually can't conclude that - we could still be getting lucky.
Is there a better approach than just running a bunch of tests?
And that takes us to a technique known as probabilistic risk assessment. If you want the details, see my video on reliability and crew safety at NASA titled "you're going to kill astronauts".
The basic idea is that you do a detailed analysis of all the components of your system, figure out how reliable every part of it is, figure out how individual failures could cause a major failure, do a bunch of math, and out pops some numbers. It takes a lot of time and money to do it well.
In NASA's analysis of the Orion parachute system, they found that the most likely value is that the system will fail in 1 out of 9,502 flights. They also note that they are 95% confident that it is at least as good as one failure in 2397 flights.
NASA has shared some of the detailed information that went into their final calculations.
I've added some red dots. Every red dot indicates a probability of either 1 in a million or one in 10 million. Those are suspiciously round numbers, and they are round numbers because NASA simply doesn't have a better estimate for the probability, so they have come up with an educated guess. Many of those guesses are probably pretty good, some of them are likely to be significantly wrong. Once again, we've run up against the issue of limited knowledge - we simply don't have enough information to do an analysis that would substitute for testing, so we need to rely on testing.
There's another topic that popped up here. There are abort scenarios where the parachute system needs to work, and the requirements there are different than a nominal reentry.
For Orion, a low altitude or pad abort totally skips the drogue deployment part of the sequence.
Crew Dragon still uses the same sequence, which gives me a great opportunity to show the crew dragon pad abort test.
SpaceX nicely shared an on-board view of their pad abort test for crew dragon.
In this first view, crew dragon has detached the trunk section so the parachutes can deploy
It also gives us a beautiful view. At the center bottom is space launch complex 40 on cape Canaveral space force station where SpaceX launches Falcon 9, and on the right is space launch complex 41 used by ULA to launch Atlas V and - soon - Vulcan rockets. In the haze of the upper right you can see pads 39A and 39B at Kennedy space center, and the vehicle assembly building is towards the middle of the picture.
Back to parachutes. It's time for the drogues to deploy.
In this view, you can see the two lines that run from the drogue parachutes to the capsule. The three lines running to the side connect to the three main chute bags - this test was before the fourth main chute was added. They are slack, but when the drogue lines are disconnected from the capsule, they will go taut and pull out the main chute bags.
Here we see the main chute bags being pulled away from the capsule by the drogues
The mains will deploy from those bags. If you watch carefully, you can see that they are initially reefed.
The mains have fully inflated, and our work is done.
Orion testing defined 4 different areas of testing required, based on the different altitudes and speeds of the capsule.
There is the nominal capsule return from orbit, where the parachute deployment is at 25,000' at a speed of about 0.45 mach, or ...
There is the pad abort below 10,000' at a much slower speed, and then there are aborts during the launch the occur at higher altitudes and speed.
NASA built a capsule-shaped parachute test vehicle to perform the tests.
They used a C-130 aircraft to carry the test vehicle aloft for many tests, but the C-130 can't fly high enough for all tests, so they used a balloon that would carry the capsule to 45,000'.
Unfortunately, the capsule-shaped vehicle has too much drag to reach higher speeds, so NASA also build a dart-shaped test vehicle that could reach the required speeds.
Two test vehicles and two drop vehicles gives four different test approaches.
Going back to our chart, we can see that the capsule and plane can cover aborts from the pad and early in the flight.
Dropping the dart from the plane almost gives us coverage of the nominal case, but not full coverage...
The balloon and the capsule give coverage of the rest of the abort scenarios...
And the balloon and dart will fully cover the nominal case.
As you can see, the complexity of abort scenarios requires a lot of additional testing.
Parachute failures are rare but they do happen.
The only fatality was on the Soyuz 1 flight in 1967, when cosmonaut Vladimir Komarov was killed.
Apollo 15 lost on main chute during deployment but landed safely with the two remaining chutes.
During the Ares 1X flight test, there was a parachute failure during the recovery of the solid rocket first stage
(video)
The failure was traced to a design flaw that led to one chute being deployed in a disreefed configuration. That produced forces that were too strong for the parachute to handle, and it failed.
And that's why space parachutes are so cranky and require so much testing.
And now, you know parachutes.
If you enjoyed this video, please buy my poster, "Capsule Parachute Testing: A Portrait in Four Colors"
Looking at rocket engines, the first thing we notice is all of the complex plumbing at the top of the engine - usually known as the "powerhead".
But the nozzle at the bottom is equally important in building an efficient engine, and that's our topic for today.
Let's start with combustion in the open air. All of the particles that comes from that combustion go in a different direction, and there's no net force. Not a very good rocket engine.
If we wrap that combustion in a chamber with a hole on one side, we force all the particles to come out mostly in the direction that we want them to come out, and that gives us a net force in the direction that we want. It's better but we're wasting a lot of energy with particles going to the side.
What we need is some sort of device to get all the exhaust going in the direction we want, and that device is a nozzle.
This design is known as a "bell nozzle" because it's shaped like a bell, and it's the most common design for liquid-fueled engines.
It's technically known as a "de Level" (day le val) nozzle, invented in 1862 by Swedish engineer and inventor Gustaf de Laval (day le val) for use in steam engines.
The nozzle converges into a tighter throat and then diverges out as the nozzle gets larger, so it's also known as a "converging diverging nozzle.
There's something more complex going on that I'd like to touch on briefly in a simplified form as the physics is fairly complex.
In the combustion chamber we have the highest temperatures and pressures, which makes sense, as that is where the combustion is taking place. The velocity is what is termed as "low" which means less than the speed of sound.
The constriction of the throat results in a reduced temperature, a reduced pressure, and an increase in velocity, up to the speed of sound.
After that, the expanding nature of the nozzle results in a drop in temperature and pressure and an increase in velocity.
Here's a nice little graph from Wikipedia that shows how the pressure and temperature go down and the velocity goes up a lot.
This velocity increase is one of the main jobs of a nozzle - the faster our exhaust velocity, the less fuel our engine uses to generate a specific amount of thrust.
And therefore, the more you can expand the exhaust, the higher the velocity, and the better your engine works.
Let's look at some engines.
This is the Rutherford vacuum engine used on Rocket Lab's Electron rocket. You can see that the throat of nozzle is really tiny, and the exit of the nozzle is really big. We can divide the area of the nozzle exit by the area of throat to calculate the expansion ratio, sometimes called the area ratio. Rocket lab doesn't publish the expansion ratio, so I've estimated it at around 100 to 1.
The Merlin 1D vacuum used on SpaceX's Falcon 9 second stage has an expansion ratio of 165:1.
The Raptor vacuum is slightly less at 150 to 1.
And finally, we have the RL-10, used as the upper stage on the Atlas V from ULA and on the SLS upper stages. It has an expansion ratio of 215 to 1.
All of these engines use high expansion ratios because that will give a high exhaust velocity and therefore a high efficiency.
Expansion ratios can be a bit misleading, because what we really care about is the actual pressure at the nozzle exit, and that depends on the chamber pressure, the expansion ratio, and other factors. Calculating the pressure is left as an exercise for the student.
I do think it's interesting to look at the chamber pressures to get a relative sense of the impact.
Raptor probably runs at around 300 bar, or 300 times atmospheric pressure, which is ridiculously high. Even with a big nozzle the exit pressure is higher than you might expect.
Merlin 1D and Rutherford run at around 100 bar.
The RL-10 runs at a very low chamber pressure of only 44 bar, so that combined with the high expansion ratio yields great efficiency. And there's a version of the RL-10 with a two-part nozzle that runs all the way up at 285 to 1, for even more efficiency.
The obvious question is "why don't the other three have bigger nozzles?", and the answer is "weight, packaging, and price". Bigger nozzles need to fit inside the rocket stage and not be too long, so it's a tradeoff. And price is something you would like to minimize - the RL-10 is a great engine but it will cost you more than $15 million a copy.
Generally with vacuum engines you fly the biggest nozzle that you can make fit.
Booster engines - the engines used on the first stages of rockets - have a different set of constraints.
Not only do we want a low pressure at the exit, we want to want to keep that nice ordered flow after the nozzle. There's a simple condition that we need to keep that happening - we need to set the exit pressure of the nozzle to be equal to the atmospheric pressure outside the nozzle.
And here's the problem.
Our booster engines start at full atmospheric pressure, at 1 bar. A minute later they are at 7 kilometers of altitude and pressure is down to 0.4 bar, and after two minutes they are 33 kilometers up and pretty much in a vacuum. There's simply no way to build a bell nozzle that is optimized at those different pressures.
We learned that vacuum nozzles were the most efficient, so lets see what happens if we choose a vacuum nozzle for our booster.
Let's say we're running our vacuum nozzle at sea level. It has a low pressure, let's just say it's 0.4 bar. The atmospheric pressure outside the nozzle is 1 bar, and that high pressure allows a reverse flow of air to creep up inside the nozzle.
The term for the situation is "overexpanded" - the nozzle expands the exhaust more that the optimal amount to match the outside air pressure.
Here's a simulation of the what happens. First, you can see that the smooth flow is detached from the wall of the nozzle - this is known as flow separation. It results in chaotic shock waves inside the nozzle.
This is a serious case of flow separation that will break pretty much any engine.
The asymmetry in the nozzle causes side forces to be created in the nozzle, and those forces cause the nozzle to deform. Here's a view of the forces on the nozzle from NASA experiments.
But it would be really nice to have a more visual representation.
Luckily we have the space shuttle main engine or RS-25.
What isn't commonly recognized is that all booster engines are prone to flow separation at engine startup because the pressure is always low during startup - they always start up overexpanded. This is one reason it is hard to start up engines smoothly.
In this video, concentrate on the inside of the nozzle. The bright lines that you see show the areas in the nozzle where the flow is separating. It's very chaotic and the flow separation doesn't go away until the engine is at full power.
The second video is the same sequence at a higher rate. Watch the rim of the exhaust nozzle during this sequence. The side forces from the flow separation are deforming the nozzle bell and it is "ringing" until the thrust is high enough that the flow separation goes away.
Flow separation limits how big of a nozzle we can use at sea level.
If we have a nozzle optimized for sea-level performance running at low pressure, we have the opposite problem
Our nozzle produces exhaust at a pressure of 0.7 bar, but the atmospheric pressure is much lower, at only 0.1 bar.
When our exhaust flow hits the edge of the nozzle, it encounters that low pressure and it immediately wants to spread out to the side rather than continuing in the direction that we want it to go.
This is known as "underexpanded" - our nozzle is too small and therefore our exit pressure is too high.
Other than the loss of efficiency, underexpanded operation does not cause problematic issues the way overexpanded can lead to flow separation.
We can see the effect of different atmospheric pressures very clearly by looking at a launch.
The Falcon 9 burns a refined version of kerosene known as RP-1 and that makes its exhaust plume very visible. Let's look at a launch.
(narrate)
There's no way to build a bell nozzle that works optimally for a booster. But it turns out that nozzle efficiency isn't the most important factor.
There is, however, another big concern that drives the size of the nozzles on booster engines.
This is the bottom of a Falcon 9 booster stage. Note that the engines are pretty tightly packed.
This red circle is the size of the Merlin Vacuum nozzle, and it's pretty clear that there is only room for one engine.
For boosters, you need high thrust to be able to get off the pad, and the small nozzles on the Merlin 1D engine make it possible to fit 9 engines in a small diameter.
You see this same decision at work on the starship super heavy booster - the engines are sized to fit 33 of them underneath the booster.
And lest you think this is just a spaceX thing, these are the F-1 engines on the base of the Saturn V booster. The nozzles are just a little bit too big to fit inside the diameter of the first stage so there are small fairings around each of the outside engines.
Total thrust matters much more than efficiency for boosters.
And that brings us to a very weird engine, the RS-25 Space Shuttle Main Engine. It's a weird engine because the shuttle was a weird vehicle, one that starts its engines at sea level and runs them all the way to orbit.
The engine needs to work efficiently in vacuum because most of the launch is in vacuum, which means it needs a big nozzle but that nozzle also needs to work at sea level.
The target pressure for the main exhaust was about 0.15 bar, low enough to give decent performance in vacuum. But that was too low to prevent flow separation.
The fix was to slightly change the nozzle right near the exit, changing the rim so that it curved back in. That produced a small layer of exhaust at about 0.35 bar, and that was just enough to make the engine avoid flow separation.
But it's not all great.
The movement that we saw on the videos caused damage to the nozzles, and they were only good for 12-15 flights before they needed to be replaced at a cost of $ 7 million in 1999 dollars. NASA did have a plan to use a different design with more longevity but never got around to building it.
Here's a cool nozzle-related phenomena
These are known as "shock diamonds"
It's commonly believed that this is a sign of a powerful or efficient engine, but it's actually just a sign that that the exhaust pressure doesn't match the outside air pressure, which we know is very common with booster engines at sea level.
There's a lot of fun physics you can dig into if you want, but it's hard to deny that it's a pretty phenomena.
Shock diamonds show up in jet engine exhausts for the same reasons - jet nozzles are designed to be most efficient at a higher altitude than sea level.
There's an interesting alternative to the Bell Nozzle...
You start with a shape that is a cut-off spike
Then you add in a combustion chambers in a ring around the shape so that the exhaust gases go against the curve, confining them on that side.
If we run this engine at sea level, we have 1 bar of air pressure pushing against that stream of exhaust gases, and that keeps it next to the curve and - more importantly - allows it to expand until the exhaust gases hit that 1 bar of pressure.
That's the pressure we're looking for to get the most efficiency.
Here's a diagram with some more detail. This design is known as an "aerospike", and at sea level it will direct the airflow in the direction we want.
As our rocket gains altitude, the outside air pressure drops. That allows the exhaust plume to expand to equal that air pressure, once again giving us exactly what we want in terms of a nozzle.
This is the "design altitude" of the aerospike, giving us optimal efficiency.
In vacuum, we end up with no air pressure to help shape the expansion. Some of the shock waves coming off the central area help keep the efficiency good.
There's an interesting variant known as the linear aerospike. This particular engine is the XRS-2200 build by Rocketdyne in 1999 as the engine for the X-33, prototype for the VentureStar single stage to orbit craft.
Here's a video.
You probably noticed that there are a series of small combustion chambers next to the central wedge, 20 of them on each side.
Aerospike engines sound like a great idea, and they are media darlings - you can find numerous breathless articles saying how revolutionary they are.
What's the reality?
If we look into the details of aerospikes in rocketry, here's what we find.
Quite a few designs
A few prototypes
No production versions
That is a general sign that something is not as great as the advocates think it is.
Let's look at three engines to see if we can garner some useful information...
The RS-25 flew on the space shuttle and now flies on SLS.
The RS-68 flies on the Delta IV
The XRS-2200 was built for the X-33 project.
All of these engines burn liquid hydrogen and liquid oxygen.
We'll start with thrust. The RS-25 puts out 1.9 meganewtons of thrust at sea level and 2.3 meganewtons in vacuum. The RS-68 is a big brute of an engine, with 3.1 meganewtons at sea level and 3.6 in vacuume. The XRS-2200 comes in at 0.9 meganewtons at sea level and 1.2 in vacuum. It's about half the thrust of the RS-25 and a third of the RS-68.
Looking at efficiency, the specific impulse of the RS-25 is a very nice 366 at sea level and 452 in vacuum. The RS-68 gets 365 at sea level and only 410 in vacuum. The RS-68 uses a gas-generator design rather than the advanced staged combustion design used by the RS-25, and it's also designed as a booster engine with an expansion ratio of only 22 to 1, so we would expect the RS-25 to be better in vacuum. The XRS-2200 gets 339 at sea level and 437 in vacuum.
This is frankly disappointing - the whole point of going with an aerospike is to get better efficiency and the XRS-2200 is roughly equal to the RS-68 and considerably behind the RS-25.
Looking at mass, the RS-25 is 3200 kilograms, the RS-68 is the big beast and 6700 kilograms, and the XRS-2200 is.. Unknown.
Try as I might, I couldn't find any mention of the mass of the XRS-2200. We can't really conclude anything from negative data, and you can draw your own conclusions about what the lack of data might mean.
Finally, if we look at the size of the engine, the RS-25 is 2.4 meters in diameter and interestingly, the RS-68 is also 2.4 meters in diameter. The XRS-2200 is rectangular and a bit larger in area than the other two.
What have we learned? Well, the XRS-2200 isn't very exciting. It doesn't put out very much thrust for the space it takes up and the high efficiency that we were hoping for doesn't appear to be present.
There was a larger RS-2200 planned that would supposedly produce similar performance to the RS-25 but come in a larger package. It of course was never built so it's not really a fair comparison to these engines.
And that's the story of aerospikes. They seem like a great idea but nobody has every built a production aerospike engine, and there's probably a reason for that.
We're now going to switch gears and look at nozzle cooling.
Our combustion chamber presents a bit of a problem. The combustion temperature is about 3500 kelvin, which is very very hot, and we need a way to keep the metal parts from getting all melty.
To deal with the heat, we're going to build our nozzle so there is an inner and outer wall and then we will flow one of our propellants between the two walls to carry the heat away. That hot propellant is then fed into the combustion chamber.
This is known as regenerative cooling and is a very common technique.
There are many different ways of building the dual wall nozzle with the cooling channels. Traditionally, they were made of individual pieces of tubing that were brazed together, but modern engines use techniques that are quicker and cheaper.
We can see this in a real engine. This is the upper part of the rocketdyne F-1 used on the first stage of the Saturn V.
The upper part of the nozzle is composed of the primary cooling tubes, and slightly below that, each primary tube splits into two secondary cool tubes.
The cooling is accomplished by a flow of RP-1 fuel down the nozzle. RP-1 is used rather than liquid oxygen because hot oxygen doesn't play well with others. At the bottom of the tubes, the fuel is collected and sent back to the top of the engine to be burned.
I deliberately hid the bottom part of the engine in the last slide. Below the section of the nozzle that is regeneratively cooled by fuel flowing through tubes there is a nozzle extension that is radiatively cooled. It's made of stainless steel and it gets rid of heat by heating up and radiating the heat away. As long as the temperatures aren't high enough to exceed the strength limit of the material, radiative cooling works fine.
But what is the weird wormy thing that's wrapping around the nozzle?
On the top-right of the engine we have the gas generator, which is a small combustion chamber that produces gas to drive the turbopump that pumps the propellants into the main combustion chamber. Gas generators run at lower temperatures than the main combustion chamber because the turbines cannot withstand high temperatures and therefore their output exhaust is also at a low temperature.
The gas is piped into the duct that wraps around the nozzle and injected into the nozzle through holes spaced around the nozzle.
The gas generator generates gas to drive the turbine, and then the leftover exhaust is injected all around the nozzle.
The gas forms a film between the nozzle and the hot gas from the combustion chamber, and this technique is therefore known as "film cooling".
If you've seen a Falcon 9 launch, you've seen this same technique in action on the Merlin vacuum engine - the output of the turbopump is injected in the nozzle to help keep it cool.
Like the F-1, the nozzle extension is radiatively cooled, but in the merlin it is made of a niobium alloy that is very heat resistant and is therefore fine even when it is red hot.
The Vulcain 2 engine used on the Ariane 5 also uses the same approach.
More advanced engines like the SpaceX Raptor used on starship do not use gas generators so there is no handy exhaust to help cool the nozzles.
However, it is still possible to use film cooling by injecting a small amount of liquid methane fuel around the nozzle.
A final cooling technique is ablative cooling, where the nozzle is lined with a material that will slowly burn away as the engine fires - very much like a heat shield on a capsule returning from orbit.
This is very common on solid rockets where there is no handy liquid to use for regenerative cooling, though it has been used on some liquid cooled designs.
And now, you know rocket nozzles.
If you enjoyed this video, please send me a Marycan hose nozzle including such features as:
Metal gun body
The metal handle
Encapsulates the knob
Saving card buckle
Package glue gun
And the all important "the pacifier"
As we move towards the first flights of SLS/Orion and Starship, I've been seeing articles talking about the space race between them...
This has me confused.
Is this actually a race?
Referencing a dictionary, in this context a race is an attempt to be the first to do or to get something.
Inherent in this definition is that there is a goal that both contestants are working towards. But that's simply not true; the goal of Artemis I is to launch Orion on a lunar test mission, and the goal of starship is to test reentry and launch some starlink 2 satellites.
Not the same goal, not a race.
If you want to talk about arbitrary milestones, then there are a whole bunch of other races we should be tracking...
It's important to remember that the goal of Artemis is to land humans back on the moon with Artemis 3 - using SLS/Orion to get astronauts to the moon, and lunar Starship to get them to the surface of the moon and back.
So it's a little weird to talk about a race when there's a symbiotic relationship between the two entrants...
But let's assume none of that matters and there's a race that is meaningful.
Artemis 1 and Starship are getting ready to launch on their first missions, and those currently look like they are somewhat aligned. Let's assume that SpaceX gets FAA approval to launch and neither SLS nor Starship see significant delays, and that they both launch in the third quarter of this year.
What happens after that?
On the Artemis side, Lockheed Martin needs to take some of the parts out of the Artemis I Orion and move them to the Artemis 2 Orion. It's surprising that this is required for a capsule program that has cost $20 billion, but that's the plan.
What that means is that Artemis 2 is currently scheduled for May of 2024, or about 21 months after the first flight.
On the starship, side, it's always useful to refer the wonderful production diagram produced by Brendan Lewis.
Booster 7 and Ship 24 are currently undergoing pre-flight testing, but you can see that there are follow on versions working their way through the production line. It won't be long until booster 8 and ship 25 are ready to fly.
On the starship side, Musk has asserted that they will be producing a starship stack every month. Let's discount that to once every three months. If that is true, they will fly another 7 times before Artemis 2 launches, for a total of 8 launches.
That's a lot more flights than SLS, and it would seem that they are set up to produce that many easily.
SpaceX is limited to 5 flights per year at Boca Chica under the current agreement, which might allow them to launch 10 in two years, but that's about all.
That's why SpaceX is working really hard on the starship pad at Kennedy Space Center in Florida. This picture show the Falcon 9 pad on the left - with the crew access tower for crew dragon flights - and the starship pad on the right, with the legs for the launch mount and 5 segments for the tower already assembled. It possible they can have that pad up and running next summer. If they have excess capacity to build rockets in Boca Chica, they can ship them to florida and launch them here. That could garner them two more launches per quarter or another 10 over the next 2 years, bringing the total to 18.
We know that SpaceX has successfully landed Falcon 9 first stages more than 125 times.
The super heavy booster used with starship is designed to be easier to recover than Falcon 9, and I'm expecting that sometime next summer they will begin reusing the boosters. That allows them to focus on building starships and could push the launch count up to 23, or pretty close to once a month for the next 2 years.
Which seems like a ton of launches, but SpaceX is on target for more than one launch per week for 2022, and this would be less than half that launch rate.
SpaceX is rapidly building a starship factory next to their Falcon 9 processing center at Kennedy Space Center in Florida, and they have begun the process to expand and more than double the amount of space they have.
And none of this assumes that SpaceX is actually able to land and reuse starship.
If SpaceX is successful - and they have a good chance of success - by the time we get to Artemis II, our discussions are going to be very different.
If you enjoyed this video, please write an essay on the mythology of artemis. Two pages, typed, double spaced.
This is the first of two videos talking about abort modes in reusable vehicles. I had originally planned on covering both the shuttle and starship together but I kept finding more shuttle information and it made more sense as a separate video.
The Space Shuttle was a very different vehicle from the ones that came before, and with those unique capabilities came some unique challenges in keeping the astronauts safe.
Early US crewed programs used modified ballistic missiles and those missiles had an unfortunate tendency to blow up, and so NASA designed escape systems that would attempt to save the crew members in such a case.
Mercury used a capsule with an escape tower on top of it that could pull the capsule free from the launch vehicle and out of danger. That's a redstone suborbital launch on the left and an Atlas orbital launch on the right.
Gemini went a different direction and used ejection seats.
The reason is that NASA originally wanted to move away from the downsides of splashdowns for Gemini - splashdowns take recovery fleets, they have weather requirements, and capsules can sink.
They instead worked on a concept based on a glider known as the "Rogallo wing" which would inflate and allow the capsule to glide to a touchdown on land. The picture on the right shows one of the test vehicles on display at the Udvar-Hazy center at Dulles airport in Virginia.
With that design, ejection seats would work as backup both on ascent and during the last stages of landing.
Because of development issues, the wing was delayed to later Gemini missions and ultimately cancelled.
Apollo went back to the escape tower approach.
The space shuttle was a very different bird - the orbiter was much, much larger than capsules - and it therefore required a different approach.
Here we have our space shuttle stack.
In a normal mission, both the solid rocket boosters and the three main engines are ignited on the launch pad. The solids burn for about 2 minutes, and the main engines burn for a little more than 8 minutes.
At that point, we reach Main Engine Cut Off, or MECO, and we're in orbit, with a little help from the orbital maneuvering system engines.
The ascent aborts are all about what happens when something doesn't go quite right. The first thing to note is that there is no abort if the solid rocket boosters fail.
The aborts are all based on one or more main engines failing or underperforming, and the options depend upon when the engine failures happen. The abort options and timings also depend upon the destination of the shuttle and the amount of payload it is carrying. The chart that I'm showing is from STS-116, which was a mission to the international space station.
We'll start by looking at the cases where a single engine fails, starting at the end of the launch and working our way back.
If the engine failure happens after about 6.5 minutes, there is no abort required. The remaining engines will burn longer and the mission proceeds normally.
If the engine failure happens from 4.5 minutes to 6.5 minutes, the orbiter can not generate enough speed to get into the desired orbit but it can get into a temporary orbit, so the Abort To Orbit - or ATO - abort option is used. This option allows time to decide what to do.
If the orbit is close to the desired - or nominal - orbit, the Orbital Maneuvering System engines can be used to raise the orbit to the nominal orbit, or at least one that is good enough for the goals of the mission.
If continuing the mission is not possible, the flight is converted to the Abort Once Around - or AOA - option. The shuttle gets into orbit and then reenters before completing an entire orbit. Depending on the orbit that the shuttle was aiming for, they would land at Edwards Air Force base in california, Holloman air force base in new Mexico, or back at the launch site in florida.
If the engine fails from 2.5 minutes to 4.5 minutes, the shuttle cannot generate enough speed to reach orbit, so the transoceanic abort landing - or TAL - option is chosen. Depending on the orbital inclination of the mission, different landing sites were used.
A mission to Hubble would be to 28.5 degrees of inclination, so their TAL landing site would be Banjul International Airport in Gambia.
STS-116 was to the international space station at 51.6 degrees of inclination, so their prime TAL landing site was Zaragosa Air Force Base in Spain, with possible backup sites in France, Spain, and Morocco, and the capability to land at other emergency landing sites.
A TAL profile simply goes from the ascent into orbit directly into the reentry and landing.
TAL aborts can also be used for systems failures unrelated to engines, such as propellant leaks, cabin leaks, or cooling issues.
If an engine fails in the first 2.5 minutes of flight, there isn't enough energy to do a TAL abort, so the return to launch site - or RTLS - option is chosen.
Getting back to the launch site is complicated. The orbiter is currently moving away from the launch site, so it will need to perform a powered pitcharound with the engines running so it is pointing back towards the launch site to kill the velocity away from the launch. Then it will perform a powered pitch down to get the nose at the proper angle for gliding.
At that point - in the highlighted box - the shuttle needs to be at the proper altitude, the proper velocity, the proper angle and be pointed at the proper direction. The main engines can then be shut off, the orbiter detached from the external tank, and the glide to the runway started.
To make it even more complicated, it's not safe to separate from the external tank if the external tank has more than 2% of its fuel remaining, so it must also use up its fuel before it reaches that point. That may require it to do a maneuver called "lofting", where it flies a high flight path to waste fuel.
It's not clear how feasible RTLS was. In 1980, STS-1 commander John Young said, "RTLS requires continuous miracles interspersed with acts of God to be successful". RTLS was improved after that, but it still remained iffy.
With two engines out, we are into what are known as contingency aborts. Quoting the NASA documentation, "Contingency abort procedures are executed when multiple main engines fail or suffer a performance degradation that results in the loss of all other intact abort options.
There are both nominal and transoceanic abort landing options, but their timelines have shrunk because of the reduction in thrust and there is no longer an Abort to orbit option as the shuttle cannot generate enough speed to make it to orbit.
DRP stands for Droop Guidance. There is a requirement that the shuttle stay above 265,000 feet (80 km) when travelling at high speeds to keep the external tank from overheating, and if the computers predict this will happen, they will enter droop guidance. The computer will rotate the shuttle into the "stand on the tail" attitude to minimize the sink rate. In this specific case, TAL should be achievable, but droop guidance may be enabled at other times.
If both engines fail before 5.5 minutes, we will enter ECAL or BDA abort. ECAL stands for "Eastern Coast of North America", and basically is a series of airports where a shuttle might be able to land. Because this flight is to the ISS, the ground path of the flight is up the eastern seaboard and therefore ECAL is the contingency option of choice; if the launch was more to the east, an abort to Bermuda (or BDA) would be the only option. If this situation is detected, the software will command an unguided 45 degree yaw (turn) in the direction of the landing sites; this is intended to make it easier for the shuttle to reach the landing site. The nose will be pointed up to prevent the vehicle from sinking too quickly.
The crew will then consult their documentation to choose an appropriate ECAL site.
Finally, if two engines fail in the first two minutes, RTLS will be initiated immediately after the solid rocket boosters separate and will proceed in the same manner as 1 engine out.
Finally, we reach the 3 engines out aborts. These are the most speculative and depend significantly on exactly when the failure happens, what the payload is, etc.
There are small nominal, Transoceanic abort landing, and ECAL/BDA options if the failures occur after 6 minutes.
Just before that we have what is known as a black zone.
The NASA documentation says, "Black zones are regions along the ascent trajectory that may not be survivable".
Black zones exist when one of the following is true:
Airspeed is too fast for the flight control system to stay in control
G loads are high enough to break up the orbiter
Three engines fail when the SRBs are firing. This may lead to failure at the external tank and SRB attach points, poor separation dynamics (the ET may hit the orbiter), or a center of gravity beyond limits that makes the orbiter uncontrollable.
Dynamic pressure during coast is too high for the reaction control system to deal with.
Control surface forces exceed their maximum or the wing or tail become too hot.
You can read all of those as "the orbiter loses control and breaks apart"
This first black zone is because the orbiter dives after the loss of engine thrust and is going too fast to level out.
Interestingly, there is a section before this where ECAL/BDA may be possible as there is less speed and therefore the speed when leveling out is acceptable.
There is a black zone in the first 2 minutes if all three main engines fail, as none of the RTLS maneuvers can be accomplished. Further, there are black zones if the remaining engines fail during RTLS for the same reason.
That covers all of the ascent aborts.
There are other flight rules that may lead to an abort; these happen when a system fails that would compromise the mission. These rules are designed to get the orbiter on the ground as quickly and safely as possible.
If the orbital maneuvering system - the two small engines above the main engines - cannot function, orbit cannot be achieved, so a TAL abrot will be used
If the auxiliary power units or hydraulics are failing, either RTLS or TAL will be chosen based on getting the orbiter back as soon as possible.
If there is a cabin leak, if the cryogenics are failing, or ig the freon cooling is failing, the same rule applies.
If the two main electric busses are failing, the rules say to do RTLS but my guess is that it's either RTLS or TAL
Finally, if one of the windows breaks, reentry is no longer possible so RTLS is the only option.
Let's talk about reentry.
Nearly all of reentry is a black zone; there is no abort possible if something goes wrong.
Originally, Columbia had ejection seats for the commander and pilot, but that was only a solution for two astronauts and they took those out after the first few flights.
After challenger, nasa added a bail out option. If you get to 50,000 feet in control and you are more than about 55 nautical miles from the landing site, you won't make it and you can bail out. Or if you get there but there's another reason to think you can't land, you can bail out.
On this chart, the red box shows the time you have to prepare to bail out by flattening out the glide and turning on the autopilot. And the green shows the 90 seconds you have to get out.
Here's a simplified sequence:
At 50,000 feet, call for bailout, flatten out the glide with wings level at 185-190 knots and engage the autopilot.
At 40,000 feet, the cabin is vented to equalize pressure so the hatch can opened
At 30,000 feet, the side hatch is jettisoned and the crew exits the vehicle.
There is a problem bailing out as the astronauts will hit the orbiter wing if they simply jump out; NASA therefore developed an extendable pole to guide the astronauts down low enough to miss the wing.
The history of aborts on the shuttle is thankfully short. STS-51F launched in 1985, and 5 minutes and 33 seconds into the flight, two temperature sensors in the center engine failed, leading to the shutdown of that engine. The shuttle performed an abort to orbit at a lower-than-planned orbital altitude, but luckily the flight was a SpaceLab mission and did not require a specific orbit.
There were no other aborts during the shuttle program, though there were a few close calls. But that's a subject for another video.
I have a few references to share for those who want to dig deeper.
The intact ascent aborts workbook and the contingency aborts workbook were used to train shuttle astronauts in abort procedures
The Crew escape systems workbook was used to train astronauts how to use the escape systems, including the bailout procedure.
NASA flew the space shuttle 135 times over 30 years, but they always flew to low earth orbit. Why didn't they fly the shuttle to the moon?
To answer this question, we need to understand the capabilities of the shuttle.
We'll start with the published capabilities, courtesy of the Wikipedia article. It can carry 27,500 kilograms to a 204 km orbit and it can carry 16,050 kilograms to the international space station in a 407 km orbit.
It doesn't tell us how much payload it can carry to the moon - that would be listed as "TLI payload", an abbreviation for trans lunar injection. To figure that out, we need a model.
No, not that kind of model, a mathematical one. The one I'm using runs in Excel, and you plug in a lot of detailed information about the rocket and it will estimate how much payload the rocket can carry to different destinations. The distance measure between places in space is known as "delta V". If you want more details on delta v, you can watch my video, "planning your solar system road trip", but if you think of the delta V between earth and the moon in the same way you think of the driving distance between Seattle and San Diego, you'll be on the right track.
The numbers my model produces are estimates - there are many details that the model ignores.
We can understand the capability of a vehicle by looking at the payload it can take to different destinations.
I'm going to start with SpaceX's Falcon 9 since most people will be familiar with it.
Getting into low earth orbit takes somewhere around 9300 meters per second of delta V, and the Falcon 9 can carry about 22,800 kilograms to that orbit.
Getting to the moon takes around 12,500 meters per second of delta v, and Falcon 9 can carry about 6000 kilograms to the moon. Note that is going around the moon - getting into lunar orbit or landing would be up to the spacecraft.
Getting to Mars is surprisingly only about 13,000 meters per second, and Falcon 9 can send 4000 kilograms of payload there.
Let's add in the line for the shuttle...
Shuttle can carry about 27,000 kilograms into low earth orbit, but the line slants down pretty quickly, and the maximum delta V that shuttle can reach - with an empty payload bay - is only about 10,000 meters per second. That's a long way from the moon.
This seems wrong...
Falcon 9 is a medium-sized rocket, with a takeoff mass of around 550 tons. Shuttle is a very large rocket, with a takeoff mass of 2030 tons.
Let's add one more vehicle to the mix. In 2009, NASA studied a shuttle-derived Heavy Lift vehicle that will help us understand what is going on. Instead of the orbiter, this vehicle had a large payload bay.
Here's the graph for that vehicle. It can lift nearly 80,000 kilograms to low earth orbit, or nearly 3 times what shuttle could lift, and it can *barely* get a payload of perhaps 3000 kilograms to the moon.
What are the differences between these three vehicles?
Starting with the Falcon 9, its payload to low earth orbit is 22,500 kilograms. The second stage that carries the payload adds a further mass of 3900 kilograms, and the total mass to orbit is 26,400 kilograms. A little math tells us that 85% of the mass to orbit is payload.
Looking at shuttle, the payload to low earth orbit is 27,000 kilograms. The shuttle weighs 85,700 kilograms and the super lightweight external tank weighs 26,550 kilograms, so the total vehicle mass is 112,190 kilograms.
The total to orbit is 139,190 kilograms and the payload percentage is 19%.
The shuttle hlv has a payload of 77,000 kilograms. The vehicle mass is 26,043 kilograms, the tank used in the studies I looked at was a slightly heavier tank at 33,172 kilograms, so the vehicle mass was 59,215. That's roughly 50% the shuttle vehicle mass. The total to orbit is 136,215 kilograms - very close to the shuttle total - and the payload percentage is 56%
The vehicle mass for shuttle is a seriously limiting factor - the heavy shuttle has to be carried all the way to the destination orbit along with the external fuel tank. The shuttle HLV is quite a bit better because the payload container is far lighter than the shuttle orbiter.
Another way to look at this is that every kilogram of mass in the shuttle is a kilogram of mass that is not in the payload.
The Falcon 9 curve is so flat because of the very high payload ratio, and that's because the vehicle mass is so small.
The low delta v was a problem for shuttle. It was supposed to replace the existing US expendable rockets - the Atlas II, the Delta II, and the Titan III - but all of those rockets had light second stages like the Falcon 9, and therefore could launch communications satellites, spy satellites, and planetary missions directly. Shuttle was unable to do those missions directly.
NASA's solution was what is known as a "kick stage" - an additional rocket stage carried into orbit and used to supply additional delta v to a payload to get it to a more challenging destination.
NASA called this the Interim Upper Stage, and it is composed of two solid fuel motors, so it's actually two stages.
It was later renamed to the Inertial upper stage when the more advanced Centaur upper stage was cancelled.
The inertial upper stage launched the Magellan probe to venus, the Chandra x-ray observatory, and many of NASA's tracking and Data relay satellites.
NASA's solution to get payloads to higher delta v destinations was a rocket carried in the shuttle bay along with the payload - these rockets are usually known as "kick stages" as they give payloads an extra kick. It was known initially as the "interim upper stage" and was then renamed the "inertial upper stage" when the more capable stage that was to follow was cancelled.
The interim upper stage used two solid rocket motors and was flown on 15 shuttle flights, including the Magellan probe to Venus, the Chandra X-ray observatory, and NASA's TDRS communications satellites.
The kick stages provides additional delta v to payloads at the cost of a significantly reduced payload.
But that wasn't really the question...
The question was whether we can get the shuttle to the moon...
The stock orbiter gives us 10,023 meters per second of delta v, and we need 12,450 meters per second to get to the moon.
We need a more powerful shuttle design.
Here's my first idea. Let's double the solid rocket boosters...
That helps a little, but only gets us to 10,734 meters per second of delta v. I explored adding up to 10 boosters but it didn't get us to the moon and I couldn't figure out how to hook them all on.
We need more tanks. This design uses three tanks and 6 boosters, and to deal with the extra mass of the tanks, there are engine pods with three engines under each of the additional tanks.
It gives us 12525 meters per second of delta v *with* a 15 ton payload.
We have achieved the ability to send shuttle on a trip to the moon; trans-lunar injection achieved.
Can we get into orbit around the moon?
The propulsion system on the shuttle is the Orbital Maneuvering System. If we use the built-in OMS propellant of about 10,000 kilograms and use all of our payload for additional propellant, we have 25,000 kilograms of propellant and that gives us about 920 meters per second of delta V.
That is just enough to put the shuttle into the low lunar orbit that the Apollo program used. Unfortunately, we are out of propellant and can't get out of that orbit to return to earth.
If we adopt the approach used by artemis, we can get into the Near rectilinear halo orbit that Orion uses and have just enough delta v to get out of that orbit and head back to earth.
And here we run into a snag...
The shuttle's heat shield was designed to deal with returning from low earth orbit at a speed of about 7800 meters per second, or 17,500 miles per hour.
Coming back from the moon the speed is more than 40% faster. A vehicle reentering at that higher speed will experience more than 3 times the reentry heating, and the shuttle would not survive a reentry at that speed.
The shuttle did not have the performance to make it to the moon, largely because the shuttle itself was so heavy.
We could maybe sorta put together a version that could make it into lunar orbit and back, but the shuttle could not survive reentering the earth's atmosphere when returning from a lunar mission.
If you enjoyed this video, please overtip your driver the next time you ride a shuttle bus.
The expansion of space exploration and the transition to permanent space settlements will require us to answer many important questions which have so far been ignored by major spaceflight organizations worldwide.
Luckily, there are space enthusiasts to keep us on the right track, and one of them posed the following question:
"How close can you get to the sun so you are able to make toast? I'd like the calculations please."
If our primary and secondary space toasters both fail, can we use to the power of the sun to properly prepare out eggos, our pop-tarts, and our toast? And if so, what orbit would be optimal for these food preparations?
This is Space Toast...
In the old days, some circus performances featured trapeze artists who worked without safety nets - where a fall would likely result in their death. "Working without a net" now means taking an action that is risky without using things that would provide safety.
I have decided to work without a net to answer this question.
Or more specifically, without "the net".
I'm going to be doing this without using the internet to look anything up. I had originally planned to allow reference books but I looked around and we don't seem to own them any more, so those are out. And I'm going to try to avoid asking my PhD chemist wife for help. The goal is to see how far I can get using only what I've stored in my head.
If you want to join with me in my quest, pause the video and go do it and then come back and we can compare answers.
My answer will not have a high degree of accuracy; I'm going to be doing what used to be called "back of the envelope" calculations to get an answer that's pretty good. I did use an actual envelope leftover from a bill I got, but unfortunately I have misplaced it. I also used my Sharp EL-515 solar calculator that I bought for college in 1983.
There's a wonderful bit of synchronicity on this page. This little bit of UI that says there is no internet access is known technically as a "toast notification", which I view as a good omen in relation to my plan.
As a benchmark, I will be using a KitchenAid KMT2115ER toaster clad in the every popular empire red color scheme.
Looking at the nameplate, I see that the toaster pulls 900 watts of power. That is spread over two slices of bread that each have two elements, so that means each side of bread gets 225 watts of power.
The toaster uses nichrome wires to convert electrical energy to heat, and like few things around, it's 100% efficient, but not all that heat goes into the toasting process - put your hand above a toaster and you can easily feel the waste heat. I have no idea how much is wasted and I'm prohibited from using google scholar to find the latest bread-toasting research, so I'm just going to assume half the heat is wasted, and that gives us 112 watts per side of bread.
I know that I'm going to need to compare this to the power of the sun, which is generally measured in watts per square meter, so we'll need to scale that 112 watts up to a hypothetical toaster that can toast bread slices that are 1 meter square.
Bread is about 4" on each side, which is conveniently 10 cm or 0.1 meters. That makes the area 0.01 square meters, or 1/100th of a square meter.
We can divide 112 watts by the bread area, and if we do that, we get 11,200 watts per square meter.
Which honestly seems like a lot; a clothes dryer is something like 4000 watts of power and this is nearly 3 times that amount.
We now have enough information about the toaster.
We next need to figure out what we can expect from our dear old sun...
I remember from the video I did on solar power satellites that the power that reaches earth orbit from the sun is 1576 watts per square meter.
We can figure out the power ratio by dividing the toaster power by the solar power, and if we plug in the numbers we get about 7. A toaster puts 7 times the amount of energy into a piece of toast than the sun does.
At this point I can hear somebody complaining about the spectral differences between solar radiation and the heat the toaster is providing along with the differences between pure radiative heating in the vacuum of space and the heating of a toaster inside a spacecraft, and those are good points. Please incorporate those factors into your solution for this problem, and remember that you can only use what you currently have in your head. Make sure to show your work - you can earn up to 5 extra credit points doing this.
We'll need to get someplace closer to the sun for our space toaster to work, and that someplace will need to provide 7 times the radiation.
Time for some planetary math. The earth is 98 million miles from the sun, or about 157 million kilometers. We need to figure out where we need to be to increase the amount of received power by a factor of 7. It's obviously going to be closer to the sun, but how much closer?
At the orbit of the earth, all the sun's output goes through a sphere with a radius of 98 million miles, so we can simply calculate that surface area, divide it by 7, and then figure out the radius for that new area.
Unfortunately, the formula for the area of a sphere isn't something that is rattling around my brain, so that approach isn't going to help.
I do, however, remember that radiation obeys an inverse-square law. Get twice as far out, and the decrease in intensity will be 1 in 2 squared, or 1 in 4. Get closer, and increase follows the same law.
This equation describes it - the distance ratio that we want is equal to the square root of 1 over the intensity ratio.
Plug in our numbrers, and the square root of one 7th is 0.38.
Multiply it out, and we find that the intensity we want is at 37 million miles, which I hereby define as one KitchenAid of distance from the sun. I don't recall the distances of the other planets, but I do suspect that 37 million miles is well inside the orbit of Venus.
There is, of course a problem - our solar toaster only toasts one side of the bread.
We can fix that problem quite handily with a little reflective magic, We put two mirrors at 45 degrees, and that allows us to reflect radiation and heat both sides of the toast at once.
And yes, the mirrors will reduce the heating slightly. Just make them a little bigger than they need to be and everything will be fine.
We've now answered our space toast question. Or have we?
Time to check my work, and for this I will be using the internet.
First, if you want to do some analysis with different spectrums, the left is what you can expect to get with a nichrome wire in a toaster, which is probably around 1000 degrees centigrade.
The right one is the spectrum we get from the sun, which is pretty close to a blackbody at 5250 degrees centigrade.
Have fun figuring how to compare these.
On to checking my work.
First, the surface area of a sphere is 4 times pi times the radius squared, which means my inverse square calculation was correct.
On the constants, I didn't do quite as well.
The earth's radius is 93 million miles, not 98 million miles. If I was smarter I would have calculated using astronomical units, where 1 is automatically the orbital radius of the earth, and that would have killed one constant.
I asserted that the solar intensity in orbit was 1576 watts when my notes from the solar power video notes tell me it's only 1360 watts per square meter. Our magic ratio then becomes 8.2 rather than 7.
Put those two changes together and that puts us at 32.5 million miles rather than 37 million miles. About 12% off.
Looking at the inner planets, we have the earth at 93 million miles, venus at 67 million miles, and mercury at 36 million miles, so the answer to the question is "about the orbit of mercury". That's where you want to get, and unfortunately it's a really long way from earth orbit.
You could simply leave your toast out 8.2 times as long, but note that as the toast heats up it starts radiating more energy so it takes longer than that. I also don't think you are willing to wait 15 minutes for a couple of slices of toast, but perhaps eggos are more important to your life than they are to mine.
If you enjoyed, this video, I am desperately in need the Dolce & Gabbana SMEG retro toaster collaboration, a steal at only $700.
It includes 6 toasting levels plus reheat, defrost, and bagel functions, and is purported to be the perfect morning assistant.
After Apollo, NASA came up with a lot of concepts that featured Space Tugs.
The idea is that - like tugboats on the water - you would be able to use a space tug to move spacecraft or satellites around...
I've recently noticed that the term "Space Tug" has been co-opted to mean pretty much whatever journalists want it to mean.
So I thought it would be fun to play a game, "Space tug, or not". We'll take a look at a number of systems that people have asserted to be space tugs, you decide whether a system is a space tug or, and then I'll tell you why you're wrong. Or right, as the case may be.
Our first contestant is Helios, a product of Impulse space. There were nicely two articles referring to it as a space tug.
The point of Helios is to act as a third stage for payloads.
You launch Helios with its attached payload into low earth orbit using whichever rocket is convenient. Once you get into low earth the helios stage ignites and takes the payload to its destination
It can send 4000 kilograms to geostationary orbit, 6000 kilograms to the moon, or 4500 kilograms to Mars.
Is Helios a "space tug"?
It might help if we talked about what makes a space tug.
Tugboats push ships and barges around, so space tugs will need to do the same in space - they'll need to move a satellite or spacecraft to a different orbit or trajectory.
So, based on that definition, is helios a space tug? Do you think yes, no, or somewhere in between?
My answer is a solid yes - it meets the definition since it moves satellites or spacecraft to a different orbit.
But then I got thinking...
When Falcon Heavy flies a mission to put a satellite directly into geostationary orbit, the second stage gets into earth orbit, coasts for a number of hours, and then performs a burn to move its satellite payload to a different orbit?
Does that make the Falcon Heavy second stage a space tug?
The SLS rocket used on Artemis missions uses the boosters and core stage to put the ICPS upper stage into orbit. That stage will then ignite to get out of the initial orbit and into the trajectory to the moon.
Does that make the ICPS upper stage a space tug?
I would argue that neither are a space tug, which means our definition is incomplete.
By analogy with a tugboat, I think inherent in the definition of a "tug" is that it's an independent vehicle that will rendezvous with an existing satellite or spacecraft.
My assertion is therefore that Helios is *not* a space tug. It's just an additional stage that is used to send a payload to a higher orbit.
If only there were a term for an third stage that gives an extra kick to a payload.
If we look at what impulse aerospace says about helios, they describe it as a high-energy kick stage, which is a more precise and meaningful term.
Here's our second contestant, talking about a company named Exotrail developing a space tug to carry small satellites to geostationary orbit.
There isn't a ton of detail on their website, but they do have a minimal description in which they talk about providing a launch and deployment service using their spacevan on-orbit transfer vehicle.
AFAICT, the scenarios are all about doing this as part of the launch process.
Like Helios, not a space tug.
Quantum space is next.
They describe their "space tug" as a delivery workhorse. I would describe it as an "adaptable satellite bus" that supports a bunch of different scenarios, where delivering satellites is only one of those.
For more info on Satellite buses, see my recent video, Hop on the satellite bus.
Once again, not a space tug.
Blue Origin has entered the market with their Blue Ring. They have a long list of features:
(read)
To the extent that they provide refueling and logistics, there's a good argument to be made that they are a space tug.
On missions where they don't do that - especially if they are just doing hosting of a payload - there's a good argument that they are just a satellite bus and therefore not a space tug.
Blue ring is schroedinger's spacecraft - you don't know whether it's a space tug or not you open the payload fairing after launch
Automos (ah-tah-mos) space is working on their quark series of vehicles, and this is what they say:
(read)
Relocation, life extension, and orbit raising put them firmly in the tug category.
NASA is currently exploring alternatives for a vehicle to journey to the international space station at the end of its mission - currently 2030 or so - latch onto it, and slow it down to deorbit it and toss it into the pacific ocean, probably in the spacecraft cemetery that already hosts 263 spacecraft.
This one is probably the easiest one in our quiz - it's very obviously a space tug.
Exolaunch announced their space tug program, but the press for it refers to it as an orbital transfer vehicle (though there's a little caption on the picture that says "reliant space tug".
Their graphics show it doing multiple satellite deployment to different orbits, and their vehicle description says this:
Read.
The first two are all about deployment, but reducing space debris could be tug like. I'll call this not a tug but with tug-like aspirations.
Firefly aerospace offers a new product named El-it-ra
Elytra has three levels. Elytra dawn is about hosting, rideshare and delivery. All useful things, not space tug things.
Elytra Dark is about on-orbit persistence and going beyond earth orbits. Also not space tug things.
Elytra Dusk is a bit of a wild card. The high level items listed here say "not a space tug". But they also talk about relocation, deorbit, and logistics, which are quite a bit more space tuggy.
I'll give it a minor in space tug, it's a "if somebody asks, I guess we could do those things as well".
Northrop Grumman has an interesting vehicle. It navigates to geostationary orbit, finds the satellite that has run out of fuel, attaches itself to the satellite and takes over the station keeping for the satellite, keeping it operational. This is a big deal for satellites like this as they are extremely expensive to replace. Here's a picture from their second mission as they sneak up on Intelsat 10-02 in geostationary orbit.
The have future versions planned that would allow one launch to service more than one satellite.
Absolutely a space tug.
What did we find out...
There are many products described as space tugs
Some are very obviously not space tugs
Some are very obviously space tugs
Some aren't really space tugs but have aspirations to do tuggy things if those markets develop
If you enjoyed this video, please send me this Imara Harbour tub boat kit, a bargain at more than 30% off.
SpaceNews Article: https://spacenews.com/chinese-rocket-stage-breaks-up-into-cloud-of-more-than-700-pieces-of-space-debris/
NASA OIG Report: https://oig.nasa.gov/wp-content/uploads/2024/08/ig-24-015.pdf
Everyday Astronaut tour of New Glenn Factory: https://www.youtube.com/watch?v=rsuqSn7ifpU&pp=ygUSZXZlcnlkYXkgYXN0cm9uYXV0
Pathfinder episode with Andy Lapsa (Stoke Space): https://pod.payloadspace.com/episodes/0104
Main Engine Cut Off podcast: https://mainenginecutoff.com/
Off Nominal Podcast: https://offnom.com/
I often come across topics that I want to comment on but which aren't really big enough for a full video, so this is the start of a series of shorter-form videos.
I'll put any links in the video description.
On of the members of the EagerSpace subreddit asked about this one.
On August 6, China launched their long march 6A rocket, and after releasing its satellites into an 800 kilometer polar orbit, something happened to the second stage and there was a "debris creation event" that resulted in 700 to 900 individually tracked objects.
That followed a similar event on March 26 which created at least 60 objects, and one on July 4th that created an unknown number of objects.
This latest one was particularly troubling as the debris is in an 800 kilometer sun synchronous polar orbit, and it can take centuries for debris to decay from those altitudes.
The question is "Will China Fix the long march 6A?"
That is an interesting question, and one that I don't have a real answer for.
I can talk about the factors that might push it one way or the other.
Starting on the "Won't Fix" side,
This is the fourth or fifth launch of the long march 6 that generated debris, and one of them was a couple of years ago, so it's not a new issue.
The group launching the rocket is likely rated on delivering satellites to orbit, not on whether they create debris or not. They may have a strong incentive to keep flying.
The launches are otherwise successful.
China regularly drops rocket stages on their own population. Being a good citizen doesn't seem to be the their highest priority.
On the "Fix" side, the debris that they generate may interfere with their own satellites.
The Chinese space program is a sign of Chinese achievement and a source of national pride.
Debris generation will generate bad PR and pushback from other spacefaring nations.
I'm hoping they'll end up on "fix" because 800 km is crowded and a really bad place to create debris, but you may have heard that hope is not a strategy.
Boeing news and it's not starliner.
When NASA started on SLS in 2010, Congress gave them a target to fly in 2016 (later revised to 2017). That short timeline would be a problem for NASA.
The core stage was supposed to be a modified version of the shuttle external tank, the RS-25 engines were shuttle leftovers, the Solid Rocket boosters were pulled directly from the constellation program and the Orion capsule was already in development. Those all looked achievable, but building a new second stage - known as the exploration upper stage - didn't fit into the schedule.
NASA therefore decided to look for an alternate second stage.
NASA looked around, and settled on the upper stage of the delta IV rocket, known as the Delta Cryogenic Second Stage.
ULA agreed to produce some modified versions of this stage, and the NASA version is known as the Interim Cryogenic Propulsion System, and that is what flew on Artemis 1.
This stage was a good fit for the Delta IV Medium rocket, but it is a comically small stage for SLS, with only a single RL-10 engine providing 11 tons of thrust, and that upper stage limits the payload to the moon to only 27 tons.
It was, however, the quickest path to get SLS flying, and ULA had no problem producing the stages, which were assigned to fly Artemis 1 through 3. This configuration of SLS is known as block 1.
The Exploration upper stage uses - "will use" is more correct at this point - four RL-10 engines and over 4 times the propellant. It is a decent match for the size of SLS and will boost the lunar payload up to 38 tons.
The exploration upper stage is being developed by Boeing. When congress told NASA to build SLS, they told them to reuse as many contracts as possible, so the Boeing upper stage contract for the Ares I rocket that was part of the constellation program got modified to encompass the core stage of SLS and then was modified again to include the exploration upper stage. If you suspect that makes it hard to tell what each of them cost, you would be right.
On August 8th of 2024, the NASA inspector general released a report on SLS Block 1B, which is the version of SLS that will include the second stage.
And no, I don't know why NBC news thought showing the core stage while talking about the second stage was a good move.
NASA requires all contractors to have a quality assurance program that complies with the quality management standard known as AS9100. This is a very common standard and is used by tens of thousands of companies worldwide. At NASA the front-line quality assurance verification is done by the defense contract management agency.
If problems are found, DCMA issues corrective action reports.
For this report, NASA OIG analyzed the corrective action reports for the exploration upper stage project for two years, from September 2021 to September 2023.
They found there were 47 level one reports, which are problems that can be corrected on the spot.
They also found 24 level two reports, which are problems that need to be addressed by supplier management.
In one instance, metal shavings were found in the liquid hydrogen tank, and metal shavings, Teflon, and other debris were found on the ladders used to access the tank.
That is really, really bad. Foreign material in tanks will get into motors and cooling channels and can cause serious issues.
It was bad enough that in may of 2023, NASA's chief safety officer for the stages project asked DCMA to draft a level III report that would deal not only with the corrective action reports but the way Boeing dealt with them. Level III reports are a significant big deal. This report was never sent to Boeing.
OIG's conclusion is that Boeing's workforce at Michoud assembly facility in New Orleans lacks sufficient experience, training, and instruction.
The other news is what we have unfortunately come to expect from the SLS program...
EUS has had lots of slips, moving from February of 2021 all the way out to April of 2027 in recent estimates. About a year of this is attributable to Boeing running out of funds for the SLS core stage and shifting funds from EUS to the core stage.
In 2017, the estimated costs were $962 million, and the current estimate is that the costs will reach $2.8 billion through 2028.
It may be worse than that. Boeing's earned value management system - the system that Boeing uses to predict schedule, costs, and perform project tracking - is not approved by the Department of Defense. It currently has open level II and level III CARS issues.
Until that system is fixed, it cannot be used to produce meaningful data that can be used to make informed decisions and produce forecasts.
NASA simply *does not know* how long the exploration upper stage will take to finish nor how much more it will cost. Maybe the projections are fine, maybe they aren't. This system is scheduled to be fixed later in 2024.
It's a lot of fun to kick Boeing while they are down and they are certainly a big part of the problem, but NASA is clearly not helping.
NASA worked on SLS Block 1B for 10 year and spent $3 billion before creating the plan that detailed life cycle and schedule commitments, violating NASA scheduling requirements.
NASA is moving assembly of the exploration upper stage engine section from Michoud to Kennedy and will move quality control surveillance from DCMA to an outside contractor. That seems unlikely to produce the level of oversight required by the program.
NASA has not instituted financial penalties for Boeing's non-compliance with quality control standards.
There are no good solutions to this situation. NASA only ordered three ICPS stages and ULA no longer flies the Delta IV so it's probably not possible to buy new ones.
NASA therefore cannot fly any missions after artemis III without the exploration upper stage.
In case you weren't one of the nearly 1 million people who watched it, Tim Dodd's tour of the Blue Origin New Glenn factory is worth your time.
I'm always happy to see real hardware and more rockets is generally better - I will be much happier with a company spending billions of dollars and flying a rocket over one spending billions of dollars and not flying.
I still have open questions about New Glenn...
How are they going to go from their current status to launching the Escapade mission to Mars in 40 days? They've never launched this rocket or any orbital rocket nor have we seen flight hardware on the pad yet. Will be impressive if they can do it.
I'm skeptical about whether as a company with a big new orbital rocket they can compete in terms of cost with Falcon 9, Vulcan, or Neutron.
The next couple of months will be interesting.
I also had somebody recommend the Pathfinder podcast interview with Andy Lapsa from Stoke Space, and that is also well worth your time. That did move the needle on my opinion on Stoke a little - Lapsa clearly thinks more like a founder than I had previously thought, but I still am skeptical about whether their approach will work and whether they have enough money - or can raise enough money - to reach their goal.
Stoke has also decided to take the software that they built to manage their hardware design and manufacturing needs - software that they call "Fusion" - and commercialize it. Given that I have worked on teams that built internal tools and teams that sold commercial software, I believe I have a bit of expertise.
What Lapsa said made perfect sense to me. Good internal software is expensive to develop and because it is a cost center - you put money into it - it typically only gets to the "tolerable" state before somebody gets tired of paying for it. And it's hard to manage the team - you might only need minimal support for 6 months but then need a lot of support for 3 months to deal with a new situation.
If what you are doing is needed by others in the industry - and I no reason to doubt that - then it will cost you more to create the commercial product, but it will quickly turn from a cost center that you pay for to a profit center that generates money for the company, and some of the requests that your outside customers make will turn out to improve your own use of the software.
You do run the risk of improving your competitor's products, but stoke seems to operate on the same "try to keep up with us" philosophy that has worked well for SpaceX and Rocket Lab.
A bit of trivia - the custom system that SpaceX wrote to do this is called "warpdrive"...
And that's all my thoughts for now...
If you liked this video, please send me this thought improvement device, as used by astronauts to improve their concentration during space walks.
https://library.sciencemadness.org/library/books/ignition.pdf
If you follow SpaceX, you've seen explosions. Quite a few explosions. That presents us with a bit of a puzzle.
SpaceX is generally considered the most innovative space company in the world and Falcon 9 is likely the most reliable launcher in the business, with 261 successes since their last failure.
And yet, explosions.
Somehow, SpaceX ends up being the most innovative and most reliable company despite the explosions...
But the truth is a bit stranger than that. SpaceX is the most innovative and reliable company because of the explosions.
To understand why this is the case will require a bit of an explanation...
It is helpful to categorize SpaceX programs by their explosiveness.
Cargo dragon had no explosions, nor has the beefy Falcon Heavy. I'm going to ignore any issues from Falcon Heavy booster recovery.
The failure of the first Falcon 1 flight involved an engine fire and crashing into the ground, and I'm going to label that an explosion.
Falcon 9 has blown up twice. Once when a strut holding a helium pressurization tank in the second stage broke and overpressurized the oxygen tank. Not a fire explosion, but still an explosion.
And once during fueling due to a different problem with the helium pressurization tanks.
A crew dragon blew up in an uncrewed test of the abort propulsion system. These three systems were slightly explosive.
Which leaves us with two more projects...
By my count, there were 8 explosions during the development of reuse for the Falcon 9, as chronicaled in SpaceX's wonderful video, "How not to land an orbital class booster".
Starship has also brought us a significant number of explosions.
Why are Falcon 9 reuse and starship different from the other projects? Why are they so much more explody?
I'm going to add another dimension to our chart, one that I'm going to label "prior art".
Prior art refers to any previous projects/programs/research that has been done in a specific area. Are you doing something that has been done many times, or are you breaking new ground? Are there books on what you are trying to accomplish, or NASA technical reports?
Cargo dragon was new to SpaceX, but satellites that need to navigate in orbit have been around for decades and they are very well understood. Falcon Heavy was based on the Falcon 9 which was very well understood, and there are other examples of clustered boosters in rocket history.
Falcon 1 had a lot of prior art in previous launch systems. Falcon 9 could build on Falcon 1 - the Merlin 1c engine came directly from Falcon 1 - and other rockets had used clusters of engines, though 9 was a large number.
Crew Dragon benefitted from Cargo Dragon and all the NASA legacy of crewed capsules, but there was a fair amount of new work due to SpaceX's choice of capsule shape, abort approach, and parachute architecture.
My guess is that you have figured out where this is going...
Falcon 9 reuse had very little prior art. Nobody had ever tried to land an orbital booster.
Starship has some prior art - falcon 9 has given SpaceX a lot of useful information, and you can look at shuttle and NASA for information on thermal protection tiles - but Starship has a ton of new stuff that has never been done before plus an innovative new engine using an uncommon fuel.
I'm going to draw an arbitrary line separating the groups of projects. For the top group, SpaceX was primarily doing development work - applying existing approaches to a specific problem.
For the bottom group, SpaceX was doing research and development - they had to figure out how to do new things, things that had never been done before. There were no existing approaches.
My assertion is that research and development is generally more explody - at least it is if you are doing it right.
It's pretty common to feel good when you come up with an insight, and it's also pretty common to find out that your insight is neither new nor original.
Back in the 1960s there was an ongoing debate about the best way to build an airplane, and by best they meant "cheapest and fastest".
There were two schools at odds with each other. The prototype folks said that you should build a prototype of the airplane so that you had actual data to feed into decisions.
The "define and produce" folks - sometimes known as the "development production" folks - said that you should do your design and definition up front and focus on starting production early. They asserted that define and produce was both faster and cheaper than prototypes - prototypes are just a waste of time and money.
The air force thought it would be very useful to know which approach was better because they bought a lot of new planes, so they commissioned the rand corporation to do a study that examined the question. Rand looked at 12 aircraft projects and concluded that their analysis did not support the assertion that the define and produce approach is faster and cheaper.
They went beyond that, and concluded that (read)
Using prototypes was considered the better way to do the things with low amounts of prior art, all the way back in the early 1970s.
If you are going to build a prototype, you need to decide what that prototype will actually do.
I believe that SpaceX is using a concept that is common in the software world, a concept known as minimal viable product, or MVP
The point is to choose a goal that meets a short set of constraints.
First, it has to have a meaningful real-world result, one that you can easily demo to people outside the team.
Second, it has to generate useful real-world data.
Third, it needs to be cheap and quick to build.
The point is to be incremental. Without prior art we aren't exactly sure where we are going so we limit ourselves to something that we know we will need that takes a step in the next direction.
For Falcon 9 booster reuse, SpaceX had an interesting opportunity. They were already going to be flying Falcon 9 rockets, and any flight that could carry the extra weight of their recovery equipment became a potential flight test.
Their first MVP was something like "survive reentry".
SpaceX put parachutes on the very first launch of Falcon 9 in June of 2010. The booster burned up on reentry before the parachutes got to deploy.
They tried again with the second launch in December of 2010, with the same result.
That was not going to work.
They were distracted for a while by the creation of Falcon 9 Version 1.1, but that new version gave them more mass to devote to reuse, and they came up with a new plan.
First, they added heat shielding around the engines of the booster, and second, they came up with the idea of running the engines during reentry to both slow the booster down and to push the reentry heat away. They called it supersonic retropropulsion, and it looks like this.
That flew on flight 6 of Falcon 9, and they successfully made it through reentry on their first try with this method, even though the booster impacted the ocean hard. They had achieved their first MVP - they knew how to survive reentry.
At the same time, they were working on different MVP- how to land a booster softly.
They built a prototype out of a Falcon 9 Version 1 first stage and a single Merlin engine and named it grasshopper. It made 8 flights, all successful.
Grasshopper was followed by F9R, built on a Falcon 9 version 1.1 first stage. It featured deployable legs and looks much more like the first stage we are used to seeing. It flew four successful flights, but on the 5th flight a blocked sensor caused the flight path to deviate and the prototype was destroyed by the flight termination system.
SpaceX had planned a third prototype for high altitude testing but decided that they had learned enough to be confident in their approach. The landing prototype stage was done.
SpaceX count get through reentry and they knew a lot about low speed control and landing. What they needed to do is hook those two together, and most of their launches provided an opportunity to try to do just that. They could afford to build lots of prototypes.
The first MVP was going from reentry to a controlled water landing.
Their next two attempts were successful; flights 9 and 10 were successful at achieving a controlled landing in the water.
Flight 13 was their next attempt and the booster ran out of liquid oxygen.
Flight 14 was the first flight of their grid fin control system that would allow them to precisely control their landing point, but it ran out of hydraulic fluid.
Flight 17 had a stuck throttle valve.
(watch video).
It seemed that SpaceX was coming up with lots of ways to crash their rocket.
After a few more tests, we finally saw this - a successful landing on land, followed by a successful landing on the drone ship.
It was impressive, but they had 6 failures and only two landing successes.
And then something happened that surprised most of us.
SpaceX landed two more missions, getting to three in a row.
One failure due to low engine thrust, then 28 in a row.
A failure due to a grid fin issue, then 17 in a row.
A failure due to incorrect wind data, then 24 in a row.
During this section, SpaceX landed 69 out of 72 attempts, for a 96% success rate. Which is far higher than pretty much anybody had forecast.
One last failure due to an engine issue on ascent - technically not even a landing failure - and that takes us to the current string of 179 successful landings, which is astonishing.
What we hadn't realized was that every failure led to a modification in the hardware, the software, or the process. They worked through virtually all the ways to fail and once they had, the system was incredibly robust.
It was their flexibility in treating every flight as a prototype they could learn from rather than trying to design things up front that took them from successfully getting through the atmosphere to landing both on land and on the drone ship in less than 3 years.
And it also resulted in a design with the minimal impact on payload because they didn't add systems they didn't need.
(keep?)
There was another time that there were lots of rockets blowing up.
That was in the 1950s, when rockets were new and therefore any new rocket was a research program.
That made them quite explody. Those sorts of failures are much less common today, and when they do occur, they are generally related to engines, which still require a lot of research.
There's one other program that comes to mind...
Which takes us to starship.
At the start of Starship, SpaceX had a grand vision but little hardware. Here's what I think their first minimal viable product was:
It needed to be shaped like the final product would be.
It needed to be the final diameter - 9 meters - because the ground support infrastructure would need to support that diameter
It needed to be build out of stainless steel, the material choice for the final product
It needed to use a raptor engine
And finally, it needed to actually fly and land, so that others could see that it did what it was supposed to do.
Those requirements drove a bunch of other requirements - you need a working engine, you need delivery of liquid methane and oxygen, you need a test area, and you need to get approval from the FAA to fly the thing. And you need to deliver finished work that actually functions well enough for your tests.
The vehicle SpaceX created came to be known as Starhopper. It was an ugly brute, made out of ½" steel that was 3 times the thickness of later vehicles and hand-welded, but it flew and it flew well, with two flights in July and August of 2019. Their success should not have been a big surprise as the technical challenge for starhopper was low.
SpaceX learned a ton about everything starship, from how the Raptor engine worked to how to work with the FAA to how to integrate systems. And it allowed them to start the project with an obvious success.
Once you have minimal viable product finished, you work on the next one by adding things that make it more real. The next MVP would move to real tanks made out of thinner stainless steel, a full size vehicle, and fins while still doing everything that starhopper had done.
Mark 1 and Mark 2 came next, with one developed in Texas and one in Florida. Mark 1 popped during pressure testing and Mark 2 was abandoned. Mark 3 (aka SN1) popped during pressure testing.
Tanks were clearly harder than expected. SpaceX stopped doing the fin work and converted SN2 to a pure tank test, which passed. They went back to full-sized tanks.
SN3 imploded when the oxygen tank was depressurized. SN4 exploded after a static fire due to a failure in the quick disconnect system. Tanks were getting better but the ground support equipment still needed work.
Finally Starship SN5 passed all the tests and hopped, and soon after, SN6 did the same.
During this time, SpaceX was doing research on how to build and operate the tanks. It took a few tries, but they learned enough so that tank construction is no longer an issue. They had use the "many prototypes" approach to both understand how to build tanks and to build enough of a factory to be able to build prototypes cheaply, and it took them roughly a year to get to where they wanted to be.
This phase was really all about the factory and the ground equipment development; the two hops didn't do anything beyond what Starhopper had done.
We can make an interesting comparison of SpaceX's approach to another approach.
ULA has been working on their Vulcan Centaur rocket for a number of years. Vulcan marries a booster with BE-4 engines from Blue Origin with a new Centaur V upper stage, much bigger than the previous Centaur III upper stage and with thinner walls.
In March of 2023, they were doing a test to qualify the Centaur upper stage for all missions. This test would generally be considered a formality - merely a box to check on the way to launching a new rocket.
But a hydrogen leak led to an explosion that destroyed the stage and caused significant damage to the test stand. Subsequent analysis found higher loads in the tank than expected and weaker welds than expected.
It's not clear how much analysis and testing ULA had done on the new tanks, but it is clear that they didn't catch this issue early, and the subsequent redesign resulted in a delay of around 6 months.
The next MVP covers the landing segment of a starship mission. It is to takeoff, flip horizontally, control the vehicle during descent, flip back, and land.
The controlled descent seems straightforward, but restarting engines and then flipping right before landing has no prior art, so we would expect it to be challenging, and therefore... you guessed, explody.
SN8 took off on three engines, climbed to 12,500 meters, flipped sideways and fell back to ground under control, but crashed into the ground due to a propellant pressure issue.
SN9 was a repeat of SN8, crashing into the ground because of engine issues.
SN10 landed successfully, but the landing was a hard one and it exploded a few minutes after landing, so we can only give partial credit for that one.
SN11 is a bit of an enigma. It flew in heavy fog, descended, and exploded either near the ground or crashing into the ground.
SN12-14 were abandoned to go with SN15, a improved design based on the four previous flights.
SN15 landed successfully. Completing this MVP had taken 5 vehicles and was completed in less than 6 months.
People often ask why SpaceX didn't fly more of these flights. One reason is that they had all the data they needed and knew how to do it successfully - they had completed their MVP. A second reason is that they would be doing landing after reentry tests on future MVPs and those would give them ample opportunity to test descent and landing under real flight conditions.
The next phase involves the whole vehicle. There are a lot of new things here and the booster is an order of magnitude higher in thrust than starship tests. If you look at what SpaceX said before and after this test, I think the MVP was "gets off the pad". They planned in case everything worked right with water landings of super heavy and an orbital reentry of starship, but I don't think they really expected to get that far - their goal was "get a ton of data" and feed that data into the next prototype.
This test did not disappoint, it did get off the pad and make it kindof near staging but it seriously damaged the pad and the vehicle couldn't even blow itself up properly.
This flight frankly confused a lot of people, who expected something like the early flights of Falcon 9, which all worked great. Falcon 9 had lots of prior art and SpaceX was focusing on just getting to orbit, while starship is mostly new stuff and SpaceX is optimizing from the start. Many people suggested that SpaceX was failing, or that they should have done a lot more work before this launch.
I mostly disagree - they are building these prototypes to fly them and even if they aren't going to be perfect their way of developing starship involves a lot of imperfect prototypes. We've already seen the power of that approach with Falcon 9 landing development and the earlier starship MVPs.
Though I think in retrospect they should have done flame diverter work before the launch and I'm pretty sure SpaceX would agree with that.
Integrated flight test 2 was up next, and I think the MVP this time was "get through staging".
To add to the fun, SpaceX modified their staging approach to add "hot staging" - starting the second stage while the first stage engines were still running. There is some good prior art with hot staging - the Titan II rocket used it, as did the Russian Proton.
The big challenge on this flight was what happened after staging. As I talked about in my starship optimization video, the booster will by flying back to the launch site as quickly as possible to use as little propellant as possible, and that means a quick turn around. And this is also the first flight of Starship into vacuum and the first flight of the Raptor Vacuum engines.
The initial part of the flight was great, and hot staging worked.
Soon after staging, the booster started the flip maneuver to head back to the launch site and soon after that it exploded. This harkens back to the early explosions during testing of the starship flip maneuver.
Starship continued on, nearly reaching orbital velocity before it too exploded, marking this as the first time SpaceX has achieved two explosions in a single test.
It's tempting to view this as another failure, but it shows amazing progress over the first flight. The pad survived, all 33 engines worked through staging, and hot stating worked. The ultimate explosions appear to have been triggered by the automatic flight termination system when the vehicle was not on the expected trajectory.
We would all like to see the booster do a controlled water landing, but at this point in the research program figuring out how to do the new optimized boost back is much higher priority than doing the already-well-understood booster landing.
Going forward, SpaceX still has some big new things to make work, and because they are new things, there's a decent chance that some of them will be explody.
But my bold prediction is that Starship is going to follow the same pattern as Falcon 9 reuse. We'll see more failures before we see full success, but soon after that starship is going to start working pretty much every time and we'll all be wondering why we didn't see this ahead of time.
Explosions are good. Learn it. Know it. Live it.
If you enjoyed, this video, I suggest you go read John D. Clark's classic book, "Ignition - an informal history of liquid rocket propellants". You can read online for free. I'll put a link in the description.
In reference to chlorine Triflouride, he writes the following: (read)
And that's the space shuttle. Shuttle was full of new technology, including the high performance RS-25 engine, the external tank, the solid rocket boosters, and the very complex orbiter.
One can make a fair argument that shuttle is similar in challenge to Starship and it's interesting to look at how NASA ran the program.
Here's the timeline
The shuttle program officially started in 1972. In 1977 the test shuttle Enterprise flew a short series of 5 flight tests to evaluate the low speed performance of the orbiter. The word "prototype" is sometime used to describe enterprise, but enterprise was not a prototype as it could not be tested on actual shuttle missions
Then in 1981, STS-1 marked the first test flight. There were a few unexpected issues that could have been significant - see my video "the near tragedy of STS-1" - but NASA got lucky and fixing the issues was relatively straightforward.
They flew three more test flights and in 1982 declared the shuttle operational.
The RS-25 main engines were considered to be a very challenging design, and NASA specified that they must undergo at least 65,000 seconds of testing before the first flight. The actually achieved over 110,000 seconds of testing.
As is typical of engine development, there were many failures during testing, and most of them were explosive. I'm particularly fond of this post-test picture.
But all of that testing allowed Rocketdyne to develop an engine that was one of the most robust parts of shuttle, with only one failure in flight and just a handful of failures on the pad.
The solid rocket boosters were a larger version of a new technology, and they were well designed with accepted practices and tested.
Unfortunately, in actual flight the solid rockets bend and vibrate in response to aerodynamic loads on the whole stack, and the seals between segments did not perform properly in that scenario, and that ultimately led to the loss of Challenger.
27,000 hours of wind tunnel testing.
(redo this, split apart)
And that's the space shuttle. Shuttle was full of new technology, including the high performance RS-25 engine, the external tank, the solid rocket boosters, and the very complex orbiter.
NASA started development work on shuttle in 1972 and spent about 9 years in design, analysis, and construction. Shuttle cost about $48 billion in 2020 dollars to get to its first flight.
At this point, NASA had a problem. They were committed to building 4 orbiters with their initial, unflown design, and committed to putting astronauts on that first flight.
That flight was thankfully successful, but it could have gone differently - see "the near tragedy of STS-1".
NASA was left with 4 experimental orbiters that they needed to fly a lot of missions on and a very limited ability to fix issues that cropped up. That is one of the downsides of trying to do innovative things without sufficient real-world testing, and it ended up being one of the reasons for the Challenger and Columbia accidents. NASA had an orbiter that had ongoing issues but needed to fly it as if it was a finished production vehicle.
Shuttle is an indication that you can get rid of some of the explody issues through analysis but some still crop up during flight, and you would much prefer that they come up during test flights.
Attach that launch vehicle to a 100 meter centrifuge inside a vacuum chamber and spin it up to 450 RPM. That gives the vehicle a speed of 2800 meters/second and about 10,000 gs of centripetal acceleration.
SpinLaunch is an interesting new idea on reducing the cost of access to space.
The open question is whether it makes sense as a project. This is not an easy thing to determine and SpinLaunch has not shared many technical details, but I've found enough information to do a project analysis and answer some of the questions I've seen about it.
The simplest way to describe SpinLaunch is simply to go watch their video, which I'll link to in the corner.
The plan is to take a simple pressure-fed two-stage rocket, and encapsulate it inside this fairing.
Attach that launch vehicle to a 100 meter centrifuge inside a vacuum chamber and spin it up to 450 RPM. That gives the vehicle a speed of 2800 meters/second and about 10,000 gs of centripetal acceleration.
Release it at the right moment so it exits the vacuum chamber and heads out of the atmosphere.
At the right point, pop off the fairing, ignite the first stage of the two-stage rocket, and head off to orbit like any other rocket.
My first questions is whether it's possible - does it require advanced technologies that we don't have?
The answer is yes, it's possible.
Is it practical?
There are a number of different factors that go into answering that question...
Is it buildable?
I'm going to start with the ground launcher, which I'm going to call the spinner for obvious reasons.
It needs a 100 meter diameter vacuum chamber that would have a volume of approximately 78,000 cubic meters.
NASA owns the largest vacuum chamber, and it has a volume of about 26,000 cubic meters. It was built about 50 years ago.
It's not clear how much it will cost, but it certainly seems buildable.
Inside the vacuum chamber, there is a long arm that will spin the payload around.
There are no exact analogies, but the wind power industry is building turbine blades that are over 100 meters in length, and fabricating the payload arm seems feasible.
The launch vehicle is subjected to very high g loads when it is spun up and those g loads vanish when it is released.
This seems like a very hard problem, but there have been a number of artillery shells with electronics on board; the current variants contain full GPS systems and guidance actuators. They tolerate 15,000 gs when fired.
The launch vehicle will require new engineering to deal with those g loads, but it seems doable.
The spinner acts like the first stage of a conventional rocket, and do to an orbital analysis, we need to know what the spinner can do.
SpinLaunch has been issued a few patents, and one of them includes this altitude/velocity graph.
The patent claims their first engine burn will start at 60 seconds after launch, and reading from the graph, that would be at 62 km of altitude and travelling at around 2000 meters per second.
That's roughly in the same range as Falcon 9, which stages at around 70 km and 2200 meters per second.
Orbital velocity is about 7500 meters per second, and at staging the vehicle is travelling at about 2000 m/s, so - ignoring gravity losses - it needs about 5500 m/s of delta-v to get into orbit.
Can we get that 5500 m/s? We'll need some more information about the rocket.
They are using pressure fed engines, which are simple and robust; just pressurize the propellant and that pushes them into the combustion chamber. This makes sense as the tanks will already be heavy to deal with the g forces, and that means they don't need to develop high-g turbopumps.
They burn JET-A fuel rather than RP-1. That's a bit of a strange choice - RP-1 was created because Jet A contains a lot of sulfur and varies widely in composition, and that causes issues in rocket engines - but it may be that it doesn't matter much for a pressure fed engine with a very simple nozzle. It's certainly cheaper and easier to get.
Because JET-A will be close to RP-1, I'm going to assume they can get an ISP of about 310 out of their engines because they only operate in vacuum, though that may be giving them too much credit.
One surprise is that the rocket has a two stage design. Starting with a three-stage design and eliminating the first stage means that you still have two engines and two stages to develop. My guess is that they need two stages to reach their payload goal, but it adds considerable complexity to the project.
https://player.vimeo.com/video/573539093?app_id=122963&referrer=https%3A%2F%2Fwww.spinlaunch.com%2F
We can now play around with some numbers...
Assuming the following:
A payload of 200 kilograms
A required delta v of 5500 meters per second
Engines with an ISP of 310.
Now we need to estimate how much of each stage's mass is propellant and much is payload and structure.
I pulled some numbers from existing rockets
Electron is about 0.7 for the first stage and 0.8 for the second stage. Falcon 9 is about 0.63 for the first stage and 0.86 for the second stage. Based on those numbers I decided to specify a mass ratio of 0.65 for the first stage and 0.75 for the second stage.
Run that through a calculator based on the rocket equation, and the total mass for the rocket is 8220 kg, or about 73% of the vehicle mass. Note that I can push this number around by changing the mass ratio estimates; push them up by 5% and the rocket loses about 3000 kg of mass.
Is this buildable?
The rocket part is low-tech and therefore likely fairly easy to develop, with the caveat that it needs to survive 10,000 Gs of acceleration and still function.
The fairing is a big question mark; it's pretty big and it needs to handle heating loads, big g loads, and protect everything inside of it very well.
I'm going to say "probably buildable", but it's a lot of new engineering which means a lot of testing, and doing it as a two-stage rocket just makes things worse.
Is the SpinLaunch organization capable of building this?
Organizational expertise is really, really important. Organizations that overreach - they try things that are much too hard for their level of experience - are rarely successful. Hubris and delusions are grandeur kill many companies.
They did, however, do this.
I'll put the full test video link up in the corner.
A subscale test - and the fact that they built a subscale tester - is a good thing. The fact that there has been no news in the 3 months of future tests at higher velocity is less good.
I am disappointed by the amount of effort put into video advocacy, and would like to see videos that focus on the details rather than being sales presentations.
So... they rate a "maybe" in answer to this question, but that's honestly about all we could expect because nobody has tried this before.
What are the failure modes?
We understand rocket failure modes pretty well, and we know how to mitigate for them - how to keep the public and the workers safe.
But spinlaunch is a different approach with different issues.
The obvious failure mode is something happening before the vehicle exits the chamber.
My first thought is that the vehicle could miss the exit aperture and impact the chamber itself. Or perhaps the clamps fail and the vehicle comes loose when it is pointing towards the ground. Or it fails and comes out of the vacuum chamber going roughly sideways.
The amount of damage it will cause will depend on the energy of the impact. But before we calculate that, I thought it might be interesting to have a benchmark
The USS Iowa was a world war II battleship which fired a very large shell, seen here with 6 propellant bags.
The shell massed 862 kilograms and left the gun at 820 meters per second.
We know that kinetic energy is equal to one half the mass times the velocity squared, and if we do the math we find that the kinetic energy of the shell is 290 mega joules.
In addition, there's 70 kg explosive charge that adds 300 MJ, for a total of 590 MJ of energy.
http://www.navweaps.com/Weapons/WNUS_16-50_mk7.php
https://en.wikipedia.org/wiki/Dunnite
That amount of energy is enough to make a crater like this, about 15 meters wide and 6 meters deep.
Now that we have some context, how does that compare to SpinLaunch?
Their video gives a launch vehicle mass of 11,200 kilograms and other materials specify a velocity of 2800 meters per second at launch. In addition, the rocket contains 1600 kg of Jet-A fuel plus enough liquid oxygen to combust it quickly.
We can now compare the battleship shell's energy to the spin launch energy.
The spin launch vehicle has a mass of 11,200 kilograms and is going 2800 meters per second at release. Plug those values into the equation, and we get 43,900 MJ of kinetic energy, and we also get another 32,000 megajoules from burning the JET-A if we burn half of it.
That gives us around 75,900 megajoules.
Or 129 times the energy in one of those battleship shells.
This is actually a fairly small amount of energy compared to other launchers; the Falcon 9 first stage has 5.1 million megajoules of energy in the RP-1 it carries, or about 67 times the energy of the SpinLaunch vehicle. Rocketry is inherently a high energy business.
If the vehicle misses the exit port, there simply isn't enough mass in the chamber walls to do more than slow the vehicle down a small amount. The chamber will be damaged and remember that there is a lot of kinetic energy in the spinner arm and drive mechanism and it could do more damage as it slows down.
If the vehicle impacts on the ground side and runs into the earth, all of that energy is going to be liberated over a very small area. It seems unlikely that any of the facility would survive.
If the vehicle comes out to the side and stays somewhat intact, it could go a long way - naval guns have ranges of 40 km and the launcher is travelling much, much faster. If it breaks up the distance would be shorter but anything inline with the possible exit vector could be at risk. Maybe you could mitigate some risk by a very strong wall around the outside of the vacuum chamber.
Traditional rockets include destruct systems to deal with rogue vehicle scenarios, but with a vehicle travelling 2.8 kilometers in a second and likely already coming apart, it's not clear that they would help here.
It will be very interesting to read the safety analysis for the full-sized version.
The next question is whether there are any legal or regulatory blockers.
I don't think the FAA is likely to let them launch the full version out of their current location in New Mexico because of the normal restrictions on overflight, so that means someplace on the east coast. There are options at cape Canaveral but few options elsewhere.
They might need to do an environmental impact because their technology is different than what is already in use.
And they will have to do enough analysis to get a launch license, and that's going to be slow because it's a different technology.
Is it practical?
Probably. Given a reasonable amount of money - say, a billion or so - you can likely get there.
One way to understand viability is to compare a new entrant against a competitor. The obvious comparison is RocketLab's Electron.
We'll start with price: The electron price is about $7.5 million per launch. We don't know what spinlaunch is aiming for, but it needs to be cheaper than that.
Both are designed as 200kg launchers.
Electron can target a wide variety of different inclinations just by changing the direction of the trajectory. We don't know what spinlaunch can do; they could conceivably build their vacuum chamber on a large turntable to point in different directions, though that hasn't shown up in any of their materials. If not, they will be limited to the orbits they can get to starting with the base launch, and that will limit orbit choice and reduce payload.
Electron is a 2 stage rocket and SpinLaunch is a 3 stage rocket if you count the spinner as the first stage.
Electron is close to recovering their first stage and reusing it, and spinlaunch would obviously reuse the spinner
Electron is a very light vehicle, massing only 1400 kg without propellants, compared to my estimate of 2860 kg for the unfueled Spinlaunch rocket.
Their launch mass is surprisingly close; 12,500 kg for electron and 11,200 kg for Spin launch.
Finally, the maximum G load for payloads is 7.5 G for electron, and 10,000 G for spin launch.
I'd like to talk a bit about how launch companies differentiate themselves.
ULA's claim to fame is extremely high reliability. They know they aren't the cheapest, but for the US government customer they target, reliability is more important.
SpaceX's claim is to be cheap, innovative, and responsive. They will give you a price that is better than anybody else and they will be able to launch your payload faster than their competitors.
Rocket lab's claim is partly to be like SpaceX, but they have recently moved towards being a one-stop shop; they will handle everything that you need to get your payload in space and functioning.
That differentiation is what allows these companies to be successful; they know what their strengths are and what market they are aiming for.
There are a lot of small launch companies out there.
Rocket lab and virgin orbit are up and running. Firefly, astra, and relativity are in development or flying.
Where does SpinLaunch fit into this very competitive marketplace? How can they differentiate?
The short answer is that I don't see much to differentiate them, and this launch market is already very crowded.
Is SpinLaunch an attractive investment?
They are at the small end of smallset, and that limits their market
There are significant technical challenges in building the spinner, the fairing, and the rocket.
They require high-g satellites. I see very little reason for satellite makers to want to do this; it's work that only makes sense for spin launch as a provider. It might make sense for large numbers of satellites, but those customers can be served by larger launchers who can carry more at once.
There is already significant competition in this market, with some competitors already launching payloads, and no obvious attribute that makes spinlaunch different. Why would a customer choose them over one of the other companies?
Finally, the economics are uncertain. It's not clear that their launch vehicle will be cheaper than the other entrants, nor is it clear that they can launch enough to amortize their fixed costs. I was unable to find any information on their price targets.
Is SpinLaunch an attractive investment?
Not really; I don't see any reason to expect their approach to be more successful than the other approaches and they are late to the market.
I think people have focused on the uniqueness of the approach rather than looking at the benefits.
You might already know the story...
With the completion of the international space station, the landing of Atlantis at the end of the STS-135 mission marked the end of the space shuttle program.
That left NASA with a problem. The only way to fly astronauts to the space station was on the Russian Soyuz, and that both limited the number of astronauts on the station and made NASA look bad.
There had been a plan to launch astronauts on the Ares I launch vehicle as part of NASA's Constellation program, but that was cancelled in 2010 and the follow-on Space Launch System was not a viable option because it wasn't ready and it would be an extremely expensive option to get to the space station.
Traditionally, NASA had a very particular view of the spaceflight world...
We fly people, and they fly cargo.
NASA worked hard to keep their monopoly on human spaceflight in the US, but with the shuttle program over and nothing on the near horizon, NASA opted to redefine what "we" meant and the commercial crew program was created.
After some up-front development work, NASA selected proposals from two companies.
With the awarding of those contracts, the race was on.
There was even a prize - this American flag was flown on the first shuttle mission in 1981 and the last shuttle mission in 2011 and was left on the international Space Station to be "captured" by the next astronauts to be launched to the ISS from North America.
It was an interesting time...
Depending on who you listened to, Boeing was ahead, SpaceX was ahead, or the race was too close to call.
Finally, on May 30, 2020.
SpaceX Demo 2 took astronauts Bob Behnken and Doug Hurley to the ISS where they claimed the flag, narrowly beating the Boeing Starliner team.
Or so we thought...
It's now November 2023
SpaceX flew the first crewed mission to the ISS in 2020
They have since flown all 6 of the missions contracted for in the original commercial crew contract and the first of 8 additional missions purchased by NASA.
In addition, crew dragon flew the inspiration 4 free-flying mission and two Axiom Space missions to the ISS.
That's 11 crewed missions carrying 42 passengers.
Starliner flew once in 2022, successfully docking with the international space station on their second orbital flight test and setting the stage for their first crewed flight.
What the hell happened to Starliner and what does the future look like?
We'll start with the first part, and begin by looking at Boeing...
I grew up in the Seattle area and worked for a few years for Boeing Computer Services out of college.
At the time, Boeing made great products and this was a common sentiment when travelling.
But since the merger with McDonnell Douglas, they haven't been the same company.
There are a bunch of different theories about why the Starliner delays are Boeing's fault and it's worth the time to examine a few.
One idea is that boeing bid on commercial crew believing that SpaceX would fail and they would later be able to extract more money from NASA.
There's certainly some truth here - in 2016 Boeing proposed higher prices for flights 3-6 of their 6-flight contract which NASA rejected as they did not align with the prices Boeing had given in their original contract proposal.
NASA made a counter-proposal, and eventually ended up agreeing to pay an extra $287.2 million towards their commercial crew contract.
When this was revealed in 2019, Boeing's response was this (read).
Yes, they got extra money for those missions to ensure that missions 3-6 would be ready to fly on time.
I'm going to vote "yes" for this assertion since Boeing did extract the promise of more money from NASA.
The idea here is that Boeing did not understand the world of firm fixed price contracts.
I have long asserted that companies are designed to do specific things and operate in specific markets. For example, ULA was created specifically to sell expensive launches to NASA and the department of defense.
I think Boeing got confused by the early part of the program, where they received $580 billion in space act agreements, which you can think of as being very much like the cost plus world that Boeing is comfortable in, and presumably Boeing spent a lot of this money coming up with a plan that they believed they could execute on.
They then signed the firm fixed price contract, one that would only pay them at specific milestones.
My confidence in this theory is high, because Boeing very recently admitted that they are struggling to make money on firm fixed price contracts for satellites, for the KC-46 refueling aircraft, and for Starliner.
Many people assert that Boeing did not build the right kind of team and spend the right amount of money to build a crewed capsule.
We frequently hear that space is hard, but the real question is whether the issues you are running into are ones you should have found or are you finding new and innovative ways to fail.
I could not find a definitive list of Starliner issues but if we look at this list, it's quite damning - if your QA is missing issues like the parachute problem or the orbital flight test software issues, there's a case to be made that you have no business developing a crewed space vehicle.
This is clearly one of the big problems, and perhaps the major problem.
I also suspect that Boeing came to the wrong conclusion about SpaceX's success with cargo dragon
The right conclusion is that cargo dragon is a fully functional and robust capsule that SpaceX developed quickly and therefore they have a head start on getting a commercial crew capsule done.
That means that Boeing would need to up their game to keep up with SpaceX.
Unfortunately, their conclusion seems to have been "SpaceX did it so capsules can't be very hard".
Those are the main issues where I think Boeing is directly responsible. But there were some other issues that they had less control over...
One unappreciated cause of delays is scheduling complexity, and that comes from the design of the international space station and a specific decision that NASA made.
This is a view of the ISS, showing us the harmony module also known as node 2.
At the part of the station facing the earth - or nadir side - there is a SpaceX version 1 cargo dragon. These cargo dragons did not dock with the space station, they were berthed - or attached - to the space station using the robotic arm.
Docking is more complicated.
Station uses a pressurized mating adapter to connect US and Russian segments and this was also where shuttle docked. After shuttle, NASA came up with the international docking adapter which connects to the PMA to support the new international standard.
Node 2 supports two of these adapters, one on the zenith or top port, and one on the forward port. You might see them referred to either as a PMA or an IDA.
If you're curious about the numbering, IDA-1 was lost during the SpaceX CRS-7 launch failure.
These adapters are the only places where commercial crew capsules can dock.
For the second phase of commercial resupply services, SpaceX gave NASA two bids...
SpaceX could keep flying the existing cargo dragon, or SpaceX could fly a cargo variant of the Dragon 2 that was being developed for commercial crew.
NASA chose the updated version.
The outcome of this decision was that there was now another vehicle that could only dock to those two ports.
Let's see how that impacts scheduling.
There are two ports on node 2, and here's how they were occupied with crew flights during this time period. There's roughly a week overlap where both ports are full, but there appears to be lots of empty space for Starliner test flights
There are, however, short visits of cargo dragon in those open time slots, and those are busy times on station because the astronauts need to prepare cargo to return to earth, unload and reload dragon, and then do whatever needs to be done with the new cargo.
There also may be short Axiom commercial visits that take up a port for a week or so.
To further complicate scheduling, there are Cygnus resupply flights. They do not take up a crew docking port, but they still require scheduling and planning.
Add in the Russian progress resupply flights and the Soyuz crewed flights - both of which impact everybody on the station - and it becomes clear that scheduling is very complicated. Starliner is not making operational flights, so it has lower priority and NASA can only give it limited windows to visit station. If there's a slip, the next open spot may be many months in the future, and therefore there are no small slips for Starliner.
SpaceX didn't have this problem with crew dragon as there was always an open port they could use and NASA really wanted crew flights as soon as possible. And it won't be an issue when starliner is operational as they will alternate flights with crew dragon. But it does cause delays now.
NASA has long had a somewhat strange approach to crew safety. For a detailed discussion, see my video "You're gonna kill astronauts - reliability and crew safety at NASA".
For the sake of this discussion, I think this is a fair description of how NASA views human rating - NASA has a process that they follow and at the end a human-rated system pops out of the process. Here's an *overview* of the entire process, and of course each step is long and involved.
With commercial crew, however, the goal is to be more contractual. The model is that NASA will send requirements to the contractor, and then the contractor will define a certification plan that details how they will verify that they meet those requirements or waiver requests detailing why it doesn't make sense for them to meet a specific requirement. After some back-and-forth, the certification plan is approved.
The point of this approach is to allow contractors the flexibility to choose the verification approach that works best for them. For example, SpaceX chose to perform an actual in-flight abort test for crew dragon, while Boeing chose to verify that ability through analytical methods.
During development, the contractor sends information about their progress and NASA maintains safety oversight over the program as a whole.
This seems like a reasonable approach, but there are a few problems.
The first is that at the time that SpaceX and Boeing bid on the commercial crew program, the requirements didn't actually exist. NASA was forced to go with the commercial option and they weren't prepared ahead of time. In some cases, the NASA requirement was aspirational - they settled on a reliability requirement of 1 loss of crew event in 270 flights without determining whether such a requirement was achievable.
NASA was also unprepared to review the certification plan and the hazard reports generated by the contractors. NASA's goal was to complete reviews within 8 weeks but the contractors had to wait as long as 6 months for some reviews. NASA did not monitor the timeliness of these reviews.
The slow responsiveness of NASA stretched the timeline directly and also increased the chance that their response would require costly rework later in the cycle.
Congress chose to underfund commercial crew.
This somewhat confusing chart from NASA's inspector general report shows the details.
For the four years indicated, congress provided significantly less funding than NASA requested, providing $1.9 billion instead of the $ 3 billion NASA requested. This had a significant effect in slowing the program down, especially in slowing down NASA's ability to move quickly early in the program.
There was a covid-19 effect
Boeing has 425 suppliers across 37 states involved with Starliner, and when the COVID-19 pandemic hit their ability to get what they needed was disrupted.
SpaceX doesn't give out supplier data but they are well-known for doing much of their work in house and that left them less vulnerable to disruption.
And now for the question many people are asking...
Why hasn't Boeing cancelled starliner?
Let's look at some possible explanations...
Prestige is a factor; Boeing views itself as a big player in defense and space and having failed at a crewed capsule would reduce their prestige.
All government bidders get ranked not only price but on their capability to execute on projects successfully. Boeing got high grades on their commercial crew proposal and that made it easier for them to get a contract.
Cancelling Starliner would make it harder for them to win future contracts as it would their management grades.
There is a possibility that Starliner can garner additional business.
The real reason comes down to money. Continuing with Starliner appears to be the best decision monetarily.
I'll get into why in a moment, but I want to address two related questions first.
The first is whether Boeing made the right decision to sign the commercial crew contract, and it's pretty clear that this point that Boeing thinks the answer is "no". They have sworn off firm fixed price contracts.
Will they make a profit off of starliner? They clearly have not made a profit *yet* and it's probable they won't, though profit is still possible.
But neither of these questions is the right one to ask at this point.
Those questions are looking backwards. They are interesting from an analytical standpoint but they aren't useful in figuring out what to do in a specific situation. They need to be looking forward.
The question they are asking - and have been asking throughout the delays - is whether there is a good chance that going forward will result in useful amounts of net cash?
And it's pretty clear that Boeing still thinks that is true. And here's why...
The fact is that the operational part of the starliner contract is a cash cow...
If you look at the commercial crew contracts you'll find that much of the interesting information - including monetary amounts - has been redacted, but there's also this pattern of either NASA's office of the inspector general or the general accounting office releasing some of that data.
The OIG estimate was that the starliner cost per seat was $90 million. Each capsule carries 4 astronauts and Boeing has a firm contract for 6 flights, so that's $2.16 billion for those 6 flights.
That is a lot of money, more than half of the total amount of the contract.
And it's actually more than that.
There is the payment that Boeing gets when they successfully complete their demo mission to the ISS. We don't know how much that is, but it is likely substantial, probably somewhere in the $250 - $500 million range.
It's also possible that the $287.2 million dollar increase Boeing negotiated for some flights is not included in the $90 million estimate.
That yields an overall estimate at somewhere between $2.41 and $2.95 billion total, or $400 - $490 million per flight.
That is why Boeing has been absorbing losses along the way and still going forward - the contract is specifically designed to put the profit in the operational phase and give the contractors a big incentive to get there. And NASA putting in a firm order for all 6 flights made the possible money bigger.
Boeing may not make a profit on the project overall, but their cash flow situation will be much better with $2 to 3 billion in revenue from it than with zero revenue. At least that's what they're betting.
If you enjoyed this video, please send me a Boeing CST-100 Starliner Air Brush Quarter Zip Shirt, size medium, on sale at only $37.99
Starliner Post Mortem:
https://www.youtube.com/watch?v=JLjR23twwDE
Boeing CST-100 Starliner Crewed Flight Test (CFT): Anatomy of the Thruster Doghouse
https://www.reddit.com/r/Starliner/comments/1eiggns/boeing_cst100_starliner_crewed_flight_test_cft/
A while back I did a video on Starliner (see link in the description).
I made two major points. The first is that Starliner delays are partly the fault of Boeing, partly beyond their control.
The second was that the commercial crew contract was deliberately structured to give Boeing a big incentive to finish development and start flying flights, and that there are at least 2 billion reasons why they have kept the program going for so long.
Before we get into the current state of starliner and what the future of starliner might look like, I have some information that I need to share...
My finances are tight.
I can barely afford the coke zero and pork rinds that keep me going, much less the fresh chicharones that I prefer.
Recent responses have been less than I have hoped for.
So I propose a simple agreement. I'll promise to keep making videos, and you promise to keep doing your part. I think that's a win for both of us.
And if you chose not to, I'm not responsible for what might happen...
Thank you for your attention to that important announcement. We now return you to your original video.
This past summer, we finally saw a crew climb into a starliner capsule on top of an Atlas V, launch into orbit, and dock with the international space station.
That was great.
What wasn't great is that Boeing has clearly not fixed the issues that they had run into on earlier Starliner flights, with helium leaks and thruster failures.
The problems put starliner in the doghouse. And the problem was with a section of the service module known as the doghouse.
There are 4 doghouses spaced evenly around the service module, and they are jam-packed with what is technically known as thruster stuff.
Why is this such a problem? It's due to a few choices made by the Starliner designers...
To understand these choices, it's useful to compare the two commercial crew capsules.
Both have an upper capsule part and a lower non-capsule part.
The SpaceX non-capsule part is named the trunk, and by analogy with a car it's a place that you can carry extra stuff that you can't put in the passenger compartment. For cargo dragon flights, it often carries large cargo that can tolerate the vacuum of space, such as these roll-out solar cell arrays. It's also a nice big surface for solar cells.
For crewed flights, the trunk modifies the aerodynamics of the capsule so that it remains aerodynamically stable - with the pointy end forward - during abort scenarios.
SpaceX chose to make the trunk simple because the trunk burns up during reentry and all the expensive stuff returns with the capsule to be reused.
Starliner follows the design choices made with the Apollo and Orion, with a complicated service module. That service module contains propellant tanks, abort thrusters, reaction control thrusters, and a whole lot of plumbing and valves to control them all. The solar cells are mounted on the bottom of the service module.
That choice makes it easier to design the capsule part because it shifts a lot of mass to the service module and therefore the capsule and heat shield can be smaller and lighter. But it has a significant disadvantage, one that came back to bite Boeing. With the dragon design, all of the important parts come back for analysis and inspection. SpaceX has had a couple of issues with thrusters on Dragon but has had the luxury of looking at the thrusters after the flight to identify the problem.
Most of the starliner issues after the first flight have been thruster issues in the service module, and since the entire service module burns up on reentry, there's no way to do post-flight analysis. That is why NASA kept delaying a decision for so long - they wanted to give Boeing every chance to gather more data from the misbehaving systems.
The situation is complicated by the fact that while Boeing is in charge of Starliner, the service module propulsion system is built by Aerojet Rocketdyne.
Aerojet rocketdyne supplies 20 orbital maneuvering engines, 28 reaction control engines, and 4 launch abort engines for every service module. In addition, they also supply 160 valves, 18 tanks, and 500' of ducts, lines, and tubing.
The problems in Starliner have led to a lot of finger pointing between the two companies.
NASA ultimately decided to send starliner home without a crew on board, and the capsule landed successfully.
I've complained in the past that NASA has grown far too risk averse and pointed people to Rand Simberg's excellent ebook "Safe is not an option".
But in this case NASA made the right call. Many people asked the question whether dragon was safer than starliner, and the reality is that Dragon has significant flight heritage and Starliner does not and by definition that makes Dragon safer even if starliner was behaving perfectly. It's the wrong question to ask.
The right question to ask is whether Starliner is operating as designed, and the answer to that question is a clear "no". NASA made the right call - you simply do not put crew on a vehicle that is misbehaving unless you are really, really sure that you understand what is going on. That clearly wasn't the case here. The point of margins and redundancy in systems is to help you out when something unexpected happens, not make it okay to fly when you know there are problems.
In the post mortem video, I was somewhat optimistic that Starliner might be successful.
I no longer have that optimism. The crew flight test was the third orbital flight of starliner and Boeing had 25 months to make it as perfect as they could, and they ended up with 5 helium leaks and 5 thrusters that were degraded enough the software marked them as failed.
I think that's a pretty good sign that there is a significant problem.
There's a good analysis of the thruster doghouse from a few months ago that asserts that the underlying problem is that the thrust doghouse overheats because the thermal analysis was done incorrectly. Link in the description.
I'm very much not a thermal engineer, but putting 11 thrusters - by my count - along with a bunch of valves and other plumbing inside a sealed box seems like a poor choice from a thermal perspective. This is also an an atypical design.
If we look at the orion service module, we see that none of the thruster bells are inside the service module.
Dragon does put thrusters inside the capsule, but the RCS thrusters are in groups of 3 and spread out. The high power thrusters are under the front cover of the capsule and are spread out.
It seems likely that the doghouse design is flawed with issues didn't get caught during design review, and testing did not adequately simulate the conditions in orbit. It's not clear what might be required to fix it. It might be as simple as moving the thrusters so most of their heat-producing parts are outside the doghouse, though even a change like that would require a lot of work and retesting. Or it could be worse.
Given that background, what does the future hold for starliner?
The first question is whether the crew flight test leads to certification of Starliner by NASA.
I pulled up the Boeing contract, and found this. The first obvious question is whether carrying crew up but not down counts as successful completion of a crewed test flight. I did a bit of searching and could not find a definition for successful completion.
If the flight does count as successful completion, there are a few relevant requirements to achieve the operational readiness review milestone before operational flights could start.
Section d requires closing all the action items from previous reviews.
Section f would seem to require that Boeing figure out what has been going on with the thruster failures and helium leaks and provide evidence that is good enough for NASA to accept.
Section J is about the risk assessment that is required to meet the 1 in 270 loss of crew probability required by the commercial crew contract.
The crewed test flight was hoped to be a demonstration of F, that the failures and anomalies from the uncrewed test flight had all been resolved, but instead in turned out worse than the previous flights, with multiple helium leaks and the same thruster issues as before.
It's not clear what sort of evidence NASA might require in this area.
In a great example of how procrastination can work for you, NASA released a commercial crew plan update on October 15th as I was writing this section.
NASA has decided that both 2025 crew flights will be flown on crew dragon, with one flight in February and one in July.
Crew 10 was already known and crew 11 is not really a surprise.
NASA also talked about the future of starliner...
(read first paragraph)
I read this as saying "we don't know enough yet to know what path Boeing will need to take to get Starliner Certified".
They also say:
(read second paragraph)
Since NASA has already awarded both 2025 operational flights to SpaceX, any starliner flight in 2025 would be a repeat certification flight. But the "options on the table" part seems to indicate that there might be a world in which Starliner could be certified without a 4th test flight.
What happens next is up to Boeing. I think there are three scenarios to consider:
The best case is that starliner can certify without reflight. Their first chance to fly an operational flight would be early 2026, which puts off the revenue they want for at least another year, though there is likely a certification completion payment. This also means they have 4 years to get in 6 operational flights before ISS is deorbited. That may cut down their potential revenue, though ISS might be extended again.
The second is that Boeing figures out what is causing the issues, implements a simple fix, and reflies the crew test flight in 2025. This is roughly the same from a program perspective except that it requires a new service module and finding another Atlas V for the launch (presumably, they could buy one that is currently allocated to Amazon Kuiper launches).
The third is that the doghouse needs a redesign to fix the issue. This is probably the worst, requiring a time consuming redesign and almost assuredly another test flight, plus a significant amount of rework to prove to NASA that the new design will do what it is supposed to do.
I generally make it a rule to avoid guessing what will happen in cases like this because it depends both on how NASA views things and how Boeing views things, and we don't really have much data about either. And it's clear that Boeing has far bigger issues elsewhere and that starliner will not make or break the company.
I've decided to make an exception in this case, with the caveat that my opinion is a poorly-justified guess.
The question is whether Boeing will continue to fund the Starliner program or whether they have had enough and cancel.
40% that Boeing stays in the program if they have a clear path to operational approval. Note that that's a clear path which means understanding what is going on, knowing what to do about it, and having a solid schedule.
15% if Boeing needs to refly a crew flight test. They really want that operational money but so far throwing money at this program has not been a recipe for success. If their fix isn't good, they could be in the same situation and a few hundred million dollars poorer.
5% if the doghouse needs a redesign. I think that puts the next flight at least 2 years out and bumps operational flights closer to the end of ISS.
My best guess is that Boeing is going to throw in the towel, but it will obviously depend on what the path to certification is.
Is it possible for Boeing to just walk away. Commercial crew is heavy on the carrot in that the operational flights can be quite lucrative, and from what I can tell it's also light on the stick; I've read the contract and I don't see any penalties for just walking away.
Boeing has so far been motivated by the prospect of that sweet NASA money and by wanting to keep their reputation intact, but commercial crew is only a sideshow when it comes to factors affecting their reputation and money to be made and I would not be surprised if they decide to just walk away.
And that's the current state of starliner.
If you liked this video, please send somebody a starliner bear.
Starship Serial number 15 flew and landed on May 5th of 2021.
Since then, all we gotten is...
Crickets...
It's all about how risk is managed.
At the start of Starship, SpaceX came up with a shopping list like this:
We'll need a big stainless steel structure, a hot new rocket engine, a way to control the second stage, a second stage heat shield, a way to land the second stage, a reusable first stage, avionics to control everything, a launch site, and, of course, a wheel of cheese.
Once you have the list, you go through each of these and estimate how hard / risky they are going to be, based on the experience that you have as a company and the prior art elsewhere in the industry. We'll use a scale from 1 to 10.
Cheese gets a 1; we know how to get it and where to get it.
A reusable first stage is probably a 3; Super Heavy is mostly just a big version of the Falcon 9 first stage, and we have a ton of experience with that. Avionics is also probably a 3, as we have lots of experience.
A big rocket structure is likely a 5. There's a lot of experience using stainless steel, but it's a new material for us and this is a big rocket. We'll need to do a lot of work but there are unlikely to be any show-stoppers. Same with the launch site; lots of big structures but probably no show stoppers.
Controlling the second stage in the air and landing it are probably around 7. We can model the aerodynamics of the fin controls that we are planning, and while the landing requires a unique maneuver, landing with engines is something we already understand well.
That leaves the new engine and the second stage heat shield as the hardest problems we have to solve. The full-flow staged combustion cycle that raptor uses has never been flown so there's not much prior art and we need it to have really high performance early in its lifetime. That's definitely a 10.
We have one example of a large vehicle with tiles in the space shuttle (okay, the soviet shuttle buran probably counts as a second), and the approach on shuttle wasn't really a success, and our vehicle is really big and needs to be light. Definitely a 10.
This defines the risk profile for the project; the higher the number, the more that item keeps you up at night. We can start work on the tanks and the engines and the avionics.
Then you start iterating. After a while, SN5 hops successfully. What did we accomplish, and how did that change our risk structure?
We've built a number of full-size tanks and they seem to be working pretty well at this point. We've learned a lot, and we now feel much more confident. So, that risk goes from a 5 down to a 3.
Our engine is working well enough for this sort of hop, though there's still a lot of work. Let's say, it goes from a 10 to a 7.
The next 4 we didn't do anything on, and let's say the avionics went from a 3 to a 2.
We call this process "retiring risk". We had risk that we are worried about, but now we are less worried, so some part of the risk has been retired and we won't worry about that risk any more. Obviously, our long term goal is to retire all of the risk.
Looking at our list, what looks scary?
There are 3 second-stage items that are very risky. We can't do the heat shield right now, but we can work on the control & landing items. Which led us to SN15, which flew to about 10 kilometers and landed successfully.
We've now flown multiple engines and they have been working okay, so let's knock the engine risk down to a 5.
Our fins work great in the lower atmosphere but are still unproven in during reentry. Let's drop them to a 4.
Our landing worked successfully. It took a few tries, but we know how to do the flip and we got the landing to work. I'm going to put that down to a 2.
We have retired a lot more risk.
Moving forward to the present day, here is roughly where I see the risks.
Our overall profile looking much better, but we have also reached a point where making progress on the highest risks is going to require an orbital flight. In particular, the heat shield is very risky - it is what we refer to as the "long pole", or the item that is most likely to cause us to be late.
Working on items that aren't the long pole has very limited ability to get the long pole done more quickly, but some of them do have the ability to delay the long pole further.
In particular, if we worked on first or second stage landings, we could damage the launch infrastructure and that could delay tests of the second stage heat shield considerably.
Y no hop?
How will starship deal with failures?
What are the contingency plans, which are sometimes known as abort modes?
Starship is still under development and we therefore don't have all the information we would like to have, but we can still use the information we do have and make some guesses, hopefully educated ones.
For this discussion, we're going to look at four different launchers and contrast their abort systems and options.
We start with the starship crew vehicle launched on top of the super heavy booster.
Next is the Orion capsule launched on top of the SLS rocket
Third is the Crew dragon capsule on top of the Falcon 9 rocket.
Last is the space shuttle orbiter on top of... well, next to the rest of the space transportation system.
The first thing to notice is two very different approaches.
Orion and Dragon use conventional capsule designs launched on top of two-stage rockets.
There's one interesting difference between Orion and Dragon.
Orion uses solid rocket motors in the escape system. It has an abort motor that fires upwards and the exhaust is then redirected down to lift the capsule up.
At the top there is an attitude control motor that is used to steer the path of the orion capsule
And in between, there is a jettison motor that is used to jettison the launch escape system when it is no longer needed.
Crew Dragon uses four liquid-fueled hypergolic super draco engines that are built into the capsule structure. Steering is handled by differential thrust on the motors. There is no need for a jettison motor because there is nothing to jettison.
The shuttle orbiter and Starship have an integrated crew/cargo/reentry vehicle design, where the crew vehicle is the second stage.
Our goal at launch is to get into whatever orbit we're aiming to reach.
The biggest concern on launch is a failure in our propulsion system that results in less energy than we had planned for. The abort options depend on when the failure happens and the design of our system.
If our engine issues mean that we can't get to our desired orbit but can get into a lower orbit, we can use the "abort to orbit" option.
All of these vehicles should be able to do an abort to orbit in some scenarios.
On STS-51-F in July of 1985, engine sensors failed on the center RS-25 engine in challenger and that led to an engine shutdown at 5 minutes and 43 seconds into the flight. The lack of performance meant that the shuttle could not reach the 385 km orbit that it was aiming for, but it was able to reach a 265 km orbit and complete its mission objectives. This was the only abort during the Space Shuttle Program.
If the engine performance isn't enough to make it into a stable orbit, it might be enough to make it around the earth once. This is known as "abort once around".
There are several different scenarios depending on the inclination of our target orbit.
If we are launching to our natural inclination - the same inclination as our launch site - it's fairly simple. This picture shows the ground track of such a launch. We launch from Florida, head around the world, and the next orbit will be close enough to the launch site so we can land at it or next to it.
If we launch to a higher inclination - say the 51 degree inclination for the space station - it gets more complicated. We've launched from Florida to the north east, and the earth rotates as we complete the first orbit offsetting us quite a bit to the west.
This is why the shuttle maintained a landing site at Holloman Air Force base in New Mexico - it is roughly under the shuttle on an abort once around mission to the international space station.
If we launch on a polar orbit to the south from Vandenberg Air Force Base in California, we have a bigger problem. Our orbital track moves to the west, so that the next orbit we are considerably out in the pacific ocean.
The air force wanted to be able to do this with the Space Shuttle - to deploy a satellite immediately and return to earth, so the space shuttle has the ability to fly a curved reentry track and make it back to the US mainland to land. This is known as "crossrange" ability, and the shuttle could travel about 2000 km to one side or the other.
This capability was never used as the shuttle never flew out of Vandenberg.
The shuttle was capable of doing abort once around from a variety of orbits. What about the other vehicles?
Capsules are built so that their mass is not evenly distributed which provides some lift and therefore some crossrange. Estimates for Dragon are around 300 km, and 400 km for Orion. This is a bit problematic; you would generally want to put your recovery fleet in the location where an ascent abort might lead to landing, but oceans are big and astronauts on either Orion or Dragon could be in for a long wait before somebody reaches them in an abort once around scenario. They will probably survive but may not enjoy the experience.
What about Starship?
It has active aerodynamic control and quite a bit of body area, so it's definitely going to have a glide ratio of...
Yeah, I don't know. I do know that it doesn't glide in the sense that the shuttle glides, but neither do orion and Dragon.
If I had to guess, I would bet that it's greater than Orion and Dragon simply because it's much less dense - it's mostly a big empty tank. It's also less dense than shuttle, so maybe it will have more crossrange than I expect, but between the capsule and shuttle is a fair bet.
If we can't get around once, we will need to land before we've completed a full orbit.
For the shuttle, this was called Trans Atlantic Abort, and NASA staffed one or more airport landing zones in Europe or Africa, depending on their launch trajectory.
The capsules can land anywhere there is enough water, which probably would mean in the Atlantic for shorter range landings or maybe the Indian ocean for longer-ranged landings. Some inclinations spend quite a bit of time over Africa and that might limit the number of water landing sites.
Starship is obviously designed to land on land, and it will probably be capable of landing in any big flat solid area. And it has a lot of delta-v, so making it across to Europe in this scenario is likely.
However, there is no reason that it couldn't land in the ocean if it had to, as Falcon 9 demonstrated that ability. Given a sealed crew section, it's probably as survivable as an Orion or Dragon landing away from the recover fleet.
This means that Starship can essentially land anywhere from an abort perspective, which gives it more flexibility than the other systems.
As things get worse, we lose the ability to travel farther.
For shuttle, that meant trying to get to any airfield that was in range for an ECAL or BDA abort.
For high inclination flights - such as to the international space station - the shuttle flies up the east coast (that's what ECAL means) and if there's a major issue it can try to use its crossrange ability to get to one of these airports. For lower inclination, the only real option is Bermuda, and there are some abort scenarios where that isn't feasible.
If something happens early in the launch and you need to land right away, things get more interesting.
The capsules very likely just trigger their abort systems and land in the water near the launch site. Easy enough, though they will need somebody to do recovery wherever they land.
Shuttle has issues. We can't do anything for the first 2 minutes until the solid rocket motors burn out, and that gives us a lot of forward velocity. Unfortunately, not enough to get to anyplace useful, so we are stuck with the infamous return to launch site abort, or RTLS. To have a chance at a landing, the shuttle needs to arrive at this circle at the proper angle and with an empty external tank so that it can separate. To do that it needs to burn off its fuel before it reaches that point, which it does by placing itself in a fuel wasting attitude - something kindof like hovering.
In 1980, STS-1 commander John Young said, "RTLS requires continuous miracles interspersed with acts of God to be successful"
What about starship? Well, assuming an early problem with super heavy, Starship has loads of delta V and can easily turn around and do a nice lofted trajectory back to the launch site - the same sort trajectory the Falcon 9 first stage uses for RTLS. With one caveat I'll talk about later.
Onto pad abort. For shuttle, there was no pad abort
Dragon can abort off the pad, and it will likely go well for Dragon as the super dracos work quite well.
It's not clear how things will go with Orion:
Here's a launch from 1997 of a Delta II rocket with 9 solid rocket boosters.
The SLS uses very large solid rocket boosters, and as we see, a failure can throw burning solid propellant high in the air, perhaps high enough to reach the capsule or the capsule's parachutes. NASA models suggest that this is unlikely though there is no consensus inside NASA about this issue.
So Orion gets a checkmark with an asterisk
What about starship? Well, it gets a bit complicated because we don't know what the crew starship configuration will be.
The design seems to have converged on 3 sea level raptors in the middle of the stage.
But it's not clear how many vacuum raptors there will be; the original design called for 3 but Musk recently tweeted that they would go with 6.
This has significant impact on the abort scenarios because the number of engines determines the power to weight ratio and that determines how quickly starship could accelerate during an abort.
This graph looks at different payload sizes and different numbers of engines. The usual caveats about everything related to Starship being an estimate apply here....
Starting with the 6 engine blue line, if we are planning a full 100 ton payload, the thrust/weight ratio is 1.002. Start up 6 engines and starship will just hover on top of Super Heavy for a few seconds before it burns off enough fuel to start moving upwards. That will heat up the top of Super Heavy considerably, probably enough to rupture the propellant tanks. Not a great idea.
If we are willing to accept less payload - 25 tons - we can get the power to weight ratio up to a little over 1.5, which will allow for a reasonably quick escape.
With 9 engines, we get a power to weight ratio of 1.5 with full payload, and 2.25 with a 25 ton payload.
For reference, the Dragon escape system will generate 4.5 g of acceleration, and the orion 7 gs of acceleration. Those systems will trigger very quickly, while the raptors on Starship will take some time to generate full thrust.
One more complication - to start the engines on the second stage before the stages are separated, the exhaust from the second stage engines needs somewhere to go. The Russians use this design on some of their rockets, including the Proton shown here, and the US used it in the past on the Titan II missile.
There's one more complication for starship.
Crew starship will start with a mass of at least 950 tons. It is designed to reenter and land with about 150 tons.
That's about 800 tons of excess mass, and while some will be used up in the abort maneuver, much of it will still be around. We'll probably see a lot of this:
Starship likely can do a pad abort, but it's going to be slower than Orion or Dragon and it's going to be spending time getting rid of a lot of fuel, so it gets a check mark with two asterisks.
What are the launch timelines and options for each vehicle?
The space shuttle burned its solid rocket boosters for 120 seconds. It is not possible to start an abort during that time period and there are no abort options if the solid rockets malfunction.
The abort options for a single main engine failure depend on when the failure happens; if it's early in the launch RTLS is required, then TAL and ATO. After about 6.5 minutes, no abort is required due to a single engine failure. Multiple engine failure scenarios get more complicated; if you want all the details, see my Space Shuttle Abort Options video.
For the Falcon 9, the capsule can abort at any point from the pad all the way into orbit, though the Super Draco abort thrusters may not be necessary later in the launch. The Falcon 9 first stage has sufficient performance to deal with one engine failure during it's flight, but there is no redundancy in the second state engines. Late in the launch there are other abort options, such as abort once around or abort to orbit.
SLS has a launch escape system for the first 3 and a half minutes, and that covers the time that the solid rockets are running, so it's better than shuttle in that regard. It does have less escape system coverage than the Super Dracos on Dragon, but there's probably no significant difference in safety. Like shuttle, it will be tolerant of engine failures late in the launch. Interestingly, for Artemis 1 at least, the ICPS second stage isn't actually used to get into orbit; it's used to get out of the orbit and head towards the moon, so it doesn't impact the ascent abort scenarios.
Super Heavy provides significant engine redundancy, with the ability to tolerate 3 engine failures and complete its mission. I also expect that starship will be able to abort during the time that Super Heavy is operating. Starship will tolerate some engine failures on ascent; how many it can tolerate will depend on how many engines SpaceX puts on the crew version, which engines fail, and when the failures happen, but my guess is that 2 engine failures will likely be okay if there are 9 installed.
Starship will also have what the shuttle program called "intact aborts" - other options if something significant happens on ascent.
One more complication...
Whatever abort system is present needs to handle a number of different scenarios; with aborts at different speeds and different altitudes depending on the kind of abort.
We can now move to talk about reentry aborts.
There aren't any.
That's not quite true. For shuttle, if you manage to make it through the hot part of reentry and end up in a controlled glide but can't land, there's a small window when you can bail out, parachute down, and maybe survive. But it's a small window.
For the other vehicles, there is no plan B.
They do, however, feature redundant parachutes or redundant landing engines. I talked quite a bit about Starship and landing in my "people on starship" video, which I'll link in the upper corner. The summary is that if Raptor engines are reasonably reliable - as reliable as the Merlin or the space Shuttle RS-25 - the chances of all three failing is very, very low.
Parachutes are interestingly more problematic. This is a chart showing testing for the Orion capsule, showing the parachute deployment altitude on the vertical axis and the speed on the horizontal axis.
The parachute deployment for a normal reentry is this small region in the middle. But when you add in aborts, those can occur anywhere in the region outlined in red, and there are some parts of that which cannot easily be tested.
Just to make it more interesting, a normal parachute deploy uses drogue parachutes before the main parachutes, but a low-level abort only uses the main parachutes.
This is probably well summarized by this Elon Musk tweet, where he notes that parachutes are way more difficult than they seem.
I'm going to cover parachutes in more detail in a separate video as there is too much detail to fit here.
Our vehicles use three different methods of landing, and it's useful to compare and contrast them.
The shuttle obviously used wheels, and they allow a nice, low energy, gentle landing.
On the con side, you can only land if there is an airport within reach, if you have enough airspeed to be flying, and if you are on a glide path that gets you to that airport.
Parachutes are a mostly passive system and there is considerable history and expertise in using them.
On the con side, they have lots of failure points, the chance of failure is higher during aborts due to higher stresses, there is a lot of testing required, and they can't land on land - at least for Orion and Dragon. And they have a limited ability to target their landing site.
Finally, engines allow your vehicle to have a lot of delta-v during abort and therefore lots of options, and they have considerable redundancy.
On the con side, their ground approach is high energy, and there are no options if the engines don't work.
One more point about safety, then I promise we'll start talking about Starship.
Let's say you are going on a space holiday for 7 days. You need to ascend into orbit, stay for a week, and then come back and land.
And let's just say that the ascent and descent both have a 1 in 500 chance of killing the passengers, and the orbital stay has a 1 in 5000 chance.
We can convert those probabilities to success rates, multiply them together, convert it back, and get 1 in 238.
Over time, we improve our landing so it's 1 in 1000. That will push our overall risk down to 1 in 412, a big improvement
Now let's change the scenario; instead of staying in orbit we are going to spend 7 days on the surface of the moon, and that part of the mission has a 1 in 50 chance of death.
That gives us a 1 in 42 overall risk.
Now posit the same increase in landing reliability, to 1 in 1000. All that gives us is 1 in 44.
Which brings up another conclusion. On risky missions, the less risky parts don't matter - they don't contribute much to the overall risk. You can spend a huge amount of effort there and make minimal gains.
Okay, now that we've covered that, time to talk about Starship.
Wait, one more topic
We need to talk about abort systems and their impact on reliability. We'll use SLS as an example.
NASA's target goal for SLS is 1 in 300 on ascent, or 99.66% reliable. Let's say that the base reliability of the launcher is 1 in 100. That only gives us 99%.
Take the Orion capsule and add a launch escape system to it. Let's assume that system will save the crew 66% of the time.
So, we can take the 1% chance of needing the launch escape system and the 66 % success rate when we need it, and figure out that we get an increase in 0.66% in the survival rate, pushing us up to 99.66% total, or the 1 in 300 we are hoping for.
Abort systems are great.
Now lets look at the nominal, or non-abort scenario. For reentry and landing to succeed, the abort system needs to be jettissoned from the capsule. If that doesn't work, the crew cannot reenter.
Let's say that works 98% of the time and fails 2% of the time.
We can take that two percent chance of failure times the 99% of the time an abort isn't needed, and that will lead to a loss of the crew 1.96% of the time, reducing our overall survival rate to 97.94%, or 1 in 43.
Abort systems are terrible.
Let's look at another example.
Finally, let's talk about the details of starship.
Like shuttle, there are two basic kinds of problems that can happen on ascent.
The first class is underperformance; something has gone wrong with one or more of the engines and we therefore don't have the thrust we expect.
The second class are major issues. The booster explodes, there is a toxic gas leak, the electrical system fails, that sort of thing.
We'll star
Our options depend upon what the performance of Super Heavy was and what the performance of Starship is; failures with either stage might cause us to explore our abort options.
The shuttle had 5 abort options.
If the energy deficit was small, they could abort to a lower-than-expected orbit.
If there is enough energy to get near to orbit, they can travel around the earth once and then land.
The next option is to land at an airport in europe.
If the ground track is convenient, they could land on the east cost or the Bahamas
And final, they could return to land at the launch site.
Assume Super heavy has a significant issue. Super heavy can either keep flying and stage when it runs out of fuel, or it can stage immediately.
If the choice is to stage immediately, the options depend upon which starship is flying. If it's a 6 engine starship, the thrust/weight ratio is less than 1 and that means it's not possible to abort while sitting on the pad.
If it's the 9 engine starship, the thrust to weight is probably greater than 1, and it's possible to abort while sitting on the pad, though SpaceX has talked about stretching Starship to carry more fuel and that might change that.
Pratt & Whitney JT9D - 747-100
General Electric GE90 - 777
If you look around, you will find a lot of explanations why airlines don't have parachutes for passengers that explain how impractical it would be and how most accidents wouldn't provide time to use the parachutes.
All those points are true.
But they largely view parachutes as pointless, and they are wrong in that.
Pratt & Whitney JT9D - 747-100
General Electric GE90 - 777
Before we dive into things, I have two links to share with you.
The first is my video on Space Shuttle abort modes, as it's very useful to understand what the options were for the shuttle.
The point here is that safety changes can be good and they can be bad. We can express this in numerical terms.
We can look at how much better our mitigation is, multiply it by the chance of that scenario coming up, and get an estimate of the safety improvement
We can look at the problems our mitigation might cause, multiply it by the chance of that scenario, and get an estimate of the safety loss
Applying this to our parachute scenario with some made-up numbers, let's assume that the parachute can save 50% of the people that would otherwise die, and the chance of that scenario is one in one thousand. That gives us an improvement of half of one in one thousand, or one in two thousand.
Let's assume that the scenario where parachutes slow down evacuation only results in 10% extra deaths and the chance of that is one in one hundred, which would results in a reduction in safety of one in one thousand.
The point being that safety losses in other scenarios can outweigh the safety gains in the scenario you are trying to address.
Here's a video of how airplane evacuations are supposed to happen, during an evacuation test of the Airbus A380.
Now ask yourself, what would happen if 10% of those people were wearing bulky parachutes and tried to make their way off the plane?
Aviation disasters are very well studied, and we know that the time it takes to get off the plane can be the difference between life or death. We also know that passengers do not follow instructions, there are documented cases where people have died because they inflated their life vests inside the plane and could therefore not get out the exits that were slightly under water.
Parachutes would kill passengers who otherwise would have lived.
We know that drone ship booster landings are the primary reuse method for Falcon 9, and these days it seems like one is happening all the time. The drone ship reuse architecture has been pivotal in SpaceX's reuse model.
And yet, for Starship, SpaceX has abandoned the drone ship model and gone solely with the return to launch site model, where the booster will be caught by a large set of arms known as mechazilla.
We know that using return to launch site with Falcon 9 has a significant payload cost, and that should be true for starship. Why would SpaceX stop using drone ships? It just doesn't add up.
I hadn't dug deeply into the marine operations side of SpaceX, and after a fair bit of research, I have come to the conclusion that rockets are easy to analyze, and drone ships are complicated.
Or, as Rocket Lab CEO Peter Beck famously said, "I can tell you 100% in full honesty, marine assets suck". And initially the plan for rocket lab's Neutron rocket was to always do return to launch site, but they have since decided to land downrange on a drone ship.
We're going to have to spend a bit of time understanding why Falcon 9 uses drone ship recovery, how it is done, and then we'll talk about why Starship - at least currently - will always fly the booster back to the launch site.
For Falcon 9, it all comes down to payload capabilities.
One of the main goals of Falcon 9 was to be able to launch communications satellites to geosynchronous transfer orbit. Those payloads are roughly in the 3-7 ton range. If we add in the payload capability of Falcon 9 to those orbits, we see that those missions cannot be flow with a return to launch site, and if you want reuse, you need to go with the drone ship.
For low earth orbit launches, the situation is a little better. Current crew and cargo dragon flights are able to sneak in just under the limit and can fly return to launch site missions, but earlier versions Falcon 9 had slightly less performance and SpaceX kept larger performance margins and therefore the dragon missions were flown with drone ship recovery. It probably made NASA happier as well.
Starlink missions carry enough satellites to be right up at the payload limit for a drone ship landing.
Starlink seems like the perfect place to use return to launch site. It would take 7 flights to carry the same number of satellites that could be carried with 5 drone ship launches, but it would be a lot simpler...
But those two more flights mean two more second stages, launch license, booster refurbishments, launch pad repairs, and propellant loads.
It would also mean - for a given flight rate - about 30% fewer satellites into orbit per month.
My guess is that those extra two flights - and especially the cost of the two extra second stages - means that RTLS costs a little more, but I think the big driver is the number of satellites per month.
To summarize, drone ship landing allowed SpaceX to do two important things with Falcon 9. They could hit specific payload targets and launch payloads that they could not carry with return to launch site landings, and they could get the maximum payload out of the second stage that they expend on every flight.
But starship is a different beast. It has sufficient performance to carry all current payloads, and it is fully reusable, so the Falcon 9 second stage concern does not apply.
To understand why starship isn't doing drone ship landing, we need to understand more about the SpaceX fleet
The SpaceX drone ships in Florida are paired with support vessels that go out to assist and also to retrieve fairings. The support ships are named Bob and Doug after the two astronauts who flew the first crew dragon mission to the international space station, not after Bob and Doug McKenzie from the great white north.
The drone ships have thrusters in the four corners that keep them on station so the boosters know where to find them. That is why they are called "autonomous drone ships" - they can stay where they need to be without people on board.
The newest droneship - a shortfall of gravitas - has onboard propulsion to take it to the catch location and back, but SpaceX has not used that capability yet.
Drone ships therefore needs to be towed. They can be towed by either Bob or Doug, but bob and doug also have fairing retrieval duties that require them to be in a slightly different position from the drone ship and each ship can carry 4 fairings so they will generally stay out for more than one mission, That means SpaceX will often pay a tug to carry a drone ship out to the proper location and back to port.
And then it's just a matter of setting up a steady cadence for all the ships and everything runs smoothly...
That of course is not true, because nothing is simple with marine assets.
This graph shows all the drone ship landings for the first half of 2024 after SpaceX has had a few years to tune their system for efficiency. Note that it's not showing the use of the support ships or tugs.
The top section shows the drone ship Of course I still love you, which operates out of long beach in California for launches from Vandenberg.
Just read the instructions and a shortfall of gravitas both operate out of port Canaveral in Florida.
We can figure out a few things. The first is that the shortest turnaround from one landing to the next is 4 days, but those are uncommon and 5 days between landings tends to be the shortest times.
There are also gaps big and small when a drone ship isn't operating.
Across the fleet average about one landing every 10 days, or about half of their maximum rate.
Remember what peter beck said about marine assets...
While it is common for ship's crew to work a month without a day off, eventually they will need downtime, and that is one factor that leads to the slower rate.
And the ship will need downtime to conduct regular maintenance, repair, and upgrades
And finally, the weather is always a factor. If it's not safe for the drone ship and support ships to be out there, the flight has to be delayed, and even with acceptable conditions, the weather can increase the time to get to the landing point or return to port.
Marine assets are also expensive...
Bob and Doug probably cost $10-20 million per year each, and the drone ships costs are likely similar, perhaps more because of the need to pay tug companies to move them around.
And they costs almost the same tied up to the dock if you aren't flying.
Marine assets add a lot of complexity and uncertainty to your life.
Let's talk about Starship.
The two big goals for starship - outside of Starlink launches - are to support NASA's Artemis program and future Mars exploration.
Both of those require what NASA calls propellant aggregation - you need to fly a bunch of missions to carry enough fuel into orbit to refuel the mission starship, which means you need to fly often.
With that in mind, let's explore what a few different scenarios would mean to super heavy. I'm going to assume that super heavy has rapid turnaround and can fly once a day.
The Falcon 9 drone ship approach can support a flight rate of about once in every 10 days. Using Starship 1 numbers, that flight could carry about 50 tons of payload if flown RTLS. If we add a drone ship to that, my model estimates we can carry about 80 tons of payload. That seems like a no brainer in terms of payload, but remember that we don't have expendable second stages to deal with and - assuming starship works the way it should - our big cost is fuel cost. Having said that, if you are flying once every 10 days the drone ship is probably a win.
But remember that both the moon and mars missions need refueling, and you are going to want to do those as quickly as possible. 100 days for 10 flights will not be acceptable, and more than 10 flights might be required.
With RTLS, we can fly a mission every day with a single super heavy. But the drone ship approach has a problem. At best, they have a cycle time of 5 days so we need 5 of them to fly once a day, but we might need 10 of them to do this on a sustained basis. 7 is a reasonable minimum if you are very organized, get lucky with the weather, and can accept downtime after a fast surge of missions.
We also need more super heavy boosters because some of them are stuck on the drone ships coming back or being transported from the dock to the launch site. I'm going to say you need 4.
Flying once a day now requires 4 super heavy boosters and 7 drone ship groups, and that's if you get lucky. That payload advantage no longer seems attractive; it's pretty likely that you can fly an extra mission a day to exceed the drone ship payload and spend a lot less than those drone ships cost. And you have to purchase those custom drone ships and find a place to store all your vessels because port Canaveral does not have room.
And let's say you have a refueling mission that can take 2 flights a day and you have two others you need to fly as well, bumping your flight rate to 4 times per day. You need four super heavies for RTLS, but you need 16 super heavies for the drone ship approach and 28 drone ship groups.
The point of these scenarios is pretty simple. Drone ships are fine at low flight rates but if you want to fly all the time - which is what SpaceX wants to do with starship - the slow cycle times just kill you.
Airlines are masters at high utilization. When the plane lands, it taxies to the gate, and then the airline unloads people and cargo, cleans the cabin, loads people cargo and fuel, leaves the gate, taxies, and takes off. That time in the middle is known as turnaround time and airlines minimize it because they only make money when planes are flying. A good airline will do that in about 60 minutes.
Most airliner are working about 11 hours a day, every day.
For Falcon 9, the time from liftoff to drone ship landing is about 8 and a half minutes, then things slow down. A couple of days to get back to port, 9 days to do the refurbishment and inspection is supposedly the fastest they've done, and then presumably a couple of days to get a new fairing attached and ready to fly.
The fastest that SpaceX has every reflown a booster is 27 days, which they have done twice. What that means is that the return time because of drone ships is a small proportion of the total cycle time, and SpaceX isn't even trying to minimize the time to refly boosters quickly. The time that boosters spend on the drone ships has pretty much zero impact on the overall cycle time.
Starship is a different beast; the SpaceX goal is for "gas and go" operation of super heavy, with inspection and propellant loading taking place almost immediately after landing. Let's say for sake of argument they can do it in 8 hours.
You have probably already figured out the issue. That 3 days spent on the drone ship is 90% of the time between flights, and because it's transit time there's very little you can do about it. You can maybe do 4 day cycle times on each booster, if the weather is good and if you don't have mechanical issues with the marine assets.
This is the real reason why Starship uses return to launch site.
It's expensive in terms of payload, but it is the only way you can get truly rapid turnaround, and that is required for the Starship missions.
What did we learn?
Drone ships are a necessary evil if you need the extra performance to reach your target market, and as long as your launch cycle time is less than the cycle time of the drone ship.
But they have high fixed costs and do not scale well to high flight rates.
If you enjoyed this video, please send me this Autel titan drone
A commonly asked question is "why don't they just fly starship and super heavy from Boca Chica to Florida?"
Let's see whether that's possible...
I did a video on shipping Starship and Super Heavy by sea.
Our goal is to get both Starship and Super Heavy from the factory in Boca Chica, Texas to Pad 39A on Kennedy Space Center without using a barge.
The distance between the launch site at Boca Chica to Pad 39A in Florida is 1660 km.
The question is really whether both Super Heavy and starship can make it there, but starship is a lot heavier with its fins and thermal protection tiles, while Super Heavy is mostly just a big tube with a lot of engines. This means that, for most purposes, the question is whether starship can make it because it has less performance than super heavy.
Can we do it with a suborbital flight, the way that the first two Mercury launches did? The first Mercury launch landed 487 kilometers downrange from the launch site.
That seems promising...
Here's my plan of attack to answer this question:
Step 1: Figure out the delta v required to make the trip. I'll put a link to a video that talks more about delta v, but it's basically the way we quantify how hard it is to get from one location in space to another.
Step 2: Compare that to an estimated delta v for starship. I have that from some earlier work.
Step 3: Go on a bike ride, because this isn't going to take more than a few hours.
The math for this is well understood; it's determined by the ballistic equation.
If we known how far away the target is we can calculate the delta-v required to get there.
Here's the equation; you can see the delta v is equal to the speed at launch...
That seems incorrect...
Here's a graph that shows a ballistic trajectory in red and a rocket trajectory in green. We can see that they are quite a bit different.
Ballistic trajectories start with their highest velocity at launch and then the only forces on them are gravity and air resistance (this chart ignores air resistance).
Rockets start at a velocity of zero and accelerate faster as they get lighter.
Ballistic trajectories have nothing to do with rocket launches, and those equations don't apply.
Time to move to plan B.
Plan B is to write a physics simulation of a rocket launch. I had been gradually moving towards this for some of the earlier videos, doing some fun stuff in Excel.
Once the simulation is there, we can plug in the values we have for starship - the dry weight, the propellant capacity, the engine thrust, etc. These are of course all *estimates*.
We then specify how full the fuel tanks are.
The simulator will then try different flight paths and find the one that gives us the maximum horizontal distance.
Note that this doesn't consider air resistance, so it's somewhat flawed. Maybe version 2 will support that.
I'm testing 200 different trajectories to see which one is best, and that's very easy to code and very hard to Excel, so I wrote a program. The simulator is written as a pretty junky Windows Forms program using C#.
This simulation is fairly simple. We want to track the time and the sideways and upwards components of the distance in meters, the velocity in meters per second, and the acceleration in meters per second squared. Here's how it works.
We start at time zero, and since we are sitting on the launch pad, we have no distance or velocity. We will start with step #3, where, based on the approach in the last slide, we figure out the sideways and upwards acceleration. At launch, the rocket is pointing straight up so there is no sideways acceleration and the upwards acceleration is 2.4 meters per second squared.
Time is now 1 second into the flight
We now go to step #1, where we update the velocity based on the acceleration the previous second. Looking at time zero, we take the side and up accelerations and add them to the previous side and up velocities (both zero), so we now have a side velocity of zero and an up velocity of 2.4 meters per second.
Finally, we do step #2, where we update the distance based on the velocity. The up velocity went from 0 to 2.4 meters per second. That means the *average* velocity was half of that, or 1.2 meters per second. In one second, that means we travelled 1.2 meters, so that is the up distance.
And back to step #3. The rocket is no longer pointed quite straight up, so we have a side velocity of 0.05 and an up velocity of 2.46.
Onto time = 2 seconds. The velocity is easy to update; we just add the acceleration to the current velocity and get 0.05 meters per second to the side and 4.86 meters per second up.
Update the distance. The sideways distance is the average of the current velocity and the previous velocity, or 0.025. The up distance is the average of 2.4 and 4.86 *plus* the previous up distance of 1.2, or 4.83 as the up distance.
Update the acceleration again, now 0.11 and 2.52.
Time equals 3 seconds.
Add the accelerations to the existing velocities to get 0.16 and 7.38, and finally the side distance is increased by 0.105, giving us 0.13. The up distance is increased by 6.12, and adding that to the existing distance gives us 10.95
This process continues every second. As I noted in the gravity loss video, the acceleration will continue as long as the engines keep burning, and then we just end up with gravity as our main force.
Here's an example of three trajectories out of the 200 the simulator evaluates.
The top blue one gets the most altitude, but that means less horizontal velocity and therefore less distance.
The bottom purple one spends the most on horizontal velocity and goes the fastest but it doesn't stay up long enough to get a lot of distance.
The middle golden one balances out altitude and horizontal velocity and manages to get a bit more distance than the other options.
A few assumptions built into the results...
First, it assume starship has 3 sea level engines and 6 vacuum engines and they are all running. Starship cannot get off the ground with a full fuel load with 6 engines, so this is a fair assumption
Second, the raptors are only putting out 1800 kN of thrust. That's at the low end of where they will be, and it does not factor in thrust improvements, higher thrust in vacuum, or higher ISP of vacuum engines.
Third, the simulation does not attempt to model drag. I figured that with the conservative choices on the engines that air resistance would not be significant.
There's a problem
As starship launches, the impact zone if something goes wrong will proceed along the flight path.
There are a number of people who would be unhappy to be put at risk, including Disney World, Universal Studios, and the 8 million people who live on or near the flight path.
It's therefore very unlikely that the FAA would allow such a flight with an unproven rocket.
Sigh
Direct flights are out, so the answer to the question is "starship can't make the hop".
Then I got an idea.
I got a wonderful, awful idea.
We need a dogleg.
No, a dogleg.
We can do this trip in two legs.
One hop south of the Florida keys ending just off the Bahamas, and then a second hop back to the north to land at kennedy space center.
The obvious way to do this is pretty simple. You take one of the mobile launch platforms SpaceX has been working on, and you park it near the Bahamas. Hop to the platform, check everything out, refuel, and then hop to the cape. Simple and straightforward.
That requires the platform to be functional, of course.
Is there a way to do this without a platform?
Maybe...
Can you do this with two hops sequentially?
Do hop 1 from Boca Chica, almost land at the position near the Bahamas, then take off and do hop 2 and land at Kennedy Space Center.
That would be about 2500 km of total distance, and that's a lot less than the 4700 km I said starship could do, so it's easy, right?
Unfortunately, that math isn't correct
Let's explore why.
The second hop is 579 km, and the simulator tells me that Starship will need 175,000 kilograms - or 175 tons - of propellant to make that hop. That's how much propellant we need left after the first hop.
Let's look at some numbers for starship. Once again, these are all estimates, and specifically this assumes starship that has not been stretched longer.
The dry mass is 126 tons and starship carries 1200 tons of propellant, making the total mass 1326 tons. That gives us a mass ratio of 10.5 and a delta v factor of 2.35.
If you want to understand more about delta v, here's a link to another video
How does this change if we need to reserve 175 tons of fuel for the second hop? That will add 175 tons to the dry mass and subtract 175 tons from the propellant mass
That gives us a dry mass of 301 tons, and that takes our mass ratio all the way down to 4.4 and the delta v factor down to 1.48. That's a lot less than 2.35.
Running the simulator, that gives us a hop 1 distance of 1443 km. That's a lot less than the 1924 km that we need.
How can we fix that?
What if the raptors have 2000 kilonewtons of thrust instead of 1800? That gets us 1562 kilometers, still not enough.
What if starship is stretched? We don't know what the details are, but that's probably not enough
What if starship can use its fins to glide farther? Maybe, bur probably not.
I don't think these work.
There's another concern...
Starship is designed to land with about 175 tons of mass
Will it stay intact at the end of the first hop if it masses 300 tons?
A few other ideas, probably stupid ones. I haven't simulated any of these...
Don't reenter the atmosphere, come down to 60 km and burn your engines to go up and change direction and the same time. The challenge with this is you have a velocity of several thousand meters per second east and you need to cancel that and go north. You probably don't have enough energy to do this.
Or you could start the reentry and use the atmosphere to lose most of your horizontal velocity, and then relight your engines to take off before you reach the surface. You have a lot of vertical velocity that you need to cancel out, and probably not enough to energy to do that.
I don't think these work either.
If we can't do it with starship, what about super heavy?
With 20 engines, super heavy can do the second hop on 235 tons of propellant, and reserving that amount leaves it with a range of 2875 km for the first hop. More than enough.
This is assuming it can handle the reentry with that much fuel, but it will have some extra fuel it could use for a reentry burn to reduce the forces it experiences.
I'm going to say that this is probably possible.
Here's a hybrid approach...
Super heavy can launch with starship on top and give it enough energy to finish the first hop. This gives plenty of margin for starship to use its engines to reenter more easily. Super heavy returns to the launch site, and then performs a later flight where it hops by itself.
This is actually practical, if you define practical as "A guy who did an amateur analysis can't come up with an obvious reason why it won't work"
What did we learn
Direct hops are possible but probably not feasible due to flightpath concerns
A hop to a floating platform and then a second hop is feasible.
A double hop with starship by itself is probably not possible
A double hop with super heavy by itself is probably possible.
A super heavy toss of starship is probably possible.
Will they do this? I have no idea.
If you liked this video please tell your friends.
Welcome to starship moonbase...
NASA's moonbase plans start with what is known as the "foundation surface habitat". Most of the designs look like this, with a habitat on top of a lander.
But we already have a design with a habitat on top of a lander. What could we do with a habitat based on starship.
There are a few issues with this approach...
First, it's a *long* way up to the payload section - 35 meters, or about 7 flights of standard stairs. The plan is to have platforms you can ride up and down, but that's not a great solution in the long term if you want to expand your habitat.
Perhaps they could attach rope ladders between the various starships to provide easy access...
There's one additional problem with the vertical approach - Radiation in the form of protons, neutrons, and gamma rays will be continuously hitting the habitat.
If the starship is horizontal, it can be covered with lunar soil - or regolith - for protection against radiation.
It's also obviously much easier to get into and out of the habitat.
If horizontal is better, now do you take a 50 meter rocket and lay it down horizontally?
My first thought is to use cables to lower the starship.
We start with a reinforced mounting point on both sides of starship.
We unroll the cable from a winch near the payload section, take it out to the side, and anchor it into the surface. This cable will be used to pull starship to the side until it starts to fall over on its own. Here's a top view.
Then we need a second set of winches to slowly lower starship down to the ground.
And finally, we need a way to anchor the bottom right point so it doesn't slide away.
Once we've gotten it tilted over, we can ignore the cable that pulls it over and concentrate on the lowering cable.
A bit lower
And we're almost down. But we have a big problem - at this point the cable stops helping keep starship up and starts pulling it down to the ground faster.
That's not going to work...
We need a way to keep the angle from getting too low.
The solution is to add a two support poles - known as "gin poles".
The poles are attached to the ground and to the cables, and serve to hold the cable up in the air.
With the pole holding the cables up, the angle stays up and we can easily lower the starship slowly to the ground.
At least theoretically. We'll need to figure out if it's practical
We will create what is known as a "free body diagram" in physis. We're going to look at the forces right before the vehicle touches down.
We'll start by adding some dimensions to our drawing. The gig pole is 30 meters high. The pivot point is 4.5 meters off the ground.
In the horizontal dimension, the pivot point is 35 meters from the base and it's a further 9 meters - the diameter of starship - to the pole.
We can pull out the triangle. It's 25.5 meters high and 44 meters long. We want to know the angle theta.
Going back to our high school trig, we know that the tangent of the angle theta is 25.5 divided by 44. We solve for theta and get 30 degrees.
Pro tip here - if you want to check your trig in an example like this, change the values and see if the changes make sense. If the vertical distance was only 12, the angle would be about 15 degrees. That's about what we would expect. And if the vertical distance was zero, the angle is zero - also what we would expect.
I'm now going to use the starship center of gravity model I developed in the "will starship fall over?" video. In it, I estimated the center of gravity of an unfueled starship with 50 tons of cargo to be at 28.5 meters, and the total mass to be 95 tons.
Right before starship touches down, some of the support of the starship is going to come from the left - or base - end, and some is going to come from the attached cable. We need to figure out how much is supported at each of those spots.
The model says that the center of gravity is 28.5 meters from the base and the mass is 95 tons. It's close to the cable attachment so most of the load is on the cable end.
It turns out that it's 18 tons on the left end and 77 tons at the cable attachment. Doing that calculation is left as an exercise to the viewer.
And that works pretty well if starship was on the earth - in standard gravity I could just switch from using tons as a measure of mass to tons-f as a measure of force.
But I've decided to use the proper unit of force, the Newton. The force in newtons due to gravity is simply the gravitational constant G multiplied by the mass in kilograms.
We can multiply by sides by 1000 to find that the force in kilonewtons is equal to G multiplied by the mass in tons.
The gravitational constant for the moon is 1.6 newtons per kilogram, so we can update our diagram with forces...
Just a little more math, and we'll have our answer.
To counter the 123 kN force pushing down at the cable attachment point, we need a 123 kN force pulling up.
But the cable doesn't pull straight up, it pulls to the side at 30 degrees, so we need to figure out how much pull at 30 degrees will be required to produce a 123 kN pull up.
Back to the trig. We know that the sine of 30 degrees will be equal to 123 kN divided by the pull. Rearranging the equation to solve for pull, we end up with a pull equal to 246 kN. That is the amount of tension on the cable.
A winch that could pull at 246 kN would be known a 25 ton winch on earth. It's a large and heavy piece of hardware. And we want one for each cable, both able to handle 25 tons so we have a good safety margin. We also need two lighter winches to pull starship enough to the side so it can be lowered by the big winches.
We don't actually need a winch with a big motor to lower starship down; we just need one with a braking system that will slow it down. That would be lighter but still massive.
We also need an anchor in the moon that we can pull against with 246 kN of force. We can easily do that by... yeah, I don't have any idea, but anchoring is a proven technology on earth. I'm sure that everything associated with it is heavy.
And then we need that 30 meter pole that can hold up to the force. We can probably just attach it to the side of starship during launch.
This all assumes only a 50 ton payload to the moon. If you want to carry 100 or 200 tons, the forces get bigger
What's the verdict on this approach? It's possible and perhaps practical, but it's heavy and complicated and requires a lot of development work. It also requires heavy equipment on the lunar surface to build anchors and pull cables around.
What we really need is a simpler approach, perhaps one that leverages things that we already have.
Which takes me to a stupid idea that spawned this video...
It is possible that lunar starship will have thrusters to use for landing to avoid any issues with using the main engines.
They will need to counter the mass of the crew lunar starship, which is 179 tons. That will require 286 kN of thrust. Adding on 25% to be able to slow down gives us 358 kN, and if there are 6 thrusters they will need to be about 60 kN each.
The arrangement might look like this...
If we are willing to add six more thrusters pointed to the side, we can do the following:
We start by firing the two left-facing thrusters, and that's enough to cause starship to topple over to the right.
And as it tips over, we fire the right pointing thrusters to control the speed of the descent.
And finally bring it to a gentle landing.
The force of gravity right at landing is 123 kN, and four of the thrusters provides 240 kN. We might be able to do it with two thrusters, one on each side.
What's the evaluation of this approach?
It uses systems already present on the vehicle
It only requires 6 extra thrusters plus the plumbing and other support mechanicals, which will not add a significant amount of weight.
It does not require any lunar surface preparation.
And finally, it will be far more entertaining to watch.
Or perhaps...
This is probably the simplest approach, and if you have thrusters strong enough to handle landing and you don't need to take off, you can just use the thruster to flip around and land horizontally.
The is the backwards version of how starship will land on earth.
In summary...
Cables could work but are a poor approach
Thruster drop could work
Flipping to land horizontally makes the most sense.
If you enjoyed this video, please study the dynamics of falling objects watching here come the brides...
It's very tempting to think of the Starship architecture as just the bigger brother of Falcon 9, with a fancy new reusable second stage. Most of the discussion has focused on that second stage, but it turns out that there are some very interesting things going on with the whole package, things that make it a very different rocket than Falcon 9.
The key difference is that Falcon 9 is a mostly normal rocket that can operate in a reusable mode, while Starship is designed and optimized to be highly reusable. This video will look at one of the ignored parts of rocketry, staging.
(show video)
This is eager space.
To understand how starship staging is different, we'll need to look at how a typical rocket approaches stage design
On the Atlas V 501 rocket - which uses no solid rocket boosters - the booster does 55% of the work to get the payload to orbit, and the upper stage does 45%.
The beefiest Atlas V - the 551 version with 5 solid rocket boosters - sees the booster doing 69% of the work and the upper stage only doing 31%.
That's a pretty typical ratio. The Atlas V stages at around 276 seconds into the flight, or roughly halfway to orbit.
The Falcon 9 is very different. It puts the beefiness into the upper stage, with that stage doing 61% of the work. It stages at around 150 seconds in expendable mode and earlier in reusable mode.
Atlas V uses a Centaur upper stage which is very efficient but low on thrust, so that required a beefier booster stage.
Falcon 9 needed a vacuum engine for the upper stage and SpaceX chose to adapt a first stage engine for the task, resulting in a high thrust upper stage engine. That worked well as they wanted to stage early to enable reuse.
Falcon 9 is the closest sibling of Starship and we'll therefore look at what happens during Falcon 9 staging next...
The first stage is thrusting along, we hit main engine cutoff (or MECO), there's a pause, the stages separate, the first stage starts to turn around, and then the second stage ignites.
Watching a few launch videos, I recorded the following times.
If we starting measuring from when the engines shut off, it's 3-4 seconds before the stages separate, and then 7-10 seconds after that before the second stage ignites, so that's 10-14 seconds without thrust.
For the first stage, it takes approximately 20 seconds after MECO for the stage to turn around and relight its engines to boost back to the launch site. That's assuming the booster is returning to the launch site - most Falcon 9 launches land on a drone ship.
Falcon 9 operates in three modes...
The first one is expendable mode - the booster fires until it runs out of fuel, the second stage takes over, and the spent booster either breaks up on reentry and/or crashes into the sea. That's traditional rocket mode, and it's used when the payload needs all the performance that Falcon 9 can provide.
The second mode is drone ship recovery. It's used when Falcon 9 has enough performance to put the payload in the desired orbit while reserving enough propellant in the booster to coast over the water, reenter, and ultimately land on the drone ship. All starlink flights are designed to use this mode as it's the cheapest one per kilogram of payload, and many commercial flights also use this mode.
The third mode is return to launch site. It's used when Falcon 9 can save enough booster propellant to be able to get back to the launch site and still put the payload in the desired orbit.
With that in mind, we can look at where Falcon 9 spends energy.
In expendable mode, the booster puts all of its energy into the second stage. In the coast phase between main engine cutoff and the start of the second stage engine there is a little bit of energy loss because gravity is slowing the rocket down during that time period. It's not particularly significant - perhaps 1% of the total energy. Falcon 9 can put 22,800 kilograms into low earth orbit in this mode.
If we look at Falcon 9 landing on a drone ship, there are two new usages of energy.
At the time of staging, the booster is travelling very fast and it will accelerate as it heads back towards the earth. A significant entry burn is required to slow the rocket down so that it will survive reentry into the atmosphere, and that takes quite a bit of energy. We also need to slow the rocket down to land, which requires some energy as well.
Looking at the numbers, we see that that energy the booster can give to the second stage is reduced because it needs to reserve energy for the entry and landing burns. That means the second stage needs to be able to give the payload more energy, and that is only possible with a smaller payload mass - around 16,500 kilograms.
Looking at the return to launch site scenario, things get more complex. We are now losing energy on staging on both the booster and the second stage. There is a large boostback burn to get us headed back to the landing site. There is a tiny entry burn because the boostback burn slows the booster down so much, and of course there is a landing burn as well. Once again, the second stage has to do more work, and the payload for this option is around 12,500 kilograms
This last scenario is the only one for Starship, and optimizing Starship will try to reduce these losses.
The entry burn doesn't do much in this scenario, so Starship will just eliminate it. We are stuck doing the landing burn, which leaves the very large boostback and the moderate staging burns as places we can optimize.
We'll start by looking at the energy going to staging.
Here's a version of our previous graph showing the portions of the trajectory with staging losses. The booster energy loss is higher because it takes a long time for the cold gas thrusters on Falcon 9 to get the stage turned around.
There is luckily an established way to reduce these losses.
This is the first launch of the Gemini program in 1964, with the Gemini capsule being carried by a Titan II missile.
If we zoom in on the rocket we can see that there are holes in the rocket between the first and second stages. The allows us to start the second stage engines before the staging has taken place. Here's a view of staging from the Titan II second stage...
This technique used to be called "fire in the hole", but is now typically known as hot staging
This technique was used on the Titan II and was also used by the Russian Proton rocket, which has a very distinctive latticework section between the two stages.
SpaceX has made this change on booster 9, with a ventilated section at the top of the Super Heavy Booster.
Hot staging allows us to mostly eliminate the coasting section the staging process, saving energy that can go to payload. It's not a lot - perhaps 300 meters/second or less of delta v.
We can now work on reducing the boostback losses.
Hot staging helps a little - we get rid of the long coast to get turned around and that means we are closer the launch site. But that's a small improvement.
The problem we have is simple - at the staging point, the Falcon 9 booster is travelling away from the launch site at perhaps 1250 meters per second, and the boostback burn needs to kill all that velocity before it can generate the velocity it needs to head back to the launch site.
That is why boostback is so expensive. What we really need is a way to reduce that velocity.
The trajectory to orbit here is a classical one - one that produces the highest payload. But it also produces high boostback energy requirements.
We can choose, however, to fly a trajectory where the booster spends more energy going up and less energy heading downrange - generally known as a "lofted" trajectory.
A lofted trajectory could reduce the horizontal velocity from 1250 meters per second down to 1000 meters per second. It is a less efficient trajectory overall but reducing the boostback cost could make it worthwhile.
Another way to reduce the horizontal velocity is simply to stage earlier in the flight.
If we can move the staging even 10 seconds earlier in the flight, the horizontal velocity could be much less, and would significantly reduce the cost of the boostback burn.
To get this to work requires a number of changes. They don't make sense for Falcon 9 because they would reduce the payload for the drone ship missions, and that would not be good.
But Starship never does droneship landings, so optimizing for return to launch site makes sense.
Elon Musk talked about this recently during an IAC 2023 session.
He mentioned shifting more of the delta v burden to starship which is exactly what happens with early staging - the booster does less of the work of getting into orbit and the second stage does more of the work.
He also mentioned that the propellant ratio between super heavy and starship was trending towards 2 to 1, which is the sort of detailed technical information that is great to get.
The current propellant ratio between super heavy and Starship is 2.8 to 1. This is already very low - Falcon 9 is 3.75 to 1, and commercial launchers like Atlas V starts at about 12 to 1 and goes up from there.
To get to a ratio of 2 to 1, there are three things we could do.
The first is a stretch to add 500 tons of propellant to starship.
The second is to reduce the booster propellant to 2400 tons.
The third is a middle ground, stretching starship to increase propellant and shrinking super heavy to reduce propellant. I chose these numbers to demonstrate, but others are possible.
All three hit the 2 to 1 ratio, and all three put more of the work of getting to orbit on starship.
Which one will they choose?
The one where they reduce the booster propellant is probably a non-starter, as it reduces the overall performance of the system *and* Musk has mentioned stretching starship to be longer in the past.
That leaves us two viable options. Just stretching Starship and leaving Super Heavy alone is the simplest change to make. It would add 9 meters to starship, and the change would look roughly like this.
The second is more complicated as it requires work on two vehicles.
Note that Musk said "trending towards 2 to 1, not at 2 to 1", so the ratio could easily be something like 2.2 to 1.
Can we figure out which one is more likely?
Musk mentioned one more thing at the IAC session, and that was that staging time is trending towards 100 seconds. That's another bit of useful technical data.
To accomplish early staging, there are only two options. You can carry less propellant, or you can burn it faster.
We have an option that carries less propellant. Let's explore faster propellant burning.
We can surprisingly figure this out fairly easy.
The staging time in seconds is equal to the propellant burned in kilograms divided by the propellant burn rate in kilograms per second. This is some wasted propellant - startup consumes some before liftoff and there is some left in the tanks at the end because running engines to empty is very bad, but this will get us pretty close.
It turns out the burn rate is also quite easy to measure. It's known as mass flow rate, and it is equal to the thrust of the engine in newtons divided by the specific impulse times 9.81. The numbers for the Raptor 2 say that it has a thrust of 2.3 million newtons and a specific impulse of 327, both at sea level.
Plug those numbers in, and we get a mass flow rate of 720 kilograms per second. There are 33 engines, so the total is 23,760 kilograms per second.
We can now plug the mass flow into our equation, and when we do that we get 143 seconds for the pure stretch option and 126 seconds for the stretch and shrink option.
That's for burning all the fuel, but of course we need to leave fuel for the boostback burn and the landing.
I'm going to allocate 2300 meters/second of delta V to that, which is probably double what it will actually be. Doing some math tells me that it will take 180 tons propellant to do that. Based on our mass flow rate, that's 8 seconds worth of fuel for the booster.
That pushes our staging times down to 135 and 118 seconds, which is looking more reasonable.
To get it lower, we need engines with a higher mass flow rate.
Our current numbers are based on the Raptor 2.
There is *supposedly* a Raptor 3 engine in the works, with a thrust of 2.6 meganewtons and a mass flow rate of 814 kilograms per second. That gets us to 120 seconds for the stretch version, 104 seconds for the stretch & shrink version.
That looks pretty good, but there's *another* problem. Musk stated that the thrust/mass ratio would be 1.3 to 1.4, and our first version is close to 1.6 and our second version is 1.75.
We can make the numbers align better if the Raptor 2 has a lower specific impulse. If the impulse is only 300, the mass flow rate goes up and therefore the staging happens faster. That gives slightly more reasonable thrust/mass numbers but still not what we are expecting.
So this is all quite confusing, and we'll need to wait to see what SpaceX actually does for the staging time of Starship.
Starship is a rocket like no other because it only does return to launch site, and that changes the rules of the game.
If you enjoyed this video, please send a card to super heavy to let it know that it's exactly the right amount of heavy.
https://www.nasa.gov/smallsat-institute/sst-soa/thermal-control
Lunar starship is the human lander for NASA's Artemis missions.
For the lunar starship to work requires that it be refueled in earth orbit before it journeys to the moon. It will either be refueled directly by tankers, or it will be refueled from an orbital propellant depot that is filled by tankers.
Nobody has built an orbital propellant depot like this before, and I'm seeing a lot of speculation about whether it's feasible to keep liquid oxygen and liquid methane liquid in orbit, or whether it will all just boil away.
One of my arguments has been that NASA clearly thinks SpaceX can do this or they wouldn't have given them the contract, but I thought it would be nice to have an analysis of how well an orbital depot could actually work.
Welcome to starship orbital depot analysis.
My goal is to estimate how many days the propellant in starship will last in low earth orbit before it is all vaporized.
What is your guess?
I'll give you some help. It's more than 1 day and it's less than 10,000 days.
Write down your guess, and we'll get started...
Before we get into the details, I need to explain my goal.
We can define an analysis quality scale.
On the left end we have conjecture. It might be total conjecture or it might be informed conjecture, but it's mostly just a guess.
It's fine to use conjecture in discussions, as long as you remember that it's just conjecture and there is probably better data out there.
At the other end of the scale, we have professional engineering analysis. This is what we would expect from NASA or SpaceX, work done by skilled engineers using advanced approaches and validated by other engineers. This analysis also comes at various quality levels - it could be an result that takes one engineer a few hours to create or it could be one that takes a team several months.
My goal is to be somewhere in this middle range. I'm hoping I will be near the top for most topics, but it's possible I will be lower. This particular topic seems endlessly complex and I've simplified it considerably, and some of those simplifications may overlook important concerns.
We are going to be talking about heat transfer in spacecraft - specifically about heat absorption in spacecraft.
When a spacecraft is on a planet such as a earth, we would need to consider the three ways that heat flows -
Conduction from direct contact, convection by the motion of molecules, and direct energy transfer by radiation.
In space, things are simpler - we just need to consider heat transfer by radiation.
Radiation consists of many different wavelengths of light, from the very short wavelength gamma rays to the longer microwaves, and the very long radio waves that are off the slide to the right.
We are obviously concerned about heat from the sun. The surface of the sun is very hot - about 5800 Kelvin - and the radiation from very hot things is concentrated at the shorter wavelengths - ultraviolet light, visible light, and the shorter part of the infrared range. Solar radiation runs from about 0.25 micrometers on the left side and about 3 micrometers on the right side.
The remainder of the important thermal wavelengths is known as the infrared, or IR range. We will use these two ranges as part of our analysis.
https://www.researchgate.net/figure/Solar-Radiation-Spectrum-Solar-radiation-is-the-radiant-energy-emitted-by-the-Sun-Fig-1_fig2_287748270
https://www.nasa.gov/smallsat-institute/sst-soa/thermal-control#_Toc120613720
The sun is putting out radiation that is hitting our spacecraft. You already know something important about sunlight. If you are wearing a black shirt in the sun you will get much hotter than if you wear a white shirt, because the black shirt is better at absorbing the energy and the white shirt is better at reflecting the energy.
In engineering terms, the amount of solar radiation that is absorbed is known as absorptability, often denoted using the lowercase Greek letter alpha. A surface that is great at reflecting solar radiation will have a low value such as 0.1, and a surface great at absorbing solar radiation will have a high value such as 0.9.
Note that it's not just about how light or dark a material appears to humans, as that is only about visible light, and a surface that reflects visible light well may not reflect ultraviolet or infrared well.
The intensity of the sun at the earth is 1367 watts per square meter. If our absorptability is 0.4, that means each square meter of the spacecraft will be absorbing 547 watts.
That added heat will cause the spacecraft to heat up.
We know that spacecraft don't just heat up and melt, so there must be something else going on. That something is the spacecraft emitting heat back out into space as infrared radiation. When the spacecraft is emitting the same amount of total heat as the sun is adding, it will be at a stable temperature.
Can we figure out that temperature?
The answer is yes, but it's a little complicated...
The first thing we need to know is that there is a analog to absorptability known as emissivity, usually denoted with the lower case Greek letter epsilon. Emissivity also ranges from 0 to 1, and it tells us how good a surface is at emitting or absorbing infrared light.
If we know the emissivity of our spacecraft is 0.6, that means we would need enough heat to be capable of emitting 912 watts per meter to be actually emitting 547 watts per meter.
Now that we know that number, we can use an equation known as "the Stefan-boltzman law", and it will tell us that the surface of the spacecraft will be 356 kelvin, which is 83 degrees Celsius or 181 Fahrenheit. Getting pretty close to boiling.
The study of this sort of situation is known as thermodynamics, or "thermo" if you want to hang out with the cool kids.
This particular analysis is significantly wrong. The sun only heats one side of the spacecraft, but the spacecraft emits infrared radiation from all sides and those two sides are also different temperatures so they emit different amounts of power. That's the sort of complication that makes thermodynamics so much fun.
In this video, you can safely ignore any slide with the "details" label in the corner as this material will not be on the test, but some viewers do want to know how the calculations are done.
For the heat input calculation, we take the total solar power and multiply by the absorptability constant to get the net input power. I didn't find any information about whether the derived value is based on the solar spectrum, but I can tell you that there are different ways to get that value with different equipment and it's not surprisingly more complex than you might think.
The temperature calculations use the Stefan-Boltzman law, which says that the power in watts is equal to the boltzman constant multiplied by the emissivity constant, the area in square meters, and the temperature in kelvin raised to the fourth power.
That fourth power in the equation means that the power output goes *way* up as the temperature increases. It's why the sun provides so much energy to the solar system.
BTW, there is nobody named "Stefan Boltzman" - it is the work of Ludwig Boltzman and Josef Stefan.
In our case we know the power that is going out needs to be the same as the power coming in, so we refactor the equation to solve for T and plug in the values for our case. Area is always 1 because we are looking at the power coming in and going out for the same square meter.
That gives us a temperature of 356 kelvin.
Now that we know a little about analysis, we can look at the orbital propellant depot.
We'll start with a stock starship in orbit around the earth. It has a tank of liquid oxygen at 90 kelvin and a tank of liquid methane at 112 kelvin.
Note that I've drawn cylindrical tanks rather than the actual shape of starship tanks. You can redo the analysis with actual tank shapes for extra credit.
There are three sources of heat gain for our propellant depot.
The sun provides the usual 1367 watts per square meter, but in LEO we're only in the sun for about half of the orbit, so the average heat gain is half that, or 684 watts per square meter. Multiply that by 0.42 - the absorptability of stainless steel - and we get 287 watts per square meter.
The second source of energy is the sunlight reflected off of the earth. We take the energy delivered by the sun and multiply it by the albedo of the earth - the proportion of light that is reflected back to space which is about 0.16. That gives us 219 watts per square meter.
Like the sun, we only get reflected light during half the orbit, which drops us down to 109 watts per square meter. Multiply by the absorptablity, and we get a net energy gain of 46 watts per square meter. We could also just have taken the net power of 287 watts per square meter and multiplied it by the albedo.
That takes us to the third component of heat gain - the infrared energy coming from the earth because the earth is hot. NASA estimates this at 240 watts per square meter. The earth is radiating on both the sun side and the dark side. Because this is IR radiation, we use the emissivity to calculate the effect of this - and we end up with 26 watts per square meter.
There are two additional sources to consider. The first is the infrared emitted by the propellant tank. The liquid methane is at 112 K, and the lox is at 90 k - that's cold enough that it's less than 1 watt per meter squared.
The second is heating from the background 3 kelvin cosmic radiation. It's less than a millionth of a watt per square meter. We will safely ignore both of these.
Based on how much heat the tanks are getting, we can figure out how much heat gain there is each day, how much propellant it vaporizes each day, and how long until the tanks are empty.
Here's a chance to see how good your guess was...
The LOX tank will last for 55 days, and the liquid methane tank will last for 50 days.
Those are bigger numbers than I expected - if you launched a tanker a day, I don't think the boiloff would be significant.
But we can do better.
Time for some more math.
To go from the amount of heat per square meter to the total heat, we need to figure out how much of the tank is exposed to that heat.
To make things simple, I'm assuming radiation hitting the side of a cylinder is equivalent to hitting a flat surface with the same diameter. This ignores the fact that absorptability depends on the angle that the sunlight hits - rays hitting near the edge are more likely to be reflected. I'm making the same assumption for reflected light from the earth. I'm also assuming that the light is hitting directly from the side, not obliquely. There are models that do not assume that, and they will likely give more accurate results.
The absorbing area is therefore just 9 meters multiplied by the height of the tank, or 112.5 square meters for the liquid oxygen tank.
To figure out the lifetime, we start by adding the three heat fluxes from sunlight, reflected sunlight, and earth IR to get 359 watts per square meter. Multiply it by 112.5 square meters and we end up with a bit more than 40 kilowatts of power.
That seems like a lot.
1 watt is equal to 1 joule per second, so that's 40.4 kilojoules per second.
Multiply that by the 86,400 seconds in a day, and we get nearly 3.5 million kilojoules per day.
The amount of liquid oxygen that much energy can vaporize is determined by taking the energy and dividing it by the heat of vaporization for liquid oxygen, which is 214 kilojoules per kilogram, giving us 16,300 kilograms vaporized per day.
The liquid oxygen tank holds 906,000 kilograms, so that means that it will take 55 days to evaporate all of the liquid oxygen.
The liquid methane lifetime is figured out the same way, using the size and capacity of that tank.
That's our baseline, the performance we get from a standard starship.
The obvious thing to do is to look for a better material than stainless steel. There are lots of coatings made for spacecraft to reduce thermal gain. I've chosen a coating known as AZW/LA-II from AZ technology.
This coating has an absorptability of 0.09. Looking at our solar energy absorption, that reduces it by 78%, down to 62 watts per square meter. We see a similar reduction on the light from the earth, down to 10 watts per meter.
This is looking great.
We now look at the earth infrared component. The emissivity went from 0.11 to 0.91, which bumps the IR heat gain all the way up to 218 watts per meter. The bulk of the heat gain is now from earth IR. Our lifetimes go up to 71 and 67 days, but that's only a small change.
This coating is designed for satellites that are warm enough to radiate quite a bit of infrared, but our LOX tank is cold enough that it will never radiate much heat, so a higher emissivity is worse for us.
Earth IR turns out to be a significant issue for earth-orbiting satellites. We want a high emissivity to radiate power back to space and stay cool, but that high emissivity means we absorb a lot of the earth IR and that warms us up.
For our scenario, a coating with a lower emissivity might be a better choice.
We find that not in the space industry, but in the solar industry.
Solec makes a coating they call LO/MIT, with an absorptability of 0.23 and an emissivity of 0.17. The higher absorptability pushes the solar and reflected gain back up by a total of 109 watts, but the IR absorption is 177 watts lower, so our overall reduction is 68 watts,
Our lifetime goes up to 90 days and 82 days.
There has been a lot of research on coatings and there are very likely coatings that are better for our scenario, but I think the current lifetimes make HLS scenarios quite possible.
Are there other techniques we could use?
http://www.solarmirror.com/fom/fom-serve/cache/43.html
We have been assuming that our tanks are sideways to the sun.
But what if we rotate the depot so the nose is facing the sun and keep it that way? That will reduce the area of the liquid methane tank from 86 square meters down to 64 square meters and push that lifetime up to 100 days.
This orientation totally shields the LOX tank from solar radiation, which means it only heats up from reflected sunlight and infrared from the earth, for a 311 day lifespan.
We now have a bit of a coating mismatch.
We have LO/MIT on the sides where it is great dealing with the earth IR and decent at dealing with reflected sunshine, but we also have it on the front that is exposed to the sun, where it's pretty bad at dealing with solar radiation.
If replace LO/MIT with AZW/LA-II on the nose, we reduce the solar gain considerably, bumping the liquid methane tank lifetime from 100 days to 167 days.
There's one more trick to play. Instead of having the liquid methane tank face the sun, we add a flat nose spaced 1 meter from the top of the liquid methane tank, with space in between.
With the AZW coating, it will absorb 123 watts per square meter of heat, and it will heat up until it emits that much energy.
Some of that energy will be radiated out to space and some will be radiated inside towards the top of the liquid methane tank.
The amounts will depend on the relative emissivity of the two surfaces. The outside one is 0.91 and the inside 0.11, so most of the energy goes towards the outside. It turns out to be 110 watts per meter squared outside and only 13 watts per meter squared for the inside.
That's a big difference in the solar gain to the liquid methane tank. It's low enough to get rid of the majority of the heat gain from the sun, and push the liquid methane tank lifetime out to 246 days.
At this point, the majority of the heat gain for that tank is coming from the earth. (do a chart for all of these to show the relative contributions?)
(math: 123 = T4 1 + T4 2 )
123 = T4 (1 + 2)
1 = 110W, 2 = 13W
Somebody is going to want the math, so here's what I came up with...
The power being absorbed is 123 watts per square meter.
That will be radiated to the front based on the emissivity to the front, and to the rear based on the emissivity to the back. We can refactor our equation and find that the power is the temperature raised to the fourth power times the boltzman constant times the sum of the front and back emissivity.
We know the AZW/LA-II is 0.91, and the raw stainless steel is 0.11. Solving for T, we get a temperature of 214 kelvin.
We can then figure out the two components, with 110 watts per square meter coming off the front and only 13 watts per square meter coming off the back and hitting the liquid methane tank.
What are our conclusions?
Based on the model, standard starship is probably good enough for HLS with a decent refueling cadence. Using simple coatings and/or a specific orientation makes things a lot better.
There are two big caveats.
This model has a lot of assumptions built into it, some of which might be problematic.
Second, heat models are hard, and there might be mistakes in this one. This information is presented for informational use only, and orbital propellant depot owners should consult a skilled professional whenever heat is involved.
And that's my analysis of starship as an orbital depot.
Some of you have probably noticed there's another heat analysis for this project.
Lunar starship will need to sit on the lunar surface for at least a week and retain enough propellant to get back into lunar orbit.
As soon as my brain has recovered a bit, it will be time to use the same principles and analyze what happens to the liquid oxygen and liquid methane with lunar starship sitting on the lunar surface. You'll probably want to subscribe so you don't miss that video, and you should also start thinking what your guess is for the lifetime in days...
If you enjoyed this video, please fact check the Wikipedia article on the Stefan Boltzmann law...
Artemis Lunar Architecture: https://www.youtube.com/watch?v=O2IBV_XSu60
Starship Orbital Propellant Depot: https://www.youtube.com/watch?v=fjWCEFioT_Y
Many of you already know the planned architecture for NASA's artemis moon missions.
SpaceX will launch a starship orbital propellant depot into low earth orbit, then fly a number of tanker flights to fill up that depot. A HLS lunar starship will be launched, refuel from the propellant depot, then proceed to lunar orbit to pick up astronauts from the orion capsule sent there by SLS, land on the surface, and then return to orbit to transfer the astronauts back to the Orion so they can return to earth.
You may look at this diagram and say, "that looks complex and risky".
And that, in fact, was what the National Team - headed by Blue Origin - asserted in this infographic in response to their loss of the HLS contract to SpaceX.
The first question to answer is "why did SpaceX choose this architecture?".
The long story is in my video on the artemis lunar architecture, which I've linked in the video description.
We need to look back at Apollo. Apollo had a very simple architecture - the apollo spacecraft included the command module that held the astronauts, the service module which held propulsion and the supplies to keep the astronauts alive, and the lunar module that would go to the moon's surface and back into lunar orbit.
For apollo 11, the command and service module massed 28.8 tons and the LEM added 15.1 tons, for a total mass of 43.9 tons.
The Saturn V used for Apollo 11 could send about 45 tons on a trans-lunar injection trajectory, so it was powerful enough to get the apollo spacecraft there.
The Artemis architecture is similar, but different in a very important way.
There is an orion command module, a service module made by the European space agency, and... that is it. Their total mass is about 26 tons.
The SLS block 1 that will be used to launch Artemis III - the first mission to the surface of the moon - can send 27 tons to trans lunar injection, and that seems fine.
The problem, of course, is that there's no lunar lander in this architecture. There was one known as Altair in the previous constellation architecture - a big one at 46 tons - but Congress told NASA to build the SLS rocket and Orion capsule and gave them money for that, but they didn't include a lander or money to build one.
There are bigger versions of SLS - the SLS block 1b has a higher-powered upper stage and it can send 38 tons to the moon, and the mighty SLS block 2 can do 43 tons. So you could fly a 17 ton lander - if you had one - on the block 2, but that actually wouldn't work because the Orion service module can barely get Orion into a lunar orbit and back out again, and it doesn't have the margin to also put a lander into lunar orbit.
NASA has a big problem. They have this "mega moon rocket" but it's only good for sightseeing to lunar orbit and they have no solution to get down to the surface.
They decided to do what we would all like to do - make it somebody else's problem. SLS would deliver orion to a lunar orbit, and somebody else would be on the hook to get the lander there. And the Human Landing System program was born.
NASA came up with this concept, the advanced exploration lander. It would have 3 modules, each massing 12-15 tons and launched independently.
There are two rockets that might work for this.
If they were 12 tons each, you could just sneak in with a Vulcan launch, times three.
If they were heavier - up to 18 tons - you could do it with Falcon Heavy, also times three.
And if you have New Glenn, that would probably work as well.
NASA isn't carrying the lander with SLS and it's going to take a big lander to get to the moon, do the landing, and get back to lunar orbit. Getting the lander to the moon requires a bigger vehicle and - with medium class rockets - multiple launches. It's inherent in what they are asking the vehicles to do.
After they lost the bid for the lander program to SpaceX, Blue Origin first offered NASA an extra $2 billion to reverse their decision, sued the government over the decision, lost their suit, and then successfully lobbied for a second program, and the Blue Origin team won that bid.
Humorously, the architecture that blue origin is planning requires not only refueling in earth orbit, but also a second refueling operation in lunar orbit. It *does* provide for possible reuse of the lander in the future, at the cost of an architecture that is more complicated and higher risk.
Both the SpaceX and Blue Origin architectures require orbital refueling to do what they need to do.
There is another good question about SpaceX's approach. Starship is a huge brute that will supposedly be able to carry 100+ tons into earth orbit.
Why not build a lander like NASA's advanced exploration lander - or Blue Origin's Blue Moon - launch it on starship, and send it to lunar orbit? Could SpaceX do that?
I think the answer is a pretty clearly yes. It would require developing a custom lander and a custom kick stage to it to the moon, and that would be fairly costly - likely more than SpaceX is getting paid on their current contract. The SpaceX HLS contract is $2.89 billion, which seems like a lot, but SpaceX is charging NASA $843 million to build a deorbit vehicle for ISS and the HLS lander is considerably more complicated than the deorbit vehicle.
Custom spacecraft are expensive, and custom spacecraft that carry humans are really expensive, and that approach doesn't align well with SpaceX's longer term goals.
The only way HLS makes sense for SpaceX is to take a version of starship and modify it to be a "mega moon lander" that would work for NASA's mission and also push them along the same path they were already planning for missions to Mars.
That is why SpaceX chose to build a starship variant instead of doing something like Blue Moon, and the starship variant invariably requires more refueling flights. But it also can carry a *lot* more mass to the lunar surface.
Is their orbital refueling approach feasible?
There are four questions to answer to determine the feasibility
Can they dock with the depot?
This is the easiest one to answer. SpaceX has flown 30 resupply flights and 12 crewed flights to the ISS. Starship is bigger but they own both vehicles, and it's pretty clear that they can dock them together.
Can they keep propellants cool enough?
I did a talk on using starship as a propellant depot that you can find linked in the video description.
My basic conclusion was that it was possible to store propellant for fairly long durations using passive techniques, including appropriate surface coatings and controlled orientation.
Those aren't the only techniques.
NASA has done a lot of development research on what is know as zero boil off technology. It uses a very specialized refrigerator known as a cryocooler that is used to keep the very cold propellants liquid.
MRI scanners need liquid helium to keep their scanner coils superconducting, all the way down at a temperature of 4.2 kelvin, and they use cryocoolers to either reduce or eliminate the loss of the helium.
The NASA designs are typically targeting liquid hydrogen at 20.3 Kelvin.
Starship uses liquid oxygen and 90.2 kelvin at liquid methane at 112 kelvin. Even if SpaceX wants to subchill those propellants as they do before launch, it's very feasible to build a system that will result in zero boil-off storage of both propellants as they are *much* easier to deal with than liquid hydrogen or helium.
Can they transfer propellants?
On the earth, it's really easy. Just connect a pump between the tanker and the storage container, and pump the fluid into the storage tank.
Or you can skip the pump and just put the tanker up higher and use gravity to make the transfer.
In either case you may need to do something with the vapor in the storage tank, as it will increase in pressure as the tank fills. If you've had a propane tank refilled, you will see the operator vent the propane vapor out to reduce the pressure.
But space gives us a problem. There is no gravity to push the propellant against one side of the tank, so your pump is mostly trying to pump vapor. That's not good for the pump and it doesn't work very well.
There's a simple way of dealing with this. You take a couple of small thrusters, put them at the bottom of the tank, and fire them. That pushes the tank up and the fluid down, setting it at the bottom so the pump can get it or it can flow into the tank "under" it.
These thrusters are typically known as "ullage" thrusters. Rocket stages in space that use cryogenic propellants use the same technique to get the propellant to the bottom of the tanks and therefore available for the engines to pump during a restart, so it's a very old technique and works fine.
The downside is the mass of the system, the mass of propellant it takes to create the thrust, and the fact that the thrust will change the orbit.
There's another approach that has been proposed.
You use a cryocooler to cool the storage system to a very cold temperature, subchilling the propellant and the tank walls. At subchilled temperatures there will be very little vapor in the storage tank and therefore very little pressure.
You then open a transfer pipe between the two tanks. The pressure in the tank then pushes the combined liquid/vapor mixture into the storage tank, spraying it in streams towards the tank walls. The tanks walls will then subchill the liquid and condense the vapor, keeping the temperature cold and allowing the propellant flow to continue.
NASA has gotten tanks 90% full using this technique, though not in orbit.
Speaking of orbit, NASA flew a mission known as RRM3 - robotic refueling mission 3 - to the ISS in 2018
It had two primary goals.
It achieved the second goal for 4 months, but the cryocooler failed and the methane had to be vented away, so the first goal was never attempted.
The thermodynamic approach seems elegant if you already have a cryocooler on your depot, but the ullage approach is more proven and quicker and that's what I expect SpaceX to choose. Or maybe they'll do both. Either way, there are workable solutions to propellant transfer.
The technology for the architecture is feasible, so the remaining question is around the number of tanker flights to support an HLS or a Mars mission.
And the magic number is... (drum roll)
Yeah, I don't know...
There seems to be a belief that this is a really important question, perhaps fueled by the early blue origin PR stating/complaining/whining that it would take 16 flights. Lots of discussions or fights about whether the number is 8 or 12 or 20 or whatever.
The actual number depends on a lot of factors. Among them are.
The tanker payload to orbit.
The mass of HLS starship
The cargo mass carried on the mission
The cargo mass left behind on the moon
The efficiency of the raptor engines
The landing engine design
I don't think I know *any* of those factors, and I think there are factors there that even SpaceX doesn't know right now...
But I also think the number doesn't really matter.
Let's say that we have an HLS launch scheduled 60 days from now and we have to complete the refueling in that time period. 60 days is easily done with current propellant depot technologies.
We'll look at two scenarios, one with an architecture that requires 5 tanker flights, and another with an architecture that requires 20 flights. The real number of flights is probably between those two numbers.
We fly once every 12 days for the first option, once every 3 days for the second one.
The target for starship and super heavy is rapid reuse, so flying every 12 days seems simple enough to do. The same pair should be able to fly once every three days, but let's just allocate 2 super heavy and starship vehicles so they can fly once every 6 days.
Neither of these look very challenging.
We also have to use a lot more propellant in the 20 flight version, but propellant is cheap.
The whole point of the starship architecture is that flights are cheap, and I would therefore submit that there's no reason to suspect that 20 flights is a lot harder than 5 flights. If you can do 5 flights, you can do 20 - you just need to turn the crank a few more times.
SpaceX flew 8 Falcon 9 flights from SLC-40 in the first 60 days of 2024, and that's with a vehicle that is less reusable than starship. 20 flights in 60 days should not be a significant challenge for Starship.
If starship isn't reusable, my estimates are that 5 flights becomes 4 and 20 becomes 14 because of the larger payload capacity of an expendable starship.
14 is obviously a lot more than 2 and that will definitely affect the economics, but SpaceX flew 19 Falcon 9 flights in the first 60 days of 2024, and every one of those had an expendable second stage. Non-reusable second stages would break the Mars scenario but probably aren't a problem for 3 lunar flights.
And that's what's going on. This has never been anything substantial behind the concerns about the architecture and especially about the exact number of launches, so don't worry about the exact number.
If starship is fully reusable, it doesn't matter, and if starship isn't fully reusable, SpaceX needs to make it fully reusable because both Starlink and Mars are relying on the reuse.
If you enjoyed this video, it's been hot in Seattle.
Mia and the martians:
https://www.indiegogo.com/projects/mia-and-the-martians-publishing-campaign#/
Off Nominal Podcase episode
https://offnom.com/episodes/166
Rocket Factory Augsburg test anomaly update: https://www.youtube.com/watch?v=ORviDaDsj7U
Rockets Behaving Badly - Pogo
https://www.youtube.com/watch?v=Fn9hAnaoDfE
I've talked about raptors and how their reliability will affect the reliability of Starship overall, but now that we have 3 successful launches there's some useful data and I've come up with what I hope is a better way of explaining things.
I need to start with a few caveats.
To evaluate reliability, NASA commonly uses an approach called probabilistic risk assessment. In this approach you look at every component in your system and predict how likely it is to fail, and then combine those together into large components, and finally into your whole vehicle, factoring in the redundancy that you have built into your system. That spits out a number.
If you have heard that the probability of loss of crew (LOC) for crew dragon is 1 in 276, that number comes from this sort of assessment.
But it's not a real number, which I can illustrate with three examples.
On June 28th, 2015, the second stage of Falcon 9 exploded during the CRS-7 mission. On September 1st, 2016, a Falcon 9 exploded during fueling before a static fire on the AMOS-6 mission. On April 20th, 2019, a crew dragon capsule exploded during routine testing of the super draco abort engines.
What caused these failures?
On CRS-7, a strut holding a helium tank was understrength and broke, letting the tank break free and overpressurizing the second stage.
On AMOS-6, a composite overwrapped pressure vessel exploded due to the creation of solid oxygen in the composite weave.
On the crew dragon test, the interaction of a titanium valve with a hypergolic propellant led to an explosion.
The commonality between all these failures is that they don't show up in a PRA assessment. "Strut weaker than the specification" didn't show up in the possible failure modes, and both of the AMOS-6 and crew dragon explosions were failure modes that had never been seen before. "Blew up because of something weird nobody had seen before" is also not in the list of failure modes.
That's the big weakness of PRA - you need to be able to conceive of a failure mode and assess how likely it is to properly factor it into the assessment. And that should make you skeptical of the numbers out of an analysis like PRA.
The alternative is to use empirical data. Test the hell out of all the individual parts and fly all the time, and you'll get a decent idea how reliable your vehicle actually is.
The problem with that most rockets don't fly very often. Delta IV flew using the RS-68 a total of 45 times in the medium and heavy variants with only one partial failure. So how reliable is it?
The RS-68 could be a 1 in 25 engine and we just got lucky. Or it could be a 1 in 250 engine, or even better than that. We just don't know.
There are only two rockets with decent empirical numbers. Soyuz has launched 1680 times in various forms, so we have a lot of information on the RD-107 and RD-108 engines that it uses. And Falcon 9 has 389 launches.
And just to complicate things a bit more, both the Soyuz and Falcon 9 engines have been modified along the way.
Returning to Raptor, we have some real empirical data to play with. IFT2 and IFT-3 both had all engines light and work until staging, and IFT-4 had one engine fail to light but the remaining 32 of 33 worked until staging.
So that's 98 out of 99, which I'm going to round up to a demonstrated 0.99 reliability. There are significant caveats on that number - we could be lucky or unlucky - but let's just assume it's the actual number for sake of argument.
If we have an engine like that, what would we expect in terms of reliability?
Let's say we're talking about the RS-68 engine that flew on the Delta IV series, and let's assume it has 0.99 reliability.
If we're flying the delta IV medium with a single engine, we don't need to do any math - 99 times out of 100 it will be successful...
But what if we are flying the delta IV heavy with 3 RS-68 engines. The chance of success is 0.99 but we have three engines, so we multiply those probabilities together - which is the same as raising the base probability to the power of the number of engines - and we find that the overall probability of success is 0.97.
That's about a 1 in 33 chance of failure, which makes sense.
There's an underlying assumption that shutdown equals failure, and that's not quite true.
During the launch of Apollo 13, the center engine of S-II second stage shut down two minutes early because of chamber pressure fluctuations.
It was, however, about two thirds of the way through the stages burn and the other engines were able to run a little longer, and it did not lead to a loss of mission.
See my video on POGO for more information.
Similarly, STS-51F had a main engine fail three and a half minutes into the flight, but the remaining engines were barely able to put the shuttle into orbit, though the orbit was lower than planned.
There is math to try to factor when an engine might shut down during launch into reliability calculations, but I'm not going to get into it here, and I'll assume that losing an engine always happens early enough in a launch for it to be problematic.
Time for a graph.
If our engine is 0.99 reliable, then we can expect that it will fail once out of every 100 uses. That's pretty straightforward.
As we add engines, the reliability goes down, and by the time we get to 8 engines, we can expect an engine failure every 13 flights.
Having a bunch of engines is clearly a bad idea.
And then we add a ninth engine, and something unexpected happens. If we use our equation of raising the reliability of 0.99 to the power of the number of engines. we should get a value of 1 in 12, way down here at the bottom of the graph.
But when we get to 9 engines, a magical thing happens, and that magical thing is engine redundancy. We have enough engines now that if one engine fails, the remaining engines are powerful enough that they can still complete the mission. Our booster is "one engine redundant"
Our chance of one engine failure is an alarming 0.086, but for the booster to fail we need two engine failures, and the probability of that is 0.0075, or 1 in 134.
That's better than our original reliability, but it's not a lot better. Hold onto that thought, I'll be back to it in a bit.
You may be wondering why the magic number is 9. I chose 9 because we know that Falcon 9 is single engine redundant
8 probably works and 7 might work, but fewer than that gets less likely - we are nearing the Saturn V example and that rocket isn't getting off the pad with only 4 working engines.
Back to our question - dealing with the hassle of 9 engines seems like a lot of work if you only get a 30% increase in reliability.
I chose a rocket engine that is only 99 percent reliable, and that's not a great engine. What if we can do better?
This is an engine that has 0.995 reliability, or 1 in 200. With 9 engines, that's 1 in 513, or 2.6 times better than a single engine. That is getting interesting...
Bumping up to 1 in 500, or 0.998, the reliability with 9 engines goes up to 1 in 3136, or 6.3 times the reliability of one engine.
I'm sure you don't expect me to stop here....
If you can built an engine with what we would call "three 9s" of reliability - with 1 in 1000 chance of failure, our 9 engine booster is expected to have a reliability of 1 in 12,445, or 12.4 times the reliability of 1 engine. That's a full order of magnitude better than the single engine solution.
The question is whether we should actually care about this.
The mighty F-1 engine that powered the Saturn V flew 13 times, so it only had 65 engine flights. If it's a 1 in 200 engine - and it probably is at least that good - the chances of an in-flight failure are small.
409 RS-25 engines flew on the shuttle and now SLS. Empirically, it's about a 1 in 400 engine.
Falcon 9 has flown 389 total launches, so that's 3501 Merlin 1D flights. And at those sorts of flight volumes, I care quite a bit about the difference between 1 in 1000 and 1 in 12,000. My analysis is that the Merlin 1D is a 1 in 1000 engine, perhaps a bit higher.
We are already exposed to the value of engine redundancy.
Normal FAA rules require that any two-engine airliner always be within 60 minutes of an airport, so they can reach a safe place within an hour if one engine fails. We therefore map out 60 minute circles and choose a flight path where we are always inside one of these circles.
That's unfortunately inefficient because we aren't flying the most direct route.
The FAA has therefore created a program called extended-range twin-engine operational performance standards, or ETOPs
If your plan and engines are good enough - if they have a higher reliability - the FAA will let you extend the 60 minutes to a larger value.
If a plane is approved for ETOPS 120 - like the airbus 320 - then it can fly the shortest route from new York to London as it is always withing 120 minutes of one of those airports.
For flights in the Pacific, this becomes more important because the distances are so much longer.
Airplanes that fly these routes might be ETOPS 180, 240, or even 330 minutes like this Boeing 777.
It's all based on engine redundancy - twin engine airliners must be able to fly with only one engine.
But you came here to hear about raptor and starship.
I'm going to assume that every 9th engine gives you redundancy. It might be a different number but it's not going to change the results much.
Here is super heavy at 0.99 reliability. We see the same pattern as before - every 9th engine gives us a nice bump in reliability - and then engines added after that reduce the reliability.
27 engines not surprisingly gives us the best result, at about 3.1 times the reliability, but adding 6 more engines bumps it down to only 1.6 times the reliability. Not very exciting for a booster that has 3 engine redundancy, where it takes 4 engines to fail to have real issues.
It's pretty unlikely that Raptor is a 1 in 100 engine. Let's explore if it is better.
If Raptor is only 1 in 200, 27 engines is 1 in 3900, and 33 engines drops down to 1 in 1850. Only 9 times the reliability of 1 engine.
But Raptor's probably not a 1 in 200 engine....
At 1 in 500 or 0.998 reliability, it starts to get silly.
27 engines gives us a 1 in 130,000 chance of failure, and going up to 33 engines drops that down to 1 in 60,000
Time to take the final step...
If raptor is a 1 in 1000 engine - if it's as good as merlin 1d - 27 engines gives you pretty close to a 1 in 2,000,000 million chance of a failure of the stage, and 33 engines drops that down to only 1 in 900,000. It is *900* times better than a single engine solution, or pretty close to 3 orders of magnitude.
At this point, I'm sure some of you have questions. You, the one in the back waving your hand around?
Decaying Flavors? Huh?
Oh, Cascading Failures.
I guess that means I have to show the video.
On August 9th of 2024, Rocket Factory Augsburg was performing a static fire and they had an anomaly that led to cascading failures when the stage failed to shut down. The analysis video is linked in the description is worth a watch.
I wanted to show this because a) it's fun to watch and b) to emphasize that this sort of thing doesn't really happen in practice. I've watched a *lot* of rocket explosions videos over the years and have never seen this sort of failure before.
Rocket engines are designed to shut down rather than blow up and take the rest of the vehicle with it.
But it's certainly true that this risk does exist in rockets with multiple engines, but there is a mitigation...
Let's say you are ULA and you are launching a Vulcan, which is powered by 2 BE-4 engines.
One of those engines starts acting hinky. Do you shut it down?
If you do, your payload will end up in the ocean.
If you let it keep running, maybe it works long enough to finish the mission.
So you set your engine shutdown parameters a little loose because the worst thing you can do is shut it down too early.
If you are flying astronauts and they have an abort system, you probably approach it a bit differently as they are safer in an abort with a shut down than one with an exploding rocket stage, but there are some cases where that isn't true.
I talked earlier about the STS-51F shuttle mission where an RS-25 engine shut down, and after that happened the crew disabled automatic shutdown on one of the other engines that was starting to act hinky, and that saved the mission.
If you have engine redundancy, the decision tree is different. If an engine starts acting hinky, you shut it down because doing so does not compromise engine success and if you keep it running it might break other engines.
This is especially true if your vehicle is reusable. Shut the engine down before it gets damaged and the fix is going to be simpler and cheaper than if you let it run longer.
It's therefore possible that the engine that didn't light for IFT-4 didn't actually have a problem but just looked a little hinky, and since they don't actually need all 33 engines running it automatically shut down.
For super heavy, it's not going to the raptors that are the problem, it's going to be everything else that is required to get through the missions successfully. Far more likely to have a problem elsewhere.
We can also spend a little time looking at starship.
Starship has 3 small nozzle sea level engines in the middle and 3 large nozzle vacuum engines on the outside.
Is this arrangement engine-redundant? It's not clear; SpaceX hasn't talked about this and the requirements for second stages are easier than first stages because they have smaller gravity losses.
Starship 3 is planned to have 3 sea level engines and 6 vacuum engines around the outside, as shown in the photoshopped image.
There will be engine redundancy with this arrangement.
Starship needs a single functional sea-level engine to land successfully.
That means for it fail, 3 engines need to fail. If they are 1 in 100 in this scenario, we need to hit the "1" part three times, so it's a 1 in 1,000,000 chance.
And that's with only a 1 in 100 reliability.
Like super heavy, we should worry about other things than engine reliability for starship landings.
And that's the story...
Multi engine redundancy will make the chance of normal raptor failures a non-issue on super heavy and we can expect high reliability on Starship 3 for both launch and landing.
If you liked this video, maybe you would enjoy a concert.
But actually, what I'd really like you to do is go to the Mia and the Martians project on indiegogo and pledge your support.
What this world needs is more cool space books for kids.
Link is in the description, along with a link to the Off Nominal podcast discussion of the book.
For the past few years, the main question about starship is "will it work?"
But there's a very good chance that in the next year, starship will become operational, and that question will change to "what can we do with it?"
We know what SpaceX wants to do with Starship...
They want to use it to launch the much larger Starlink Version 2 satellites
They want to finish the HLS lunar starship mission as part of NASA's Artemis mission to return to the moon.
And ultimately, they want to use it to build a mars colony.
But how will everybody else use Starship?
What can you do with a super-heavy lift rocket?
It's a fair question, as super-heavy lift rockets are rare. I'm going to use the Russia benchmark of at least 100 tons to low earth orbit as otherwise this class includes Falcon Heavy and that is just wrong.
The Saturn V was built purely to get to the moon and its only other use was to launch Skylab. The Soviet N1 never flew successfully but likely would have been purely dedicated to the moon mission.
The Soviet Energia is one of the more disappointing rockets around as it only flew twice, once carrying the Buran Shuttle. After those flights it was discontinued. Tons of potential to do interesting things, but the Soviets didn't come up with any uses.
And of course, the space shuttle. It was super heavy lift if you look at how much mass it took into orbit, but if you look at the payloads it actually carried, it was mostly a medium-lift rocket. So we'll have to take it away.
Based on these rockets, there's not much guidance for what we might use starship for.
What we need is a framework to compare what Starship can do to rockets that are currently operational.
What we're going to be talking about is payloads to various orbits, and we'll be doing comparisons to Falcon 9 and Falcon Heavy, both in their fully expendable modes. Adding in the reusable versions adds a lot of complexity because of the different reuse modes so I've decided to skip it.
Starting at the bottom, Falcon 9 can carry 22.8 tons into low earth orbit, but only 8.3 tons to geosynchronous transfer orbit. The GTO orbit requires a lot more energy and cuts the payload down considerably. I didn't find any published numbers on Falcon 9's payload to geostationary orbit, but one of my models suggested that it's about 3 tons, so we'll go with that. Atlas V is a little less than 4 tons, so that's in the right ballpark.
Falcon Heavy can lift 63.8 tons to low earth orbit and a healthy 26.7 to geosynchronous transfer orbit. It can also carry about 9 tons all the way to geostationary orbit, which is the main reason Falcon Heavy exists - the NSSL launches for the government require SpaceX to be able to launch 6.6 tons all the way to geostationary orbit.
Note that any number higher than about 18 tons is a theoretical number - if you want to use Falcon Heavy to lift 63 tons into low earth orbit, you'll need a redesigned upper stage and a whole new payload adapter to do it.
So what about starship?
Starship gets you 150 tons in low earth orbit.
Voyager Space - a holding company for a number of space-related businesses - and airbus have teamed up to create starlab, a commercial space station in low earth orbit that will launch on a single flight of starship.
That's the only project I found planning to use all of starship's capacity at once.
Here's some data for starship on higher orbits.
We can expect it to only deliver about 21 tons to geosynchronous transfer orbit. That's only about 14% of the payload to leo and quite a bit less than Falcon Heavy can do.
What's going on?
It's all about empty mass.
The Falcon 9 second stage is very light - only 3.9 tons. That means when it's going into low earth orbit, 85% of the mass is payload. Going to higher energy orbits, the empty mass obviously remains the same, so a reduction in total mass only comes from the payload. For geosynchronous transfer orbit, we're down to 68% payload, which is still quite good.
Starship is a very different vehicle. Because it is designed to be reusable, it's carrying a lot of extra mass - flaps, heat shield, motor protection, and landing fuel. That bumps up the empty mass and therefore knocks down the payload at the same time. So at LEO we end up with a payload percentage of only 54%. For GTO we're still carrying all that extra mass, and our payload percentage goes way down, to 14%. A little harder orbit and the payload goes to zero.
Much better than shuttle, which had an atrocious payload percentage, but reusable second stages are always going to have the problem of high empty mass.
Are there ways to get to higher orbits efficiently using starship?
The obvious solution is to add another stage. I talked about Impulse Space's Helios kick stage in a previous video, and their goal is to put payloads directly into GEO from LEO. A conservative estimate says that they can get 5 tons to GEO with a 15 ton vehicle, so 20 tons total mass.
Do a bit of math, and we find that you could carry 7 kick stages and payloads in a single starship launch and that would get you 35 tons to GEO in a single launch. Though not a single 35 ton payload.
Adding that to our graph, if we can carry 7 Helios kick stages with payloads, that puts our starship payload to Geostationary orbit to 35 tons.
That's pretty good for the cost of 7 kick stages which are supposedly going to be relatively cheap.
K2 space is a new company taking a new approach...
Traditional wisdom assumes that it's expensive to launch a kilogram of mass and it's expensive to launch big satellites.
That has driven satellite bus designs to be light, elegant, and compact. And also hugely expensive.
The first ViaSat-3 satellite was launched in 2023 on a Falcon Heavy. The satellite cost over $500 million dollars, not including the cost of the Falcon Heavy launch, probably in the $100 million range.
If you want the ultimate communications satellite, that may be the only way to do it.
The market is in a very strange position. Launching on Falcon 9 costs about $69 million
A communications satellite costs $300 million and up, partly because it's constructed to be as light as possible.
What if you could build a satellite using less expensive materials that would do 90% of what your expensive one could do but it only cost $100 million?
That seems interesting, but the satellite manufacturers are in the business of selling the most expensive satellite possible.
Let look at a constellation example in low-ish orbit.
Let's say you build a wonderful constellation satellite at $100 million each using the same tech as in the geostationary satellites. They are small and light, so you can fit 5 of them in a Falcon 9. That gives a total price to build and launch 5 satellites at $569 million, or $114 million each.
Now let's say you build simple heavier satellites, and that allows you to build them for $50 million each, but because of their size and mass you can only fit 3 in a Falcon 9. That gives you total cost of $219 million for 3 satellites, or $ 73 million each. About 25% cheaper.
That is K2 Space's bet. Because the cost of launch is so much lower, you can build capable satellites that are bigger and heavier but will end up being cheaper.
Their target is $15 million per satellite. If they could achieve that in this scenario, the cost per satellite drops to $38 million each.
Highly optimized expensive satellites make little sense when launch is so cheap.
K2 is working on two products.
The mega satellite bus is a flat-pack design that will fit 10 satellites in a Falcon 9 fairing, stacked the way Starlink Satellites are stacked. It has a one-ton payload and a power system that can supply 20 kilowatts of power, an immense amount of power for a satellite this size.
Their second product is the giga satellite bus using the same concept but sized for starship, with a payload of 15 tons and 200 kilowatts of power.
K2 will be worth watching in the next few years.
The other option to deal with starships low payload to harder destinations is pretty obvious.
Inherent in the starship lunar architecture - and the starship architecture in general - is the concept of orbital refueling. It simply does not work without orbital refueling, and SpaceX is working on hard on that capability.
So, if we have that capability, what does it do for us with normal payloads?
We can use refueling to turn starship into its own kick stage. If we take a starship and refuel it with 4 tankers, we have enough oomph to take the entire 150 ton payload to GEO.
It probably does not leave enough margin to do it in a reusable mode, so the starship stays out there. An expendable starship with more than four tankers of propellant will likely do a lot more payload, perhaps 250 tons.
Note that if you are launching from Florida or Texas, the cost to get a payload to geostationary orbit is pretty close to the cost to get it to the moon, so that means 150 tons to the moon as well.
There are two really obvious customers for this sort of capability - the US Space Force would love to have that capability in geostationary orbit, and NASA could certainly use it for its moon plans.
These are the options for payload to harder orbits. Use a kick stage like Helios, or refuel and use starship as its own kick stage. Either give very impressive payload to GEO and beyond.
I'd like to share one final perspective with you.
Noted science fiction authors Jerry Pournelle and Robert A. Heinlein were discussing the plot of a story, and Pournelle remarked that if you can get a ship into orbit, you're halfway to the moon.
Heinlein corrected that if you can get a ship to orbit, you're halfway to anywhere.
The ability of Starship to get very heavy payloads into low earth orbit and beyond opens up a whole new era in space exploration.
If you enjoyed this video, give this a listen
People have been asking me to talk about stoke space for a couple of years...
I long resisted it because I didn't think I had a useful perspective to add and I had no way of answering the question that I thought people really wanted answered.
But I was looking at it recently and I realized that I had a way to add something to the conversation and do the sort of analysis that I like to do.
When I'm analyzing a company, I'm trying to evaluate the chances that it can become a commercial success. There are a bunch of different ways of doing such an analysis - my starting point is usually looking at 4 things that I will call my pillars of commercial success.
Pillar #1 is a good idea. Note the emphasis on the word "good". It's not a cool idea, it's not paradigm breaking idea, it's a good idea. Good ideas are often simple - remember that Amazon started with the idea "maybe we could sell books using this internet thing?"
Pillar #2 is a team that can execute on the good idea. There are the obvious technical skills required to do whatever you want to do, but you also need what I call a "founder class entrepreneur". You need somebody who understands the technical side and goal deeply, somebody who can deal with investors, and somebody who can make a dollar scream. These people are not surprisingly rare.
Pillar #3 is a meaningful market. You need people that are going to be paying you cash money. This is about having a market where your new company can be competitive enough to make money.
Pillar #4 is runway - you need time to bring your good idea to reality. An idea that is attainable and a management team that can keep your cash burn as low as practical is critical, because pretty much everything costs more and takes longer than you expect.
Now let's talk about Stoke.
We'll start with Stoke's idea, which is all about second stage reuse. SpaceX is working on that with Starship, but nobody else has even gotten to first stage reuse.
Stoke's second stage is a modified capsule shape and is made out of stainless steel. It is fueled by liquid hydrogen and liquid oxygen. Their engine uses a traditional expander cycle, where the very cold fuel is run through the rocket nozzle and the combustion heat produces a pressurized output is used to power the turbopump. This is the same cycle used by the hugely popular RL-10 upper stage engine.
There are a few different variants of the expander cycle, and it's not clear to me which one they have chosen, and that choice probably doesn't matter for the analysis.
At this point, a traditional design puts the engine with a large nozzle at the bottom of the stage, and you are off and running.
Stoke has a different idea, however. Instead of a single big combustion chamber, they are putting 30 small combustion chambers spaced equally around the outside of the second stage. These combustion chambers can be independently controlled to provide differential thrust and therefore steering.
In 2023 Stoke posted this video of their second hopper flight test, which shows they are able to get their engine up and running and have enough control to achieve a hop with 15 combustion chambers.
Note that this is a very early test - a short hop is a long way from a second stage engine that runs for a full cycle and meets performance requirements.
At this point, I hope you are saying to yourself, "this looks very complicated and I don't understand the point".
The point of their design as all about getting the second stage through reentry and back to the ground. Their approach is a new one, using active cooling to keep their second stage from melting.
Build into the base of the second stage is a heat exchanger, and it will take in cold liquid hydrogen, absorb reentry heat, and use that heat to drive the expander cycle turbopumps the same way nozzle heat is used on ascent.
It's not clear what happens to the heated hydrogen, but my best bet is that it is burned in the rocket combustion chambers with some liquid oxygen, helping protect the rocket chambers from the heat of reentry the same way Falcon 9 starts its engines to protect them.
The system that handles both ascent and descent and uses active cooling is Stoke's idea.
Will it work?
I've thought long and hard about this question, and my very firm answer is "I don't know".
Reentry heating is notoriously complicated to model, so this isn't an easy question to answer.
The best summary of what has been tried for active cooling appears to be this NASA memorandum from 1994.
My quick summary is that it turns out to not surprisingly be very complicated, though I will note that they were working primarily on winged structures which are much more complicated than the capsule shape that Stoke is envisioning.
This paper references 55 different papers, so enjoy your time exploring them and figuring out your answer on whether it will work.
To figure out whether active cooling might be better, there are many things that you would want to consider.
Which one has a greater payload penalty? Is one 10% and the other 40%? Or are they pretty close?
How do their development costs and times compare?
How much will it cost to construct a vehicle using each technology?
Is one more robust than the other? What failure modes do they have?
Weighing all of these factors would be a long and challenging project for any engineering team, and it's not something we can do looking from the outside.
It's often very useful to look at prior art.
If we look at existing systems, the shuttle orbiter used heat shielding tiles. The soviet "Buran" shuttle used heat shield tiles.
The air forces X-37B "minishuttle" uses heat shielding tiles. Sierra Nevada's dream chaser cargo system uses tiles.
And of course, SpaceX's starship uses heat shield tiles.
5 systems already deployed or entering testing, all using tiles.
And on the other side, a new idea that has been explored a bit but never built.
Conventional wisdom is always right except when it isn't, but you need to really understand what is going on before you decide that you are right and everybody else is wrong.
My next analysis point is a team that can execute on the idea.
Stoke has two co-founders...
Andy Lapsa is the CEO. His past experience is as the director for the BE-3 and BE-3U engines at Blue origin and as an engineer for the BE-4.
His co-founder Tom Feldman is CTO for the company, and he brings specific experience from the BE-4 program.
Engine development experience is obviously critical to building a new launch system, especially one with new engine concepts.
What I don't see at stoke is a entrepreneur founder type - the person who will figure out the critical path to operation, the person who will deal with investors, the person who understands efficiency in both time and money. I also don't see the distilled vision I would hope for - there is a big long term vision but what is the path to take a small company in Kent Washington to one that is successful with customers around the world? The entrepreneur/founder is the make or break person for most startups.
My other concern is about both founders coming from Blue Origin. Blue Origin is likely the new space company that is most like the old space companies and therefore least likely in my opinion to create entreprenuers.
The next analysis point is a meaningful market.
Looking at Stoke's website and their materials, what I see as their market description is "any orbit. Any time."
The US launch market is very complex - I can come up with 9 distinct markets. Some are competitive, some are very hard to compete in. If you want more detail on these markets, see my video "who will compete with SpaceX"
To compete in one of these markets, you need be a credible company and be able to compete directly with the three US companies launching now - Rocket Lab, ULA, and SpaceX.
All three of these companies have advantages - existing launch sites, lucrative fixed-price contracts, reasonable flight rates. These are structural advantages - ones that are very hard for new entrants to counter.
You need a compelling reason why a customer would choose to fly with you over an existing competitors.
And just to make things more fun, all the existing launch companies are not selling services at their most competitive price, they are selling at the price that maximizes their revenue. Aim to undercut them by 25% and you may find out that they can undercut your best price.
Finally, you need to have runway - enough money to keep your company running for long enough for you to reach your goal and get a chance to compete. Lack of enough runway is the killer of startup companies in many different industries.
We interestingly have some numbers we can throw at this. Falcon 9 1.0 was a simple launcher using the already developed Merlin 1C engine plus a new vacuum variant of that engine. It contained no real innovations.
SpaceX spent $443 million developing that simple rocket, including the first two flight vehicles.
Later versions of Falcon 9 deployed two new ideas - a significantly improved version of the Merlin engine, and all the tech for first stage landing. That cost them a further $1 billion.
Stoke is developing a first stage with a brand new advanced engine and they plan to land it like Falcon 9. And they have the brand new second stage reuse tech to develop and test.
That's a lot of new stuff, and a reasonable price tag is somewhere between $1 and $2 billion dollars.
So far, Stoke has raised $170 million in total. Will that be enough for them to finish their idea? Will they be fast enough compete with Starship and Neutron?
What does my analysis suggest?
Stoke has an interesting idea, but I don't have enough information to evaluate whether it's a good idea compared to the alternatives.
I'm concerned that they don't have the right leadership team - they are high on engineering but lack the driving founder who can do all the things that need to be done.
I'm not sure what market they are trying to target. Launch is very much a cutthroat business and the existing companies are entrenched. You need a good market plan
Stokes plan is quite complex and my guess is it's $1-2 billion to execute on their idea. They currently have less than a tenth of that.
I love different ideas, but my analysis is that Stoke is unlikely to become a commercial success.
If you enjoyed this video, please send this T shirt, size medium, Heather Midnight Navy.
https://www.youtube.com/watch?v=rJr360r_LfQ
Proton M launch failure, July 2013
https://www.youtube.com/watch?v=ycRVAcZC5R4
https://www.youtube.com/watch?v=EJ5__1PPgNQ
When the super heavy booster from flight 5 came casually cruising in and was caught by the tower, the space internet went wild
And like few space things do, it leaked out of the space community and showed up big in the mainstream news, such as
The associated press
Reuters
BBC News
and even Jeff Bezos' Washington post.
My response was a "cool". Which will probably annoy some of you, but this was never one of the most challenging parts of the starship program, and my prediction was an 80% chance of success if they got to the start of the landing burn.
So, "cool and important for the program going forward" sums up my opinion.
But that would make a very short video, so I've decided to compare this accomplishment with something from SpaceX's past.
I want to talk about how the super heavy catch compares to the first droneship landing of Falcon 9.
To do that we'll need a scorecard.
We'll start by looking at vehicle scale, and super heavy is a super heavy lift rocket, while Falcon 9 was a medium lift rocket. A larger vehicle is harder to build, so that's one for super heavy being harder to do.
The precision control required is what most people were paying attention to for the super heavy catch; it needed solid control within a meter to be caught, while the broad deck of the drone ship gives at least a 10 meter margin of error for a successful landing. That's another win for super heavy in terms of difficulty.
But maybe not.
The drone ship thrusters are capable of maintaining location within 3 meters, and both the drone ship and the booster are using GPS which is accurate to perhaps 1 to 5 meters. And wind gusts may push the booster around.
That means the droneship platform isn't huge compared to the available precision, and you can see from this picture that sometimes the landing is quite a bit off center. There's a reason the droneship platforms are big.
This starship image shows the super heavy support pins sitting on top of the landing chopsticks. The booster needs to have the proper orientation and be pretty close to the right position - say within a meter - for the landing to be successful. That requires much more precision than the Falcon 9 booster.
GPS might be up to 5 meters off, so the position fix that the booster is getting could be anywhere within the yellow circle.
That's not going to work.
Luckily, there's a more advanced technology available.
There is a fixed GPS receiver on top of the tower that knows exactly where it is and it also knows where GPS is saying that it is. It can compare those two values and send an error correction to the booster that essentially says "take what GPS is telling you and add the following corrections to get your actual location".
This technique is known as "differential GPS" and is commonly used in applications that require high precision - if you've seen surveyors in your area, they are very likely using differential GPS.
Differential GPS can give positional locations plus or minus 1 to 3 centimeters.
What this means is that Super Heavy doesn't really have a harder job getting itself in the proper position than Falcon 9 does.
So it turns out that super heavy can pretty easily get the precision that it needs because it's being caught by a fixed tower, so I'm going to say that getting the required precision is equally difficult for both boosters.
Next comes the landing approach.
We'll look at two videos. The first is the catch of super heavy from flight 5, and the second is the first droneship landing of falcon 9 on the CRS-8 mission way back in April of 2016.
I've synchronized the videos to make the differences more apparent. Both will start one booster-length above the final position.
(run video)
You might want to rewind and watch that a few times.
The Falcon 9 lands at a relatively constant rate. It has to do this because the minimum thrust of the Merlin 1D engine on the Falcon 9 is greater than the weight of the booster when landing, and therefore the booster cannot hover. This trajectory is known as a "hoverslam", as the booster has to slow down at exactly the rate that allows its vertical velocity to hit zero when the legs touch down, and that limits any positional adjustments that it can make.
Super heavy can hover, so as it gets close to the catch location it slows *way* down and spends about 3 seconds fine-tuning its final location.
The hoverslam requirement makes landing the Falcon 9 harder than catching the super heavy.
The hoverslam is the more difficult approach and Falcon 9 wins the landing approach difficulty competition.
There was also a big difference in development effort.
SpaceX built 2 development vehicles for Falcon 9 reuse.
The grasshopper test vehicle flew 8 flights, and the full sized F9R dev vehicle flew 5 flights.
SpaceX continued development with boosters that had already completed their flight missions for customers. They had a lot of work to do, and it wasn't until the 11th attempt that they landed on the droneship successfully.
Adding in the 13 flights of the two development vehicles, that's 24 flights to get to the goal. Along the way they developed a host of technologies that were required to achieve the final goal, and many of these were things that had never been done before.
Super heavy leaned heavily on everything SpaceX learned from Falcon 9, and it only took two successful ocean landings and then a successful catch, for a total of 3 flights.
SpaceX didn't do any tests like grasshopper or F9R Dev with super heavy because they didn't think they needed them, and they were correct.
It was much, much harder to get to the first Falcon 9 droneship landing than the first Super Heavy catch.
I'll label the development effort for Falcon 9 high and the development effort for Super heavy low, and Falcon 9 wins this one easily.
Falcon 9 had the added constraint that they had a requirement to keep flying missions without failure to bring in the money that was keeping SpaceX going, while the Super Heavy team has the luxury of flying pure development vehicles where a failure is much less impactful. Falcon 9 wins that one.
Finally, if we look at expertise, we find that the Super Heavy SpaceX team is more experienced than the Falcon 9 team was, and Falcon 9 wins again.
With the exception of vehicle scale, Falcon 9 has tied or won on all of the difficulty measures, and that's why I wasn't surprised that Super Heavy was successfully caught. It just wasn't that hard for a team that is as good as the SpaceX team is.
My issue with the flight 5 coverage is that most people are focusing on the wrong vehicle.
But before we get into that, a short message.
Choose wisely...
Like and subscribe, or
Instead of super heavy, we should be looking at starship.
They rate about the same on vehicle scale and on required vehicle precision, so there's no clear winner for either of those.
Landing approach. We known that Falcon 9 does the difficult Hoverslam...
Starship comes in flat in what is called a "bellyflop" position, and then as it gets close to landing, it performs a landing burn to reorient to a vertical position.
Here's how it looked on the SN15 test flight...
The orbital version will at some point be caught by the same tower than caught super heavy.
Starship comes in with the bellyflop, flip, and catch approach. Like superheavy it can hover, but the maneuver that it does is much more complex and makes the vehicle design more complex - that's why it took a few flights to figure it out. I think these were roughly equal to develop.
I said that the development effort for Falcon 9 was high. What about starship?
Starship is a weird vehicle that falls sideways using fins for control unlike shuttle which flew like a plane for reentry and had traditional aircraft control surfaces, plus the somewhat weird body flap.
Starship is really more like a capsule than shuttle, but capsules are engineered to be dynamically stable and Starship - like shuttle - requires computer assistance to stay in control.
Nobody had ever built a reentry vehicle like this.
Starship uses heat shield tiles like shuttle, but they are much bigger than shuttles and mechanically attached unlike the glue that shuttle used.
And they have a harder job; the starship tiles are mounted directly on the skin of the propellant tank and that gets extremely cold due to the cryogenic propellants inside, so the Starship heat shield tiles have to deal with a much wider temperature range.
One of the big problems with shuttle is that it was too heavy.
The orbiter was about 85 tons empty and it carried the 27 ton external tank nearly to orbit as well. The heaviest payload it ever carried was the 23 ton Chandra Telescope, which means the total mass to orbit was 135 tons and only 17% of that was payload, and that was with throwing away the very large external tank.
Starship aims to reuse the whole second stage and have a much higher payload percentage, but that is hard to do. SpaceX has taken an interesting approach to optimize the heat shield mass.
They have a mathematical model that tells them how much heat shielding they need to have for the vehicle to barely survive.
If we were building a bridge or a building, we could take what our model tells us and multiply that by a factor of 2 or more, and that would give us a design that is slightly more expensive because of extra materials but will still work.
You can't do that in rockets. If you use a factor of 2 in your design, your rocket will be perfectly safe because it will never reach orbit, but that tends to upset your customers, so you need a smaller safety factor. Rockets that carry humans are generally designed with a 1.4 safety factor, and uncrewed rockets might be down at 1.2.
Starship is designed to have crew, so we'd expect it to be designed with a 1.4 safety factor.
The problem is that your models are never perfect, so your safety factor is only a guess. It might be 1.6, which means that you are wasting mass on more heat shielding than you need and have reduced payload. Or it might be 1.2, in which case you do not have the safety factor you need.
Neither of these are good places to be, so SpaceX makes a different choice - they set the safety factor pretty close to what the model predicts but maybe just a little less, closer to the flamey side of things. They do this to validate that they know the point where there are problems - they are essential testing the failure scenario.
That's where they started in flight 4, and we saw that their model was pretty good - their design was *barely* able to survive though it wouldn't be reusable. That was pretty much the perfect result. And it was no coincidence that they had a camera watching the fin that failed; the camera was added to give the engineering team direct data on the part that they expected to fail.
SpaceX gathered a lot of data and then deployed better tiles in flight 5, and that worked better, but still not quite good enough. The real fix will come in the future when their plan is to move the fins up a bit so that the hinge area has less hot plasma hitting it.
Taking this approach is not a new thing for SpaceX - Falcon 9 landing was also developed with designs that they thought would barely work. It is an uncommon approach in the rocket world as flying prototypes is expensive, but it's a great way to understand the true safety factors of your design.
I'd classify the development effort for starship as epic.
All these brand new things being done with a giant reentry vehicle makes it super hard.
And finally, I'll add in the Margins and Expertise measures, which I think are the same as before.
And that's the scorecard comparing the difficulty of starship and falcon 9 reuse.
Which is more impressive?
Is it Falcon 9's long road to reusability and the quick establishment of routine reuse?
Or is it starship's demonstrated capability in Flight 4 and Flight 5?
I love all of my SpaceX vehicles equally, but if you have an opinion one way or the other, head to the comments and tell me which one is more impressive.
And then you can write a comment complaining that I'm shamelessly trying to generate a ton of comments for this video.
I couldn't find a starship or falcon 9 plush toy, but Puffkins created Swoop the Falcon and Swoop flew under the stage name "Aurora" on Crew 9, in the role of "zero gravity indicator".
Swoop is no longer available from Puffkins but ebay and other sites have an adequate supply, so maybe you could send one to somebody.
Description - links:
http://epizodsspace.airbase.ru/bibl/inostr-yazyki/tsiolkovskii/tsiolkovskii-nhedy-t2-1954.pdf
If you're interested in rockets and space, you may have encountered the rocket equation.
It looks complicated and it *is* rocket science, so it must be hard to understand....
But it's actually fairly simple, and you probably understand some of the basics already.
We'll need to take a bit of a side journey first...
Rocket engines are reaction engines, which means they work by throwing mass out of the rocket nozzle, causing an opposite force pushing the rocket forward. We call that force "thrust".
It's pretty simple to define - it's equal to the amount of mass we are throwing out the engine every second multiplied by how quickly we are throwing that mass.
For example, we might be throwing 1 kilogram of mass out every second, and we might be throwing it out with a velocity of 1 meter per second. That gives us one kilogram meter per second squared.
That's a mouthful, so we write it as 1 newton. This is why thrust is measured in newtons, though you can also measure it in pounds of force if you prefer.
Since thrust depends on both the mass of the exhaust and how fast it is moving, we have a choice.
If we toss out 100 kilograms per second at a velocity of 1 meter per second, our thrust is 100 newtons
If we toss out 1 kilogram per second at a velocity of 100 meters per second, our thrust is also 100 newtons
The difference is that the first example burns 100 kilograms of propellant each second, and the second one burns 1 kilogram per second, so our propellant burn in the first case is 100 times the second case.
We therefore want the exhaust velocity of our rocket to be as high as possible.
How does this relate to actual rockets? It helps to look at an example.
Let's say we want to figure out how long it takes an rocket to burn through the propellant it carries.
The SpaceX Merlin 1D vacuum engine has a thrust of 981 kN. We know that thrust is equal to the mass per second times the velocity of that mass. In rocket terms, we would call that the mass flow rate times the exhaust velocity.
But all we know is the thrust, so how can we use this equation?
It actually turns out that we do know the exhaust velocity. Or - we know how to figure it out.
To explain I need to switch to a historical example
On the left, we have the rocketdyne f-1 engine used on the Saturn V, and on the right we have the NPO Energomash RD-170 engine used on the Energia.
If you ask the rocketdyne engineers what their exhaust velocity is, they say 8465 feet per second. Ask the NPO Energomash engineers, and they say 3031 meters per second.
We know that we prefer a higher exhaust velocity, but which one of these is better? Our units don't match, and neither group wants to switch to the other one's units. And converting could lead to mistakes.
What we need to do is get rid of the "feet" and "meters" in the measurements. Hmm... we know that the acceleration due to gravity uses those units, with 32.2 feet per second squared in imperial units and 9.81 meters per second squared in metric units.
Take the F-1 exhaust velocity and divide it by gravity, and we get 263 seconds. Do the same operation for the RD-170 and we get 309 seconds.
This is of course the specific impulse of the rocket engine, and since the RD-170 has a higher specific impulse, it has a higher exhaust velocity and is therefore more efficient when it comes how much mass it uses to generate a given amount of thrust. Specific impulse - often written as Isp - has the units of seconds, which we often ignore when writing the values.
If you are wondering why the RD-170 is so much more efficient, it is an advanced staged combustion design and the F-1 is less efficient gas-generator design. Note that the RD-170 is one engine with 4 separate nozzles.
What this means is that if you know the specific impulse of an engine, you can always figure out the exhaust velocity - just multiply it by gravity.
You'll find some discussions about the "seconds" in specific impulse meaning something like this: (read)
Yeah, I don't get it either, and all everybody cares about is exhaust velocity, so just remember that specific impulse is a simpler way to compare engines than exhaust velocity and you'll be fine.
If you want to know more about specific impulse, see my video "what's up with specific impulse?".
We can now return looking at the merlin 1d vacuum, and we have enough information to figure out the mass flow rate. After refactoring the equation to solve for mass flow rate, we plug in the values we know, and come up with 287 kilograms per second.
Before I learned this, I used to think that people who could figure out mass flow rates were doing some advanced work, but it turns out that it's really simple if you know the thrust and specific impulse.
BTW, if you know the mixture ratio for an engine, you can use that to figure out the mass flow rate of each propellant. If you care.
Now that we've talked about thrust, we can move towards the rocket equation.
We're going to look at the merlin 1d vacuum engine in a falcon 9 second stage, which has an empty mass of 3900 kilograms. We'll assume a payload of 10,000 kilograms. We just figured out that the mass flow rate for the engine is 287 kilograms per second.
If we put 1 second worth of fuel in the stage, that gives us a total mass of 14,187 kilograms. We know that acceleration = force divided by mass, so we take the force in newtons and divide it by total mass and we get 69.1 meters per second squared. For the last second of the burn, that's how much acceleration we will get.
Now we can look at the last 5 seconds of the burn. When there is 5 seconds of propellant left, our acceleration is 64 meters per second squared.
This isn't at all surprising - we would naturally expect the rocket to accelerate more slowly as we add more propellant mass.
Here's a chart if we put 20 seconds of fuel in the stage. With 20 seconds of fuel in the stage, we only get 50 meters per second of acceleration.
Each additional second of propellant provides less benefit.
There's enough propellant in the stage to burn for 387 seconds, and if it's full it has over 111,000 kilograms of propellant, and the acceleration is only 7.8 meters per second squared.
Doing this for the whole burn gives us the following graph.
On the right side when the stage is full, it's very heavy and the thrust of the engine can only accelerate it slowly.
On the left side when the stage is nearly empty, it's very light and the thrust can accelerate it much faster.
It's a very simple concept.
If we want to know the total acceleration for this payload, we can add up the accelerations for each second.
That gives us 7460 meters per second. It's known as "delta v", where delta is the term used to mean "change" and v means velocity.
You might also see it written using the Greek letter delta - that's what the cool kids like to do.
It's not really very convenient, however, to add up all those numbers.
If you took calculus, I'm hoping at this point you're thinking "this looks familiar - I remember something about the area under a curve..."
And the answer is - of course - integration.
If you don't know or have forgotten how to do integration, don't worry. Wikipedia is happy to give you the details if you care about them.
The thing to remember is that the following formula is equivalent to adding up all the accelerations for each second because math.
It says that the total delta v is equal to the exhaust velocity multiplied by the natural log of the mass including fuel divided by the mass when all the fuel is gone. That part is often known as the mass ratio.
You will also see the exhaust velocity replaced with the specific impulse multiplied by the gravitational constant.
The rocket equation is really quite simple and the derivation is also simple - it's the sort of problem you might see in a physics class.
The rocket equation is generally credited to Konstantin Eduardovich Tsiolkovsky, from his 1903 work, exploration of the universe with reaction machines.
The derivation was, as I said, fairly straightforward, and had in fact been derived by British Mathematician William Moore in 1810 in his work, "on the motion of rockets both in nonresisting and resisting mediums". Tsiolkovsky gets the credit because he did a lot of other interesting work on rockets, including coming up with the concept of staging.
The equation would be re-derived by the American rocket researcher Robert Goddard in 1912, and finally, re-re-derived by German engineer Herman Oberth in 1920.
Let's return to our study of the Falcon 9 Second Stage...
We found that the full load of propellant gave us a delta v of around 7500 meters per second with a 10,000 kilogram payload.
But say that we need more delta-v. We somehow expand the stage so it carries twice as much propellant. How much delta v do we get out of the second 111,500 kilograms?
I'm hoping you aren't saying double. It is in fact only 1740 meters per second, for a total of 9240 meters per second.
If you keep going farther with this exercise, the third increment of 111,500 kilograms of propellant only nets 920 meters per second of additional delta v.
Also note that if you double or triple the propellant, you increase the size of your tanks and that reduces your mass ratio, so these numbers are optimistic.
The diminishing returns you get as you add more propellant is what people are referring to when they talk about "the tyranny of the rocket equation". It's really hard to push delta v numbers for a given stage higher.
Since we can't get more delta v by adding more propellant to our stage, we need a different solution...
We take our stage that can generate 7500 meters per second of delta v, and it becomes payload for our first stage, all 125,400 kilograms of it.
But that first stage only needs to generate about 1900 meters per second of delta v, and that's fairly easy even with the high payload mass. Easy enough that the first stage has enough delta v left over to land so that it can be reused.
We can compare this to a different rocket, the Atlas V with a centaur upper stage using the RL-10 engine.
The RL-10 burns hydrogen and oxygen and has a specific impulse of 465, which is great.
The final mass is the mass of the centaur stage - 2247 kilograms - plus the 10,000 kilogram payload
The initial mass is the final mass plus the amount of propellant the stage carries, which is 20,830 kilograms.
We have what we need to do the rocket equation. The delta v is 465 * 9.81 * the natural log of 33,077 kilograms divided by 12,247 kilograms). Do the calculation, and we get 4532 meters per second. That's only about 60% of the delta v that the Falcon 9 second stage produces? What is going on?
If we look at the Falcon 9 second stage numbers, it becomes clearer...
The Merlin vacuum specific impulse is only 348, or about 25% less than the 465 the RL-10 gets. The two stages are about the same dry weight, but the Merlin is burning RP-1, which is much, much denser than the hydrogen used by the RL-10 and the Falcon 9 second stage has slightly bigger tanks than the Centaur, so it packs in 111,500 kilograms of propellant versus the 20,830 kilograms the centaur has.
That gives the Falcon 9 second stage a mass ratio of 8.6 and the natural log of that is 2.1. The Centaur has a mass ratio of only 2.7, and the natural log of that is 1.0. The 25% specific impulse advantage of the RL-10 is overshadowed by the 210% advantage in mass ratio of the Falcon 9.
The same factor that gives you high specific impulse - light exhaust particles - means that you can fit less propellant in a given tank and therefore end up with a worse mass ratio.
Here's a graph that better illustrates the differences...
You can choose the delta v that you need and the specific impulse of your engine, and it will tell you what mass ratio you need.
We can add in the two examples, with the blue circle on the right showing the Falcon 9 second stage and the green circle on the left showing the Centaur. The Centaur has the more efficient engine, but the mass ratio it achieves is significantly inferior to the Falcon 9 and it only achieves 60% of the delta v of the Falcon 9 second stage.
It's important to note that part of this is an architectural choice - the Falcon 9 is designed to stage low so that the first stage can be reused and the Atlas V was designed to use the existing Centaur second stage and therefore needed a big beefy booster and solid rockets to give enough delta v to the centaur.
Here's another interesting graph, comparing the delta v that both stages can achieve with a given payload.
With a 1000 kilogram payload, the Falcon 9 second stage is only about 18% higher in delta v for a given payload, but as the payload mass goes up the centaur falls off quickly. Once we get up to 20,000 kilograms of payload, the Falcon 9 second stage is producing twice the delta v of the Centaur.
The centaur simply cannot overcome the low mass ratio despite having a specific impulse that is 30% better.
If you take one thing away from this video, let it be this.
If you are talking about specific impulse and you don't address the impact of the mass ratio of the whole vehicle, you simply do not know what you are talking about.
I see this all the time with discussions of nuclear thermal engines. Here's one that I liked:
"if the same mass fraction that was used for the chemical rocket engine case were used for the nuclear case, the travel time to Mars would be only 124 day".
It's a true statement, but tremendously misleading. Here are the numbers that we got from the centaur example. A comparable nuclear engine has a mass 2950 kilograms heavier than the RL-10, so our empty mass goes up to 15,196 kilograms. Since we fill the tanks only with hydrogen, the propellant mass goes way down, from 20,800 kilograms to 4,100 kilograms. The mass ratio goes from 2.7 down to 1.27.
The specific impulse is an impressive 900, but if you run the numbers, you get 2108 meters per second.
You need to consider the whole stage + payload when you look at specific impulse.
Here's another application of the rocket equation.
Let's look at an asteroid rendezvous mission. The internet says that the asteroid Davida is worth about $27 quintillion US dollars.
If you want to learn why that's not even close to true, see my video "asteroid mining buzzkill". But we'll pretend we want to do this mission and understand what sort of spacecraft we need to design.
511 Davida requires 6.17 km/s of delta v to get to, assuming we depart in July of 2035. Timing is very important when we are trying to journey beyond the moon; that is why missions to Mars launch every 26 months or so, when the delta v cost is the lowest.
It takes 6.17 kilometers per second to get from LEO to Davida. This in the same ballpark as what it takes to get a probe to Jupiter.
The usual way to do this is to use your launch vehicle to toss you wherever you want to go, but let's assume you want to start in LEO, because reasons.
We can refactor the rocket equation so that we can plug in the delta v we need and get back the mass ratio. E in the equation is the natural number e, which is about 2.7.
If we choose hypergolic propellants, they have a specific impulse of about 311. Plug that into the equation with the delta v we need, and the answer that we get is 7.6
That means we spend 12% of our mass on the spacecraft - the structure, the engines, the electronics, and whatever payload we are carrying - and 88% of our mass on propellant.
This illustrates why the usual method is to use the launch vehicle to provide the velocity to get away from earth.
And that's all I have to say about the rocket equation.
Learn it. Know it. Live it.
If you enjoyed this video, please send me a model of the F-1 engine made out of peanut M&M's
The first mention of the three booster "Falcon 9 Heavy" rocket comes way back in in October of 2005. Those of you who know SpaceX history might note that this was 5 months before the first successful flight of Falcon 1, so it was a long time ago.
It was over 12 years later that Falcon Heavy would first fly, pretty much tying NASA's SLS rocket for the winner of "longest rocket development time in the US". They are both soundly beaten by the Soviet Angara, conceived in 1992 and not flying until 2014, a 22 year long development.
The long development time of Falcon Heavy seems out of place given SpaceX's "move fast" philosophy, and it's sometimes been used by detractors as an example of a SpaceX failure.
The reality is actually quite a bit more interesting. We'll need to go back to 2010...
The year is 2010 and SpaceX has a problem. They have a contract to develop Falcon 9 for the NASA commercial resupply program, but that was only one market and they need more launches to make the Falcon 9 a success.
SpaceX wants to enter the lucrative market of geosynchronous satellite launch, but the payload capacity of Falcon 9 V1.0 for those launches to geosynchronous transfer orbit - or GTO - was a disappointing 3.4 tons. If we look at the communications satellite payloads that were launched by Ariane 5 in 2010, we see that only 2 of them could fly on Falcon 9, and that would leave no margin for first stage reuse. And those satellites were getting bigger over time.
The fix was clear. You're gonna need a bigger rocket.
When McDonnell Douglass was designing the Delta IV, they had the same problem, and they came up with the Delta IV Heavy, which roughly tripled the GTO payload of the rocket. That's why SpaceX announced the rocket they called Falcon 9 Heavy; it gave a much higher GTO payload, one which would cover all the current satellite sizes. It's a cool idea but is a significant engineering challenge.
Falcon 9 Heavy was actually plan B for a bigger rocket, and we'll need to spend some time looking at how plan A turned out...
Falcon 9 Version 1.0 used the Merlin 1C engine, a version that first flew on Falcon 1. SpaceX used it because it was the quickest way to get Falcon 9 up and flying and they needed to start flying cargo flights as soon as possible. The Merlin 1C was moderate in terms of efficiency but it was nice and light, with a thrust to weight ratio about 90, more than pretty much any engine burning RP-1 kerosene and liquid oxygen. It generated 420 kilonewtons of thrust or about 43 tons.
But the Merlin team wasn't finished, and they soon coughed up a new version, the Merlin 1D. It pushed the thrust up to 620 kilonewtons, or 63 tons, pretty much a 50% increase. More thrust means you can lift a bigger rocket, and the Falcon 9 version 1.1 was 40% taller and 50% heavier. That upsize pushed the GTO payload up to 4.9 tons, an increase that actually made it reasonable for SpaceX to start flying some of the smaller geosynchronous communications satellites and to start working seriously on reuse.
This new rocket covered more of the target market and made Falcon Heavy less critical from a business perspective. It was a big investment in time and engineering resources to build Falcon Heavy.
The Merlin team still wasn't finished. Their next version was naturally called Merlin 1E.
Just kidding. The SpaceX team seems to delight in breaking naming conventions on a whim, so this version was Merlin 1D full thrust, raising the takeoff thrust by another 17%. But now they had a bit of a problem. Falcon 9 couldn't get any wider and still travel by truck and it was already very tall and thin, so they didn't want to make it taller.
So they adopted a trick that had been researched in the past known as densified propellant or sub chilling. If you cool the propellants down below their usual temperatures, they become denser and you can therefore fit more into a given volume, about 8% more in this case.
The rocket is sometimes known as Falcon 9 V1.2 but is more commonly known as "full thrust". It can carry 3.9 tons to geosynchronous transfer orbit with first stage reuse and a full 6.8 tons in expendable mode. That covers *almost* all the market.
Which brings us to the final version of the Merlin, producing 845 kilonewtons or 87 tons, a little over double the thrust of the Merlin 1C, which is honestly an astounding increase in performance.
At this point SpaceX reveals that they have a different naming convention for their internal versions using "block" numbers to tell the revisions apart. This naming approach has a long history in aerospace dating back to Apollo and earlier.
This is the engine for block 5, so it's typically known as the Merlin 1D Block 5, though I have a fair bit affection for the "fuller thrust" label that is sometimes applied. The rocket is now the Falcon 9 block 5 that we are so familiar with. This final version pushes the reusable payload up to 5.5 tons and the expendable payload to 8.3 tons when flying to geosynchronous transfer orbit. That covers the vast majority of geosynchronous satellites.
While Plan A has been running - Falcon Heavy has been on the back burner, as it makes little sense doing a lot of engineering work when the rocket keeps changing underneath you. Plan A was hugely successful, getting rid of the main original justification for Falcon Heavy, but there are now some interesting other markets...
The evolved expendable launch vehicle program was created in 1994 to provide the US Department of Defense assured access to space for bigger payloads than could be launched by the Atlas II and Delta II vehicles. They also hoped to reduce the price of launches.
In 1988, the air force awarded contracts to Lockheed Martin for the Atlas V and McDonnell Douglas for the Delta IV. It turned into a very lucrative contract for both companies, and when they merged in 2006 to form United Launch Alliance, they had a monopoly on department of defense launches.
See "spies, lies, and enterprise - the strange story of ULA".
This situation led to ULA getting about $140 million for either an Atlas V or Delta IV launch and $430 million for a delta IV heavy. It was very profitable for ULA, and SpaceX wanted in on that business.
The first step was to get certified so they could be allowed to bid on those launches.
SpaceX felt that the Air Force was dragging its feet in certifying Falcon 9 and that the Air Force's plan to do a bulk buy of 36 rockets from ULA did not allow SpaceX to compete for the launches that Falcon 9 was capable of, and was therefore illegal under procurement laws.
SpaceX won an intermediate victory and the presiding judge sent the matter to mediation, and an agreement was reached in January 2015.
The big thing that SpaceX got out of this was a certification to launch Air Force payloads on Falcon 9, and a total of 7 launches in the phase 1 batch.
For the next set of launches - known as NSSL Phase 2 - the Air Force would be doing open bids and needed to come up with requirements for the launch companies. They decided that to be one of the two companies that did launches, you had to be able to launch all possible payloads.
Most of these can be covered by Falcon 9, either in reusable or expendable mode, but there's one glaring problem. The GEO 2 orbit requires putting 6,600 kilograms of payload all the way into geosynchronous orbit, and the Falcon 9 can only do about 3,000 kilograms. If you can't do that, you can't bid on this contract and therefore can't fly any of these payloads.
You can view this requirement in two ways...
The first is that it's designed to keep as much of the existing monopoly as possible up and running and the money going to ULA.
The second is that the Air Force is tired of paying $430 million for these launches and really wants SpaceX to build something so they can save that money.
I think it's probably a little bit of both.
The solution is obviously Falcon Heavy, and the air force was quick to certify Falcon Heavy after its first successful flight, thereby allowing SpaceX to get 40% of the launches. The remaining 60% of course went to ULA.
The other market was NASA, who were overjoyed to have access to such a big and cheap rocket, with a payload capacity of 63.8 tons to low earth orbit and 26.7 tons to geosynchronous transfer orbit. Those high payloads meant that Falcon Heavy had enough oomph to send the relatively light NASA probes on their high energy missions.
The Psyche mission to an asteroid was 2608 kg when launched in 2023, and NASA paid $133 for the launch.
NASA has 4 other missions scheduled to fly on Falcon Heavy.
The 5000 kg GOES-U weather satellite will fly to geostationary orbit in 2024 at a launch cost of $152 million.
Europa clipper will take a very long journey to arrive at Jupiter's moon Europa, and launching the 6065 kg satellite in late 2024 will cost NASA $178 million. Europa Clipper was originally designed to fly on NASA's SLS rocket which would have gotten the probe there faster, but lack of available SLS rockets and their $2 billion per launch price caused NASA to move to Falcon Heavy. This is a case where Falcon Heavy is a little small for the job.
Falcon Heavy will launch the 4166 kilogram Nancy Grace Roman Telescope to a halo orbit similar to the one the James Webb space telescope currently occupies in 2027, and NASA will pay $255 million for that mission.
And finally, Falcon Heavy will launch the first two modules of the lunar gateway space station into a weird elliptical orbit for $331 million. It will mass at least 14,000 kilograms, and it will put itself into the target Near Rectilinear Halo Orbit around the moon. This launch is currently scheduled for late 2025.
These NASA launches send just over $1 billion to SpaceX for 5 flights, though it's important to note that NASA requires extra services for their launches and therefore SpaceX does more work for those mission.
The Department of Defense and NASA both like Falcon Heavy, so why doesn't SpaceX use it Starlink? It would seem to be the perfect rocket for that mission.
It turns out that there are a few reasons why that isn't true...
If we look on SpaceX's Falcon Heavy page, we see this table, and it lists the Falcon Heavy payload to low earth orbit or geosynchronous transfer orbit. Those are big numbers.
Unfortunately, they aren't real, or at least they haven't been demonstrated...
This is a Falcon 9 payload adapter, taken from the Falcon 9 User's Guide. The payload adapter is the interface between the rocket and the payload.
One of the things the user guide tells you is the center of gravity limitations of the adapter, and your payload will need to comply with these limitations or you might break something. Looking over at the right side of the graph, we see that there is a hard limit at 18,800 kilograms, which means the "standard" Falcon 9 cannot carry payloads greater than that amount.
You can ask SpaceX to do special work to carry more, but if you want to get to the GTO payload of 27 tons you are likely going to need strengthening elsewhere in the rocket and if you want to lift 64 tons to low earth orbit, you are definitely going to need a lot of extra work. So view both of those payloads as theoretical.
This doesn't matter for the department of defense or NASA because their goal is to take light payloads to hard-to-reach orbits rather than big payloads to easier orbits, though there are reports that NASA's gateway modules have hit this limit and that has caused some issues.
But this means that the standard Falcon Heavy couldn't carry many more Starlink satellites than Falcon 9 reusable can.
Even if Falcon 9 could carry more satellites, the current crop of satellites fill the Falcon 9 fairing to capacity - they are designed to fit together so they can fill the fairing. You could only fit a few more satellites in that fairing if you tried, and even the taller fairing under development for both NASA and the department of defense payloads could not carry enough satellites.
A third issue crops up related to reuse. SpaceX only has two autonomous drone ships available on the east coast, and that means they could only recover two of the cores. SpaceX has recently mentioned they are building a third drone ship for the east coast, but it's not currently available.
All these issues explain why Falcon Heavy does not fly starlink missions.
And that's the story of Falcon Heavy. Originally planned to make up for lack of performance on Falcon 9, then repurposed to launch difficult missions for the US government, and currently the worlds largest operational commercial rocket.
If you enjoyed this video, please send me this Gyrfalcon.
The Gyrfalcon is the world's heaviest true Falcon, though at 2 kilograms it is outmassed by the Falcon Heavy by a factor of 1 to 700,000 .
https://medium.com/@ToryBrunoULA/the-secrets-of-rocket-design-revealed-e2c7fc89694c
This is the third video in the series looking at ULA PR, looking at an article authored by ULA CEO Tory Bruno titled "The secrets of rocket design revealed".
We're going to be doing something a little different, which is why this talk is titled, "the real secrets of rocket design revealed", because I don't think the picture that he paints is accurate.
But don't worry, there are plenty of visual aids in this one and we'll be looking at those as well.
The article starts out with this paragraph: (read)
I mostly agree with it, though I would talk about markets rather than missions.
What I don't agree with is the assertion that the design of the rocket flows directly from the mission the rocket is intended to do and that efficiency is of high importance.
You really should go read the whole paper - it's linked in the description - but since you've been spending too much time on insta and not enough on homework, I'll summarize what I think the main argument is.
(read)
Before we dig into the paper, we need to do a little work.
We need to understand where rockets come from, and we're going to explore this by looking the birth of two rockets.
We'll start with SpaceX. They were working on the Falcon 1 when a great opportunity fell into their lap.
The shuttle was retiring and NASA needed a way to carry cargo to the international space station, so they were funding the commercial orbital transportation services program to help companies develop that capability, and after that was done they would hire companies to actually carry the cargo.
SpaceX's reaction was "we have to win a contract". The development contract would be for 100s of millions of dollars, and the ongoing cargo flights would be just the thing for a small rocket company with growth aspirations.
SpaceX would need a new rocket to get this contract, a much bigger rocket. Whatever the design, it has to be a rocket that not only can win the NASA contract, it needs to be a design that will win the contract. So the design comes down to what NASA wants.
NASA wants a high chance of success, and by success I mean "the rocket and capsule get designed, built, tested, and actually fly missions to the space station". This is not a technology development contract, this is to achieve a useful result.
NASA wants it to be timely - they need it ASAP
NASA wants it to be affordable. They aren't terribly price-sensitive because their only other option is to fly cargo with the Russians and that looks bad, but if it can be cheaper than resupply using the space shuttle, that would make them happy.
To sum it up, NASA wants boring. They are looking for the Ford Transit van of rockets, a utilitarian vehicle for carrying stuff around.
Now we get into what are known in engineering as "trades", or tradeoff studies. You look at a problem, come up with a set of options, and compare them to each other.
Rocket design always starts with engines, and typically with the second stage engine.
The Kestrel engine used on the Falcon 1 is an option. It's a pressure fed engine with no pumps, burns liquid oxygen and RP-1 kerosene, and has a specific impulse - or mass efficiency - of 317 and a thrust of 3 tons.
The RL-10 is made by aerojet rocketdyne and has been flying since the 1960s. It's a very efficient expander cycle engine that burns liquid oxygen and liquid hydrogen. It has a specific impulse of 450 and a thrust of 11 tons
And then there's the engine to be named later, an upper stage engine that SpaceX would develop themselves.
The RL-10 looks like an excellent choice - it's well understood, very reliable, and very efficient. It's also very expensive, at something like $10 million per engine, and SpaceX would be constrained by the capability of Aerojet rocketdyne to provide engines. SpaceX also has not flown a stage using liquid hydrogen as a fuel, and that would require more research and development. And the thrust is marginal - you need a big first stage if you choose the RL-10.
So the RL-10 isn't a good choice. The cost is too much, and buying engines isn't the SpaceX way.
Kestrel isn't very exciting technically; pressure fed designs mean thicker and heavier tanks and that's especially a problem for second stages. The specific impulse is low, and the thrust isn't close to sufficient. Kestrel doesn't work.
Which means SpaceX need a new engine for this rocket, and that's troubling because engine design takes a long time and is expensive and that's not going to make NASA excited. They need an engine that can be developed quickly
SpaceX decides to take the Merlin 1C first stage engine from the Falcon 1 and modify it to be the second stage engine for their new rocket. Merlin is a gas generator design that is less efficient than the other alternatives, burns RP-1 kerosene, and has a low specific impulse, though that is mostly because it doesn't have a large vacuum nozzle on it. It has a thrust of 49 tons
This is a really weird choice from a rocket design perspective. It loses efficiency both because it's a gas generator design and because it burns RP-1 kerosene, and it's really big for a second stage engine which means it's kindof heavy.
It's not a very *efficient* choice from that perspective.
But it's a great choice for SpaceX - they already have the Merlin 1C functioning and it will be *relatively* straightforward to modify it into a second stage engine, and it will also be quicker to do that than to try to develop a new engine from scratch. NASA will be happy and that is the main requirement.
It ends up looking like this without the large vacuum nozzle attached. The big nozzle gets the specific impulse up to 342, and the thrust drops a little to 42 tons, but it's still a beast.
For the first stage, there is a similar analysis.
There's the RS-68 from aeroject rocketdyne. It's kindof like the Merlin 1C except that it runs on liquid hydrogen rather than kerosene and that gives it a specific impulse of 365 at sea level. It's a big engine, putting out 306 tons of thrust.
A second option is a new SpaceX engine, much bigger than the Merlin.
And a third option is a cluster of the existing Merlin 1C.
The RS-68 isn't very exciting. Because it uses hydrogen, you need big tanks and those tanks are heavy and expensive to build. And it has the same issue as the RL-10 - it's expensive, probably at about $15 million an engine.
A new engine has the same issues we ran into for the second stage - you don't have the time and money to do it and you're already doing the vacuum version of the Merlin.
So a cluster looks interesting, and it was used quite successfully on both the first and second stages of the Saturn V moon rocket. The downside is that to get the thrust that is needed, SpaceX will need to use 9 of those engines. Nobody has every done 9 clustered engines and that means they will need to be able to build those engines quite cheaply.
Given the constraints of their situation, the cluster of 9 Merlin 1C engines is the only reasonable choice.
In reality, trades are more complicated than this, because trades will look at the whole vehicle together, not the stages by themselves, and changes to a detail in one stage will force changes in the other stage.
To oversimplify, SpaceX chose a Beefy second stage and a "stage early" philosophy because it made sense to build a high-thrust second stage engine. On a launch to geosynchronous transfer orbit, the second stage has a thrust/mass ratio of 0.88, which is quite high for a second stage, but it needs this high ratio because it stages early.
The other thing driving the trades for Falcon 9 was SpaceX's desire to reuse the first stage. They didn't know how to do it yet, but they did know that the slower and lower the rocket staged, the easier reuse would be, so that also argued towards a beefy second stage architecture.
All of their constraints resulted in the architecture of the Falcon 9. You can say that it is optimized towards low earth orbit missions, but it's really just a rocket designed to accomplish the NASA mission simply, easily, and in as boring of a manner as possible.
Or, to put it another way, you design your rocket to serve the customers that you care about within the business constraints that you have.
And that's why Falcon 9 is the way that it is.
Our second story is about United Launch Alliance's Vulcan, and for that we need a bit more context.
ULA had been gifted a virtual monopoly on US Government launches by the Department of defense when they brokered the creation of ULA from the rocket division of Lockheed Martin and Boeing. If you want that story, go watch spies, allies, and enterprise - the strange story of ULA.
With that monopoly, ULA launched payloads using three different rockets. The Atlas V, the Delta IV M+, and the delta IV heavy.
The Atlas V and Delta IV M+ were roughly the same price - around $160 million - but the delta IV Heavy was a mind-blowing $420 million per launch. The government paid those prices because they had no alternative.
That worked for a number of years and ULA made large profits. Then they ran into two problems.
The Atlas V first stage was powered by the highly efficient Russian RD-180 engine. This was initially viewed as a good thing, as it kept Russian rocket folks employed building engines rather than working for less friendly countries, but US relations with Russian deteriorated, and with the annexation of Crimea in 2014 it became clear that relying on the RD-180 was not a long-term solution - congress would simply not allow it.
So the Atlas V was out.
ULA could have switched all their launches over to the Delta IV, but the M+ variant was less capable than the Atlas V, which would have required more Delta IV Heavy missions and higher costs. Congress was already asking about the high cost of those missions, and when Falcon 9 showed up with much lower prices than the ULA launchers, just flying delta IV was not a winning strategy.
ULA needed a new rocket...
ULA had a specific set of goals...
First, it had to be able to fly all the government payloads that Atlas V, Delta IV Medium, and Delta IV Heavy could fly. Those missions are the bread and butter for ULA, and you have to be able to fly them all to get the contract.
Second, it needs to be done reasonable quickly.
Third, it needs to be reasonably price competitive with SpaceX for space force launches.
And finally, it needs to make the space force happy.
Like SpaceX and the Falcon 9, this is all about keeping a specific customer happy.
Once again, we'll start with the second stage.
One of the significant disadvantages that ULA has is that they do not build their own engines, so their engine choices are limited to what they can buy from other companies, companies that are in the business of making money selling expensive engines.
Aerojet rocketdyne bid the RL10, which we already talked about. A great engine but with low thrust and very expensive.
Traditionally, everybody flew the RL10 because there was no other choice. But Blue Origin decided to bid their BE-3U engine. It's also a hydrogen oxygen engine, but a much more powerful one, with a projected thrust of 70 tons. We don't know what the engine would cost.
That "projected" status is important; at the time, Blue had the BE-3 engine that they used on New Shepherd, but the BE-3U is really a very different engine, which made the timing iffy, and ULA needed an engine that could support the advanced scenarios required for government launches.
ULA decided to stick with the engine they knew and understood, the RL10. They would need two for the bigger and more powerful upper stage they needed to fly the difficult missions that Delta IV Heavy flew, and even then their second stage would be quite underpowered.
For the first stage, there were the same bidders, with Blue Origin pitching the BE-4 engine that would power their New Glenn rocket and aerojet rocketdyne pitching the AR1, an upscaled version of the RD-180 engine that flew on the Atlas V. In this case, the BE-4 was under development already and the AR1 was just a proposal. These are the only US engines that would work, and it would take two of each of them to power the first stage.
In this case, ULA chose the BE-4 as the first stage engine.
The design was settled, with two RL10 in the Centaur V second stage and two BE-4 in the first stage of the Vulcan rocket.
But that wasn't quite enough to handle the tough government missions they needed to handle, so the Vulcan centaur can add two, four, or six GEM 63XL solid rocket boosters to provide more liftoff thrust and higher performance.
The low thrust of the RL-10 engines meant that Vulcan would have a weak second stage and therefore it would need a beefy first stage and it would need to stage relatively late. The thrust to mass ratio launching a GTO payload with 6 solid boosters is only 0.29, much lower than what we would see on the Falcon 9 second stage.
Note that this wasn't a choice driven by rocket architectural concerns but one driven by engine availability, and also ULA's experience and comfort using this approach.
And that's two very different rockets with their differences driven very much by rocket engine availability and company history. Neither are what we would call "clean sheet" designs where you get to design the rocket you've always wanted.
Now we can return to the ULA content.
The first graphic we come to is this one, which is entitled "how hard is going to space?" It's a modification of...
This graphic, from the 2022 article entitled "how hard is going to space?
The arrow heading up to the right says "Increasing difficulty and complexity". I think this is a fairly good graphic, though it breaks rule 25: always start your bar charts at zero, and that's very annoying on this chart where you have lots of cases where it looks like you have zero values.
It's not the worst graph I've seen, however, and it expresses the concept pretty well - missions that require a higher energy trajectory or specific capabilities from the second stage add difficulty and complexity over missions that don't require those things.
This new version has the same information as the last one, but it's grouping the launches into two categories, low energy and high energy. The text says (read)
This is a particularly annoying thing to do.
The problem is that the term "energy" already has a use in rocketry, with high energy orbits ones being defined as ones that require more energy to get into, as you can see from this example from ULA's very own "cubesat launch to high energy orbits" document published in April of 2019.
This is about brand differentiation; ULA is redefining "high energy" to mean "difficult things" and - more importantly - things that ULA is uniquely qualified to do. And it's true that the launches on the right require a very capable launch system, but redefining a term that is already in use is not a helpful thing to do.
One of the main points of the paper is that the requirements of those high energy launches are best served by a specific rocket architecture, the one that Vulcan Centaur uses.
This is where the high energy redefinition gets messy. The slide is supposedly about rocket architecture but has a lot of items that are capabilities.
For example, payload fairings are a capability, so they don't belong on an architecture slide. We can get rid of them.
There are other things listed that are also capabilities - burns, thermal management, etc. They may be great to have in a rocket, but we are talking about architecture...
The rocket on the right side is obviously Vulcan Centaur, but it's not clear what the left rocket is. This is pretty obviously a strawman argument, where you argue about a hypothetical rather than your actual opponent, but we can work with it anyway.
This rocket stages early and can do fly home reuse. Note that when ULA says "fly home" it means either a drone ship landing or landing back at the launch site.
Here are the candidates I came up with - it could be the SpaceX Falcons, SpaceX starship, Blue Origin's New Glenn, or Rocket Lab's Neutron.
All of these are by companies that make their own engines and therefore they can choose whatever architecture makes sense for them. They are likely to be fairly close in architecture because propulsive landing means you want to stage fairly low and early.
I'm going to eliminate new glenn and neutron because they aren't yet flying and we therefore don't have much information about them.
And I'm going to eliminate starship because it is not in the same class as Vulcan and because - without refueling - it can't do the commercial GTO and interplanetary missions that the slide says the commercial optimized rocket can.
Which leaves us with Falcon 9 and Falcon Heavy, and wouldn't it have been easier to just draw one of those?
Which means we are really comparing the Falcon 9 and Falcon Heavy to Vulcan Centaur.
At which point the comparison falls apart.
As we saw at the beginning, Falcon 9 and Vulcan have different architectures because of the constraints that they were designed under; it's not at all about the missions that they can do.
Despite having a "commercial optimized" design, Falcon 9 and Falcon Heavy have the contract to fly all of the us government missions that Vulcan Centaur can and has in fact flown 12 National security space launch missions while Vulcan has yet to be certified and therefore has flown none of those government missions.
I'm not saying that Vulcan Centaur is a bad rocket - it appears to be a good rocket that is capable of doing everything ULA needs it to do at a cheaper price than Atlas V or Delta IV, and more good rockets is better.
It's just more than a bit disingenuous to do a comparison slide against a hypothetical rocket that is designed for a different market. You can build a rocket with a weak second stage and put a big hefty booster under it or you can build a stronger second stage and put a weaker booster under it. Both can get you to where you want to go, even for difficult "high energy" missions.
And now we come to my favorite graphic in the paper, the graph that launched 1000 tweets.
ULA wants to differentiate the Vulcan Centaur approach as better because of the high energy architecture, and they have decided to highlight the difference in how they stage.
The graph is pretty good; it's certainly true that Vulcan stages much higher than Falcon 9, and it also stages at a higher velocity, which is something I would have worked into the graph as well. We already know that this is mostly because the Falcon 9 has a big beefy second stage and wants to reuse the first stage, and the Vulcan has a wimpy second stage.
Then somebody comes along and slaps that title on it and you have a PR issue on your hands.
The immediate question is "are they actually asserting that the first stage of Vulcan Centaur will fly all the way to low earth orbit? That's not what the graph says, but it is what the title says.
If we look elsewhere in the text, we find this. I've added the highlights.
(read)
I think there's a good argument that neither of these are true - I'll show you why in a minute - but what primarily and nearly mean is open to interpretation.
And it's really obvious that the text of the paper does not say the first stages actually fly into low earth orbit.
We also have this graphic, and if you look at the parts I censored earlier, you see that the big label says "upper stage transits from LEO" but the detail says the second stage is carried nearly to orbit by the first stage.
So I guess there's some ambiguity there, but I'd still argue that "nearly" is the operational part.
BTW, does anybody think that it's weird that this graphic uses both "second stage" and "upper stage"? Yeah, me too.
This slide has come up a number of times and Tory Bruno has been asked about it.
The response is pretty simple to make; you say that the graphic is purely about the staging altitude and that the title was a mistake, and if you look at the rest of the document our position is more clear. And then, if people ask what "primarily" and "nearly" mean, you can have a nuanced discussion about that, but you've mostly defused the situation. That is what a good PR person would tell you to do.
Instead, Bruno tweets this.
(read)
My response on reading it was two-fold. The term LEO is well established to mean low earth orbit and nobody uses it to mean anything else. I looked at 6 space acronym sites and it means low earth orbit. Search for LEO on the nasa technical reports server and you get the same result. So it's pretty clear that Bruno is factually wrong, and very obviously factually wrong.
My second response was, "Oh, Tory, why didn't you do the really obvious thing and just say the title was wrong and it got missed during review?"
He later added this tweet, which only makes things worse.
(read)
You had an out but you decided to add fuel to the fire and very obviously didn't provide a reference to support your assertion.
To those of you who will assert that ULA said their boosters flew to low earth orbit, I'm sorry to say that you are wrong. But that "nearly" is something we can look at.
We can build ourselves a little model
We're going to build ourselves a very simple model using the rocket equation that can give us an estimate of how much work the second stage needs to do, and we can therefore estimate how much work the first stage does and compare that to what it takes to get into low earth orbit.
You can find more details in my video, the care and feeding of the rocket equation.
Based on the efficiency of the engine and the ratio of the mass of the stage full of propellant and the mass of the stage empty of propellant, we can figure out what the stage can do.
ULA has been nice enough to tell us how much propellant the centaur V carries and how much its empty weight is.
We're going to be trying to get a payload all the way to geostationary orbit, because that is the highest energy orbit we've been talking about. We'll be flying Vulcan with two boosters in the VC2S configuration, and ULA tells us that its payload to that orbit is 2.5 tons.
Plug some numbers in - the specific impulse of the RL10 is 450 - and we get a delta v of 9030 meters per second. Which is really great. We can subtract that number from the total required, and we find that the first stage needs to supply about 4130 meters per second of delta v.
This is of course an approximate number - I'm ignoring a number of factors - but it's close enough.
Using that, I've created a chart showing the relative delta v contribution of the first and second stages on a mission.
It takes about 9300 meters per second to get into low earth orbit, so I'll put a line on the graph at that point.
Once again, we're going all the way to geostationary orbit, requiring about 13,000 meters per second of delta v.
Here's the data for Vulcan, along with the payloads they carry. None of the first stages generate more than about 5500 meters per second of delta v, so they clearly don't take the second stage "nearly" to low earth orbit. Note that as we add more payload, the second stage can do less work and therefore the first stage adds solid rocket boosters so it can do more work.
Adding data for Falcon 9 and Falcon Heavy, along with some speculative payloads since SpaceX doesn't publish them.
Given what we know about the architectural choices that were driven by the engine choices, this isn't surprising at all. Falcon 9 has the beefier upper stage and therefore that stage does more work than the centaur upper stage in most cases, though the VC2S Vulcan configuration is pretty close in the split between the two stages as it has the least powerful first stage.
It's also interesting that the 6 booster Vulcan and falcon heavy are pretty close in terms of the ratio of work done between the two stages on this mission.
So does the Vulcan Centaur fly nearly to low earth orbit? Not in my book, no.
Pop quiz time.
Here are the time versus speed graphs for two different rockets launching into low earth orbit, from Declan Murphy's excellent FlightClub.io website.
Vulcan Centaur is absent today, so Atlas V has agreed to be a substitute, showing up with 5 solid rocket boosters in the 551 configuration.
Falcon 9 graciously agreed to be represented.
Here are the graphs. Which one is which?
The Atlas V is on the left; you can see the solid boosters separate at about 90 seconds and you can see centaur take over at about 270 seconds. The low thrust to mass of the centaur is very obvious here by how flat the green line is - the stage can only accelerate slowly even though the booster took it high and fast.
I guess you can make an argument that the staging point is nearly to LEO since it's about 80% of the way there, but it takes centaur 360 seconds to get into orbit.
The Falcon 9 launch is quite distinctive because of the booster path back to land on the drone ship. The staging is at a much slower speed, but the beefy Falcon 9 second stage gets to orbital velocity about 30 seconds quicker than centaur despite having to provide nearly 4 times the delta v.
Falcon 9 gets into orbit nearly 3 minutes earlier, and that's 3 minutes where the stage didn't have to waste energy on gravity losses, which you can argue makes it a more efficient rocket.
The important point is that both work.
We finally come to the last chart.
The text says the following:
(read)
It then lists three architectures:
A typical three core LEO optimized rocket
A high energy rocket
An extreme LEO optimized "super heavy leo lifter"
These are nicely specific descriptions - there is only one 3 core reusable rocket, and that's falcon heavy. High energy is obviously Vulcan, and I'm not sure, but the super heavy leo lifter might be starship.
Let's start looking at what it says for Falcon Heavy and Vulcan.
This dotted pink line shows the performance SpaceX gets out of Falcon 9 with drone ship landing, and this solid one is Falcon 9 in expendable mode. It's pretty clear that the gray line is not Falcon Heavy.
Falcon Heavy operates in a number of modes. In fully expendable mode, SpaceX numbers put it roughly here. I made up the MTO number as there are different MTO orbits and SpaceX doesn't talk about them, but it will be roughly the same as GTO.
There are no reusable numbers from SpaceX. If we assume that the LEO payload drops by 25% with drone ship landing - the same as F9 - then we end up with about 48 tons for landing the two boosters on the drone ships. GTO is likely about the same as Vulcan, and we know that GEO is at least 4 tons but is less than 6.7 tons, so I put it roughly where Vulcan is, but it might be a bit less.
I do want to make one clarification. The payload adapter for Falcon is designed to support up to 18.8 tons of payload, so it cannot carry the full payload for expendable Falcon 9 or any of the Falcon Heavy variants. To do so would require a new payload adapter and possibly a strengthened second stage. SpaceX would be happy to do this work if a customer wanted to pay for it, but there aren't any low earth orbit payloads of that size.
That's all for Falcon Heavy. You can draw your own conclusion about whether ULA correctly showed the data for the Falcons.
Let's move onto starship. I did a whole video on this topic titled "starship - what can we do with it?", so I'm going to keep this section short.
Starship is a very different architecture than Vulcan, and frankly to compare it to a rocket like Vulcan is to miss the point.
But it's on the chart.
The booster is designed to only fly back to the launch site, which means it needs to stage as low and close to the launch site as possible.
The second stage is not only huge but it's designed to be reused, so it has flaps and all the mechanisms to control them, heat shield tiles, and an enclosed payload section so it doesn't throw away a payload fairing every flight. All of those are heavy and they push the empty weight up quite a lot, which means it's not a very good second stage for higher energy orbits.
However, a number of people have done calculations and they agree that starship has a GTO payload of 20-some tons. That is probably a one-way trip. Would it have useful payload if it reserved enough propellant to get out of GTO and back to earth? I don't know the answer to that question.
Here's what ULA had to say about this scenario...
That is clearly not true and the calculations to estimate the GTO payload are straightforward.
Assuming starship refueling works - and you are willing to pay for the flights - the graph is quite a bit different; starship can take as much mass as it can get into LEO to any of these destinations.
And there are also options for non-reusable second stages that would be lighter and therefore would have better performance to higher energy destinations.
I talked about reuse at length in part 2 of this series, and frankly it feels like ULA is talking about it only because they have to...
The quick summary has no surprises; the low energy rocket stages lower and that makes flying the first stage home less challenging - I would say "practical" - where staging high means that it's too challenging to do that. And I think that's a fair summary of what's going on...
But this is missing one of the big drivers - all of the groups working towards propulsive landing build their own engines, and this is an enabler to do propulsive landing as you need a cluster of engines - 5, 7, or 9.
ULA buys their engines which both limits their choices and makes the economics very different, so they are stuck with their current architecture.
ULA makes one point in the reuse section that needs to be addressed.
They say that low energy reusable rockets must often fly expendable on the most challenging missions.
That is true, and it's a feature.
A communications satellite launched with a drone ship landing currently costs $70 million. Well, actually, like a car, it's priced at $69.75 million, but I'm going to call it 70.
If you need less performance than that, you might be able to get by with a return to launch site landing, and that will cost less.
If you need more performance, you will need to expend the first stage, and that will cost more.
You can tune the cost of your rocket based on what you need it to do...
In the same document where ULA lists that as a disadvantage, they talk about a Vulcan feature they call dial a rocket. The text describes it this way.
(read)
Four different first stages. That *kindof* reminds me of the way the Falcon 9 has three different first stages depending on how much propellant it has to save for landing.
What are the real secrets of rocket design?
(read)
And those are the real secrets of rocket design.
Please discuss in the comments.
If you enjoyed this video, please send me this KTEL album from 1979, with a list of hits that just don't stop.
I especially appreciate their audacity in putting Foreigner and Captain and Tennille on the same album.
NASA integrated Program Plan 1970 : https://www.wired.com/2012/04/integrated-program-plan-maximum-rate-traffic-model-1970/
ISS Deorbit analysis summary white paper: https://www.nasa.gov/wp-content/uploads/2024/06/iss-deorbit-analysis-summary.pdf
And with that, the first two modules of the international space station - the unity module from the US and the Zarya module from Russia - were attached, and the space station was born
The international space station has been a fixture at NASA since that first assembly flight in 1998, but with some of the elements in their mid-20s, at some point it will no longer be safe for Astronauts and Cosmonauts to occupy.
This is the story of the space station - why it was built and what will happen when it's no longer safe.
This is the rise and fall of the ISS.
Back the late 1960s, NASA had really big plans.
Plans for bases on the moon and mars, space stations around both of them, space tugs, and shuttles, both conventional and nuclear.
But Congress cancelled the moon program, and Apollo 17 was the last flight. The US had won the race to the moon and that was the source of most of the backing for the program, and it's clear that the government had other issues to focus on, including the war in Vietnam.
There a story from this time that president Nixon was presented with three options for NASA going forward:
A manned mission to Mars, a manned space station, and reusable space shuttle, and he chose the shuttle.
It's pretty clear that that story is not true.
There was no way that a mission to Mars would happen. It would be more costly than Apollo, and there was zero chance that congress would allocate money for that mission.
A space station was also out, because with the cancellation of Apollo, NASA had no way to get crew to a Space Station.
Which means the shuttle was the only choice, but it took a long time for Nixon to decide to go ahead with it. NASA had their shuttle program, but the space station would have to wait until some time in the future.
NASA did get something out of the cancellation of Apollo.
There were leftover Saturn V rockets and a number of Saturn IB rockets and Apollo spacecraft.
NASA eventually came up with a design for a station they named Skylab, which was built inside the third stage of one of the surplus Saturn V rockets. It was launched into orbit in May of 1973.
There were three missions flown to skylab, confusingly named skylab 2, 3, and 4 since the launch was named Skylab one. There was a fourth mission scheduled for spring of 1974 but it was cancelled.
NASA had plans to reactivate skylab when shuttle was flying, but a combination of shuttle delays and increased solar activity resulted in skylab reentering the atmosphere in July of 1979.
NASA was still planning a space station, which would be named "Freedom".
The Power tower configuration was announced by President Reagan in 1984, but - as we know - presidents don't control NASA's budget, congress does, and congress was not impressed.
Three years later there was a revised baseline, and in the meantime NASA had recruited the European space agency and Japan's JAXA to participate, but as the projected costs kept increasing congress was unwilling to fund the program. NASA needed a way to reduce the costs.
The solution came from a very unexpected source.
In December of 1991 the Soviet Union officially dissolved.
The Soviet - now Russian - space agency Roscosmos now had their own problem
They had been planning for many years to launch an updated space station to replace the aging Mir station, but they now also had monetary issues.
Take the Russian plans and the US/European/Japanese plans, put them together, shake well, and you end up with the international space station.
This was truly an excellent idea.
The Russians would get:
a bigger station at lower costs
access to some of NASA's technology
Additional resupply and support for Mir until the ISS could be launched
Recognition for what Russian space had accomplished in the years since Apollo.
NASA/ESA/JAXA would get
Years of Russian experience in space stations
An established pipeline to carry crew and cargo to the station
Cheaper hardware costs
The opportunity to get US astronaut space station experience on Mir
A way to keep Russian aerospace engineers doing things that the US liked rather than working for countries the US didn't like.
Both groups would get:
Redundancy in crew and cargo transport
An international program, which means that national prestige is involved making it harder for either participant to back out.
There's another US benefit that is less obvious.
NASA had a shuttle problem...
Originally, the shuttle was going to be flying all the time, but post-challenger it wasn't going to be launching commercial satellites and over time the DoD wasn't really interested either.
If we look at the shuttle manifests, they flew 22 times in 1995-1997, and 7 of those are flights to Mir as part of the ISS agreement, so it's clear that the manifest is pretty thin.
If we look at what shuttle did *after* ISS assembly started, it's a bit more clear.
From the start of ISS assembly in late 1998 until the loss of Columbia in 2003, shuttle flew 16 missions to ISS, 2 to service hubble, the launch of the Chandra X-ray telescope mission, and radar topography mission, and the last flight of Columbia, a spacehab mission. That's a full 75% of the missions dedicated to ISS, and 3 of the 5 non-ISS missions were flown on Columbia, which could not perform ISS missions because it was heavier than the other shuttles.
After Columbia, all the remaining shuttle missions were to the ISS, with the sole exception of that last hubble servicing mission.
I have talked a lot about why shuttle was cancelled in the past, but it wasn't until I looked at the manifests closely that I realized that one of the reasons was that the only big mission left for shuttle was delivery of cargo and crew to ISS and it was a very expensive way to accomplish that. All the shuttle infrastructure was sized to fly about 8 times a year, and flying only a few times a year would cost about the same overall, bumping the per-flight cost up significantly.
And that's the story of how the international space station was created.
The ISS modules were originally designed to last 15 years, and we know that NASA has robust designs and can therefore expect that they will last longer with proper maintenance. We are at 26 years for the first modules.
At some point, NASA will be done with ISS. When will that be?
There's already a well-established pattern...
In 2009, ISS was extended to 2016, in 2014 it was extended to 2024, and in 2021, it was extended to 2030. The big reason for this is that maintaining the status quo is easier and more predictable for everybody - everybody at the NASA centers, the contractors, the politicians in states that benefit from the program, and the international partners - all know exactly what keeping the current program for another year will entail. This is a very powerful force - there's a reason NASA flew shuttle for 30 years - and I would not be surprised at all to see them extend beyond 2030. There are NASA studies that are exploring that option.
On the other hand, there are developing factors that are becoming more prominent
The station is getting crankier, with more ongoing problems. Air leaks in the Russian segment have become persistent and so far have resisted attempts to fix them.
With Artemis becoming more real it is asking for more funds and additional money for NASA from Congress seems unlikely.
NASA has actually come up with a deorbit plan and given SpaceX a contract to develop a deorbit vehicle, though it is not fully funded.
How will these two sets of forces balance out?
I don't think the decision will be based on a date; as long as station is functional enough it will be extended.
The decision will come from other factors, and I see a few likely possibilities:
Congress could set a firm date and say "no more money after that". The status quo argument says this is unlikely to happen, but political winds are always hard to predict
The Russians could quit. The function of the Russian segment and the materials that come up through progress resupply and the reboost are critical without some quick thinking on NASA's side, and AFAIK there is no contingency plan. It's not clear how likely this is; the ISS is pretty much all that Russia has left of their space program but they could chose to spend money elsewhere.
The station may become that old car that you are putting $500 into every month and you can no longer rely on it to do what you need it to do.
Something significantly bad happens - there's a big debris strike, a critical system breaks down, there's a hazardous fluid leak - and the astronauts and cosmonauts have to get in their capsules and head for home. That could easily be the end for human habitation of station.
My guess is that it will be one of those. Let's hope that it doesn't take killing astronauts or cosmonauts for the decision to be made.
NASA's plan of record is to deorbit ISS.
Why are they doing this? Isn't there a better alternative?
NASA did a detailed analysis of 7 alternatives and summarized them in this white paper.
The alternatives are:
There is a government standard that says you can only do uncontrolled reentry if you can show that there is less than a 1 in 10,000 likelihood of public risk.
In 1979 when Skylab reentered, it missed the planned disposal location in the Indian ocean and instead dropped all around the port town of Esperance in western Australia.
Skylab had a mass of 76 tons, and the ISS has a mass of 400 tons. "Hope it doesn't hit anybody" isn't a responsible strategy.
Three of the options would require disassembly.
There are no disassembly instructions available.
It might appear that one could just take the assembly order and perform it backwards, but remember that the station has to remain operational and habitable at every step in taking it apart, and it has been modified over the years to make that more complex than you would think.
There is this weird thing called "cold welding", where metals that are in contact in vacuum will spontaneously bond to each other. Some of the space station modules have been held together for 25 years; they may not want to come apart.
Disassembly and return to earth.
I actually started a video on doing this with starship called "bring ISS Home", but then this report came out and I lost interest.
I did do some calculations using the dimensions of the modules and found that you can fit them within starship's payload volume, and with some modules you could stack them and return them together.
It would probably take about 6 missions and you'd probably need a custom starship.
And then once you get them back you would need to decide what to do with them.
This is one of the Space Shuttles that NASA flew into orbit is a very impressive sight.
This is the space station toilet is a little bit less impressive.
Reuse sounds like a nice plan.
However...
The modules were design in the 1990s or earlier and are old
NASA doesn't own all of them
You need to figure out how to take them apart
Their orbit is at 51.6 degrees of inclination, so you are stuck there, and it's more costly to get there than the 28.4 degrees native inclination you get from cape Canaveral. Unmentioned is the fact that station is at 51.6 degrees because it needed to be there so the Russians could access it, and nobody in the US is going to touch working with them on new projects in the foreseeable future. It *is* true that the higher inclination covers more of the earth's surface if you care about that.
And finally, no commercial companies had feasible proposals
These are all good reasons, though the last one need a caveat:
The caveat is that Axiom space has an ongoing project where they will build one or more modules and connect them to ISS, and then when ISS is decommissioned, those modules will convert to a free flyer.
NASA has kicked in some pocket change with a firm fixed price contract for up to $140 million over 5 years.
If you can't place the name, Axiom is the operator for the Crew Dragon missions that visit the ISS, and they also have the contract to build the Artemis Lunar spacesuits and - presumably because Collins Aerospace has dropped out - future space station suits as well.
Disassemble and deorbit in smaller pieces.
The benefit of this is (read)
The downsides are the same as the other disassembly options, plus the requirement for a number of deorbit tugs rather than one big one.
Thanks for sitting through all the optinos up until now; we have finally reached the one you are waiting for, boosting to a higher orbit.
NASA has done a really good job with the explanation in this section.
We start with this chart that shows the delta V and propellant mass required to move ISS to a higher orbit. None of the current launchers can carry that sort of propellant in a single launch when you include the mass of the boost ship. Maybe you can get there with a heavily modified Falcon Heavy, or perhaps with a New Glenn.
NASA mentions starship as an option, and it could certainly do the job, but you would need to modify it to be able to give enough thrust to ISS without tearing it apart. A single raptor may be too much thrust.
NASA also talks about Solar Electric propulsion. That only takes 15 tons of propellant to get to 850 kilometers, but it also takes 3 years and you need to keep station operational for that whole time period.
NASA has an orbital debris model known as ORDEM that can be used to predict the amount of debris at specific orbital altitudes.
Station lives down here, at a little more than 400 km. At that altitude, ISS has to perform about 1 debris avoidance maneuver per year. If there isn't time to maneuver, the astronauts and cosmonauts move to their capsules and wait for the danger to pass.
A 1000 year orbit has a debris flux that is 6 to 7 times worse. And remember that you would be putting it there without the ability to dodge debris.
Here's a scary picture...
At the current altitude, the estimated mean time to failure for ISS due to debris strikes is 51.3 years.
Bump it up to 800 kilometers, and that number drops to 4 years.
ISS is a huge target, and if you put it at 800 kilometers, it is going to get hit and those hits are going to generate fragments that increase the debris risk, and more of those new fragments are in long-term orbits.
I don't generally think that Kessler syndrome - where you get so much junk in low earth orbit that it creates a self-perpetuating debris cloud - is a major risk, but putting ISS at 800 kilometers is probably the worst thing you can do when it comes to space debris.
If you get out to 2000 km you start getting back to the low risks sections, but it's not a no risk area - station still has to do active avoidance at that density - and you really need starship-class tugs to move it there.
At this point you should be asking "what's the point of trying to preserve ISS in a higher orbit if the debris is just going to tear it apart?"
I have no idea why NASA included this as an option - deliberately creating space debris even at low orbits is a pretty dumb idea - but my guess is that somebody thought it was a good idea and NASA therefore decided to include it.
I'm not sure how you would actually do this, but even if you could, some of the fragments are going to end up with more energy and in higher orbits.
This is the second-worst option, and it's only second because the previous one is so bad.
Transitioning to a commercial operator.
NASA notes that they had discussions with the US industry in regard to their interest in repurposing any parts of station or taking over operation, and they received no feasible proposals.
NASA is currently spending $4.2 billion per year on ISS. You need to cover that, either convince the Russians to keep working with you and flying Progress missions or come up with your own reboost and attitude systems and detach the Russian modules. In return, you have to figure out how to bring in enough money to make the investment worthwhile.
I don't see any world where that would make sense, and the industry agrees.
And that's why NASA plans to deorbit ISS; none of the other options are workable.
What comes after ISS?
NASA's current plan is known as commercial low earth orbit destinations, or commercial leo.
Commercial companies will build and operate space stations and NASA will buy space station time from those companies.
I'll cover that in detail in a future video
And that's the rise and impending fall of the international space station.
If you liked this video, please send me this surf beach 17 foot dual lane water slide featuring a splash pool at the bottom, currently available for only $234 dollars a month.
I've been asked a few times whether I think the ISS was worth it.
To do that, we'll need to do a cost/benefit analysis to weigh the cost of ISS versus the benefits that we got from.
We'll start with cost.
Every part of the ISS is custom made for the ISS program to meet the needs of the program, to last for 15 years in the hostile environment of space, and to be as light as practical.
The majority of these parts are carried into space on a shuttle flight that costs perhaps $500 million a flight.
The parts are then very carefully assembled by highly trained astronauts.
The astronauts and cosmonauts that make up the ISS expedition crew tend to get a bit cranky if they don't have air to breathe, water to drink, and food to eat, so the station requires a constant stream of cargo vehicles to bring up consumable and take away waste, along with crew vehicles to bring up new crew members so the old ones can return to earth.
That is not - unsurprisingly - a cheap thing to do, so there's a lot of cost there.
And there are two control rooms that are continuously staffed, the Russian one in Korlyov (car-la-yov) outside Moscow and the NASA one in Houston Texas.
Add all that up, and you get $100 billion.
Actually, you get a bunch of people each with their own number, but $100 billion is the right order of magnitude and that's all we need to answer the question.
We now know the cost.
Now we can look at the benefits.
I'll note that there's an implicit question that comes out of the cost/benefit analysis which is "what could we have done with the money if we didn't do ISS?" "Stuck in LEO"
The first benefit is pretty obvious - ISS has taught us a huge amount about what it takes to actually operate a space station, and much of that applies to other situations - lunar bases, mars missions, and mars colonies. It's quite useful learning your lessons in LEO rather than in a location that's much farther away.
The second thing is international cooperation in space. More countries cooperating in space is a good thing, and the ISS model has been transferred to the Artemis accords.
In the present day, US and Russian relations are about as frosty as they have been since the collapse of the Soviet Union, but both countries are still sending both crew and cargo to the station and still flying on each other's rockets to get there. Putin removed Roscosmos head Dmitri Rogozon 2022 when his anti-western position was threatening the cooperation in the space
From the start of ISS assembly in late 1998 until the loss of Columbia in 2003,
Chandra (Columbia)
to the end of the program, shuttle flew 37 missions to the ISS, 3 to hubble, and 3 other missions.
Shuttle existing almost purely to put the space station together, and one of the reasons it was cancelled was that it had finished assembly of the space station and there was no other good job for a vehicle that cost $4-5 billion per year.
People often ask about how much ISS cost to build, and since nobody could agree on the costs during the program it's hard to find good numbers now. A conservative estimate for the cost of the modules is $40 billion and the cost of the shuttle flights is about the same, so I'll say $80 billion for the US contribution.
Other people have estimated that the total cost across all the countries is closer to $150 billion.
Either way, ISS is probably the single most expensive thing ever built, and that explains why NASA keeps extending it.
Well, not really. Remember that NASA walked away from all the Apollo stuff and later walked away from all of the shuttle era stuff.
It really comes down to politics.
To explain I'm going to need to inflict on you the current NASA org chart, which looks pretty typical if you've seen corporate org charts. It has the important people at the top, and the working people in the NASA centers at the bottom.
I've recently realized that this is the wrong way to look at NASA, and if you look at it this way, you'll ask lots of "why don't they just?" questions.
I have a new model that I hope is more enlightening.
The way I would draw the org chart is a view that puts the NASA centers in the middle.
For ISS, the most important NASA centers are Johnson Space Center in Texas and Marshall space flight center in Alabama.
They get their marching orders from congress, specifically from the space subcommittees, and they also get the money to support those programs from congress.
Johnson's big part of the pie for ISS is astronaut training and ISS mission control, and Marshall's is ISS science operations, though Marshall have worked on building ISS modules in the past.
Below that are the contractors that handle the transportation of crew and cargo to and from station, those that manage operations on station, and those that build or have built modules for station.
The drivers for the politicians deciding which NASA missions get funded are the benefits that it brings to them and their constituents, so the money that goes to the NASA centers in their states and the contractors wherever they may be located.
What this means is that the NASA centers are the drivers of what happens because they have the connections that control funding, and NASA HQ mostly plays an organizational role off to the side.
Let's ask some questions and see if this model helps answer them...
Learn from the Russians and keep them busy.
The Russians had serious amounts of on-orbit experience that the US lacked and that was the point of both the shuttle visits and the seven NASA astronauts who carried out long-duration missions on Mir. And the Russians brought that experience to the space station, and their participation was critical during the time period between the retirement of shuttle and the first SpaceX crew dragon flight - without the Russians, the space station would have been uncrewed for 9 years.
The second goal was to keep the Russians busy working on space station stuff rather than leaving to work on rocket projects in countries that might not be friendly to the US.
It's fairly clear that the endpoint between the Russians and the US in space isn't what we had hoped for, but I would argume that it worked pretty well for quite a few years.
I was asked a while back whether ISS was a good investment.
This is complicated because the answer depends on what you think the purpose of the space station should be, and that differs from person to person. I'll give a list of what I think the goals are for the space station and whether they accomplished them.
The first two are the ones I already talked about - keeping shuttle busy and keeping NASA centers busy. It did a good job at both, and from NASA's perspective, those were important.
Keep the Russians busy. The idea was that keeping Russian aerospace engineers working on peaceful space things was better than turning them loose to work on other things, and that's why the US paid for the construction of the Zarya module. It was also to bring the astronaut and cosmonaut community and their group training communities closer together. I would say it achieved all of those, and though the situation in Russia isn't what we had hoped it would be, that doesn't mean that the effort wasn't worth it.
Cooperate with other countries. In addition to Russia, JAXA in Japan, CSA in Canada, and all of the ESA countries in Europe have worked closely with the US on the station. 280 people representing 23 countries and five international partners have visited ISS. The international nature of station and relatively easy access makes visits possible for many countries that don't have large space budgets.
NASA talks about the human research on ISS a lot, and it's certainly a good research topic, but I suspect that we are well past the point where the benefit of further human research is worth the cost.
Identify opportunities for commercial activity in space.
NASA created this list of the top 20 breakthroughs from 20 years in science aboard the station.
I'm not going to read them all, but I do want to highlight a few. Disease research and especially drug development seem like useful research, but as far as I can tell there haven't been any breakthroughs in this area.
Stimulating the low-earth orbit economy is pretty obviously not *science*, but it would be a good thing if it were true, but I'm skeptical in this area.
And a lot of these just don't seem very important.
The problem is that NASA has spent a lot of time and money searching for a killer app for ISS without really coming up with one. I'll talk more about this in the video on commercial LEO space stations, but it could be that ISS is too expensive for commercial applications, or it could be that the humans are the problem.
Humans need enough space, oxygen to breathe, food, lighting, ventilation, temperature control, waste disposal, entertainment - the list goes on and on. You need a big station with periodic resupply to keep that going, and everybody expects that you will do your best to keep the astronauts alive. And a lot of your astronaut time goes towards maintenance to keep those astronauts alive.
In return, you get some very useful capabilities - they are very adaptable - but they mess up the microgravity you seek by moving around, by exercising, and by the vibration from the equipment to keep them alive.
Humans are messy.
So this remains an open question for future space stations.
Orion is a critical component of NASA's upcoming Artemis missions to the moon, but it's not discussed very much.
It turns out there is a very interesting story
Our story starts in 2002, on a NASA project known as Orbital Space Plane. The plan was to build a small spaceplane - or even a capsule - and use it to launch astronauts or cargo to the international space station on top of commercial rockets. Here we see an orbital sciences design that would launch on the Delta IV Heavy, and a Lockheed Martin design that would launch on the Atlas V.
The intention was for this to supplement the shuttle fleet and provide complementary capabilities.
But then this happened:
After the Columbia accident, there was agreement in Congress that the shuttle should fly until the international space station was completed and then be retired.
One of the unfortunate byproducts of NASA being a government agency is that changes in administration and/or NASA administrators can change the path of the agency. This is one of those cases.
The idea of a new vision sounded to NASA like a great chance to belly up to the all-you-can eat buffet of federal money, so they dug out a bunch of proposals they had lying around, massaged them into a plan, and published it.
They called it "the vision for space exploration"
There were four high-level goals (read).
I'd like to highlight two important phrases:
The first is "sustained and affordable". The point being that this is not a program like Apollo, to visit a few times and go away.
The second is "commercial".
Both of these will come up later.
There are two sub goals that are relevant to our topic:
Develop a new crew exploration vehicle to provide crew transportations for missions beyond low earth orbit.
Acquire crew transportation to and from the international space station.
Note that these are separate goals, and note the use of the word "acquire" rather than "develop" for crew transportation.
There's a popular belief that the president is in charge of the government, and therefore in charge of NASA.
This is not true; it is congress that decides what NASA works on and how much money they can spend.
For more information on this, see my video "budgets at NASA - I'm so confused..."
We therefore need to see how Congress responded to the plan.
https://www.govinfo.gov/content/pkg/PLAW-109publ155/pdf/PLAW-109publ155.pdf
The exploration goals showed up in the NASA authorization act of 2004 and roughly aligned with the vision. It specifically instructed NASA to:
(read)
These are both about going beyond earth orbit.
There's also an interesting requirement around commercialization
Read
This is clear direction for NASA to use commercial solutions unless there is a really good reason to do it themselves.
This was amplified in the 2005 authorization.
(read)
Work with new space companies to encourage them, and specifically contract with the private sector for crew and cargo services to the international space station.
I couldn't find a official logo for the crew exploration vehicle, so this was the best I could do...
To understand the crew exploration vehicle it's important to understand the overall mission architecture, and to do that I'll need to rewind 50 years...
Apollo used an "all at once" architecture; the Saturn V launched a single stack that included the command module, the service module, and the lunar module all at once. This was a great way to get there fast, but it had some limitations...
The first is that the service module needs to have enough oomph - what we call delta v - to be able to brake the command module and the lunar module into lunar orbit *and* then have enough delta v remaining to get back out of lunar orbit and back to earth. That puts a big constraint on the mass of all three of those components; you can't, for example, increase the lunar module mass by 50% as the service module simply can't put that much mass into lunar orbit. You could make a bigger service module, but it might be too big for the Saturn V to lift. Apollo just *barely* worked, and the engineers were sweating grams of mass by the end to make it work.
Actually, this was the 60's, so they were sweating 1/28th of an ounce, but you get my point.
The architecture for this new vision - known as Constellation - is a more flexible architecture that relies on multiple launches.
There are different versions; in this version there are 4 individual launches. One for the crew vehicle, one for the lander, and two earth departure stages.
Dock the earth departure stages with the lander and crew exploration vehicle, and they take the vehicles from earth orbit all the way into lunar orbit, where they dock.
The big advantage here is that this makes it much easier to design the crew vehicle, since all it has to do is get itself back from the moon, and that's a small amount of delta V.
Essentially what it's doing is adding two more stages to the architecture, and more stages makes things easier, at least from a delta v perspective.
Now we can look at the requirements that NASA had for the crew exploration vehicle.
First, it needed to mass less than 20 tons at liftoff. For reference, the Apollo 11 command and service module massed about 29 tons, so how are they going to make this version so much smaller and lighter?
The Apollo CSM had a much harder job to do.
Coming from earth, it had to slow down its own mass and the mass of the lunar module, for a total mass of 43.9 tons. Getting into lunar orbit takes about 1000 meters per second of delta v.
I'll skip the detailed delta-v calculations this time, but they show that the first fuel burn required 12.3 tons of fuel.
Then after the mission, we need to get back to earth. That takes the same 1000 meters per second of delta v, but we don't have the mass of the lunar module and we don't have the mass of the fuel we burned getting into orbit, so it only requires 4.6 tons of fuel to get out of lunar orbit and back to the earth.
Subtract the fuel burn, and that tells us that the dry mass of the Apollo command and service module will be about 11.9 tons. That's all the structure, fuel tanks, engines, electronics in the service module, plus the entire command module. That whole package is about 59% fuel.
That was very hard to build, especially with the technology in the mid 1960s.
https://web.archive.org/web/20090117083711/http://prod.nais.nasa.gov/cgi-bin/eps/sol.cgi?acqid=113638
Let's do the same set of calculations for the crew exploration vehicle. It starts at 20 tons.
The lunar orbit insertion is done by the earth departure stage, so the first fuel burn is 0 tons.
Getting back to earth only takes 5.6 tons of fuel, leaving a dry mass of 14.3 tons. This means only 28% of the mass is fuel.
The dry mass of the crew exploration vehicle is considerably more than the apollo command and service module, and remember that it only needs to carry 5.6 tons of fuel instead of the 16.9 tons the apollo CSM carried, so it has smaller tanks.
And there are many more lightweight materials now.
All of that means the crew exploration vehicle will be much easier to build.
The second requirement is that the crew exploration vehicle perform it's mission requirements. It needs to hitch a ride to orbit, hitch a ride to lunar orbit, and then perform the "get home from the moon maneuver" and reenter into the earth's atmosphere and land.
Keep the astronauts alive for 16 day missions. That seems like a good thing to do.
Be capable of uncrewed operations. Unlike apollo, the intention is to take all the astronauts to the lunar surface.
Support 1 to 4 crewmembers with 3.54 cubic meters of habitable volume per person. That's 125 cubic feet, and is an example of NASA thinking in imperial and talking in metric.
Apollo had about 2 cubic meters per person.
The shuttle *supposedly* had 10 cubic meters per person with 7 people on board. I suspect this is the full pressurized volume and doesn't include the space taken up underneath the middeck, in the avionics bay, or in the storage lockers. But its still a lot of space.
So the crew exploration vehicle needs to roughly double the space of Apollo.
Remember how I said there is more mass available in the crew exploration vehicle? One place it will go is to more space for the crew.
NASA also directed designers to consider other scenarios. I won't read all of these, but I do want to highlight the one that says "CEV support for transfer of crew to and from the international space station"
There was also a different development approach
They would use Spiral development, where needed capabilities would be developed incrementally in a series of spirals.
The first spiral is to test crew transport in orbit, with an objective to test this by 2014.
The second spiral extends the capability to support extended lunar exploration, where extended means "up to 4 days"
The third spiral extends that to several months, to serve as an analog for possible Mars missions.
Very much a "build a little, fly a little" approach.
They also planned what they called flight application of spacecraft technologies - or FAST - where multiple companies would develop full prototypes and there would then be a "fly off" to decide who would get the contract. This had been done in the past for military programs, but not for NASA.
Spiral from statement of objectives, document 4.
https://www.govinfo.gov/content/pkg/PLAW-109publ155/pdf/PLAW-109publ155.pdf
How does this align with NASA's mission?
I think the answer is "pretty well"; there's lots of exploration and a great chance for some innovation. And some nice competition. This feels like NASA was very much trying to do something different and something for the long term.
NASA is gearing up for sharing this with industry in 2005, then in December of 2004, Administrator Sean O'Keefe announced his resignation. His official reason was that the $500,000 salary he would get as the chancellor at Louisiana state university would work better to send his kids to college than the $158,000 NASA was paying him. Others have speculated that he was pushed out due to some comments he made on climate science, but the reality is that 3 years is a decent run for a NASA administrator.
Deputy administrator Frederick D. Gregory became acting administrator starting in February, and it was during this time that NASA released their request for proposal for the crew exploration vehicle with the requirements I already discussed.
There were two groups that submitted bids.
https://www.govinfo.gov/content/pkg/PLAW-109publ155/pdf/PLAW-109publ155.pdf
Northrop-Grumman and Boeing submitted a very Apollo-like design - a capsule for the astronauts and a service module with solar panels. Pretty much all I could find about it is this picture and a few notes saying there is very limited information.
This group had a tremendous amount of history in crewed flight - they had been involved in a lot of NASA projects, including the Lunar Module during Apollo.
Lockheed Martin came up with a proposal that is directly based on their orbital space plane concept, along with add-ons that would allow it to go beyond low earth orbit.
One of the interesting things about this design is that its shuttle-like in the sense that it reenters like a plane, but when it has slowed down enough, it will pop drogue chutes, main chutes, and then land on air bags, just like a capsule.
I'm calling it a Shapsule.
I think the orbital space planes deserved funding, and this approach does as well; it aligns well with what NASA should be doing. It pushes the state of that art but is making a small leap rather than a big one.
Given that you know that Orion is a capsule, you may have guessed that NASA chose the capsule design, but in fact, NASA awarded $28 million to both bidders to begin design work.
While this was going on, Michael D. Griffen was nominated as the new NASA administrator on March 11 and confirmed in early April.
This would lead to a major change in trajectory for the program. Ha ha.
https://www.govinfo.gov/content/pkg/PLAW-109publ155/pdf/PLAW-109publ155.pdf
In January of 2006, NASA released a revision of the previous RFP. What NASA creatively called a "Call for Improvements"
Here's a summary of what changed:
Up to 6 crew members to ISS, four to the moon, and support future Mars missions.
Will be used to select a single contractor later in 2006. Both the spiral approach and the FAST fly-off is gone.
The spacecraft will be an improved, blunt-body crew capsule shape. No Shapsules allowed.
Requirements based on exploration mission needs and a desire to fly as close as possible to 2010
Phase II proposals due march 20, 2006. So, two months to "improve" your proposal.
NASA was asked in congressional hearings about the change of approach, and they gave a lot of reasons for the change. I won't read them all, but a good summary is
We are NASA, we are the experts, we know what we are doing.
http://commdocs.house.gov/committees/science/hsy29949.000/hsy29949_0.HTM
That's not the real reason.
The real reason is that NASA wants to maintain their US monopoly on flying humans.
It's pretty obvious that the Lockheed concept is going to have a version that is light enough to fly on Atlas V, and it's a fair bet that the Northrop Grumman / Boeing approach would be able to do the same thing - especially since NASA asked them specifically to try to support that scenario.
That is precisely what some people at NASA - including their new administrator - do not want to happen. The point of the changes are to keep that from happening. I'll talk about how that worked in a bit.
What we got out of the Phase II request was not only a capsule, but a big capsule. Here's what it would look like.
The Orion designation would come the next year, giving us the Orion Crew Exploration Vehicle.
In a big surprise, Lockheed Martin was awarded the orion contract in August of 2006. It was a cost-plus contract for development and two flight test capsules, in the amount of $3.9 billion.
I think it's fair to say that many people were not happy.
Read Bart Gordon
Read General account office
But Congress decided to let NASA continue.
http://commdocs.house.gov/committees/science/hsy29949.000/hsy29949_0.HTM
Now we have a new capsule, and it's a big capsule.
If we compare it to Apollo, we can see that it's a meter wider and a bit taller. That means a lot more space inside.
If we add in the two commercial crew capsules - both of which could carry 6 passengers to the ISS - we can see it's still a very big capsule.
We can find some of the justification for this in the NASA exploration system documents.
The size of Orion was based on sending 4 crew members to the moon. If you look at these highlighted numbers, the total vehicle masses about 23,000 kg and has 1724 meters per second of delta v. That's more than enough delta v if you are doing the constellation architecture, but probably not enough if you need to put yourself into orbit. You are at least 300 meters per second short.
Michael Griffin called this design "Apollo on Steroids"; it is a very big capsule but it's not really a very capable system.
Remember where I talked about NASA wanting to be the only ones who flew crew?
That shows up here in the ISS section. The ISS vehicle is only 253 kilograms lighter than the lunar one, and it has a ridiculous amount of delta v; probably 4x what you need for the ISS mission. Why would they do that?
Time for some numbers. Given the total mass and the delta-v, we can figure out how much fuel they are carrying to get that delta-v. For the ISS version, it's about 9 tons, and for the lunar version, about 10 tons.
So the ISS version that can do 1544 meters per second of delta-v has a mass of 22.9 tons. What if we relax that delta-v required to something more realistic, like 400 meters per second.
The fuel mass drops *way* down, and the total mass is now 15.6 tons.
https://www.nasa.gov/pdf/140649main_ESAS_full.pdf
Remember that congress has mandated that NASA look at commercial solutions.
There were two possible options - the Atlas V from Lockheed Martin and the Delta IV Heavy from Boeing. At about this same time, the parent companies would merge the rockets into a new company known as United Launch Alliance. That's another strange story; see the link to my video that talks about it.
NASA needs to come up with reasons why neither of these are practical. Here's what they come up with:
(read)
Note that they are arguing against small solid rocket motors when they will in the very next page argue that using a large solid rocket motor is fine.
(read)
This is a fair concern; there will need to be work done to crew rate these vehicles.
(read)
A new or upgraded upper stage would be required. Just like the new upper stage required for the NASA-preferred approach.
(read)
This is why they switched to a capsule, and this is why the ISS version of Orion is so damn heavy. If you take that extra fuel away, these performance shortfalls evaporate. It's all about keeping crew away from commercial launchers.
https://www.nasa.gov/pdf/140649main_ESAS_full.pdf, page 40
Of course, that Atlas V that was not suited for human flights to the space station would be chosen by Boeing to be used for commercial crew flights of their Starliner capsule to the space station.
NASA ends up specifying a design that became known as the Ares I. I'll talk about the details in a future video, but suffice it to say that it's a shuttle solid rocket booster with a new hydrogen/oxygen second stage.
NASA asserts in this slide that it's a $1 billion / flight launcher.
This one on the left is the triple-core Atlas V Heavy. Atlas V is - by itself - somewhere around $200-$250 million. It may be kindof pricey to develop the heavy variant - SpaceX spent a lot of effort on Falcon Heavy - but it seems unlikely that the cost per flight goes up 4x.
Next to it we have Delta IV Heavy. It needs a new second stage, and it might need engine modifications, but the changes seem a lot easier than Atlas V. Somehow that new second stage triples the cost of the existing launcher, one that is very expensive right now because so few are built.
Anyway, that finishes the discussion of launchers and explains why NASA needed such a big and heavy capsule; they needed one that couldn't be easily launched by commercial launchers.
Time passed...
In 2010, a review of constellation found that it was unaffordable, and it was cancelled by the Obama administration.
That's not really allowed, and congress quickly revived it in a different form.
The net effect was to cancel the Ares I launcher and move Orion onto the Ares V launcher, now renamed "Space Launch System".
In addition the three Orion variants were merged into one Orion to rule them all.
There was also a rebranding, from the Orion crew exploration vehicle to the Orion Multi-purpose crew vehicle.
Which is a good name if you consider the purposes to be:
Going to the moon
Coming back from the moon
Everybody knows there is no chance NASA will use this launcher to send a capsule to ISS, and there's no Mars plan, so it's all about the moon.
The same act that gave us SLS and continued Orion gave us the commercial crew program that would see both Crew Dragon and Starliner capsules dock with the ISS.
Time passes.
In 2012, the European space agency agreed to construct the service module for Orion, thereby sharing the costs with NASA. This slightly reduced the amount of delta-v available.
Time passes.
In 2014, Orion flew on Exploration Flight Test 1.
And yes, that's a Delta IV Heavy, the same Delta IV Heavy that was an unacceptable choice as a launch vehicle.
Though to be fair, this was an uncrewed test flight.
And now it's time to talk about one of Orion's little secrets, an outgrowth of the constellation architecture.
You remember this chart from earlier, which said that Orion would have 1724 meters per second of delta V.
Orion has gotten heavier and the service module has gotten smaller, so that figure has dropped to 1350 meters/second. If we reserve some delta v for safety, that takes it down to 1250 meters/second.
To get into and out of a reasonably low lunar orbit takes at least 1800 meters/second.
Orion was designed for the Constellation approach, but with the Artemis approach, Orion can get into a low lunar orbit, but it cannot get out of it. Which would seem to be a drawback for a crew vehicle.
https://ntrs.nasa.gov/api/citations/20150019648/downloads/20150019648.pdf
The Artemis missions therefore use a Near Rectilinear Halo Orbit; it costs less than 900 meters per second of total delta v, and that's comfortably within the capability of Orion.
The downside is that the delta-v needs to be made up by the lander; it has to make a transition from the halo orbit to low lunar orbit and land from there, and do the reverse to return to the orion capsule.
That pushes the delta-v requirement from about 4000 meters per second up to 5500 meters per second. Much harder to build a lander that can do that than to do what the lunar module did.
SpaceX won the Human Landing System contract with lunar starship, which is able - with refueling in low earth orbit - to make it to the near rectilinear halo orbit, down to the lunar surface, and then back to the near rectilinear halo orbit. SpaceX's bid was relatively low because they are already building starship, but the other bids were very high because the landers needed to have so much delta-v.
So how much has Orion cost?
If we look purely at the payments to Lockheed martin, the initial contract to build and fly two test capsules was $3.9 billion.
The cost up to the exploration test flight in 2014 was $7.2 billion.
And through August of 2022 they have been paid $14.2 billion.
Orion has been a tremendously lucrative program for them.
https://www.usaspending.gov/award/CONT_AWD_NNJ06TA25C_8000_-NONE-_-NONE-
NASA of course adds overhead on top of the Lockheed Martin costs.
Through the exploration test flight in 2014, NASA spend $10.8 billion on Orion.
Through August 2022, they have spent more than $20 billion
There has been a lot of flak aimed at Boeing for cost overruns in the development of SLS, but nobody is asking Lockheed Martin what they are spending $1 billion / year on.
https://www.usaspending.gov/award/CONT_AWD_NNJ06TA25C_8000_-NONE-_-NONE-
Thus concludes the story of Orion. A tragic story of how we could have had an interesting and innovative Shapsule to carry humans to low earth orbit and beyond, but instead ended up with a big heavy capsule that isn't a great fit for the mission that it is being used for.
But it did send $14 billion to Lockheed Martin, and that at least means something.
If you enjoyed this video, please make a donation to the "Save the Shapsules" fund.
The year was 2014, and Tory Bruno had a problem. More specifically, the company he worked for - United Launch Alliance - had a problem, but since Tory Bruno had just taken over as CEO of United Launch Alliance, it was his problem.
He had too many rockets...
To understand why that was a problem, we'll need to back up a few years...
ULA was created in late 2006 as a forced marriage between Lockheed Martin and Boeing
If you want *that* story, see "Spies, Allies, and Enterprise - the Strange Story of ULA"
Each of the companies brought their rocket to the new company.
Lockheed Martin brought the Atlas V and Boeing brought the Delta IV Medium and Delta IV Heavy, both originally developed at McDonnell Douglas before their merger with Boeing.
And business was good - as the only significant US launch provider, they launched pretty much all the government payloads, be they NASA payloads or department of defense payloads.
Duplication was a problem - they had one assembly line for the Atlas V rockets and one for Delta IV rockets and of course workforces for each rocket.
These two rockets needed four launch pads - an atlas V and delta IV pad at cape Canaveral space force station in Florida, and an atlas v and delta iv pad at Vandenberg space force base in California.
The duplication was expensive.
Both of these rocket designs came out of the Air Forces Evolved Expendable Launch Vehicle program, which is now known as national security space launch. The goal of that program was to come up with a new launcher that could launch both government payloads and commercial payloads and therefore would be cheaper for the government.
Lockheed Martin bid Atlas V, McDonnell Douglas bid Delta IV
After the project was out to bid, it was decided that the Air Force would choose two winners, which sounds like a good thing for the companies but was actually a bad thing.
If you are designing your rocket to launch say 10 times a year, your economics are much worse if you only launch 5 times a year. Both companies involved called "foul" and stated they wouldn't make money if there were multiple providers.
The Air Force decided to deal with this by throwing money at the problem, and the "launch capability payments" system was born. The government would pay the companies a retainer to keep the capability to launch if there weren't enough launches to make it profitable for the company.
These payments continued after ULA was formed and ended up being approximately $1 Billion per year.
Business was, as I said, good.
But this upstart named SpaceX came on the scene and had the audacity to sue the air force over a big launch award that was given to ULA without a competitive bid process and the air force had to settle because they would very clearly lose in court. It was clear that the days of the ULA monopoly were numbered.
ULA had two rockets, which was one rocket too many. The obvious solution was to choose one.
The Atlas V seemed like an obvious choice - it was cheaper and flying a lot more often than the Delta IV.
There were two problems, however.
The first was that for ULA to continue to get EELV/NSSL contracts, they needed to be able to reach 9 different orbits. The GEO 2 orbit at the bottom is an extremely challenging one - with a requirement of lifting 6,600 kilograms all the way to geostationary orbit.
A quick look in the Atlas User's Guide shows that there is a problem - the beefiest Atlas V can only carry 3,904 kilograms to that orbit and therefore it cannot meet all 9 orbits.
Atlas V cannot be ULA's only rocket.
The Atlas V also relies on the RD-180 engine, which is highly efficient and relatively cheap but unfortunately is made in Russia.
That made use of that engine a political issue, and it went from being banned to being allowed back to being banned and the finally, having Russia say they would no longer supply any engines.
And therefore the RD-180 was not a long-term solution.
If not the Atlas, then use the Delta IV. It could do what the Atlas could do and the Delta IV Heavy variant with three boosters could meet that difficult geostationary orbit requirement.
The problem with delta iv was not performance but price.
Atlas V launches are in the $125-150 million range. Delta IV medium prices are hard to come by because it rarely flies, but it's likely in the $175 million range.
Delta IV Heavy launches were an eye-popping $450 million. This price raised a lot of eyebrows when ULA was the only game in town and with SpaceX touting Falcon Heavy at $90 million, it was going to make it very hard to justify.
Delta IV could do what was needed but it just cost too much.
ULA needed a new rocket and that means ULA needed a new engine, and it would need to come from somebody else because ULA doesn't make engines.
The first option was surprisingly the RD-180.
Pratt & Whitney had formed a joint venture with NPO Energomash in Russia, the manufacturers of the RD-180, and out of that agreement had obtained a license to manufacture the RD-180 in the US. That seems like a great idea, but the number being passed around by Pratt & Whitney was $1 billion.
The second option was from Aerojet Rocketdyne, creatively named the AR1. It's a similar design to the RD-180 but only has about 60% the thrust. Aerojet Rocketdyne had gotten Congress to pass a law requiring the air force to develop an RD-180 replacement. The "billion" number was mentioned for this engine as well, though the bulk of that would be funded by the air force and aerojet rocketdyne.
The third choice was the BE-4 engine from the upstart Blue Origin. It was in the right thrust class for an Atlas V replacement and the development would be paid for by Blue Origin. It was also farther along in development than the AR1 and likely would be cheaper as well. It would require ULA to work with a new propellant as the fuel for the BE-4 is liquid methane rather than the RP-1 used by the RD-180 and AR1. ULA decided on the BE-4 for their new rocket.
It is often remarked that ULA took a big risk going with Blue Origin who had never developed a large rocket engine, and they have certainly had to deal with many BE-4 delays. But it's important to remember that the last high performance engine that Aerojet rocketdyne developed with the RS-25 for the space shuttle all the way back in the 1970s and the AR1 was only a concept. Tory Bruno has repeatedly said that, despite the delays, they made the right choice going with the BE-4.
Once ULA had an engine to design around, they could design a rocket to use it, and they came up with Vulcan. Vulcan is the god of fire in Roman mythology and that's certainly a great choice for a rocket and better than his Greek counterpart, Hephaestus.
Conceptually, Vulcan is like a big version of Atlas V.
The first stage has gone from 3.8 meters in diameter to 5.4 meters. The Atlas V used an RD-180 engine burning RP-1 refined kerosene with liquid oxygen and produced 3.83 meganewtons of thrust. Vulcan has two BE-4 engines burning liquid methane and liquid oxygen and producing 4.9 meganewtons of thrust.
Like the Atlas V, Vulcan's performance is tunable by adding solid rocket boosters to the first stage. Atlas V can use up to 5 GEM-63 boosters to add 8.35 meganewtons of thrust.
Vulcan can use up to 6 GEM-63XL boosters to add up to 12.36 meganewtons of thrust. Add the solids to the main engines, and Vulcan's 17.26 meganewtons is 40% greater than Atlas V's 12.18 meganewtons.
All of this extra thrust is in service of a much larger second stage. The Centaur III used on the Atlas V has a single RL-10 engine and only 21,000 kilograms of propellant. It's a very efficient stage but not a very powerful one.
The new Centaur V used on Vulcan has double the thrust with 2 RL-10 engines and 2.5x the propellant, so it is a much more capable stage.
Vulcan can fly much heavier payloads than Atlas V, with 27,200 kilograms to low earth orbit, 15,300 kilograms to geosynchronous transfer orbit, and 7000 kilograms to that pesky geo 2 orbit required by NSSL.
Vulcan will meet all the needs of ULA and will allow them to fly a single rocket.
The world of rockets has changed since Vulcan was conceived, with Falcon 9 making first stage reuse routine, the obvious question for any new rocket is "Can it land like Falcon 9?". And the answer for Vulcan is "no"
This is a chart showing a Falcon 9 launch, with the staging point indicated by the green circle.
At that point the first stage is travelling at 2150 meters per second. Falcon 9 has a big beefy second stage that can do a lot of work, so that allows it to stage early in flight.
We don't have all the numbers for Vulcan, but Atlas V uses the same philosophy
Atlas V has a less beefy second stage, and therefore stages later and at a higher velocity of 5870 meters per second.
Kinetic energy is proportional to the square of the velocity, and the Atlas V first stage has 7 times the energy that the Falcon 9 first stage has at staging. Dealing with that energy during reentry into the earth's atmosphere would require a thermal protection system, a significant entry burn, or both. Either of those would result in a big performance penalty.
The low engine count of Vulcan is also an issue.
Falcon 9 can run a single engine at 40% and that lets it throttle down to 4.4% of full thrust when landing. That gives it a thrust to mass ratio of 1.2 to 1; a little bit too high to hover, which is why it does a "hoverslam" landing.
Vulcan has two engines, so the best it can do if the BE-4 can throttle down to 40% is 20% of full thrust, which gives a thrust to mass ratio of 3.2 to 1. Falcon 9 has done three engine landings at that thrust/mass ratio, so it is possible but much more difficult and Vulcan would also need to somehow deal with the off-axis thrust as it only has two engines.
This two issues together mean that Vulcan is not designed to be able to do propulsive landing.
ULA is planning on implementing a different reuse technology.
Their plan is to put the BE-4 engines into an engine "pod", and jettison that pod after staging. The pod will have an inflatable heatshield to make it through the atmosphere, then it will deploy a parachute that will ultimately be caught by a helicopter. It will then be returned to their factory, refurbished, and flown again.
It's very much like a capsule approach.
There are three main technical challenges to this approach...
The first challenge is the building an engine pod that is easily attached to the rest of the first stage and easily detached when desired. That will require a fair bit of engineering but seems straightforward.
The second challenge is getting the engine pod through the atmosphere. NASA has flown a test mission using an inflatable heatshield to return from low earth orbit, so this also seems practical.
The third challenge is catching the parachute with a helicopter and bringing the pod down to a ship.
RocketLab tried doing this to enable reuse of the first stage of their Electron rocket. They were able to successfully attach but found that the resulting combination of helicopter and stage was not aerodynamically stable and they have since abandoned this approach.
The Vulcan engine pad will be smaller than the Electron first stage but likely 5 times the mass. This part of the architecture will be hard to do in the dark and presumably puts a human crew at risk during the catch operation. I think it is the riskiest part of their reuse approach.
How does Vulcan reuse compare with Falcon 9?
Let's compare Vulcan with SMART reuse to Falcon 9 landing on a drone ship, and assume our goal is to get 150 tons into low earth orbit. ULA claims that Vulcan is optimized for high energy missions - like going to geosynchronous transfer orbit or direct to geostationary orbit, so this will probably show it at its worst...
The Vulcan with 6 boosters and SMART reuse will do around 25,000 kilograms to low earth orbit, and therefore will take 6 flights to lift 150 tons. Falcon 9 has a payload in drone ship mode of about 16,500 kilograms, so it takes 9 flights.
We can now look at what it takes for each rocket.
The Vulcan needs 6 booster airframes - everything but the engine pod. Falcon 9 needs 0.1 assuming you only fly each booster 10 times.
Vulcan needs 36 solid rocket motors.
Vulcan needs 6 second stages while Falcon 9 needs 9.
Vulcan needs 6 recoveries and refurbishments while falcon 9 needs 9
Vulcan needs to integrate the engine pods into the new airframes and test them, and it also needs to ship pods and boosters around.
Looking at the negatives for Vulcan, they need to be building the new booster airframes and putting used engine pods back into them while SpaceX doesn't need to do any of that work. They also need to be buying solid rocket motors for their flights.
SpaceX has 3 negatives. They need to build 3 more second stages, but each of those stages uses a single Merlin Vacuum engine that SpaceX makes while each Centaur V uses two RL-10 engines that they buy off the shelf. RL-10 engines are rumored to cost $10 million each, but ULA has ordered 116 of them from Aerojet Rocketdyne and that is going to reduce the cost a lot, though only ULA knows how much. SpaceX has built a ridiculous number of Falcon 9 second stages and their design is highly optimized.
At best, I'd expect the 6 Centaur V stages to be similar in cost to 9 Falcon 9 second stages but my actual guess is that the 6 Centaurs cost at least twice what the 9 Falcon 9 Second stages cost.
The second SpaceX negative is having to do 9 recoveries versus only 6 for Vulcan. SpaceX needs to get both the drone ship and the support vessels out out in the ocean, while ULA will only need a single ship with their helicopter, but it will be much farther from land. Assuming SMART recovery lives up to its name, I'm going to say that SMART is as cheap as the drone ship which means Vulcan will be cheaper on recovery.
SpaceX also has to refurbish more boosters but their process is simpler and they don't have to deal with shipping. They also have obviously done this a *lot*. I think SpaceX is going to be cheaper even with more boosters to work on.
ULA has been talking a lot recently about how Vulcan is a high energy rocket, which means it is better at orbits that are harder to get to than LEO.
Is that true? We can look at a couple of different scenarios..
Mars gives us an opportunity to do a direct comparison as ULA and SpaceX both publish payload numbers...
Starting at the low end, the Vulcan VC2S configuration with 2 solids can send 3,600 kilograms to Mars.
Just a little bit higher is Falcon 9 expended, at 4,020 kilograms. We are just getting started and Falcon 9 has reached its limit. Falcon 9 is not a great high energy rocket, it can only get 0.7% of its takeoff weight to Mars.
Adding on solids, Vulcan gets 6,000 kilograms with 4 solids, and 7,600 kilograms with 6 solids. These are healthy amounts of payload, and good support for the assertion that they have a design that works well for high energy missions. With 6 solids, they get 1% of their takeoff weight to Mars, and that's about 40% better than Falcon 9.
Unfortunately for this argument, there's another SpaceX rocket, and it can toss 16,800 kg to Mars in fully expendable mode, with lower payload options for side boosters landed on drone ships or back at the launch sites. It can throw about 1.2% of its takeoff weight to Mars.
Vulcan is therefore a nice choice for high energy or planetary missions if they are on the light side but can't compete with Falcon Heavy for big payloads.
We can also look at geostationary transfer orbit performance.
ULA gives us some great information for Vulcan showing its performance to two separate GTO orbits and all the way to geostationary orbit. You can see that the closer you get to geostationary orbit, the less payload you can carry.
SpaceX doesn't publish information on this so the information we have is reverse engineered from what is launched, and here are the Wikipedia numbers. The problem we have is we don't know what the actual destination orbit is - is it the nominal GTO-1800 orbit, or was the customer willing to accept GTO-2000 or wanted GTO-1600?
There are complex tradeoffs at work.
The satellites do the work to get from the transfer orbit to their final orbit and on current generations that takes quite a bit of propellant and time - typically 6 months or more.
Let's just say we have a 5,500 kg satellite we want to launch. On the Vulcan side, we could fly 2, 4, or 6 solids which would get us anywhere from about GTO-1200 or all the way to geosynchronous orbit. More solids costs more but it gets us operational quicker and preserves fuel.
On the Falcon 9 side, we could fly drone ship recovery and be - probably - somewhere around GTO-1800, we could fly expended and be a lot closer, or we could fly Falcon Heavy and get all the way to GEO.
Each option on both rockets has a different price.
Sometimes customers want their satellites in the destination orbit as soon as possible. Sometimes they are launching spares that likely won't be needed for a couple of years.
Factor in how soon you can launch after your satellite is done - where you can fit into a provider's schedule - and that makes things even more complex. And customer requirements may change along the way - they were okay with the cheapest launch but they've had a failure and they're burning cash until they get a replacement up there.
Both Vulcan and Falcon 9 / Falcon Heavy are solid offerings in this market and it's hard to come to any definitive conclusion.
Time for some market analysis. The first big competition for ULA was NSSL Phase 2.
ULA was allocated 60% of the launches from NSSL Phase 2. They have 1 Atlas payload early in 2024, and then the remaining 25 flights will be Vulcan flights. They *do* need to get Vulcan flying as soon as possible as they need 2 successful flights before they can start launching NSSL payloads. Assuming they can do this, NSSL will remain a reliable source of high profit missions for them.
SpaceX got 40% of the launches, using either Falcon 9 or Falcon Heavy
NSSL Phase 3 is for the next batch of launches, and the DoD has decided to change things around by introducing the concept of "lanes".
Lane 2 has the traditional NSSL rules and the traditional 60/40 split, with the 60% going to ULA though it is possible that SpaceX will get the 60%.
Blue Origin did some effective lobbying and the second draft of the phase 3 rules added room for a third provider who would get 7 launches, so the launch numbers are likely 26, 17, and 7, a reduction of 4 launches for ULA and 3 for SpaceX.
There is also this new thing called Lane 1. It provides a way for other providers to bid for lower risk missions and all you need to participate is to get certified and be able to carry at least 1000 kilograms to low earth orbit. This is a short-term program; they will do annual awards. If you are familiar with the NASA launch services program, it's like that.
Lane 1 may start stealing payloads from lane 2 as the shorter scheduling period and likely lower price will be attractive to many programs.
The prices charged for NSSL launches provides an interesting window into how competitive Vulcan will be.
In NSSL phase 1, ULA ended up launching 10 Atlas V rockets at an average of $137 million and 6 delta IV heavy at an average of $422 million.
SpaceX joined in after the lawsuit with 7 falcon 9 launches at $52 million each and 2 Falcon Heavy at $90 million each.
There's obviously a big discrepancy in the prices, but SpaceX very much wanted to get into the NSSL world and low prices made it easier for them to do it.
In Phase 2, ULA needs to compete with SpaceX. They will be flying 26 Vulcans at an average price of $120 million.
SpaceX will be flying 22 Falcon 9 or Falcon Heavy at an average price of $114 million.
The ULA price reduction is drastic - their average Vulcan price is cheaper than their cheapest Atlas V launch, which is a good indication that Vulcan is a *lot* cheaper than the previous options.
SpaceX no longer needs to be the cheapest out there and their previous prices were leaving a lot of money on the table, so they have raised their prices considerably to capture that revenue.
This is what we would expect with a duopoly - without any external competition, the cheaper provider will raise their prices.
Vulcan and Falcon are about the same price, but only in this artificial market...
In the commercial market, things are different. Can Vulcan compete with Falcon here?
I'm skeptical. Falcon 9 is a cash cow right now and I think that SpaceX is reaping solid profits on every commercial flight that they launch. Falcon 9 is regularly reused and SMART reuse for Vulcan is at least a few years away.
There is, however, a market segment that could be quite lucrative for ULA, and that's the "anybody but SpaceX" market segment, which currently means Amazon's project Kuiper constellation.
Amazon has announced launch contracts for Vulcan, New Glenn, and Ariane 6. I also expect that Neutron will be in the mix.
Of those, I expect that New Glenn and Ariane 6 will have issues getting to a high flight rate and that leaves Neutron as the only competitor for Vulcan. I expect Vulcan to do well in this segment.
Vulcan is an important new launcher for ULA.
It fixes their problem of having two rockets, they don't have to fight over Russian engines, and the modern design of Vulcan makes everything easier.
They have a partial reuse approach on the horizon - it's not as good as what SpaceX has in Falcon 9 but its likely the best they can do.
It's a great fit for NSSL launches, and that is the most lucrative market for ULA, so that had to hit that well, and they have.
The open question is whether it's enough. Many people question whether they should have aimed for a design more like Falcon 9 or even a fully reusable design like Starship, but it looks like Vulcan will fly very soon and most of the other new rockets have at least a couple of years before they launch.
If you enjoyed this video, please send me this Vulcan S Café ABS.
Though to be honest, I'd prefer this one instead...
A while back I got the bright idea to ask my viewers for questions that they had about SpaceFlight.
I got 252 top level comments, which was both an opportunity and a problem. How would I deal with so many questions?
The first step was to get them out of YouTube. Nicely - and perhaps surprisingly - YouTube provides a really nice .NET interface.
I opened up Visual Studio, created a C# project, and wrote myself a YouTube comment extractor that would pull out all the top level questions and write them to a CSV file that I could import into Excel.
Then, after about ten minutes of looking at that file, I did two things.
The first was that I broke it into starship and non-starship questions. And I took out the CSV backend and put in a PowerPoint backend, so it would walk through the questions and then create a powerpoint slide for each question.
But even with that automation, I can't answer all the questions. Which makes me feel a little bit bad.
In working on some of the answers, I realized that answering them in a video is a lot of work and that inherently means I'm going to choose to answer fewer questions there. I spent 3 or 4 hours looking at cargo shipping cost models to try to understand how much the SpaceX recovery ships cost for the question that became the drone ship video, and that just doesn't scale.
So I came up with an alternative... I created a subreddit on reddit named, not surprisingly, EagerSpace. I will post video links there - at least when I remember - as I currently do on TwiX, but the real point is that I'm more likely to answer questions there because it's cheap and I'm there quite a bit anyway. And other people can answer questions as well.
I should also state that this is the first of a few videos that answer questions, so if your question doesn't show up here, it may show up in a later one. That's why it's labelled as 2.1.
Onto the questions...
There were four really good questions about Nuclear thermal rockets.
My weasely answer to these questions is that it will depend on what they actually end up building...
But that's not very satisfying, so I'll see what I can do. If you've watched some of my other videos, you already know that I call myself a skeptic when it comes to nuclear thermal rockets. If you haven't', you should go watch those now...
And I should probably enumerate why I'm skeptical...
First, AFAICT, nobody is self-funding development of NTRs, which generally means companies are more excited about the development dollars than they are about the commercial or military applications.
Second, I can read the rocket equation and do simple calculations. Pretty much universally, the nuclear rocket advocates talk about how great the specific impulse is and totally ignore the mass ratio. I complain about this a lot when it comes to hydrolox engines, but it's a lot worse with NTR because the propellant is only liquid hydrogen so the tanks are giant and the engines made out of very heavy uranium are not surprisingly very heavy.
Third, everybody ignores all the other problems. The core is hugely radioactive once you turn it on which means nobody wants to be around it, you want to run it hot to get the highest specific impulse but that means you might melt your core, and you can't actually turn it off when you want to because you have to keep running hydrogen through the core to slowly cool it off. And your engine is low thrust; you basically get RL-10 level thrust out of most designs.
Fourth, NASA cheats whenever they make mission-based proposals; they come up with these gossamer tank designs that weigh nothing but somehow are good enough to hold in liquid hydrogen and they compare it to standard designs.
The problem is that there isn't enough reality.
Which brings us to the DRACO program, which AFAICT is a continuation of the NASA program but is now run by DARPA, and, not surprisingly given the way NASA is running things these days, there's been no status update on the program since it was first started. I've thought of doing a freedom of information act request but my record getting useful information out of DoD is pretty poor.
I'm on record saying that I support the NASA program because I just want someone to build and fly something so we have a real stage that I can talk about rather that basing my opinion on a bunch of less trustworthy sources.
That's not a very exciting opinion, so I will make three predictions...
The first is that DRACO will fly a real stage and test it in orbit, though it won't be in 2027. I'm really going out on a limb here predicting that a space program will be late.
The second is that draco will be unimpressive; it's not going to make anybody want to use it instead of chemical engines. I'm predicting this because the parameters set in the original NASA proposal pretty much assured that the resulting engines will be boring. I'd love to be wrong here and see somebody step up with the advanced technology advocates have talked about for so long, but there are good reasons that we probably won't see that.
The third is that advocates will use the results as a talking point to assert that draco shows the concept has promise and it just needs a lot more money to reach its goals.
Add in a bonus prediction - somebody is going to go crazy about the prospect of flying a nuclear engine. There are lawsuits over simple RTGs, and this will be a full reactor.
I already talked about nuclear thermal so this will be about nuclear electric. Conceptually, it's pretty nice - you use a nuclear reactor to generate a bunch of electricity and then use that to drive a system based on ion thrusters. You need a whole bunch of ion thrusters to get a useful level of thrust but that appears to be possible and they have crazy high specific impulses, which means they are very fuel efficient.
The problem is heat. We generally think of space as a very cold place, but it's really a place where the usual methods of heating and cooling do not work well - no convection or conduction, so we are just stuck with radiation. The international space station has these large radiators to get rid of waste heat.
The problem with nuclear electric is that there is a ton of waste heat. The efficiency of your radiators depends on how hot they are; if you want to use water-based radiators you are going to need a ton of them, like in this NASA design.
The alternative is to use a much hotter fluid, with liquid sodium being a common choice. This has been used in fast nuclear reactors, but they use a very special design.
Sodium will be much hotter and therefore give you smaller and lighter radiators, with the slight downside that if your system springs a leak in any place with air it will immediately catch fire, and if the system gets just a little too cold the sodium solidifies in the pipes.
The concepts after that get even weirder, including liquid droplet radiators where you spray the hot fluid in a stream and then collect it after it has cooled.
As usual, Winchell Chung's Atomic Rockets site has all the best information on radiators.
So my summary is that the nuclear electric heat problem seems like a really hard one to solve and for the inner solar system, solar electric might be a better choice.
That's a fun paper from 1994. The concept is normally known as LANTR, for lox-augmented nuclear thermal rocket.
You take a normal nuclear thermal rocket, and then when you need extra thrust you inject LOX into the nozzle and it functions as an afterburner. It's a multi-mode rocket - low thrust and high specific impulse with pure hydrogen, high thrust with lower specific impulse with oxygen, or in between. The big issue I see - beyond the usual NTR disadvantages - is that with a lot of oxygen the specific impulse is down to only about 500 and you have an engine that weighs about 7 times what an RL-10 hydrolox engine weighs but gives equivalent thrust, which means you get a worse mass ratio and therefore less delta v.
I don't see much benefit over a hydrolox engine.
When I was 5 years old, I wanted to be an astronaut. By the time I was 12, I switched to engineer which seemed more practical, especially given my eyesight.
Then I went to a high school in 1978, and had a math teacher who was ahead of his time and got exposed to an Alpha Microsystems computer with ADM3A terminals, and the die was cast.
Programming was both easy and fun and it was pretty clear that I was going to make it a career. I have a BS in computer science with minors in math and business. I worked at a number of places during my career, including a long stint at a "major redmond software company".
So the obvious question is, "How do you know all this space stuff?"
The first explanation is simply time. If you are interested in a topic and you study it for 4 decades, you tend to accumulate a lot of useful information. Or at least I do.
The second explanation is that I'm a generalist. I do have in-depth knowledge in some areas, but I have a decent level of knowledge in a bunch of areas. For this channel, I know a fair bit of physics, chemistry, engineering, material science, business, organization-ology (not sure what to call that), and psychology.
That gives me the ability to see connections, which is good. It also means I run out of depth in a lot of areas - I know a little about rocket engine design but nothing compared to the people who do it for a living.
There's another old word that kindof applies, and that's polymath, in the sense that I have wide ranging interests and know a lot about a bunch of topics. Like I know what a polymath is....
If we look at the polymath scale, I'm down here, a long way from the greats like DaVinci, Franklin, and Martin.
(read)
There can only be one choice. I think it's both the safest and the coolest.
If you watch my video about how spacex might have failed, I talk about it there.
The short answer is that President George W. Bush wanted a legacy and he chose "a really cool space program", and that required cancelling shuttle because there wasn't enough money to do shuttle, the space station, and the big exploration program.
Turns out there wasn't enough money after shuttle was cancelled, but that's another story.
It was funky and different and probably the worst thing I did in terms of getting subscribers to the channel.
But take a look at the end of this video and tell me if you think it works as an outro...
Unlikely.
SABRE is a concept for a hybrid mode rocket engine where you take in outside air, cool it down a bunch using a heat exchanger, and then that cooled air is injected into the rocket engine and burned with liquid hydrogen.
The concept has been around in various forms since the 1950s.
The reason I think it's unlikely is that a) they were initially talking about a single stage to orbit vehicle and that made very little sense and b) nobody really wants to put money into it. That second one is usually a good indication of problems with a concept as companies have to disclose a lot of information to get funding.
Note that the opposite is not true - people putting money into a project is not a sign that it's a good bet.
I've also found it strange that reaction engines was targeting sabre towards single stage to orbit when that is so obviously a really, really hard thing to do.
It would make more sense to target it towards hypersonic flight where the constraints are easier and you may garner interest and money from the military.
If you go to the reaction engine website you can find information about SABRE if you dig deeply enough, but their tagline is now "thermal management for a cooler, cleaner world".
Probably not.
But if you are asking whether the Chinese will beat Artemis to the moon, I think it's pretty likely. One of the reasons that Apollo succeeded was that the program was deeply tied to American pride. The Chinese are very proud of their space program and have - from what I can tell - a decent moon program.
Artemis is this weird program that came about because congress told NASA to build a big rocket and work on a capsule and ended up spending $45 billion on it and it seemed like going back to the moon was something you could do with it. That's why we're in this weird situation where Orion can get to lunar orbit but there's no NASA-developed lander and the architecture to use starship and Blue Moon is so weird.
NASA of course has wanted this for years, and DARPA is working on their LunA-10 architecture.
For anything like this to happen, we need a combination of two things.
The first is cheap enough transport for people and cargo to the lunar surface. I'm not sure what cheap enough actually means, but starship might get you there, and the Blue Origin mark II lander might get you there.
The second requirement is for there to be somebody who cares and can pay for it. And that somebody pretty much has to be congress - NASA cares but they don't control their budget and DoD cares but they would also need congressional approval.
I'm kindof thinking the answer is "no", but it's going to depend on how congress reacts to China getting there and how cheap transport ends up.
As for economy, there's lots of talk about a cis-lunar economy but nobody has a real application where it would make sense, and by that I mean a market that somebody will actually pay money for.
Rotating detonation engines have been getting a lot of hype. A video on them is on my list but I'm only interested if I can answer these questions and I don't think there's enough data yet.
Is it a real innovation, or is it another aerospike? I don't know.
I tend towards skepticism, but I could be wrong.
I suspect the answer you are looking for is "in the works", and I'm afraid I'm going to have to disappoint you.
When I released the propellent depot video, I was already at work on two other videos and those took priority for a few weeks. Then when I came back to look at an HLS video, I ran into some immediate issues.
The depot analysis was pretty simple to do and I could come up with reasonable approximations. It's a big round tube of metal and the factors are reflected sunlight, infrared from the earth, and direct sunlight.
The HLS lander is a lot different
(read)
One problem at the time was we had one render of what the lander would look like without much detail and SpaceX was - and still is - quite tight-lipped around what the lander will actually look like.
The second problem is the last one listed. The lander will get heated by the sun, presumably coming roughly from the side because it's landing near the south pole, and it will be heated by sunlight reflected from the moon's surface and also by infrared coming off the surface. There were things there that I don't know how to model, that differ depending on where the landing site is and when the landing occurs, and I suspected that the things I didn't know would have a big impact on the result.
So, at the time I decided I wasn't really up to it. And in the last month or so, I had a discussion with somebody who knows more than I do and based on the approach they said they would take, I'm more sure that I'm not up to it.
This was a single comment but there are two distinct questions.
(read question)
RTG means radioisotope thermoelectric generator, a way of using the heat from decaying radioactive materials to make electricity.
There are two big factors. The first is that RTGs are a big pain in the butt to deal with from a legal standpoint. This paper says the following:
(read)
Even after that process, you will very likely get sued and have to deal with that.
There is also both national and institutional policy that says that radioactive solutions can only be used when they are a big win.
NASA of course has a study from 2019 that compares missions that used radioactive or solar power sources.
This graph shows the mass of the electrical power system compared to the mass of the whole spacecraft bus (masses are compared without payload mass). The nuclear (RPS) missions are coded in blue. I mostly expected the solar ones to be heavier because solar panels are heavy, but the radioactive elements used are very heavy and they tend to have heavier spacecraft busses.
That probably has an effect on launch cost or the mass of the payload.
This second graph looks at the cost of the electrical power system, and it turns out the nuclear systems are considerably more pricey. Or, to put it another way, you spend more of your money on power than if you used solar.
Solar is a pretty obvious choice for inner solar system missions, but I was surprised to find this result from the Europa clipper mission all the way out to Jupiter, where solar won pretty resoundingly.
I do expect that missions to Saturn and beyond will still rely on radioactive power sources because the sun is so dim at those distances.
For Mars you can get away with solar panels and batteries. The lunar night is 14 days long, so you will need a lot of batteries to make solar work.
In either case, I think you'll be a lot happier if you have a few kilopower reactors to give you a constant power source.
I read a lot, so this is a really hard question to answer.
In the non-fiction space realm, I recommend
Escaping Gravity by Lori Garver, which is a great look inside NASA during commercial crew
Liftoff by Eric Berger, which looks at SpaceX during the Falcon 1 era
I'm going to "pre-recommend" reentry, which is Eric Berger's book about Falcon 9, scheduled to release in September of 2024.
Everybody should read Rand Simberg's "safe is not an option". It's very insightful and the kindle version is only $5.
And finally, Apollo Remastered is just a stunningly great photography book about Apollo. If you've ever wondered what a talented photographer could do with the Apollo originals, this is the book for you. It's not cheap but you really want the hardback.
I'll also recommend the classic "surely you're joking, Mr. Feynman" and the sequel is pretty good.
James Burke's Connections is also a classic, along with "the day the universe changed".
For science fiction, I have three series that I think you might like.
Trading in Danger by Elizabeth Moon, Valor's choice by Tanya Huff, and Live Free or Die by John Ringo
But that didn't really answer the question...
My favorite book is pride and prejudice by Jane Austen
If you prefer to watch a video, by far the best version is the 1995 BBC miniseries starring Colin Firth and Jennifer Ehle (ee-lee).
I might.
I've heard that this playlist is okay...
Happy to share my process which I think should make the answer clear.
This is my main topic board. I haven't looked at that for a while, and there a few that no longer should be there.
The tear in the paper is because of my cats jumping onto the top of the door and then tearing things when they slide off.
That leads to the overflow of topics on my workbench, which are post-its that my cats knocked down or fell off on their own.
And then, finally, what is closest to my current backlog is another set of post-its on top of my desk, though the main pile was actually hiding under some envelopes.
The SpaceX 1990 and LunA 10 Darpa topics are new ones that came out of the viewer questions. The droneship landing video came out of the questions as well and there are a couple more questions that might make their way to a separate video. I'm not sure yet.
I'm generally working on one video, with maybe a second one in introductory stages when I get tired of the first one. Sometimes I'll reach a point where a video just isn't working, so I'll put it to the side and maybe pick it up later.
While I'm finishing this one I'm working on a new one about the future of NASA and working on another one based on questions.
That's a short way of saying that my process is very ad-hoc and therefore I can't answer the question.
No way can I equal Tim Dodd's video "Rocket engine cycles: how do you power a rocket engine?"
Go watch that.
This video is a little dated but I mostly still agree with what I said here...
I have been thinking of contrasting the reuse approaches of Falcon 9, Neutron, and Starship, and that might show up in a video.
Are we done with the questions?
Yes, yes we are.
If you enjoyed this video, I could do with a cold beverage.
I can't drink the real stuff very often because I'm kindof allergic to alcohol, but this is a decent substitute IMO.
nd that's why starting a launch company is a bad decision...
Spaceplanes are great.
Who wouldn't want to start on the ground and take a single vehicle all the way into orbit?
The media has written many articles about spaceplanes and how great they will be.
And yet, none of the spaceplane designs have made it to actual hardware, though a design without wings did:
(show video)
The DC-X demonstrators made 12 flights in the 1990s. The highest flight reached 3 kilometers in altitude, but an explosion after landing on the 12th flight led to the cancellation of the program.
Where are the spaceplanes?
This is...
To understand why there aren't any single stage spaceplanes, we need to understand how fuel capacity, payload, and range work for rockets.
For cars it's pretty easy...
The Citroen 2CV requires 3 liters of gas to travel 100 kilometers.
It has a fuel tank that holds 25 liters, and a little bit of math tells us that it can travel 833 kilometers on one tank of gas.
That's an impressive range that covers all of France, the Netherlands, and Switzerland, but you can't get to Copenhagen, Berlin, Prague, or Madrid.
That's carrying only a driver. If we were carrying more payload - 3 passengers for example - our mass would go up and our range would go down.
For space, we don't use kilometers as a measure of distance; we use a measurement known as "delta v", which is measured in meters per second. Sometimes we write it using the lowercase greek letter "delta".
To get from the surface into low earth orbit requires around 9400 meters per second of delta v - that's the distance we need to cross.
Like a car, a specific rocket has a delta-v capability that varies based on the amount of payload it is carrying. This Soyuz FG rocket can carry 6,900 kilograms to low earth orbit.
If you want to know more about delta v distances in the solar system, see my video "planning your solar system road trip".
At this point to talk about the physics of rocketry.
I'll be doing all the math and the physics will not be on the test, so just pay attention to the basic concepts.
It is possible to estimate the amount of delta v that a rocket can generate using the rocket equation.
If you want all the details of the rocket equation, go watch my video titled "the care and feeding of the rocket equation". Here are highlights.
The first term of the equation is the specific impulse, often written as ISP. The specific impulse depends on the design of the engine and the propellants that it uses. A higher value is better as that results in a higher delta v - the rocket can travel farther on a given amount of fuel. Its like the fuel economy of a car.
The second term is known as the mass ratio and has two parts. We take the initial mass when the rocket is fully loaded with propellant and we divide that by the final mass when all the propellant is gone. The ln term is the natural logarithm, which is there because physics.
We want the mass ratio to be as high as possible, which mean carrying as much propellant as possible and making the empty stage as light as possible.
Higher payload increases both the initial mass and the final mass, reducing the mass ratio and resulting in less delta v. Like our car, if we have a heavier load we can't go as far.
It's common to find people asking whether a specific rocket stage can make it into orbit as a single stage. Let's look at the Falcon 9.
Note that all the numbers I'll be using are estimates and I'm ignoring some other factors, so the conclusions we reach are for entertainment purposes only.
The specific impulse of the merlin engines used on the first stage is 297.
The final mass is the mass of the empty stage, which is 27,200 kilograms.
The initial mass is the final mass plus the mass of propellant, giving a total of 445,900 kilograms.
Plug that into the rocket equation, do the math, and we end up with the number 8,150 meters per second of delta v. That does not reach the threshold to get into orbit, so this stage cannot operate as a single stage to orbit vehicle.
It's tempting to look that delta v number and say, "8150 is pretty close to 9400 so we should be able to get there". This is not the case - this stage would need to add 60% more propellant or reduce the final mass by 60% to get to 9400 meters per second. That's just not possible.
Looking at the two-stage version, we can use the second stage numbers with it carrying a payload of 15,600 kilograms and determine that the second stage can generate 6,500 meters per second of delta v.
The first stage carries the fully-fueled second stage as a payload, and when we run the numbers we find that it can generate 3,770 meters per second of delta v, for a total of 10,270 meters per second of delta v. That's enough to get into orbit plus leave enough extra for the first stage to return to earth and land on a drone ship.
Our single-stage version couldn't get into orbit at all, and our two-stage version can carry over 15 tons into orbit.
That's why orbital rockets pretty universally use at least two stages.
But the point of this video is to talk about spaceplanes, so we'll need to define the problem in more detail.
If we are going to build a single stage to orbit rocket, we will need to somehow squeeze 9400 meters per second of delta v out of it.
We can plug that into the rocket equation, and then solve for the mass fraction, and we get this equation because math. E is a numerical constant equal to about 2.72 and shows up as the inverse of the natural log. Note that the mass fraction depends on the specific impulse.
Using that equation, we can come up with a graph that tells us what mass ratio we need to achieve for a given specific impulse.
The higher the specific impulse, the lower the mass ratio required to get the delta V we need to get into orbit.
We can add some sample engines here.
The SpaceX Merlin used on Falcon 9 burns RP-1 fuel, and has a specific impulse of 311. That translates to a required mass ratio of 21.8.
The SpaceX Raptor used on Starship burns liquid methane fuel and has a specific impulse of 363 and requires a mass ratio of 14.
The RS-25 used on the Space shuttle and SLS has a specific impulse of 431 and requires a ratio of only 9.3.
The high specific impulse of liquid hydrogen appears to be the obvious choice for a single stage to orbit design. To explore further we need more specifics, and that means we need a rocket...
We can use the SLS Core Stage as an example, as it conveniently uses 4 RS-25 engines.
The information we need to know about the SLS core stage is published but it's a little squishy. These are the best numbers I could find.
The empty mass of the stage is 99,300 kilograms.
The stage holds 979,500 kilograms of propellant.
We can compute the mass ratio using the following equation.
Plug the numbers in, and we get 10.9. Our threshold for this fuel and engine choice was 9.3, so this scenario gets a big green checkmark as the mass ratio is high enough to get us into orbit. We can do a few more calculations and figure out that we could carry 17,500 kilograms of payload.
This looks great...
But it turns out that we forget something...
The RS-25 engine creates a thrust of 1.86 Meganewtons at sea level. Newtons is the proper unit of force in the metric system, but for our use kilograms of force is more convenient, so a quick bit of math shows that this is 190,000 kilograms of force.
There are 4 engines on the core stage, so the total is 7.44 meganewtons or 760,000 kilogram of force.
We know from our earlier numbers that the total mass is 1,078,800 kilograms, so we can do a quick division and and we find the thrust to weight ratio at launch is 0.65.
It turns out that our single stage to orbit rocket lacks the thrust to get off the pad.
What we need is more engines. Seven engines pushes the thrust up to a little over 1.3 million kilograms which looks like it's in the right ballpark.
There is a downside - each of these new engines adds about 3177 kilograms in mass, for a total of 9531 kilograms. We also need to make the stage bigger to hold those new engines, so I'm going to arbitrarily assume that changes to the stage to support the additional engines costs the same mass, so our total mass increase is 19,062 kilograms. That increases our empty mass.
Run the numbers again, and we find that our mass ratio is reduced to 9.28. Our thrust to weight is 1.2, so that's good.
At this point we have a marginal single stage to orbit vehicle. We can probably get it into orbit but without any payload.
Note that so far we've just been talking about an empty stage, but if we are building a spaceplane, we need to add in wings, landing gear, and some sort of thermal protection system.
I wasn't able to find any proposal that gives overall vehicle masses so it's hard to know how much it would add. The space shuttle orbiter was made of aluminum and massed about 78,000 kilograms and it was considerably smaller than the stage we are considering here. Titanium would be lighter if you could afford to use it; carbon fiber could be very light but would make the thermal protection system heavier.
There's a problem with developing single-stage-to-orbit vehicles, and that's the impact of increases in mass.
Let's say that our target mass was 90,000 kilograms for the vehicle and a 10,000 kilogram payload.
But it turns out that our estimates were low, and our stage is 5% heavier. Every kilogram of extra mass in the vehicle is one less kilogram of payload, and that 5% increase nearly cuts our payload in half.
If we're running a two-stage rocket and our second stage gets 5% heavier, we only lose a small amount of payload because the second stage doesn't have a lot of mass. Most of the mass is in the first stage.
If the first stage gets heavier, it doesn't matter than much because the first stage does not go into orbit; in our scenario here a 1360 kilogram increase in empty mass only results in a payload reduction of 200 kilograms - it takes an extra 6 kilograms in the empty mass of the first stage to reduce the payload by 1 kilogram.
Keeping the single stage to orbit vehicle light is absolutely critical, as 10% mass gain could eliminate all the payload. That means expensive materials and expensive development, and even with that it's not sure that the vehicle will have meaningful payload when it's finished. That makes it a risky investment.
It's much less important for two-stage vehicles. There isn't much mass in the upper stage so growth there doesn't matter much, and mass growth in the lower stage isn't very important either. Extra mass matters but it's unlikely to be enough to kill the program.
Are there any ways we can cheat and improve this situation? The answer is yes...
What we need is more engines. Two extra engines is marginal, but if we add 3 engines, that pushes the thrust up to a little over 1.3 million kilograms which looks like it's in the right ballpark.
There is a downside - each of these new engines adds about 3177 kilograms in mass, for a total of 9531 kilograms. We also need to make the stage bigger to hold those new engines, so I'm going to arbitrarily assume that changes to the stage to support the additional engines costs the same mass, so our total mass increase is 19,062 kilograms. That increases our empty mass.
Run the numbers again, and we find that our mass ratio is reduced to 9.28. Our thrust to weight is 1.2, so that's good.
At this point we have a marginal single stage to orbit vehicle. We can probably get it into orbit but without any payload.
The Falcon 9 uses subchilled propellants that have a higher density so they can stuff more propellant in a given stage and therefore improve our mass ratio.
We can do that with our stage - we can subchill the liquid hydrogen and liquid oxygen so that they are approximately 8% denser, and that will allow us to put more propellant in our tanks, and that will improve our mass ratio.
If we can get 8% increase in density for both the liquid hydrogen and oxygen, we would theoretically get 9,500 kilograms of payload increase. However, the increased liftoff mass will either require an 8th engine that makes the stage heavier or will result in increased gravity losses that will consume some of that payload increase. It's worth doing but it does have some side effects.
That is assuming that it's practical to use subchilled hydrogen. NASA has done some research but normal liquid hydrogen is already very challenging to deal with and nobody has built a rocket or engine that uses subchilled hydrogen.
If it's only possible to subchill the oxygen, the increase in payload is much smaller, only 3000 kilograms.
Radian Aerospace has an interesting idea.
Rather than launch the spaceplane from a runway, you can put it on a rocket-powered sled that accelerates it to approximately 200 meters per second before releasing the plane.
That by itself reduces the required mass ratio from 9.3 to 8.8.
That will make the spaceplane much easier to build, but the rocket sled limits your launch locations, requires more complexity during the development process, and failures during the rocket sled run could be problematic.
Reaction engines has an engine concept known as SABRE that they are hoping to use in their Skylon spaceplane.
SABRE is a hybrid engine. At low altitude it functions like a jet engine, burning liquid hydrogen with oxygen pulled out of the air, and in vacuum, it burns liquid hydrogen and liquid oxygen.
It's an interesting concept...
Being able to burn atmospheric oxygen provides a great advantage at lower altitude and results in a very high specific impulse.
But the hybrid engine is going to be much heavier than a conventional rocket, and the atmospheric oxygen content drops off quickly as you go higher, so the weight penalty will hurt the mass ratio when in rocket mode.
It's also a complex system, having to deal with hypersonic airflow, liquid hydrogen, liquid oxygen, and a complicated engine system.
Of the cheats, subchilled propellants are useful and sled launch and SABRE are maybe practical, maybe not.
I'm sure somebody is wondering about starship. It's already equipped to come back and land, so can we operate it in single stage to orbit mode?
Because the Raptor engine burns liquid methane, it has a considerably lower specific impulse than the RS-25, so that means Starship will need a much higher mass ratio, all the way up at 14. That seems like a significant disadvantage - we chose the RS-25 because the high specific impulse meant it required the lowest mass ratio.
But there's more going on here. The starship tanks are only about half the length of the SLS core stage tanks and are roughly half the volume.
Despite that, starship manages to stuff 20% more propellant mass into its tiny tanks than SLS can stuff into its tanks.
The problem is the low density of liquid hydrogen. The low density of hydrogen is one of the main factors that results in a higher specific impulse but that low density also means that the SLS tanks are roughly 2.3 times the size of the starship ones, which makes the empty stage heavier and reduces the mass fraction. If we look at the propellant density we see that it is much higher in starship.
That's not the question you wanted an answer to, but unfortunately starship is still in development and we don't have good numbers for the mass of the production version or even the current test version, so I can't give a yes/no answer to whether it can make it to orbit.
But I can talk about a few scenarios...
Based on the amount of propellant I used on the last slide, if the starship empty mass is 120 tons and there is no payload, it can't get into orbit or land.
If its mass is 80 tons, it can get into orbit with a 10 ton payload but will not have enough fuel to reenter and land, and if it has a mass of 70 tons, it can carry 10 tons into orbit and return to land.
So if it gets light enough, it can sorta maybe function as a single stage to orbit vehicle. That 70 tons seems quite low, but SpaceX has talked about stretching starship to make it taller so it could carry more fuel, and that would improve the mass ratio.
However, if starship is light enough to carry 10 tons into orbit by itself, it will be able to do around 200 tons when launched by super heavy so the SSTO version doesn't seem very exciting IMO.
Why don't we have spaceplanes?
It's pretty simple... It's because physics hates humans.
It's hard enough to build a two-stage rocket that can make it into orbit when you play on normal mode, but trying to build a single stage vehicle that can land back on earth is playing on legendary.
It's just too hard to do and has become even less interesting with SpaceX showing how to reuse the first stage of a rocket.
If you enjoyed this video, please join the movement to repeal the rocket equation
Pad 39A at Kennedy Space Center was built for Apollo, heavily used by the shuttle program, and then leased to SpaceX for their exclusive use in 2014.
The relationship between SpaceX and NASA has been good. SpaceX launches crew dragon missions from this pad, and this is currently the only pad where SpaceX can launch those missions.
Then SpaceX started building a starship launch tower at 39A, and the role of the complex became more complex.
NASA has expressed concern that an explosion of Starship could damage the Falcon 9 launch infrastructure, which - until Starliner flies - is the only way of getting astronauts to and from the space station.
Different news outlets have viewed the situation differently, and I've seen lots of opinions. The opinion most often expressed is that NASA has told SpaceX that they can't do that. .
"Who is in charge Pad 39A?" Is it NASA or is it SpaceX?
The real answer is "it depends on what NASA and SpaceX agreed to when they signed the lease agreement". Without that agreement, all we have is speculation.
And here's the lease. I got it through a FOIA request that only took a single day, with the NASA response saying "that document is already public on our website - here's the link".
A link to an file that is - afaict - not indexed or linked anywhere is a strange definition of the word public, bringing to mind a section of the hitchhiker's guide to the galaxy, where Arthur Dent finds the plans to destroy the earth on display in the bottom of a locked filing cabinet stuck in a disused lavatory with a sign on the door saying 'Beware of the Leopard."
The full lease is linked in the description.
I *will* attempt to answer the question I posed, but I'm going to highlight some other parts of the lease along the way. Let's dive in.
The lease consists of 67 pages of exciting legal verbiage. It's similar to apartment leases that I've signed in the past, and my guess is that it's pretty typical for a commercial lease. For example, NASA can enter the premises only during normal business hours after 48 hours of notice, except in cases of emergency. And there's a strict rule against keggers in the flame trench.
Before we get to the main question of who is in charge, I was wondering how much SpaceX is paying to use Pad39A every year.
Here's the answer - SpaceX reimburses the government for the costs that NASA incurs due to SpaceX using the premises.
SpaceX does have to carry a $1,000,000 in liability insurance, but it's very likely that they carry much more than that.
That seems like a rotten deal for NASA - they don't get anything extra from the lease of one of the premier launch pads in the world.
They do get a small fiscal benefit - they no longer need to spend $1.2 million per year maintaining the pad, but that is a small fraction of the amount they spend on ground systems.
The real reason why they do leases like this is that the Space Act of 2010 says the following:
Congress declares that the general welfare of the United States requires that the administration seek and encourage, to the maximum extent possible, the fullest commercial use of space.
That is the basic guideline and shows up verbatim in the lease.
There's an interesting question about how NASA has managed to skirt that language in the past, but that's another video.
With the retirement of shuttle and the expected flight rate of SLS, NASA did not need both pads in complex 39 and decided that they could encourage commercial use of space by leasing pad 39A.
NASA went with a "clean pad" design for pad 39B for the same reason, so that they could share it with commercial companies. Orbital ATK was planning to launch their OmegA vehicle there, but that was cancelled when Northrup Grumman bought Orbital ATK, and there are currently no commercial users of pad 39B
A requirement of the lease is that SpaceX actually uses it for commercial activities that align with the reason that NASA was leasing it in the first place. And they need to do it on a timely basis - the lease mentions "expeditious conduct of commercial launch and reentry activities on the premises"
This is likely one reason why SpaceX was chosen for the lease over Blue Origin - they simply had a plan to use the launch pad earlier. Blue Origin filed a formal protest which was denied by the government accountability office.
The lease contains a list of things that SpaceX can do with the site...
Design, manufacture and test hardware. I've highlighted the phrase "necessary or desirable". I am not a lawyer, but my interpretation is that this means that SpaceX gets to decide what uses they want to do, rather than having to meet some definition of "necessary".
They can do training of crew and participants.
They can run mission control and any other support activities
They can design and construct improvements, once again "necessary or desireable".
They can conduct launches and reentries. I'm not sure what "appropriate coordination and approvals means", but I think it means they need to have the usual FAA launch approvals and coordinate with NASA.
And if SpaceX and NASA agree, additional uses can be added to this list.
Note that there is no language here that gives NASA the power to decide what rockets will fly from this pad, and the lease does not mention Falcon 9 or Falcon Heavy by name.
Note that "kegger in the flame trench" is not an allowed use, though I guess it could fall under "administrative support activities".
Who has priority if there is a conflict?
It's pretty clear that NASA gets priority if there is conflict in scheduling, assuming NASA *wants* priority, and SpaceX has to suck it up and deal with any ramifications of that.
In addition, SpaceX has to manage what it does to not get in the way of NASA's programmatic needs or operations.
My guess is that somebody is going to make the argument that commercial crew is a programmatic need and therefore SpaceX can't do anything that will get in the way of it, but I think that is a stretch. This section is not about allowed uses, it is about priority. It's about NASA saying "we need SpaceX to delay doing
What if something blows up?
If the premises are damaged by fire or other casualty, SpaceX will need to rebuild them to the state before the damage.
However, if it's in the last 5 years of the agreement, SpaceX can terminate the lease and walk away, only repairing the NASA property that was damaged.
So, who is in charge of pad 39A?
SpaceX has pretty much complete control of what goes on there, subject to a lot of reasonable limitations you can read about if you'd like.
But it's not quite that simple...
There is a section on cooperation. The second paragraph basically says that NASA and SpaceX agree to try to not get in each other's way. A starship accident could certainly get in the way of commercial crew flights at 39A, and there is therefore an argument that this section compels SpaceX to move those flights.
But note that the language says a partner must minimize, not eliminate impacts or all interference. And I'm not sure that commercial crew is considered a NASA activity since SpaceX is contracted to provide crew transport. And remember that Starship is *also* a NASA activity because of the HLS tie-in.
But it's probably a moot point...
SpaceX can only fly 6 starship flights per year from boca chica in Texas.
To fly starship often - and they want to fly it all the time - they simply must move as many Falcon flights as possible from pad 39A to other sites.
There are two kinds of flights that currently require 39A.
Crew and cargo dragon require the access arm that is only present at 39A, but those could move to space launch complex 40 on cape Canaveral air force station if SpaceX builds a crew access arm there, as they are said to be doing.
Second, falcon heavy flights are only supported at pad 39A, but some of those flights - those that are to the moon or planets - might be possible from the future falcon heavy launch location at Vandenberg in California. That would involve a slight reduction in delta v - about 400 meters per second - which might not be practical for some payloads.
Regardless, there aren't *that* many Falcon Heavy flights around, and obviously none involving crew.
Given SpaceX's starship flight goals, they need to deconflict pad 39A *anyway*, and that will eliminate the scenarios that NASA has expressed concern about.
And that's the story. SpaceX is in charge, but their interests are aligned with NASA's.
If you enjoyed this video, please write a two-page, single-spaced essay supporting the idea that "kegger in the flame trench" is an important administrative process.
hello welcome to who will compete with spacex a talk about markets and moats. To review a little history spacex first flew the falcon 9 in 2010, they first landed a booster in late 2015 and they did their first crewed launch in 2020 so they've been really flying this booster for at least 10 years
The obvious question is where's the competition? Why isn't anybody else doing the kinds of things that spacex is doing and that's one form that this question comes up in the other form is an assertion that a specific company will be competitive with spacex as soon as they are flying without without a lot of reasons
So I thought it would be interesting to look at the sort of advantages that spacex currently has and has had along the way and do a business analysis and this is the same sort of business analysis I'd do if i was looking to invest in a company, you know if i was looking to buy their stock i want to understand what their competition is.
I'm going to start by talking about moats and motes are just a fun way to think about competition so here's a nice moat in tokyo and on the left side we have a castle and then we have a big moat in the middle and on the right side we have the other side of the moat.
In terms of this analogy the company is the castle and the competitors are off on the other side of the moat and for them to be able to compete with the company they must cross the moat to the other side, so conceptually a moat is anything, any advantage that the company has that makes it hard for competitors to compete with them.
There are different kinds of moats. some moats are big and really hard to cross; there's a lot of water there and it's really hard to get to the castle on the other side.
some moats are very easy to cross so just because there's a moat does not mean that this is a significant impediment to competing with a company so let's start by looking at existing contracts.
Spacex has a number of existing contracts:
they have a contract for commercial resupply of iss and that contract is for nine flights running from december 2020 so last year throughout through October 2023. That might be extended - it's likely iss will be extended so this contract might be extended or it might be a new contract at the end of that period and those are costing around 150 million per mission.
There's commercial crew - spacex in the initial contract has up to six flights we don't actually know how many nasa has chosen but it's pretty likely they've chosen six and then once again a chance for renewal after that, and those are probably 200 million plus
per mission. Obviously flying people in the crew version takes more work than flying cargo in just the cargo version.
Then finally there's national national security space launch and spacex gets 40% of these launches through 2027. ULA gets the other 60 so they were the two winners and they split up 60/40 the award, so that'll be somewhere around 12 or 15 launches as part of this award.
Spacex also got 316 million dollars to develop the ability to launch vertically integrated payloads at pad 39a - some of the NSSL payloads don't tolerate being down on their side they need to be vertical so this is the government paying spacex to build themselves a new capability so it's another nice thing that often comes with these sort of things.
Then they have an award for a couple missions kind of at the start of this time period.
None of these launches are open to competition at this point so you can add that up and there are 20 tomaybe 25 flights here that nobody else is going to fly them they will only be flown on spacex unless something weird was to happen, and with the chance of extending all ofthem and you can expect that spacex would be a significant competitor in rebids for any of these.
This is a pretty significant moat and all of these pay more than spacex's commercial rate - if you came and said to spacex i just want you to launch a satellite, it's not going to cost you this much, this kind of money. Part of that is for additional services so for the space station launches obviously they need a dragon spacecraft and that costs some money to build and refurbish and fly and for NSSL there's a lot of extra paperwork actually for both of them there's a lot of extra paperwork but still these are nice lucrative contracts.
If you want to compete with spacex you can't expect to pull any of these contracts away - they aren't launches you can compete on.
There's also what's called the nasa launch services program, and nasa categorizes launch vehicles based upon their flight history, so you can join this program and when you first join with the new launcher they'll say okay you have no history you don't have any previous fights so that's how we class you with no history.
You fly for a while and i list one of the criteria here there's actually a a long series of things that you have to do to qualify with limited flight history but you have to have flown some that puts you in the limited flight history category, and then finally if you've flown a lot you get put into the significant flight history category and the requirements to get in there just keep going up once you get there.
The reason nasa does this is they want to both encourage additional commercial providers to develop launchers and they want to make sure that their most important missions go on the launch systems that are the most reliable ones so from the nasa perspective if i'm a program in nasa and i need launch services, they will do what's called a payload class evaluation, and class a b c and d are the simplest ones and nasa calls them high risk. What that really means is they can accept high risk, so they're saying this program we're willing to accept launching on a rocket with no flight heritage even though we realize that there might be a problem, and then as we get to the more important ones - the b and c ones - then you need at least some fight history before nasa will allow you to bid on that sort of thing.
Then if you get up to what nasa would call the flagship sort of flights - things like james webb space telescope, anything that people are likely to hear of - that will be payload class a and you need significant flight history to be able to fly that.
You can only bid on payload classes you're certified to carry and spacex currently has a certification for class A, so they can pretty much fly everything so that means if you want to compete with spacex for launching nasa payloads you need to be able to fly enough to get into that significant flight history, or at least if you want to compete with all of those flights
The next topic is looking at fixed costs and how that relates to your flight rate. Every launch company has fixed costs - you need a factory where you build rockets, you need some sort of launch pad you either own or lease, you have a recovery fleet if you're doing reusable rockets and then you have the salaries for all the people for engineering and operations.
You need to figure out how to pay for these and you need to pay for these if you fly once a year and you need to pay for these if you fly 20 times a year
Let's assume you're in a company and your fixed costs are about 100 million dollars a year, and let's further assume you're doing a new launcher and you think initially you're going to fly a couple times a year, so what does that mean?
Well that means the amount of fixed cost allocated to each flight is 100 million divided by 2 or about 50 million per flight. If you think of this from a profit perspective for each of those fights you need to pay for the rocket, you need to pay for the 50 million, the part of the fixed costs allocated to this launch, and once you've done that then you can pay for profit. What that means is these fights are going to cost a lot because you aren't flying very much.
If for example you're at a company that flies 10 fights per year, we're able to spread out that fixed cost across far more flights, so 100 million, 10 flights, 10 million per flight, so very basically the more you fly, the less your fixed costs apply to each launch and the less you can charge for each launch.
This makes it very useful for an existing launch provider who flies; they already know what their fixed costs are and if they fly a lot, they can allocate them across a lot of flights.
If you're trying to be a new company you have to try to spread them across a relatively small amount of flights and that limits the amount you can charge or i guess that actually increases the amount of charge
The next topic i'd like to talk about is what's called the first mover advantage, and first mover is really really misunderstood.
The concept of first mover is that if you are the first company to go into an area, you end up with this advantage but being first is not an automatic advantage.
Being first may give you opportunities to take advantage of the market opportunities that exist now and those market opportunities will not exist for later entrants, so by being first i can get an advantage now,
It may not be a durable advantage but it may also give me a chance to establish moats, so just in the very obvious case it may give me a chance to work with a specific set of customers and establish them as regular customers.
Now can other people take those customers away?
Well, definitely and the ability of them to do that depends on all the other factors that i'm talking about.
First mover does apply in the case of spacex and i want to do that by talking about the launch markets that
What did it cost to launch in 2010?
These are rough estimates - all the numbers i show you are going to be rough estimates but they will get the basic idea across the arion 5; upper berth it's about 108 million and the lower one is about 72 million. That's probably not exact but that's probably in the right ballpark. Proton about 65 million and spacex was a firm number - they actually published their number and said we'll launch your satellite for 62 million - so what we see is that spacex is fairly competitive with proton and they look pretty good compared to ariane.
I'm going to add some coding to this chart and basically what these numbers show is the money you could save by choosing spacex and what we see is if you were going to fly proton you'd save a little bit of money by choosing spacex. If you're going to fly ariane as the lower payload you'd save a fair bit of money flying spacex, and perhaps more importantly you wouldn't have to do all that excess design work and you wouldn't have to coordinate schedules so that you launch with the upper one so that is a little more convenient, and if you're going to fly arion upper stage the upper berth you could save quite a bit of money by going with spacex assuming that spacex could handle your satellite, so this is the reason that spacex generated so much business so quickly when they started with falcon 9 it only took them a few years and they were launching a lot of satellites.
Now part of that was they were considerably cheaper than arianne, part of that was proton has never been a particular particularly reliable launch system and about this time they had some significant reliability issues so spacex managed to take some business from proton that they might not have in other circumstances and the other thing that was going on was there was actually more demand for launch than the existing providers were able to meet, so you kind of put all those three together and that is the way that spacex managed to essentially start launching a lot of satellites very quickly and the point for talking about this is that that particular market condition only existed in 2010.
If you're trying to go into commercial launch in this same market now you have to compete against SpaceX.
In 2010 there were six us government satellite launches three by atlas, two by delta four and then one by delta iv heavy.
All three of these vehicles are operated by united launch alliance and at that point united launch alliancehad a monopoly essentially on this sort of government launches at
least on what we'd call nssl now - it was called eelv the previous program - so launching things for the department of defense ula had a monopoly.
So what did these ones cost? These are also hard numbers to get so don't take them as verbatim or as necessarily true but they are in the right ballpark, so atlas 5 about $110 million. Atlas 5 uses an approach where they have a base system, a base launcher and then to get more capability they add solid rockets to the outside so that means you aren't always at 110 - you might be 110, you might be higher.
Delta 4 $164 million it also uses that solid rocket boosters added. You can think of those as competing with falcon 9 at $62 million.
If we look farther over we see delta 4 heavy which is $350 million - perhaps even more than 350 million - a very expensive rocket and that roughly competes with falcon heavy which is about $90 million. Of course falcon heavy didn't exist in 2010, but you can look at that number compared to now i guess.
If we add in the color coding what do we see well we see once again there's a lot of money that could be saved by choosing spacex. Now the department of defense is notoriously not terribly price sensitive when it comes to launch - they're actually much more concerned about reliability than they are about money - so instead of looking at this and looking at those numbers as money that the department of defense could save, you should look at it as margin that spacex can use to increase the cost of their bid.
So you can see if you're competing against atlas 5 you can push your push your bid price up. If you're competing against delta for heavy launch you can push your bid price up a lot and you'll still be cheaper so these ones. Unlike the commercial ones, the commercial ones have gone away, because if you're trying to launch commercial you're going to be competing with spacex. Here it's a little different because ula still has half of the contract here so they are still trying to fix this
Summary on first mover.
In the commercial market spacex was cheaper and they had some other market conditions that made it easy for them or easier for them to gather a lot of business. In the us government market spacex could charge a premium above their commercial price and still get business because they were so much cheaper than the existing solutions. These are the first mover advantages - the inefficiencies in the market made it easier for spacex to establish themselves.
If you're a new entrant you aren't competing with those high priced options, you're competing with spacex with a bit of a caveat - programs like nssl choose two providers so we have the relatively cheap spacex provider and we have the much more expensive ula provider and you could still be a new entrant and you can compete for this nssl launch and compete against ula now.
Note that ula has been doing this for a long long time and they are right in the process of building their new vulcan rocket and that will allow them to do all the launches they want with a single system - they won't have both atlas 5 and delta 4 heavy and that is going to reduce their per launch cost and it's going to reduce their fixed costs considerably so if you want to compete against them it's going to get harder than it was in the past.
Vertical integration is the next topic.
I'm going to talk about vertical integration only in the context of engines because I think that's probably the biggest impact but this applies with engines, it applies with valves, it applies with avionics, applies with pretty much any kind of
advanced system that goes into your launcher.
Go back to the atlas 5. Its base engine of the first stage is an rd180 from npo Energomash in Russia, and its centaur upper stage engine is an rl10 from aerojet rocketdyne. These are both wonderful high performance engines. As i mentioned earlier for some mission they will add on solid rocket boosters - they're currently using the GEM 63 solid rocket booster from northrop Grumman. I went and looked forprices on these and of course ula doesn't say what they pay for these so we just have estimates...
The rd-180s are estimated to be about 20 million dollars per engine, and the rl-10 is also estimated to be about 20 million dollars per engine.
I went looking for estimates of the solids and after i spent a fair bunch of time I realized nobody was talking at least publicly i about how much these cost and this is a new one. I also looked for the old solid they used to fly - I couldn't find any more numbers there so i don't know what it is - $500,000, a million, something like that.
So if we add these up uh we see the atlas 5 is looking at 40 million dollars for engines for any given flight and that is a lot of money and if you look at you know the everything else has to come out of that price so 40 million dollars.
And so the question is why are these engines so expensive and well, the simple fact is the engine manufacturers listed here want to get as much money as possible for each engine - that is their business and they need to make a profit to make
to stay in business.
Now because a vehicle like the atlas 5 only flies a few times a year they are only going to be buying a few rl-10s from aerojet and they're only going to be buying a few rd-180s so what does that mean? well that means the engine manufacturers need to allocate
their fixed costs to a small number of engines.
Now another factor going on is that any improvements in manufacturing will show up as profits to the engine manufacturer, they will not show up as cheaper prices to ula, and the rl-10 has been flying for more than 50 years now and clearly the people building it have had a lot of time to optimize their build process, so you can figure they are doing fairly well on it assuming that they're fine with their fixed costs.
So this is why these engines are so expensive ula would like them to be cheap but frankly aerojet doesn't have any competition for the rl-10 - the centaur upper stage is designed for an rl10. You'd have to build something else and as long as they keep making centaurs they need to buy rl-10s for pretty much for whatever they cost and there isn't really a lot of competition in the rocket engine realm anymore
There used to be - back in like 1960s and 1970s lots of people made engines and they've mostly been bought up and combined and merged so we're left pretty much with aerojet rocketdyne.
This is a case where ula does not have vertical integration - they need to buy their engines from somewhere else.
looking at the falcon 9 in the base it has 9 merlin engines and then in the second stage it has a merlin vacuum engine, and both of those are developed in-house by spacex.
we have guesses on what it cost based on some stuff musk has said - the base merlins are about a million dollars a piece. They're probably a little bit less but we'll say about one million and they have nine of them so that's about nine million in engine cost.
The merlin vacuum is a much more complex engine than the one used in the first stage. It's probably a two million dollar engine - something like that - so what does that give for our total engine cost? About 11 million, or well over 20 million dollars cheaper than what ula is paying for their atlas engines, and the reason is what very simple. The spacex internal engine team has no incentive to make expensive engines and they have every incentive to make cheaper and better engines so they want to make them cheap so the engine team goals align with the rest of spacex's goals, and that is the advantage of vertical integration.
So it's really no coincidence that all the new launch companies build their own engines; if you look out rocket lab builds their engines, blue origin is building their own engines ,all this sort of stuff happens simply because it's much much harder to compete and it's especially much much harder to start up a company if you buy engines
Another moat around engines is simply engine creation - I usually just say engines are hard - it takes a long time to develop them - you need factories like the Blue origin factory on the right, you
need test facilities - that's the spacex test facility in mcgregor texas on the left - and you need an experienced team.
So if you're trying to compete with spacex who has years of engine development you need a team that can compete in engine development, so it's going to take you a while to create that team.
Reusability
This one seems fairly obvious... Spacex invested a lot of money in developing reuse and they now reap that benefit every time they fly a booster. So if we estimate that a new booster is about 30 million - it's probably in the right ballpark - so if we fly a new booster every time back before they were reusing boosters it would take 30 million dollars - that was 30 million dollars in their cost
If you take that same booster and you can launch it five times, that's about six million dollars of cost allocated to each of those launches. Now you have to go and recover it - you need to bring it back, you need to inspect it, you need to do some refurbishment and I just threw a number at this it takes your costs probably up to around 10 million, and it might be 10 million it might be 15 - I don't know what the exact number is but the important point is it's much cheaper to refly boosters than to build new boosters
If you're a competitor and you don't have reusability you're going to find it hard to compete, so spacex has somewhere between a 10 to 20 million dollar advantage per launch merely because they can reuse these boosters.
The last moat i want to talk about is related to people and this one i think is really not understood very well.
There seem to be a lot of people who think that engineers are just kind of interchangeable and if i want to start up a new project i just need to find some engineers and they will perform about as well as any other engineers.
And this partly depends on people and it partly depends upon the environment in which they are working. There's this great book by daniel pink called drive which is all about motivation and i'll summarize it.
Pink's theory is that there are three big things that motivate people and the first is autonomy and this is i get to decide how to do my job - i can be self-directed. The second is mastery - i can get better at what i do i can develop my skills. and finally purpose i work on something that is important and that gives us meaning and that gives us contribution.
So these are the things that actually drive people, especially when they are engineers and they're going to be getting paid a fair bit of money -money is much less of a driver. So if we look at spacex, spacex is notorious perhaps for providing all three of these to their young engineers. There's a nice video that talks about pink's drive and some of the evidence so go to youtube and search for it.
So spacex is great at providing autonomy, mastery, and purpose and its employees value these three things over things like work life balance - it's very well known that if you go to spacex and work you're not going to have a lot of free time - you're going to spend a ton of time working. The people who go and work there choose to do that because of these other things so essentially spacex by doing this gives themself a nice way to attract these sort of people - they have a recruiting advantage on the sort of people who do well in this environment.
Another way to say that is they have first call on new graduating engineers who really want to do this kind of thing and this kind of thing is exactly what you need to build a new company when you're trying to compete with spacex.
you can look at all of that nice web streaming that spacex does as a pr thing, purely to recruit people to come work for the company, and that's why - it's very obvious - most of their online hosts are young engineers, and clearly very knowledgeable and very passionate engineers. So merely by producing their web streams in the way they produce them they reinforce that spacex is a place that those kind of people want to come and work.
Now having said that, this is the sort of advantage that can go away easily and it could be a case where spacex just stops hiring people, they stop expanding, and it could be that something inside happens and the corporate culture changes and these three things that are motivating people go away and people start leaving spacex because they've lost these things, so this would be something to watch, to understand why people are leaving spacex.
so that's pretty much what i wanted to cover the big point is the large aerospace
companies like boeing like lockheed they aren't competing with spacex
because there are too many moats it's not worth the investment to try to
cross them you can look at the problems it would take what would it
take for ula to become competitive and that would take a bunch
bunch of money for them to do that and then they would just be competitive but
they might not actually get more contracts than they are currently getting
national governments are an exception so we see efforts in europe
to try to do something similar to falcon 9 and part of that is national pride or i guess
national pride for many companies for many countries part of that is trying to get back
some of their competitive advantage and then we see it in china also very
much a national national pride sort of thing small and new aerospace companies will
try to compete with spacex simply because that's what they need to do to grow and we see that we see rocket
lab has announced they're building a new launcher called neutron which is kind of like a smallish
slightly smaller version of falcon 9. and they will have problems with these
spacex moats they expect that there's enough new business in
big constellations to keep them in business and they also expect that they can be cheap enough
because they're already a very small lean company kind of like spacex to be able to compete
there will also be a chance for a company like like rocket lab to conceivably start
competing in contracts like the nasa resupply ones
and maybe even in the nssl ones and since those contracts are not winner take all
they don't have to compete with spacex they just have to compete with the next best company
starship which i didn't talk about at all will work will clearly make all of this harder assuming starship is
successful and i think that's another thing that's pushing these big
companies uh to not invest
so that's my talk i hope you enjoyed it
English (auto-generated)
AllFor youRecently uploadedWatched
That's pretty much what i wanted to cover.
The big point is the large aerospace companies like boeing like lockheed they aren't competing with spacex because there are too many moats - it's not worth the investment to try to cross them. You can look at what would it take for ula to become competitive and that would take a bunch of money for them to do that and then they would just be competitive, but they might not actually get more contracts than they are currently getting.
National governments are an exception so we see efforts in Europe to try to do something similar to falcon 9 and part of that is national pride for many countries. Part of that is trying to get back some of their competitive advantage and then we see it in china also - very much a national pride sort of thing.
Small and new aerospace companies will try to compete with spacex simply because that's what they need to do to grow and we see that we see rocket lab has announced they're building a new launcher called neutron which is kind of like a slightly smaller version of falcon 9. They will have problems with these spacex moats but they expect that there's enough new business in big constellations to keep them in business, and they also expect that they can be cheap enough because they're already a very small lean company kind of like spacex.
To be able to compete - there will also be a chance for a company like rocket lab to conceivably start competing in contracts like the nasa resupply ones and maybe even in the nssl ones, and since those contracts are not winner take all, they don't have to compete with spacex, they just have to compete with the next best company.
Starship which i didn't talk about at all will clearly make all of this harder assuming starship is successful and i think that's another thing that's pushing these big companies to not invest.
So that's my talk i hope you enjoyed it
What, exactly, is up with Starship?
Why is a company that has done so well with Falcon 9 and both cargo and crew dragon having so much trouble with starship?
We got Starhopper in 2019 and then in 2020 a series of tank tests, and then in late 2021 we had a great series of high flight and landing tests, culminating in the successful flight of SN15.
And then lots of activity and talk about flights for a couple of years, but it wasn't until April of 2023 that we got to IFT-1, the first test flight, which tore up the launchpad, had significant engine problems, didn't make it to staging, and couldn't even blow itself up properly.
It was a bit of a fiasco.
IFT-2 in November was a bit better - it got through separation and was able to blow itself up.
IFT-3 in March made more progress, with the booster making its way back towards the launch pad and the ship making its way on its trajectory though both went out of control.
Progress, but this was coming from SpaceX, the company that developed booster landing and is launching far more Falcon 9s than the rest of the world is launching rockets.
The final straw was this chart that Elon presented in an employee update after IFT-3 where it was disclosed that the current starship didn't really have a specified payload, but Musk said that it was about 40-50 tons in the current configuration.
That's a far cry from the 150 tons that is commonly talked about, and a lot worse than the Saturn V, a smaller rocket.
How is this not a failure?
Now that I've pushed you out on the ledge, let's see if I can talk you down.
I asked my TwiX followers why they thought starship was late, and got a bunch of different responses that I'll paraphrase here.
(add them here)
Many of these responses align to what I used to think.
But there were a few responses that gave what I now believe is the best explanation.
We've all heard the simple saying that space is hard. It has quite a bit of truth in it, but it tends to mush together things that have high technical challenge with those that are hard just because an organization is inexperienced.
Falcon 9 version 1 was hard for SpaceX because they had never built a rocket that big, but it was a boring rocket design with a boring engine, and that was one of the keys to their success.
Falcon 9 booster reuse was technically challenging because it was this brand new thing, but it wasn't as hard from an organizational perspective because SpaceX had a good handle on Falcon 9 in general and they could work on it incrementally.
Then starship comes along, and it feels like it's harder than booster reuse but not that much harder - at least that's where I put it in my mind. And that seemed to be reinforced by the early starship flight tests - they felt a lot like the Falcon 9 booster reuse tests. They had an design, flew it, and iterated until they had it working. Starship would need to nail reentry but shuttle had done that 40 years ago. And this was SpaceX, who made Falcon 9 and capsules look easy.
It turns out that I was wrong. Quite wrong.
Starship belongs over here. It is quite literally "off the charts" in terms of the technical difficulty, and once I realized that, there were a number of things about starship that had always confused me that suddenly made sense.
But I'm getting ahead of myself. We need to talk about the physics first.
We'll start with our old friend, the rocket equation.
If you want the whole story, see my video "the care and feeding of the rocket equation".
What we are calculating is the delta v that we can get out of specific stage with a given payload. If you think of delta v as a measurement of how far you can go, you'll have the right idea.
It's actually fairly simple. The ability to send payload to a given destination depends on how good your engine is, how light your rocket stage is, and how much propellant you can stuff into it.
Isp is the specific impulse of the engine, which is a measure of fuel efficiency. That is set by the propellant that you use - liquid hydrogen, liquid methane, and RP-1 kerosene all have different specific impulse profiles - and by the kind of engine that you have. If you want a higher specific impulse out of a given fuel, you will need an engine design that can accomplish that. More on that later.
The fraction inside the natural log is known as the mass ratio. The denominator is the mass of the empty stage plus the payload - it's the mass that is leftover when all the propellant is burned.
The numerator is that same mass plus the mass of the propellant.
If we can put more propellant in a stage we get a better mass ratio and can do more. If we can reduce the inert mass or fly with a smaller payload, we get a better mass ratio and can do more. If you want to know why the natural log is there, the answer is "because math" or "because physics". Go watch the video I linked to.
For Starship, we already know what engine and fuel we are using, so the specific impulse is pretty much a constant. Changes in engine design might move it a little, but not a lot.
The mass ratio therefore matters a lot.
Let's look at an example. I'm going to use Falcon 9 in fully expendable mode because there are good numbers there, and we're going to only look at the second stage. We know that the empty mass of the stage is 4 tons and it holds 112 tons of propellant.
Let's say that we are trying to launch a payload to low earth orbit. That will take about 9400 meters per second of delta v total and for the Falcon 9, about 5600 meters per second of that needs to come from the second stage. We can run the rocket equation backwards and we'll find that we need a mass ratio of 5.2.
We can then plug the ratio into the bottom equation - which tells us how much payload we get from a given ratio - and we get the expected 22.8 tons. The payload to low earth orbit is about 16.5 % of the total second stage mass, which is good performance. The inert mass is only 3% of the total.
Now we can try a harder destination...
Flying a payload to geosynchronous transfer orbit requires about 11,800 meters per second of delta v, and 7900 meters per second needs to come from the second stage. To achieve that, the ratio goes all the way up to 10.1. The only thing we can change is the payload, and to get that ratio, we have to reduce it all the way down to 8.3 tons. The payload percentage drops to 6.7% and notice that we took a 64% payload hit to only increase our delta v by 41%. The relationship between delta v and payload is not linear.
Flying all the way to geostationary orbit requires 13,600 meters per second of delta V, and the second stage needs to provide 9700 meters per second. That pushes the ratio up to 17.2 and the payload down to 3 tons, or 2.5%
Looking at those numbers, we can see that GEO only requires about 75% more delta-v, but the ratio has to go up by 230%, and the payload is cut by down to only 13 %. That is the way the rocket equation works; adding delta v is hard.
The falcon 9 second stage is very good; it is light and it can carry a lot of propellant.
It might make more sense with a graph.
This is a graph of the payload at different delta v values for the full Falcon 9 vehicle.
On the left we have the expendable payload of 22.8 tons to LEO. If we go with drone ship recovery, the delta v required goes from 9400 to 10,055 meters per second, and that drops the payload down to about 16.8 tons. And then a return to launch site requires about 10,700 meters per second of delta V and therefore can support only 13 tons of payload.
The line is pretty flat because the second stage is so light and the payload fraction is so high.
Now let's look at starship
Reference:
(Starship Orbital Delta v spreadsheet)
Falcon 9 - 2.6 kg reduction in payload for every m/s in additional delta v. 0.01%
Starship - 50 kg reduction in payload for every m/s of additional delta v. 0.1%
Starship is very different. Based on some reasonably decent numbers, it has an empty mass of 130 tons and a propellant mass of 1200 tons. Because it is fully reusable, it needs to devote delta v to getting the first stage back to the launch site and to the reentry and landing of starship.
That's a pretty big tax; my numbers suggest it's around 11,000 meters per second in total, and 7270 meters per second from starship. That gives a mass ratio of 7.7 and a payload of 50 tons. That 50 tons is only 3.6 % of the second stage mass.
The problem is the inert mass. Starship has a full payload section, a tail section to cover the engines, fins to control it during reentry, and a heat shield to cover all of the bottom so that it will survive reentry. That inert mass is about 9.4% of the full stage mass.
As for GTO and GEO orbits, you can forget about them. You may have heard me state that starship could send 20 tons to GTO. Not with this starship, and maybe not even with future ones.
It's a little clearer if we look at a visual representation.
It looks like this, with the 50 ton payload highlighted. Starship is hugely sensitive to changes in delta v because of the poor inert weight, but that inert weight is what you are stuck with if you want a fully reusable second stage.
I first did a chart like this for the video on starship and heavy gravity, and I'm a bit embarrassed to admit that I never did a chart like this before, because it makes so many things make sense. The combination of the high required delta v and the mass cost of reuse just make it ridiculously hard to do full reuse, and that is what is has been driving starship all along.
Musk often gets accused of overhyping things, and there's some truth there. But this is a case where he has been seriously underhyping the true difficulty of what starship is trying to do.
Before we move on, I want to add this little data point at 9400 meters per second of delta V, which shows us that the 50 ton reusable starship can carry 160 tons to low earth orbit if it flies in fully expendable mode. And that's carrying around the reusable fins and heat shield; if you remove those the mass ratio will improve a lot and it's probably closer to 200 tons.
Once internalized what this graph meant, a few things made sense.
Reference:
(Starship Orbital Delta v spreadsheet)
Falcon 9 - 2.6 kg reduction in payload for every m/s in additional delta v. 0.01%
Starship - 50 kg reduction in payload for every m/s of additional delta v. 0.1%
Staging on Falcon 9 is inherently a wasteful process.
The first stage is thrusting hard, and then it hits MECO and shuts off all of its engines. At that point the big force acting on the rocket is gravity, pulling it back towards the earth every second.
Then there is a pause, and the second stage separates, and after it gets far enough away to be safe, the second stage engine starts and the second stage continues merrily on its way to orbit.
If it takes 10 seconds, the gravity loss is probably on the order of 100-120 meters per second of delta v. On the Falcon 9, that probably costs around 250 kg of payload - or only around 1% - because the payload to delta v graph is so flat.
The same loss on starship is worth about 5000 kg, or about 10% of the payload, because the payload to delta v graph is much steeper.
There's also the problem of the first stage, especially for return to launch site reuse. After separation, the booster is pointing in the wrong direction and you need to spin it around so it's pointing in the right direction before you relight the engines. Not only is it falling towards earth during that time, it's getting farther away from the launch site.
Both of those mean that the booster needs to use more fuel to get back to the launch site, and that's fuel that it can't use to give more energy to the second stage.
It's not a big deal for Falcon 9 because the graph is flat and also because Falcon 9 only flies return to launch site when it has very light payloads and delta v isn't very important with a light payload.
Falcon 9 can afford the casual staging approach, but Starship can't.
Which is of course why they have gone with hot staging.
My initial thought was that this was just an optimization they did because they were trying to get things right from the start, but since I've seen the delta v payload chart I realized that SpaceX is fighting for every little improvement they can find just to make full reuse *possible*.
Which brings up to Raptor, and what I used to think was the most confusing part of Starship.
Musk tweeted about it all the time. In fact, it seemed like he might be obsessed with it.
The obsession seems to be about chamber pressure.
The record chamber pressure for a rocket engine was the Russian RD-191 engine, which ran at 263 bar, or about 3800 pounds per square inch, which is very high. The RS-25 space shuttle main engine was the current US title holder at 206.4 bar, and the Merlin 1D flown on the Falcon 9 came in a very conservative 108.
Raptor 1 was testing at 300 bar in 2018, Raptor 2 is likely capable of around 325 bar, and Raptor 3 has a goal of 350 bar.
All of these are far larger than the RD-191, and developing such high chamber pressure engines is very difficult and therefore expensive in both money and time.
Why didn't SpaceX stick with the "tame" 300 bar engine and save the other versions for later?
The answer is simple. The Raptor 1 engine that powered IFT-1 has a payload to orbit of pretty close to 0kg.
That is a fairly astonishing result. The lower thrust of the raptor 1 means that you have to reduce the propellant you carry to get the same thrust to mass ratio, and the lower propellant load reduces the stage mass ratio and therefore reduces the delta v you can create.
We want high thrust so that we can lift more propellant. That's why Super Heavy has a ridiculous number of engines.
The equation for thrust is simple - it's the amount of mass that flows into the combustion chamber multiplied by the exhaust velocity. We would generally write that using 9.8 times the specific impulse.
So if we want to build a higher thrust engine, we need to either increase the mass flow rate or we need to increase the specific impulse. If we increase the pressure feeding into the combustion chamber, that gives us a higher chamber pressure and increases both the mass flow rate and the specific impulse.
A higher chamber pressure burns hotter, which gives the exhaust particles more energy and therefore a higher exhaust velocity and higher specific impulse. But as you can see with this graph, it gets pretty flat at high chamber pressures, so the specific impulse gains are fairly small.
We also can open up the size of the combustion chamber throat a bit; that makes it easier for the exhaust to escape and increases the mass flow rate, but by itself it will reduce the chamber pressure. Or you can do both.
SpaceX continues to push for more chamber pressure and thrust because that will improve the delta v capability of the rocket. Every rocket designer always wants more thrust.
I had a thought about the raptor design and while I don't know whether it's true or not, it's an interesting thing to think about.
In an engine like the RD-180, some of the RP-1 kerosene is burned with liquid oxygen in the preburner, and that drives the main turbine to produce the power to pump the propellants. Both the RP-1 fuel and liquid oxygen pumps are driven with that same turbine, and that means that the ratio between those two propellants is fixed.
The full-flow staged combustion design that raptor uses has separate liquid oxygen and liquid methane preburners and turbopumps, and that gives them the opportunity to vary the mixture ratio between the two propellants on the fly.
Rockets generally fly fuel rich so that there isn't any hot free oxygen in the exhaust and they generally use a fixed ratio. Raptor supposedly runs at a ratio of 3.6:1 for oxygen and methane, but with the full-flow setup SpaceX can easily experiment with different ratios including the hotter ratios that are closer to 4:1. They could even vary the ratio during flight.
I have no idea how much they are doing with this, but it seems possible.
Note that the RS-25 space shuttle main engine also had two preburners so it theoretically could have done this but AFAICT, they always ran at a fixed ratio.
I now understand why SpaceX has been so focused on raptor performance and why they decided to do hot staging. It's the only way to get the performance they need.
One more weird thing about Starship that makes more sense now...
Remember IFT-1 and how much of a fiasco it seemed to be, with a severely damaged launch pad area, multiple engine failures, and flight control issues. At the time, this didn't seem like a SpaceX thing to do, to launch a vehicle that was so flawed and to have delays fixing the launch pad issue.
At the time, Elon estimated a 50% chance of success and said that if the rocket got far enough away from the launch pad before something goes wrong, that he would consider the test to be a success.
Which seemed like a really low bar at the time.
What we didn't know is that the IFT-1 flight wasn't their first prototype flight because to actually fly the first flight the way they wanted to, they needed the additional performance of raptor 2 in a vehicle designed for it.
So they chose to launch a vehicle that they knew was going to have issues to see what they could learn from it. It was either to launch that vehicle or just scrap it. And that's why they seemed to be so unconcerned about the issues on the flight.
Knowing that, I would argue that IFT-1 was a very SpaceX thing to do. Though I don't think they counted on how much reaction the launch pad concrete issues would bring.
However, even with all that work, the current version of Starship has a performance problem. It needs to carry more payload.
There are two things to do...
Reduce the delta v losses during launch.
Increase the delta v available from the vehicle by improving the specific impulse or the mass ratio.
The first is further improvements to raptor. Raptor three will push the thrust up another 20% - allowing the same number of engines to lift more propellant and increase the mass ratio and therefore the delta v and therefore the payload.
They are also continuing the simplification that they started on raptor 2; they are routing the extra plumbing they need internally in the parts. Nobody has done that before and it will be a lot of work, but the hope is that it will allow them to both reduce the mass of the engine itself and eliminate the heat shielding the engines currently wear. That will reduce the inert mass and improve the mass ratio.
Elon showed this chart in the employee update after the IFT-3 flight.
Based on engine thrust improvements, starship version 2 will get a bit taller in both the ship and the booster to carry more propellant and that will increase the mass ratio enough to put it about 100 tons of payload. It won't take a lot because the line on the graph is so steep.
I consider starship 3 to still be speculative based on Musk's comments. The big change there is going to 6 vacuum engines on the ship to give it much higher thrust and therefore a big bump in propellant and delta-v, and then a small increase in the booster to help it carry the much heavier ship. This depends on further engine thrust improvements; if those aren't there I'd expect to see a bigger ship than starship 2 but not as big as starship 3.
There's another small optimization going on here that isn't obvious. We can figure out how many of seconds of propellant the booster and ship carry, and for flight 3 we get 505 seconds. Note that some of the booster fuel goes to boostback and landing and some of the ship fuel also goes for landing, and also note that they don't fly full throttle the whole time, so this number is wrong in actuality but okay for comparison.
For starship 2, the higher thrust means that they burn the fuel faster and even with bigger tanks, the total burn time is less. This means they get into orbit more quickly and *that* means that they waste less energy on gravity losses.
For starship 3, it's even more profound because of the addition of three more engines to starship means starship will really fly. They will likely have to throttle down - or perhaps shut down some engines - more on starship 3 as it gets lighter near the end of its flight, so the improvement won't be as big as these numbers indicate.
You can also think of this in terms of thrust to mass ratios; better thrust to mass ratios means you get into orbit more quickly and spend less time fighting gravity.
What did I learn about Starship?
It is hard to overestimate how hard it is to build a reusable second stage. They need to be robust enough to survive reentry and light enough to carry a useful payload.
Landing the booster at the launch site makes it harder. There's a non-trivial cost to doing that.
Even small optimizations help. Approaches like hot staging help.
And, most important, Raptor is the absolute key to starship. There may be other fully-reusable approaches that might work with a tamer engine, but for the starship architecture you need an absolute screamer of an engine.
Musk is obsessed with Raptor not because it's a cool engine, not because SpaceX is showing off, but because without a ridiculous level of performance out of Raptor, Starship cannot reach its goals.
And that's why Starship is late, or at least why there was such a big gap before they flew a real prototype. Its goal is ridiculously hard and it takes incredibly good engineering to make it possible.
The good news about this is that SpaceX finally has an engine that is good enough for Starship, and that's why they have made so much progress in IFT-2, 3, and 4.
We are back to the SpaceX we expect, and I am less confused. Which is nice.
If you enjoyed this video, please send me this wonderful poster from despair.com
I've had many people ask about using other materials than stainless steel on starship.
The three options that are commonly mentioned are titanium, aluminum, and carbon fiber.
Note that for titanium and aluminum I'll be talking about alloys, and different alloys behave very differently. I've tried to pick the most common ones used for aerospace but the specifications aren't always available.
The analysis I'm going to be doing will make a little more sense if you've watched my engineering tradeoff analysis video.
If you want to go watch it now, we'll wait for you here.
We'll start by talking about starship
Titanium is our first candidate.
Titanium is widely reputed to be as strong as steel and 45% lighter but it looses that advantage at higher temperatures. 304 stainless is heat resistant up to about 870 degrees centigrade, and the best titanium alloys only go up to about 600 degrees. That is a significant disadvantage for starship because of reentry heating. You may be able to keep the titanium from getting too hot, but that will likely involve a heavier thermal protection system, which will likely negate any weight savings.
There are also issues with contamination in high-heat environments; under high heat, titanium wants to grab onto other elements and that can degrade its strength. You don't want a material that sees high heat and appears to be fine but actually has significantly reduced strength.
One of the sources I came across described titanium as "notoriously difficult to machine". Welding is possible, and NASA has a 236-page document that tells you how to do it properly. It says the following about the process
(read)
If you don't weld it exactly right, the impurities you get lead to embrittlement and the weld is not as strong as it should be.
Cost is also an issue.
If we assume that starship has 60 tons of steel in it, we can make a guess at material costs.
I found suppliers of 304L stainless in rolls for $1500 per ton - I'm sure SpaceX pays less with the quantities they use - and that means that starship has about $90,000 of stainless steel in it
If we assume titanium is 45% lighter for a given strength - which many not be true at high temps, that would mean that we would need 33 tons of titanium
My research suggested that it would probably an alloy known as Ti-6A-4V.
It's hard to find prices for coils of titanium, but the the sources I checked suggest that it's $50 per kilogram, or $50,000 per ton. That's $1.65 million in materials, or over 18 times the cost of the stainless steel.
With stainless the material is cheap and machining and welding is well understood and therefore also relatively cheap. The titanium alloy is very expensive, machining is hard, and welding must be done very well or things break.
And, depending on the temperatures the material encounters, it may not be any better than stainless in terms of strength per kilogram, and overtemp behavior is probably worse. A titanium starship probably doesn't survive the IFT-4 reentry conditions.
If you know the heat environment really well and you are okay with the hassle of building with titanium, then I could maybe see using it, but I'm not excited about a less robust vehicle that is much much more expensive..
What about aluminum alloy? The biggest reentry vehicle before starship was made of aluminum and also used tiles to protect it during reentry, and all the big jetliners and many rockets use aluminum alloys, including SpaceX's own Falcon 9.
It's widely available, well understood and easy to weld and machine.
Starship would only require 22 tons of aluminum and its pretty cheap at around $2500 per ton, so it would be a little cheaper than stainless.
There are good reasons that aluminum is so widely used in aerospace
And here's the problem with aluminum for starship. The red line shows the point where a material has lost half of its strength. We see that stainless has 50% of its strength up to about 800 degrees centigrade.
Aluminum alloys lose half their strength before the temp reaches 400 degrees. Any issues with your thermal protection system and your structure melts away. In fact, SpaceX originally flew aluminum grid fins on the Falcon 9 but they had a tendency to melt even with the relatively small amount of heating experienced by the booster.
Not a great material for the high temperatures that come with reentry. The fin that survived the OFT-4 flight would have melted off the starship body very early in reentry and the vehicle would not have survived.
Which brings us to carbon fiber.
Back before the name "Starship" existed, SpaceX called their project BFR, and it was going to be full carbon fiber. SpaceX had two tank domes made and put together into this short tank, and they had this full-sized mandrel built that would be used to wind carbon fiber on to build the straight sections of the tanks.
So, somewhere, there is a design for how to build a starship-class vehicle out of carbon fiber.
Which is interesting...
Carbon fiber is nice and light, so you likely only need 11 tons of it, and even at a relatively pricey amount of $10,000 per ton, that's $110,000 in materials.
That seems promising.
There are three big issues with carbon fiber for starship.
The first is that it has terrible high temperature properties; it loses about half its strength when it gets up to boiling water temperatures, or only about 100 degrees centigrade. There may be versions of carbon fiber that are more robust, but it is essentially very strong fibers held together by plastic or epoxy
You would need to have a really good thermal protection system to keep it cool enough, though it is fair to note a carbon fiber starship is going to be very light, which means a lot less energy in orbit and therefore a lot less heat to deal with.
The second issue is that you have to build molds that the carbon fiber is applied to, and it is those molds that define the shape. Creating the mold takes a lot of time and money, and it's therefore not well suited when you aren't sure what you are building. Steel is great for that - if you want need more reinforcements in a section, you can just weld it in. If you need to mount new equipment, just weld it in. If you need a tank to test, just build a short version and you can test it, and go on.
This is a huge disadvantage if you aren't sure about what sort of design will work, which is certainly the case for starship.
The third issue was the logistical aspects of the plan.
The plan was that they would build large carbon fiber tanks and other sections in seattle, ship them to LA to build the vehicles, then - presumably - ship them through the panama canal and then up to boca chica for testing.
The logistics would have been absolutely horrible - the cycle time would have been long and with making new molds to make changes, development would have been slow. Carbon fiber is absolutely the wrong choice if you don't have a design you know will work, especially if you aren't building your rocket at your launch site.
Would carbon fiber work with a mature design, one that is already thoroughly debugged?
I guess I can maybe see it, but SpaceX is not loaded with carbon fiber expertise right now and they are investing a large amount of money to be able to build a lot of starships out of stainless. Switching to carbon means that they need to rewind their development process back a few years and build a new factory, and that seems unlikely.
What I've wondered about is whether they'll consider a smaller vehicle for crewed flight. You can fly people on starship, but you don't need a 150 ton payload vehicle for LEO flights. Even if starship is very cheap, there's probably a niche for a Falcon 9 / Crew Dragon replacement, and that might be a carbon fiber solution.
Or maybe Rocket Lab jumps up to do that, since they're the carbon fiber experts.
Stainless steel is a great material for starship. It's heavier than the other options but none of the other materials have the combination of high heat performance, ease of construction, and low cost.
But super heavy also loves stainless steel.
Before we talk about materials for super heavy, we need to cover one very important point.
When we are talking about second stages like starship, saving a kilogram of mass adds a kilogram of payload, and therefore second stage mass is very important. That is the biggest issue with the current Starship 1 design - it simply has too much mass.
That is not true for boosters. As a rough rule of thumb, you need to save about 6 kilograms of mass on the first stage to add a kilogram of payload to orbit.
That makes lighter materials much less interesting for first stages, but it's still worth looking at.
I see very little reason you would want to build a booster out of titanium. It's a lot of extra money and hassle for what would probably be small gains.
Aluminum is more interesting; it's cheap enough that you could afford to do it.
We know that the Falcon 9 booster is made of aluminum and it survives reentry, but we also know that it needs an entry burn to reduce its velocity and that super heavy does not do an entry burn.
That means there is penalty that the Falcon 9 pays because it's made of aluminum. Let's see if we can quantify it
On the transporter 2 launch, the booster was travelling at 4700 kilometers per hour at the start of the burn and only 2500 kilometers per hour at the end. That burn takes off about half the velocity of the stage and therefore about 75% of the kinetic energy.
We can do some rough math on the cost of the burn.
The merlin burns about 280 kilograms of propellant every second and this burn is 3 engines and goes for 25 seconds. That means it burns about 21 tons of propellant.
That 21 tons is about 70% of the empty mass of the booster. If we scale that up to super heavy masses, that's going to be a lot of extra propellant and my guess is that the extra propellant masses as much as the mass savings you would get with the lighter structure of aluminum.
So aluminum is also unexciting for super heavy.
Carbon fiber will follow the same pattern as starship, right?
Well, not quite.
The first stage of rocket labs neutron will be made fully from carbon fiber, and carbon fiber is roughly 5 times stronger than steel in terms of mass, so neutron might be only 20% the weight of a stainless steel rocket the same size.
That means that it is - to use a technical term - a lot more "floaty" than a stainless rocket, and that means it will come down more slowly and experience a lot less heating during reentry. Rocket lab believes that it will survive without an entry burn.
It's a fairly sure bet. Rocket lab has recovered their electron booster - also made of carbon fiber - without an entry burn, and SpaceX regularly recovers the falcon 9 fairings without entry burns.
So it's likely practical to build a super heavy out of carbon fiber.
But does it make sense to do so?
Let's do a little math. My usual disclaimer applies - I'm going to make a lot of simplifying assumptions, so the numbers will be wrong but the overall answer will be correct in spirit.
I'll start by assuming that super heavy has a mass of 180 tons empty. It has 33 raptors, which are probably a couple of tons each, so that's 66 tons of raptors. That leaves us with 124 non-raptor tons.
And let's just assume that all of that is structure made out of stainless. It's not, but that will give us an upper limit. If carbon fiber is 20% the weight of stainless steel, that gives us a mass of carbon fiber structure of only 25 tons, a savings of 99 tons.
That is huge. Nearly 100 tons lighter.
*But* we have to remember that it takes roughly 6 kilograms of weight reduction on the first stage to get an extra kilogram of payload to orbit, so we divide that by 6 and get an expected payload impact of 17 tons.
That's a whole lot of investment to build a brand new factory to end up with only 17 tons of extra payload.
So I'm going to say "yes, you could do this, but I think there's pretty much zero chance spacex will do it". It's just not worth the effort.
In summary, I don't find the other materials compelling enough to make them worth it, and I expect that SpaceX will stick with Stainless for starship.
If you look at scenarios where starship isn't coming back to earth, then alternate materials look a lot nicer, but the amount of work to build a factory to create them and test flights to verify that they work seems unlikely to be worth the cost. Easier to just stick with the normal starship and just fly more flights. Propellant is cheap.
If you enjoyed this video, I need me some tops. I'm obviously especially interested in the stainless steel ones.
Analysis of changes in the estimated cost of the skylab program: https://www.gao.gov/assets/b-172192.pdf
Commercial LEO destination concept of operations: https://ntrs.nasa.gov/api/citations/20230002770/downloads/ATTACHMENT%201%20CLDP-WP-1101_ConOps_Final.pdf
International space station transition report: https://www.nasa.gov/wp-content/uploads/2015/01/2022_iss_transition_report-final_tagged.pdf?emrc=4c4497
CLD requirements and standards: https://govtribe.com/file/government-file/80jsc023cldprequirementsandstandards-1-cldp-req-1130-and-annex-1-dot-pdf
ISS vibration: https://gipoc.grc.nasa.gov/wp/pims/handbook/
As you probably already know, the international Space Station is getting old, and it is currently scheduled to be deorbited in 2030, though that might be extended.
That will leave NASA with no space station in earth orbit, and NASA has some things they want to keep doing there, including:
Continuous US presence in orbit
Further scientific investigation related to deep space human exploration, such as crewed missions to mars.
Sponsoring advancement of technology and systems for deep space human exploration.
The obvious solution is ISS version 2
There are some problems with building a new version of ISS...
The first problem is that the ISS operational budget is about $3 billion a year, and with Artemis heating up and moon landers to pay for, NASA simply doesn't have the budget to support building and launching a new space station, and congress is unlikely to give them more money.
You, in the back. Put your hand down - I'll talk about the lunar gateway space station in a future video.
More importantly, as is often the case, congress would not choose to build a new international space station.
They actually don't say that, what they say is that extending ISS beyond 2028 will significantly slow down the scheduling of crewed missions to Mars.
How cute. Congress actually thinks that we have a crewed mission to Mars program, or that NASA can come up with such a plan in the 2030s.
The act specifically requires the following:
(read)
The goal is to build something NASA has started calling "the low earth orbit economy". And that's something I think most of us would support, at least as an abstraction.
And so we have the commercial low earth orbit development program. Here's what NASA is planning...
They plan to buy time for at least two NASA crewmembers per year aboard commercial space stations. This is an "anchor tenant" concept, where the assured NASA business enables the providers to build, launch, and operate their space stations profitably.
Their plan is to continue the microgravity and biomedical research that they have done on ISS and to add in technology development for exploration, which you can read as "Moon bases and Mars missions".
They also assert/hope that private astronauts would visit for research or tourism.
Early proposals have generated some very impressive slides, but since then, there hasn't been much progress and in fact, in 2023, Northup Grumman abandoned their own plans and decided to join the Nanorack Starlab team.
But there's a deeper problem, and it's pretty simple. These commercial stations do not make financial sense for the private companies that would launch them.
Businesses exist to make money, and what is true for your local store is also true for space stations
It's time for a quick look at investment and returns.
Let's say that your brother-in-law is starting a new business and wants to borrow $1000.
He says that he will pay you $200 each year for 5 years. Do you give him the money?
You probably have a gut feeling based on the numbers, and my guess is that your answer is probably "no", unless you are especially fond of your brother-in-law or camels.
You put in $1000 dollars and in the best case, you only have $1000 after he makes the last payment. From a business perspective, this doesn't look very good.
If he offers $400 a year, that's obviously a better deal - you invest $1000 and get $2000 back. But is it good enough?
What we need is way to evaluate projects in a more objective way.
There are two common ways to look at this.
If we assume that inflation is 10% a year - that our money is worth 10% less every year - we can calculate that those $400 payments are worth $1500, so you will make money. This is called the "net present value" approach.
I prefer a different approach, known as the internal rate of return. You can think of it as "how bad would inflation have to be for this investment to break even?". In this case, the internal rate of return is 29%, which makes it seem like a comfortable investment. It would be better than an investment that returns only 22% and worse than one that returns 37%.
But there's a big scary factor that we haven't considered, and that is the concept of risk. How confident are we that we will get that $400 every year?
That is where things get difficult. Perhaps the market in camels is saturated in his area and he should consider focusing his love of camelids such as the alpaca or even the vicuña. Unexpected things might happen - feed or insurance prices may go up, or there may be a breakout of camel flu.
Generally speaking, investments that have low risk have a low rate of return and investments that have high risk require a high rate of return.
If we want to evaluate the investment and return on commercial space stations, we'll need to understand the details of what NASA is proposing.
What is NASA's model for commercial space stations?
The first step is that the provider builds the space station. What kind of space station?
NASA details that in the 41 page CLD concept of operations documents, the several hundred page requirements and standards document, and the long list of references in those two documents.
This program is about doing everything "the NASA way", and to put things simply, your space station must meet ISS module standards. I'm sure many of these standards make a ton of sense, but this is not "design a commercial station", this is "design a NASA station", and NASA will be validating that you have met all the requirements. If you have other ideas about how things should be done, you'll run into the same sort of issues that SpaceX had with commercial crew.
In return for following this standard, NASA will presumably kick in money to help pay for the design construction of the station, but the partner will be required to invest a lot of their own money.
The provider pays to launch the station in orbit, and then NASA is interested in using it. The responsibilities between NASA and the provider are clearly defined.
NASA provides astronauts, money for the astronaut's visit, and oversight of station operation.
The provider provides literally everything else.
NASA enumerates this as:
(read)
This is the all-inclusive exclusive island hideaway space station model.
NASA says the goals of this approach are to maximize partner authority, autonomy, and control over mission operations and to minimize NASA operational involvement.
For crew transport, the vehicles must be certified by the commercial crew program, the FAA, or some future agency.
#2 and 3 don't currently exist, so the effect of this requirement in the short term is to limit crew transport to crew dragon and perhaps starliner if/when starliner is certified.
Cargo will be like the commercial resupply program.
Just to emphasize, NASA isn't buying either of these services; they are part of the all-inclusive orbital package that NASA wants to buy.
Following the all-inclusive space retreat model, the station must serve as: (read)
There are a long list of systems that need to be provided for this, and they all must meet NASA specs.
The roles and expectations for the NASA astronauts and crew members are fairly clear and quite a bit different than ISS.
(read)
The ISS model has 4 astronauts who do everything, and that makes it very easy to load balance across the team. If you need to unload a vehicle, the whole team can pitch in and get it done quickly, and it's easy to deal with rest days and time off.
The model they are proposing has a really obvious issue. Everybody on station has to have time off to sleep and exercise and days off. That means that the provider needs not 1, but 2 crew members to keep the station up and running. The cost of these crew members will be factored into the costs of the station.
I think we now know enough to play around with some numbers to see if we want to be part of the program. These numbers are for illustration purposes only as we do not have any real details.
Our station costs $1000 million dollars to develop and $500 million to launch. Just to keep things simple, I'm going to assume that NASA pays $500 million for the launch, leaving us with an easy to understand $1000 million investment that is spent at $250 million per year over four years.
If we've decided that we only want to be in this business if our model says we can get a 30% return on investment for 5 years, we are going to need a net profit of $630 million each of those 5 years. That is a *lot* of money considering the NASA budget for ISS is only about $3 billion per year, and remember that this is the profit we need above our yearly costs. Also remember that NASA wants to have more than one commercial station.
Maybe you think that 20% should be a good enough rate of return. That bumps the required net profit down to *only* $450 million per year.
What does the yearly cash flow look like, and is it feasible to net $630 million per year?
We have fixed costs - we need a mission control with a full-time crew, we need backing engineers, we need people to work with NASA and other customers. I'm going to throw the number $50 million at that. And honestly, that doesn't seem very bad - in aerospace terms, $50 million is not a ton of money. Compared to the other costs that we are going to discuss, that amount doesn't really matter.
Looking at the costs for the 6 month missions that NASA wants to fly, that's going to take 1 crew dragon and 2 cargo dragon flights. The crew dragon costs around $288 million and the two cargo dragons add another $280 million, so the total is $568 million.
We are only interested in doing this if we make our required profit of $630 million, so we add that in and get about $1200 million.
That's what we would charge NASA for 1 astronaut spending 6 months on research along with the two support astronauts.
The obvious thing to do is to send 2 researchers, as that will cost pretty much the same amount and half the per researcher price.
Can we save money on the transportation? Maybe, but note that our flight rate is less than what SpaceX is currently flying to the ISS and they have to prove to NASA that they are still crew certified, so there's probably not a lot of room there. Worse, we are absolutely beholden to our transportation company. If they bump up their prices, we are in trouble. If they are grounded, we're in trouble. Etc., etc.
Can we make the numbers better with commercial clients?
First, will anybody buy that $600 million dollar extra seat to send somebody to space for 6 months? The answer to that is probably "no", so we need shorter missions. Let's see how two week missions work out for a 3 month timespan.
For our first launch, we fly with two of our astronauts and two tourists for a two week stay. The next group launches with one astronaut and three tourists, and we start doing astronaut rotation to always keep two on station. We do four flights with 3 tourists, and then the last flight only has two so that we can bring both astronauts home.
That gives us 16 tourists across 6 flights.
We are only going to be flying Crew Dragon on the assumption that the flights are short enough that we can carry both people and expendables. The cost for those flights is $1728 million. Add in our profit, and we get $2358 million, which comes out to about $148 million per seat.
We will somehow need to have SpaceX figure how to fly a crew dragon flight every 2 weeks.
Oh, and why did I limit it to tourists?
Well, I can see some countries springing for this mission, and I guess I'd call them "governmental tourists". What I don't think we'll see is any researchers because you can do a ridiculous amount of research on the ground for $148 million.
What if we can support a station that has 6 months with NASA astronauts, 6 months with tourists. This is a bit complicated, so bear with me...
The 6 month NASA visit requires 1 crew dragon and 2 cargo dragons, for $568 million.
The 6 month tourist season requires 12 crew dragons for $3456 million - and when we add that to the NASA cost, we get $4024 million. Add in our profit requirement, and we get $4654 million.
We somehow need to allocate that cost across the two programs to come up with a price. I'll just do it proportionally, and we end up with $328 million for each NASA astronaut that is there for 6 months and $125 million for each of the 32 tourists.
That's the plan. But is it a good plan?
Let's compare the tourist cost of a space station visit to the alternatives.
We decided that the cost per seat for our space station was $125 million.
The obvious alternative is a free-flying crew dragon mission.
The cost for inspiration 4 was less than $200 million, which puts the cost per seat at around $50 million. That's for only 1 flight per year.
The inspiration 4 flight is only 4 days in a small capsule, while our space station is a full 2 weeks in a much bigger station. But is it $75 million per person better? That seems like a hard sell.
Inspiration 4 was a one-off mission, and while SpaceX might be subsidizing the cost, we know that the Axiom flights to ISS are perhaps $55-60 million per seat. Crew dragon flights is a very good model; SpaceX is reusing the infrastructure that is already paid for through the NASA flights and can set their price based on the incremental cost to do another flight and still make money.
If nobody wants to do a dragon flight in a year, it doesn't cost them anything, and if there turn out to be 4 flights of people who want to do this every year, they can scale up.
Our space station tourism model is horrible; the $125 million number depends on us finding 32 people who want to fly with us on 12 flights a year.
If we only fly one mission a year, we need to get $245 million a seat to make our required profit. That *is not* going to happen.
This tourist model is broken. I don't see any world in which we can get even 16 people a year to fly for those prices.
Which brings us back to the NASA model.
For the ISS, NASA has optimized their costs well by maximizing the utility of the very expensive crew transport flights.
NASA flies two 6 month missions with 4 crew on each mission. Roughly half their time is spend on maintenance and operations, giving then two crew years to do research every year.
For the commercial stations, NASA wants to fly one astronaut for a 6 month stay. Given that adding an extra astronaut means a small increment in cost, I'd expect they would fly 2 , and they would be mostly free to do research, which gives NASA 1 crew year of research.
The cost for the ISS is about $3 Billion, while the charge to NASA for our station is $1.2 Billion, given us $1.5 billion per research year for the ISS and $1.2 billion per research year for the new station.
This ignores the development cost that we said NASA would be paying part of.
So maybe it's a little better, but it's not a lot better. More than half of the ISS cost is transportation, and while NASA spends a lot operations, they have an existing station and don't need to make a return on the construction costs to make the investment worthwhile.
You may have noticed that I didn't talk anything about commercial research and manufacturing and how they might occur on our low earth orbit space station.
The problem is that there is no killer app in these areas at the current transportation costs, and we know this because if the app existed, it would be flying already.
But that is a subject for a future video...
I had originally intended to talk in a lot more detail about what each of the 4 - now 3 - groups are planning with their stations, but as I dug deeper I realized that the concept didn't really make much sense.
The big issues that I see are:
NASA's specified level of participation is insufficient to make the economics of a private space station work.
ISS costs a lot, and NASA already needs to fund the deorbit vehicle. It's likely that commercial funding will be slow, and that makes a poor case for the providers.
Transportation is just too expensive. If you are NASA you can pay those kinds of prices but it's very limiting.
Building a station and operating it are very different businesses, like building a hotel and operating one. The companies who can build a station do not have that sort of experience.
NASA wants the providers to build an ISS-like station and take on the bulk of the financial risk.
The reality is that the current approach isn't workable, and reportedly NASA has already iterated on it without getting to a better place.
The cost structure, the NASA requirements, the overhead of running a station, and the commercial need to make a profit all seem very hard to reconcile. The non-NASA market does not exist and the current transportation costs seem to make it unlikely.
Predicting the future where congress is involved is always difficult, but I see two likely outcomes when ISS is gone.
The first is that NASA builds an ISS version 2. It would likely be smaller than the ISS, but since your current transportation system flies four astronauts, it needs to be sized for that.
The second option is that NASA says "we've spent years and years studying humans in low earth orbit; let's take that money and spend it doing something more impactful".
The decision to move the money to exploration would obviously require the approval of congress, but it's pretty clear that the current exploration budget is not sufficient for the plans that NASA is showing here, despite what NASA says about the fiscal year 2025 request.
If you liked this video, you should know what to do...
This set 9030 piece set with hand-made items is available for only $759.
With the recent announcement of more details on Neutron there has been considerable discussion about this new vehicle. I've seen one question come up repeatedly - how will they compete with SpaceX?
I've been planning a Neutron video for a while, but there hasn't been enough technical meat to dive in and do the kind of analysis I like to do. However, the question of competition is a fun one as it revolves more about markets and overall design than specific technical details.
Before I start, in the interest of full disclosure, I own a few thousand dollars of RocketLab stock, so it's fair to say I have a small vested interest in their success...
I'm going to start by talking about competition in general.
If we are thinking of entering a market, we need to know a few things.
Who are the customers?
What are they looking for?
How do they segment?
When we look at cars, we know some people are looking for basic transportation, some are looking for hauling capacity, some are looking for performance, and some for prestige. Those same sorts of concerns apply to launch customers.
If we don't understand the customers, what they care about, and how they differ, we can't understand the market
Second, we need to understand the existing companies that serve that market - the competitors.
What are their strengths?
What are their weaknesses?
How do they differentiate themselves?
Let's look at some details on three companies that you might be familiar with:
McDonald's, Qdoba, and Taco Bell. They are all in the food business, but each is in a slightly different food business.
McDonalds has three things they focus on:
Cheap prices
Speedy customer service
Universality of taste
If you've visited McDonalds, you can tell that this is their focus, or to use a business term, their brand identity.
Qdoba is very different. They focus on:
Freshly prepared, flavorful food made in-house
Menu differentiation
Hospitality
And finally, Taco Bell focuses on:
Cheap prices, high value
Product innovation
Social media engagement
Try to compete with McDonalds on low prices, they will walk over you. Try to out-social-media Taco Bell, their customers will laugh at you.
This leads to a strategy that the business world calls "product differentiation", but I refer to it as "go where they aren't". The goal is to avoid the areas where your competition is strong and focus on the areas where they are weak.
I talked about where SpaceX is strong in my video on competing with SpaceX - see the link - but to summarize, they have 3 major areas of strength.
The first is Dragon - with cargo dragon and crew dragon for ISS, they have contracts with NASA that nobody is going to take away for the rest of the ISS lifetime. And they are selling some non-NASA flights as well.
The second is the national security space launches for the Space Force. NSSL is very hard to qualify for because it requires high reliability and high vehicle performance.
The third is geosynchronous satellite launches. The barriers here are lower because there are no long-term contracts, but big communication satellites inherently require big launchers and you need to compete with both SpaceX and Arianespace.
Conversely, if we look where SpaceX is less strong, we find a few opportunities:
On small/medium payloads, they are less exciting. They have flown rideshares, but for many companies the constraints of the rideshare aren't great - they want control over their orbit and their launch date. And most don't need the performance of a Falcon 9.
There are constellations like OneWeb and Project Kuiper that are in direct competition with SpaceX's Starlink and those companies would rather not give their launch business to SpaceX. That is why Neutron is branded as "the mega constellation launcher".
As NASA transitions away from ISS and towards commercial space stations, NASA is going to want multiple suppliers for those stations (the way they do for commercial resupply now), and the pure commercial companies will want the same. This could be a big growth area. There are also opportunities for flying crew.
In addition, there are other NASA payloads; NASA's Launch Services Program is specifically designed to make it attractive for new companies and new launchers.
Why is there an opportunity for competitors?
Falcon 9 was originally designed as a rocket that could garner a commercial resupply contract from NASA and compete with existing launchers on price.
SpaceX also hoped they could use it to develop reusability
After they got Falcon 9 working for reuse they ended up freezing it at what is called block five, and they froze it for a couple of reasons:
The first reason is that NASA wanted a fixed design so that they could understand the safety aspects for commercial crew
the second reason is that SpaceX had decided that they were going to invest their engineering resources on Starship
The important point here is Falcon 9 is not SpaceX's best partially reusable design; it's a modification of an expendable design. They got very good benefit out of it, but if SpaceX did a second partially reusable design, it would be a much better partially reusable rocket.
This is what is known as a market opportunity; as a business you look for competitors that are distracted.
But what about Starship? How will Neutron compete with Starship?
Here are my top 6 reasons why everybody should ignore starship for now.
What starship is changes ever month or so. You cannot come up with a business plan to compete with a company that is a burger house one week and then a seafood restaurant the next.
The launch market doesn't like monopolies. Everybody wants second suppliers. The real question we should be asking is whether there is anybody who can compete with Neutron, because they will become the default second supplier
We have no idea what Starship will cost per flight nor how much it will carry. How could you possible make plans when the price might be $50 million or it might be $10 million.
Starship is arguably the hardest technical thing ever done in space, harder than apollo or shuttle. It is stretching SpaceX to the limit, and it's not clear if/when they will be successful.
SpaceX cares about Mars more than anybody else, and Starship is designed for that environment. Do you think the market needs a second Mars rocket? Or something else?
Do you honestly think you can compete with SpaceX on something like Starship? "Hubris" means "excessive pride or self-confidence". And yes, I'm looking at you, project Jarvis.
Why Blue Origin keeps taking their current projects that aren't progressing well and making them harder baffles me.
Let's talk about the cost structure for reusable rockets. I've come up with three cost rules:
The first rule is that you can never be cheaper than the cost of the parts you expend. if you have a second stage and it's expendable your cost can never go lower than the cost of your second stage
The second rule is that you can never be cheaper than your per-flight recovery costs. However you recover the reusable parts, you need to pay that cost every flight.
The third rule is that you can never be cheaper than the cost to refurbish the parts you recover so that you can fly them again.
We can write that as an equation; the cost floor for a launch is the cost of the expendable parts plus the cost of the recovery plus the cost of refurbishment.
On the other hand, the cost of the reusable booster is spread over the number of flights it completes, so you can reduce its cost simply by flying it more.
This means that the ultimate cost per flight is determined by the cost floor, not by the cost of the booster.
What are the main drivers for the Falcon 9 cost floor?
There is the cost to make the second stage, which is expended on every flight
There is the cost to recover the booster and fairings. This requires drone ships, support ships, dock space, etc. , which have significant fixed and incremental costs.
There is the cost to refurbish the booster engines, which use a propellant mix (RP-1 and liquid oxygen) that deposits soot in some parts of the motor.
These are significant costs that impact the cost floor, and therefore the overall cost of the system.
A Falcon 9 booster likely costs $30 million or so, and if it flies 10 times, the capital cost of that booster is $3 million per flight
The flight costs are therefore dominated by the cost floor, which is likely three to four times that amount.
Before discussing the specifics of Neutron, a brief digression on aerospace materials will be useful
Aluminum alloys have been used for quite some time in rockets. They have a high strength to weight ratio, decent heat resistance, but manufacturing and assembling them is challenging.
The upper right shows a ULA rocket tank using what is known as an "orthogrid" design. It started as a big flat slab of aluminum and the inside of each rectangle was machined away, and then it was bent into the appropriate curve.
The lower right shows the inside of a Falcon 9 tank; SpaceX uses sheet aluminum that then has strengthening stringers attached to it. Both of these methods are slow and expensive.
To weld together aluminum tanks, a technique known as friction stir welding is often used, where a spinning tool with a pin is applied to the seam with enough force to melt the two sides of the seam and they then cool and are welded together. This produces very strong welds but the process is slow and the equipment is expensive.
Stainless steel has very high strength, is cheap, has great heat resistance, and is easy to machine and weld. It is unfortunately quite heavy.
The right picture shows the inside of a centaur stage during manufacturing, and this stage has been built for more than 50 years, dating back to the 1960s.
The left picture shows part of the SpaceX's Starship vehicle being welded out in the open air.
A third option is carbon fiber composites.
They have great strength/weight ratio - perhaps 20-40% lighter than the alternatives - and this can be easily tuned to meet the expected loads. They also allow great design flexibility.
However, they require a significant amount of equipment to fabricate, changes are harder to make, and they have poor heat resistance.
They are very widely used in aviation - the Boeing 787 is about 50% carbon fiber by weight, and Boeing has used the design flexibility to produce a very distinctive and efficient upswept wing.
In rockets, they have mostly been used as simple tubes, with some solid rocket motor casings and of course RocketLab's Electron rocket.
Bicycles provide a good example of what can be done with carbon fiber.
This bicycle from the 1980s; it uses steel tubes of various diameters, that are then joined together with brazed connections.
The bicycle on the right is one of the first carbon fiber bicycles. Note that it looks very much like the steel bicycle; the big difference is that it is using tubes made of carbon fiber, but it does not take advantage of any of the design options that carbon fiber provides. This is exactly the way that carbon fiber has been applied to rockets; a lighter material for big tubes, but still big tubes.
This is the bike that I currently ride. Note that very few of the frame sections are tubular and most of them taper. The frame is designed to provide strength where strength is needed and compliance/flexibility where that is needed. Not only is the size and shape of the tubes designed, the alignment of the carbon fiber in each section is controlled to be strong enough and yet very light.
The bike on the right is a time-trail bike, designed to be as aerodynamic as possible, and therefore as fast as possible. The designers have really used the flexibility of carbon fiber to produce a bicycle that is optimized to go fast.
That is the sort of design that RocketLab will be applying to Neutron; it won't just be a "better and lighter tube" design, it will be a "use all the advantages that carbon fiber allows" design.
To build the Neutron, RocketLab will use what is known as "automated fiber placement" - it essentially uses a robot that places the carbon fiber on top of a mold, building up both the appropriate thickness of carbon fiber and the appropriate fiber alignment.
This video shows Lockheed Martin using automated fiber placement to build a small airframe. The Neutron approach will be similar but scaled up for the size of Neutron.
RocketLab has made a number of design choices to make Neutron as cheap as possible. Before we look at them, I want to review the three design rules, expressed slightly differently:
Minimize the cost of the parts that you throw away
Minimize the cost to recover the parts that you reuse
Minimize the cost to refurbish the parts that you reuse
Do that, and you will minimize the per-launch cost
Neutron features a fairing that is integrated into the booster. This means that fairing recovery is automatic - the fairing ends up wherever the booster ends up, and there is no refurbishment required.
There is a support structure between the second stage and payload, and this support structure will attach to the booster at the base of the fairing petals.
This approach moves the mass of the structure that supports the second stage and payload from the expendable second stage into the reusable booster, and it also moves the fairing mass to protect the payload from aerodynamic forces to the booster as well.
This allows the second stage to be much lighter, cheaper, and better performing.
The neutron will always return to land at the launch site. That eliminates the expense and complexity of recovery at sea - landing platforms, support ships, docks, cranes, and transport vehicles.
It also reduces the time it takes to get a booster ready to fly the next payload, as it would normally take days for a ship to return, and this may allow fewer boosters to fly the same number of payloads.
Landing on a drone ship downrange is a *performance* optimization. Returning to the launch site is a *cost* optimization, and RocketLab has chosen the cost optimization over the performance optimization.
This is exactly the choice that SpaceX has made with Starship and Super Heavy.
Neutron uses integral carbon fiber landing legs that are part of the airframe and will be lighter and require less refurbishment than the folding legs used on the Falcon 9.
Something is missing from the Neutron launch site - there is no launch tower, merely a short launch platform. Both the booster and second stage will be fueled through the booster. This will limit the amount of pad maintenance required after a launch.
It's not clear yet how Neutron gets on the launch platform, nor how the second stage and payload are placed inside the booster. One option would be the approach that ULA uses on the Atlas V, where the rocket is assembled and the payload encapsulated inside a building and then rolled out to the pad.
Neutron takes a page from the Falcon 9 design when it comes to engines.
It uses 7 engines in the first stage, as that is enough to do propulsive landing. The new Archimedes engine uses a tried-and-true gas generator approach, just like the SpaceX Merlin used on the Falcon 9, but it will be burning liquid methane and liquid oxygen (known as methalox) instead of RP-1 and liquid oxygen (kerolox). Liquid methane burns much cleaner and therefore reduces the amount of engine refurbishment considerably.
It is reportedly a lightly stressed engine, which will make development quicker and easier and means less refurbishment. It has an ISP (add link) of 320, which is probably the sea-level engine in vacuum. That's better than the 311 that the Merlin gets, but much less than the 348 that the Raptor gets.
The best comparison of this engine is probably the Merlin 1A that SpaceX developed for the Falcon 1. A light, simple, reliable engine that can be easily reused.
The second stage will use a vacuum variant of the same engine.
From the details that have been disclosed, Neutron will be an optimized version of the Falcon 9; it will waste less money on the second stage and fairings, spend far less money on recovery and refurbishment, and it will benefit from both SpaceX's experience with Falcon 9 and RocketLab's experience with carbon fiber and Electron.
The answer to the question of whether Neutron is Falcon 9 Mark II is "Yes", and that's why Neutron will win in the market...
Starship is a tall and thin vehicle, and it appears to be much tippier than the Apollo lunar module.
Will it fall over when it lands on the moon? Can we figure out how stable it really is?
To answer our question, we will need to build a model of the center of gravity of the starship lander.
When building a model, it's often useful and/or required to make simplifying assumptions to make the calculations possible or - in my case - simple enough that I don't lose interest.
But it's also possible to go too far and end up with a model that is divorced from reality, such as a spherical cow of uniform density.
I'll be pointing out some of my assumptions along the way. Some will be wrong because we don't have enough details about starship, some wrong because I needed to simplify things, and some wrong because I don't understand things well enough.
The goal of this exercise is to create a model that is useful enough to get a sense of the right answer, without using a spherical starship.
Here's my simplified model...
Starship is a 50 meter tall tube that is 9 meters in diameter.
If we unrolled the stainless steel used to make that tube, it would be 50 meters by 28 meters, for an area of 1400 square meters.
It is made out of stainless steel that is 3 mm thick, and that means it uses 4.2 cubic meters of material.
301 stainless steel has a mass of 7880 kilograms per cubic meter, and therefore the total mass is 33,000 kilograms or 33 tons. The center of gravity is right in the middle of the tube.
There are two other significant mass concentrations.
There is a payload section that starts roughly 32 meters off the ground, and I'm going to assume the center of mass for the payload is a few meters higher, at 37 meters. The payload is 50 tons, which will leave plenty of room for the most important payload items.
There is engine mass right near the tail, with a center of gravity at around 3 meters. Six engines at about 2 tons each would be 12 tons. And yes, the raptors might be lighter than that, but there's other mass back there as well.
How do we figure out the center of gravity? There are multiple methods, and I've chosen one that I hope is understandable.
We lay the rocket down on its side and treat it like a balance - each of the masses that we calculated are like weights hanging from the rocket at a specific point.
There are 12 tons of engine hanging 3 meters from the left end, 33 tons of structure hanging at the middle at 25 meters, and 50 tons of payload hanging at 37 meters.
Let's say that we add a support point - a fulcrum - 12 meters from the left. Will this balance? Instinctively, we would say that the right side is much heavier. But can we come up with a mathematical solution?
We can express this in terms of torques.
To the left side, we have a weight of 12 tons 9 meters to the left of the fulcrum, giving us a torque of 9 * 12 = 108 ton meters.
On the right side, we have 33 tones 13 meters to the right of the fulcrum, giving us 429 ton meters, and 50 tons 25 meters from the fulcrum, giving us 1250 ton meters.
Clearly this does not balance.
If we define C to be the distance of the fulcrum from the left side, we can write the contribution of each mass in terms of that distance.
For the first term, it is 12 tons * 3 minus C meters. Note that this gives a negative torque because the weight is to the left of the fulcrum.
The vehicle is 33 tons * 25 minus C meters, and the payload 50 tons times 37 minus C meters.
Add these up, and we get the net torque - how much the forces are twisting the rocket. Our goal is for all those forces to balance and for the net torque to be zero.
I'll skip simplifying the equation to solve for C - I did it by hand but wolfram alpha would be happy to help you out - and the answer is 28.5 meters.
The center of gravity of starship is 28.5 meters up from the bottom of the rocket.
Now that we've figured out the rough center of gravity, we can figure out how stable starship will be.
This is actually pretty simple - as long as a vertical line from the center of gravity to the bottom of the rocket is inside the base of the rocket, it will be stable.
As soon as it moves outside the base, bad things will happen.
We can find the threshold - what I'll call the critical angle - by drawing the following triangle. The vertical distance is 28.5 meters, the horizontal distance is half the diameter, for 4.5m.
I'm sure everybody remembers from their high school trigonometry that the tangent of the angle is equal to the length of the opposite side divided by the length of the adjacent side, and from that we can calculate that the angle is 9 degrees.
Which would seem to validate our belief that starship makes a tippy lander.
I'm guessing that some of you noticed that I forget the landing legs, which was deliberate.
All we know about the legs is renders like this one and that's hardly definitive, but based on what we see here, I estimate that the base diameter of the legs is 16 meters and that the legs raise the body of the rocket by 2 meters.
That changes the side of the triangle to 30.5 meters and the base to 8 meters. We can redo the calculation, and we get an answer of 15 degrees.
And therefore we can definitely say that Starship can lean at up to 15 degrees before it falls over, and we have finished our discus...
Yes, you in the back, with your hand up. What do you want?
What do you mean "what about the propellant?" Yes, yes, starship needs to have propellant to get back into orbit, but how could we possibly figure that...
The rocket equation? Okay, we'll try that.
We need to figure out how much propellant starship will have in its tanks when it lands on the moon. If you want to understand more about the upcoming calculations, see "planning your solar system road trip".
The required delta v to get from the lunar surface to a 100 km orbit is about 1800 meters per second, and getting to near rectilinear halo orbit is another 450 meters per second, for a total of 2250 meters per second.
Here's that rocket equation.
We refactor to solve for the mass ratio.
Plug in the delta v that we need and the specific impulse of the raptor engine, 363.
And we find that the mass ratio needs to be 1.88
In our model, the dry mass is 95 tons, so to get that mass ratio we need 84 tons of propellant.
We know the mixture ratio of the raptor is 3.6 parts liquid oxygen to 1 part liquid methane, and that comes out to 78% liquid oxygen and 22% liquid methane, or 65.5 tons of liquid oxygen and 18.5 tons of liquid methane.
That is the minimal amount of propellant that the lunar starship will need to get back into orbit.
How much does that fill the tanks?
65.5 tons of liquid oxygen at a density of 1142 kilograms per cubic meter means that we have 57 cubic meters of liquid oxygen, and in a 9 meter tank that is 0.9 meters tall.
18.5 tons of liquid methane at a density of 439 kilograms per cubic meter means we have 42 cubic meters of liquid methane, and in a 9 meter tank that is 0.66 meters tall.
Basing the tank locations on drawings of starship, that adds 66 tons of mass at 5.5 meters and 19 tons of mass at 17.3 meters.
Do the calculation, and we find that the center of mass is now 19 meters off the ground, or nearly 10 meters lower.
Redoing our calculations with the lower center of gravity, we find that a starship without legs can tilt up to 13 degrees without falling over, and one with legs can tilt 21 degrees.
How does this angle compare to Apollo?
The Apollo LEM was quite resistant to tip-over, requiring a full 40 degrees before it would tip over. Which is quite a bit higher than lunar starship.
But the LEM had a takeoff limit of either 12 or 15 degrees depending on who you listen to. Which seems pretty high, but on Apollo 15 the LEM ended up at an angle of 11 degrees, which is close to the limit.
Will lunar starship tip over?
Starship has a low center of gravity and would need to land at an angle of more than 20 degrees to be in danger, and that seems unlikely.
If you enjoyed this video, conduct your own stability experiments using an empty paper towel tube, four quail eggs, and 135 cc of chicken broth.
The current starship is 9 meters in diameter. SpaceX has plans to stretch both starship and super heavy to make them slightly taller to carry more fuel, but Musk has talked about versions that would be bigger in diameter.
Starship is already the biggest rocket ever and handling it poses some unique challenges. Why would you want to go bigger?
It turns out that there's a pretty good reason to consider going bigger, and it's all about propellant tanks, because rockets are mostly tanks with engines stuck on the bottom.
Real tanks have domes at the top and bottom to better deal with pressure, but it's simpler to just consider the tanks to be cylinders. Doing the same calculations with domes is left as an exercise for the viewer.
What we want to do is figure out how efficient a tank is - how much material it takes to create a tank that can store a given volume of propellant
We'll be looking at cylinders based on their height and their radius, which is half of their diameter
The volume is equal to the area of the end of the cylinder, which is pi times the radius squared multiplied by the height h.
The surface area is composed of two components. The area of the end - pi times the radius squared - multiplied by 3 because the tankage has two ends plus a wall between the two propellants plus the side of the tanks if you unrolled them which is the perimeter of the tank multiplied by the height, or 2 pi times the radius times the height.
We want to maximize the amount of volume and minimize the surface area to minimize the weight of the tank. There's an assumption there that the tank material is the same thickness throughout and I'm going to ignore the presence of any stiffeners in the tank.
How do the volume and surface area relate? What size tank would be the lightest for the volume that we need?
Rocket people often publish the mass of propellant for a stage, but they often do not publish the volume.
We can figure that out with some rocket math, and I'm going to walk you through it so that can make appropriate calculations when you are designing your next rocket.
We will start with our usual example, Falcon 9.
The Falcon 9 first stage carries - according to the figures I found - about 419,000 kilograms of propellant. It carries that in two separate tanks - one for the liquid oxygen and one for the RP-1 kerosene fuel. Those two propellants are different densities so we need to know how much of each we are carrying, and that's going to depend on the ratio of those two propellants that our engine burns.
That is known as the mixture ratio of the engine, and we can generally look it up. For the Merlin 1D engine, it's 2.36. Note that mixture ratios vary widely between different fuels because of the underlying chemistry.
What that number means is that for every 1 kilogram of fuel, the engine will consume 2.36 kilograms of liquid oxygen. That's a total of 3.36 kilograms of propellant, so we divide that 419,000 kilograms by 3.36 and get 125,000 kilograms of RP-1 and the remaining 294,000 kilograms is liquid oxygen.
We now know how much mass each tank holds but we want to know how much volume that takes up, so we will need to know the density of each propellant.
Liquid oxygen's density is easy to find - it's about 1141 kilograms per cubic meter. RP-1 presents a challenge - it's not one molecule but a mix of a whole bunch of different molecules so the density can vary - but I'm going to use 806 kilograms per cubic meter.
A simple division and we find out that we need 258 cubic meters to store the liquid oxygen and 155 cubic meters to store the RP-1.
That would be the right answer for most rockets, but SpaceX has to be different. They take advantage of the fact that if you chill the propellants, they get denser and you can therefore stuff more mass into the same space.
This works pretty well for liquid oxygen and you can get about 10% more density, up to about 1255 kilograms per cubic meter. RP-1 is less chillable as if you cool it too much it turns to gel, so we only get about 2% more density, up to 822 kilograms per cubic meter.
Redo the math, and we get 234 cubic meters for the liquid oxygen and 152 cubic meters for the RP-1. It's about 6.5% savings in volume.
Here's a summary of the improvements you can get through subchilling. The liquid hydrogen one looks useful but liquid hydrogen is so cold it's a huge effort to just keep it liquid so I'm not sure if subchilling would be practical.
Anyway, we now have a way to estimate how big the propellant tanks need to be for any rocket stage.
Now we can calculate how much material it will take to hold the propellant for a specific stage.
Excel coughed up this chart, starting with a very thin 2 meter rocket on the left and a fat 10 meter rocket on the right.
We can see that the total material required is at a minimum at around 5 meters in diameter.
The actual Falcon 9 is 3.7 meters in diameter, which means that the tanks uses about 13% more material than a 5 meter one would.
The obvious question at this point is "why isn't Falcon 9 a 5 meter rocket?"
The usual explanation is that it's because of the US road system.
Falcon 9 stages are made in Hawthorne California and need to be shipped either to nearby Vandenberg air force base or all the way across the country to Florida, and you can't practically truck rockets bigger than 3.7 meters.
5 meter rockets are generally shipped by ship.
So SpaceX accepted a slightly heavier rocket so that they could make transporting them much simpler.
That's the story why Falcon 9 is so tall and thin, but there's a little more nuance than that.
This is the same chart for the second stage tanks. Note that the minimum area is pretty much exactly at the 3.7 meter stage diameter.
So the SpaceX Falcon 9 has a second stage designed to be as light as possible - and therefore as performant as possible - and the first stage is then sized to be capable of doing what it needs to do to put the second stage in the right location. Efficiency is much more important for the second stage.
Interestingly, if you look at Falcon 9 version 1.0, you'll find that the second stage carried only about half the propellant of the V1.1 version. A 3 meter rocket would be better for performance for the V1.0 version. That suggests that the V1.1 size was what SpaceX had planned all along and just needed to wait for a higher thrust version of the Merlin engine to start flying it.
My usual explanation for Falcon 9 using an weird propellant combination for a second stage - RP-1 kerosene and liquid oxygen rather than the far more common liquid hydrogen and liquid oxygen - has been that it was easiest and quickest for them to adapt the merlin engine to be a vacuum engine, but this shows another reason - you can build a nice beefy second stage for a 3.7 meter rocket if you use RP-1 as the fuel.
Because SpaceX chose a small diameter rocket, using RP-1 for both stages was an excellent choice. If they had chosen to use hydrogen, the stages would have been optimal at 5 meters in diameter and therefore poor performers at 3.7 meters.
Speaking of hydrogen, let's look at the Vulcan rocket from ULA.
Here's the second stage for Vulcan. It's 5.4 meters in diameter, so it's about 10% bigger than the optimal size.
Its optimal size is actually pretty close to the Falcon 9 diameter, but because Centaur burns hydrogen, the centaur tanks are 50% bigger than the Falcon 9 ones but only hold half the propellant by mass.
If we look at the first stage - which burns liquid methane and liquid oxygen - the diameter of Vulcan is pretty much the optimal size.
With Vulcan, 5.4 meters is great for the first stage but a little big for the second stage. My guess is that they made that choice because they wanted the very large 5.4 meter fairing size, and to put that on top of a small second stage on top of a big first stage would cost more in mass than just making centaur a little bit wider than it would like to be. Vulcan also stages much later than Falcon 9 so the first stage does more of the work and the second stage is less important.
But what would you do if you had a case where you needed to have efficient sizes for 3 stages?
Like, say, you were building a moon rocket.
The first stage burns RP-1 kerosene and liquid oxygen, and the minimum material for the tank is pretty much exactly at the 10 meter size of the stage.
The second stage switches over to liquid hydrogen as the fuel, which naturally means that it is going to need more space for the fuel and a largish diameter.
The second stage stays at the 10 meter diameter - it could have worked a tiny bit better at 8 meters but the difference is small and 10 meters makes the overall rocket a little shorter.
The third stage also uses liquid hydrogen but carries less propellant.
It not surprisingly prefers a smaller diameter of about 5 meters, but it's pretty good at the actual 6.8 meter diameter. If it had maintained the 10 meter diameter, it would use 50% more material in its tanks than it did at 6.8 meters
The third stage really needs to be light and efficient, so the tank size choice needs to be close to optimal.
And that explains why the Saturn V has that weird step between the second and third stages.
Wasn't this talk about bigger starships?
Well, yes, but I needed to cover how tank sizing works I got carried away because it's an interesting topic.
We'll start with the same chart for starship
This is the Starship 1 variant that flew in IFT1-4.
The 9 meter tank is a little big, but it's a small difference.
I don't show the newly proposed versions on the chart.
The Starship 2 version carries just a little more fuel and is pretty much perfect.
The Starship 3 version would prefer to be 10 meters in diameter, but it's only a 2% difference in tank area.
This suggests that SpaceX had the Starship 2 version in mind but needed to make sure they had enough engine thrust to make it work. Sounds similar to Falcon 9.
For the super heavy first stage, it's the same story as the Falcon 9 first stage; it's smaller than optimum so that the second stage can be close to optimum. The best size would be about 11 meters.
The starship 2 and 3 versions would be happier at 12 meters, but they stay at 9 meters.
Back to the original topic, which is whether physics prefers rockets of unusual size and therefore would a larger starship be better?
Or, to put it another way, how do our calculations change as the diameter gets bigger.
I've rearranged the equations a bit so that that things are a bit clearer.
If we look at these two terms, both have pi times r squared in them, but the volume is h times that and the surface area is only 3 times that, so if h is more than 3, increases in the radius will cause the volume to increase faster than the surface area.
If we look at these two terms we can see that the first one uses radius squared while the second one uses the radius, so the once again the volume will increase faster.
But what if you put the two surface area terms together?
There are mathematical ways to answer that question - you calculus folks can figure it out - but I prefer a good graph.
If we arbitrarily choose a 10 meter high tank, we see that the volume increases faster than the surface area, which means that bigger tank diameters are better and therefore bigger rockets are better.
But it's not quite that simple...
We know that the optimal diameter depends on how much propellant we want to carry, so we want a different graph that is specific to starship
This is a graph that shows us how many tons we can put into a tank per square meter of tank skin at different propellant amounts for a rocket burning liquid methane. And this is aimed at Starship, not at Super Heavy, because this measure is far more important for Starship.
We see two things.
The first is that you do need to choose your rocket diameter based on the amount of propellant you want. If you are only carrying 500 tons, you get far better performance from a 6 meter rocket than the larger sizes.
The 9 meter diameter is not surprisingly the best choice for 1200 tons of propellant carried by the current starship. It's also the best choice for the 1500 tons that starship 2 will carry. Interestingly, the 2300 tons that Starship 3 is currently projected to carry is right at the transition point where a 9 meter and 12 meter size are roughly equal. You can still get gains at 9 meters and larger propellant amounts but you would do better with a bigger rocket.
The other thing to notice is that as you increase propellant, a large diameter becomes a better choice at a specific propellant range. I'm a bit surprised to find that 12 meters is still the best choice even if you go up to 4800 tons of propellant, which is 4 times what starship currently carries. That is a much bigger rocket.
And because I know somebody is going to ask, 15 meters reigns from 4800 to 9200 tons, and then 18 meters takes over
And of course there are intermediate curves for diameters between the ones I list, but 5 lines on a chart is already too many.
It is, of course, more complicated than that.
We also need to talk about tank pressure.
In pressure tank design, tank wall thickness goes up as tank size increases, so the gain that you see in terms of more propellant per square meter may not help if you need to increase your tank wall thickness.
The question is whether thinking of them as pressure tanks is the right thing to do. Liquid propellant ullage tank pressure - the pressure above the propellant - is fairly low for pump-fed engines - somewhere from maybe 15 to 60 psi, or only 1-4 bar. That's not a lot, and these tanks have to carry significant loads through their sides when the rocket is operating, and those loads likely set the tank thickness more than the pressure.
And just to make it more confusing, the pressure at the top of the tank is just the gas pressure but the pressure at the bottom of the tank is dependent on how tall the tank is the density of the propellant. The pressure is equal to the pressure of the gas at the top of the tank plus the density of the liquid multiplied by the height of the fluid in the tank multiplied by the acceleration due to gravity. That means there is higher pressure at the bottom of the tank and on the sides right next to the bottom than at the top.
And to add another factor, the acceleration due to gravity goes up during the flight of the stage which will increase the pressure on the tank but the height of fluid in the tank goes down during the flight which will reduce the pressure, and that will complicate the calculation of the maximum pressure in the tank.
It's complex, but since the engineers seem to believe that bigger diameter is still better from an engineering perspective, I'm going to go with them.
There are, of course, other considerations...
That huge factory that SpaceX built at Boca Chica is sized to deal with 9 meter rockets, both from a diameter standpoint and from a height standpoint. The buildings would have to be expanded to deal with a bigger rocket.
It's no surprise that super heavy doesn't get much longer because there's not much room left in the high bays for a taller stage.
Starship and super heavy are transported on these large self propelled transporters from Roll Lift. Even at 9 meters, they require the whole width of the road plus the shoulders. Bigger means bigger infrastructure.
And everything at the launch pad needs to get bigger.
There are also environmental concerns. A bigger rocket means higher peak sound levels at both Boca Chica and Kennedy, and it also mean a larger blast damage footprint and larger "keep out" zones.
It's not clear what kind of vehicle Starship is at this point.
Is it a regional jet like the bombardier CRJ700 that is used primarily on short routes.
Is it like the 737 or Airbus' A320, the absolute workhorses of the industry.
Or is it like Airbus' A380, a plane that ended up too big for the market.
At this point we don't know. If starship is the 737, maybe it doesn't get any bigger.
We therefore end up with a big question mark. There are some advantages to bigger vehicles but there are also some significant costs and other issues that would come with a larger diameter.
My guess is that they will likely stick with the 9 meter diameter for the near future - say the next 5 years - until they really understand the system well. After that, I don't have a real opinion. Build hundreds or even thousands of the 9 meter variant so that they are cheap, or upgrade to a 12 or 15 meter vehicle so they are individually a lot more capable.
We'll just have to wait and see...
If you enjoyed this video, please send me this poster....
Let's say that you want to build a rocket. There are lots of different rockets out there, with different sizes, different fuels and engines, different payload capacity, different materials.
How do you figure out what the best choices are?
You need to do a tradeoff analysis
You may have done this when purchasing a car.
You come up with a list of requirements - how much it costs, how many seats it has, how many cupholders are available. These are the things that you care about in a car.
You then create a list of cars and rank each car based on the requirements.
That gives you a nice matrix of data and - based on how important each requirement is to you - you can figure out which car is the best choice.
The same basic idea works in aerospace.
If you are developing an uncrewed aerial vehicle - the fancy name for a large drone - you come up with a list of requirements and, in this case, a value that tells you how important each requirement is.
You then spend a lot of time coming up different approaches that might meet your requirement. Unlike cars - which already exist - these vehicles are proposals, and you need to spend enough time on analysis to be confident that the designs will work the way you expect them to work.
Then finally, you can rank the designs on the requirement. In this case, it's a heat map that shows you what measures are good and what ones are bad.
System analysis tradeoff studies is a bit of a mouthful, so this is often abbreviated to trade studies, or simply "trades", as in "we were looking at different cushioning materials for landing and the trades said that marshmallow was superior to chocolate ganache".
If we look at the process, trades might take one person a few days, or a significant team might spend a few months
Trades use the best information available, often proprietary information that is only available inside an organization
Trades often require developing a sophisticated model to predict behavior.
Trades are done by experts
All that means that trades give the best guesses available. They are still guesses, just highly educated ones. That is why plans change - a company learns more and it turns out that the conclusion from the trades changes.
Let's see how I stack up to these.
I might spend a couple of days working at something, but never more than that
I only have publicly available information to work with
My models are simple and designed to get a feel for the behavior of a system rather than real numbers
I'm not an expert
So we can expect that the results I get aren't very good; I simply don't have the information or training to do a great job.
But it's still interesting to try to understand why certain tradeoffs are made, as long as we keep in mind the limits of our analysis.
I also highly recommend you watch my video "you suck at space", which is all about why humans tend to be poor at understanding the limits of their expertise.
And that's how trades work.
My meta point - the reason that I made this video - is that companies very rarely make decisions without a lot of hard work and analysis, so if their decisions seem strange to us, it's almost always because we are lacking data.
If you enjoyed this video, please send me this poster....
After the starship orbital test, there were a lot of strong opinions flying around the subreddits, the twitters, and the facebooks.
I read a bunch of remarks and it was pretty clear that there were a lot of people out there who did not have - how can I say it nicely - a good grasp of the underlying issues.
Hence the title of this video.
However, I didn't think anybody would interested in watching a video about my shocking assertion that there are actually people on the internet who are ill-informed, so I've chosen to do something that I hope will be more constructive...
I'm going to talk about three questions...
Number 1: Why do people behave this way. How can they be so certain and so wrong?
Letter B: How can one self-identify being in this state?
Roman numeral III: What can one do to be better informed and move out of this state.
We'll start with the first question, which gives me the opportunity to talk one of my favorite studies in the field of psychology.
This study was published way back in 1999, and the topic of the study was humor.
Two researchers came up with 30 jokes that they believed exhibited varying degrees of humor. (Humority? There's strangely not a good word for this in English).
They were not experts in humor, so they then recruited 8 professional comedians who rated each joke on a scale of 1 to 11, with 1 meaning not at all funny and 11 meaning very funny.
The paper is silent on why they chose that range, but my guess is that it had something to do with Nigel Tufnel.
That gives them 30 jokes with expert ratings of how funny each of them is.
They then recruited 65 psychology undergrads from cornell university, and asked them to rate the jokes.
That gives us one expert set of ratings, plus a set of ratings for each student.
Those of you who know about research methods can predict the next step - they took the expert ratings and the students ratings for each joke and compared them, and produced a correlation number that indicated how closely each student joke rating correlated to the expert joke rating.
Those correlations are averaged, and that gives a rating on how well the student aligned with the experts - a sort of "humor quotient"
They then took the students and divided them into four groups, or quartiles, based upon those quotients. The 25% of the group that got the lowest scores went into the bottom quartile, the next highest scores in the second quartile, and so on for the third and top quartiles.
This is merely to group the students together into four groups based on humor quotient. We can then draw this graph.
It doesn't really tell us anything except that different people have different abilities, which is hardly a surprising outcome.
But there was something else they did, one other bit of data that they collected for each of the students.
They asked each student to please compare your ability to recognize what's funny with that of the average Cornell student.
They gave them a 0 to 100 scale, with "I'm at the very bottom" corresponding to zero, I'm exactly average corresponding to 50, and I'm at the very top corresponding to 100.
That number is their humor quotient perceived ability, and we're going to compare it to the rating they got when they rated the jokes.
Here we see the actual performance of the bottom quartile in blue along with their average self-rating in orange.
The bottom quartile had an average performance at the 12th percentile, but they believed their ability was at the 58th percentile for an overestimate of 46 points. They were very poor at determining what was funny, but they thought they were better than average.
Moving up to the second quartile, this group had an average performance at the 37th percentile, and estimated they were at the 60th percentile, 23 points above their actual performance.
In the third quartile, the difference shrinks to 8 points. This group is pretty good at evaluating themselves.
In the top quartile, we see a reversal. Their actual test score is 87 and their perceived ability is only 75, so this group underestimates their actual ability by 12 points.
The paper says the following about these results.
(read)
But maybe, just maybe, humor isn't a good topic of study.
So they did another test looking at logical reasoning. It shows the same effect, but the bottom quartile here overestimates their ability by 56 points.
Some of you have probably figured out what study I'm talking about.
Here's the title of the paper.
(read)
You can see the authors were David *Dunning* and Justin *Kruger*, and this is the paper that popularized what is known as the "dunning kruger effect".
When I talk about this, I get a very common reaction to the bottom quartile.
It's something like, "I work with people like this and they are *the worst*."
Which amusingly, is people demonstrating that they are in the bottom quartile when it comes to their understanding of the Dunning Kruger effect. It's very meta.
The Dunning Kruger effect is about ability in a *specific area*. You might be world-class in one area, not only in the top 25% but in the top 5%. And you might be at the absolute bottom in another area.
Dunning Kruger is a general result - whenever you are working in an area where you are in the bottom quartile, you are not capable of evaluating your proficiency in that area.
It's just part of being human.
And therefore, what I'm saying is that those uninformed opinions that I was seeing are mostly coming from people who are in the bottom quartile of a specific topic.
Hundreds of people suddenly became experts on flame deflectors and concrete repair despite having no real experience or expertise in the area.
If you are in the top quartile of any area, you will notice the people in the bottom quartile. Just remember that you are assuredly in the bottom quartile in other areas.
Which takes us to the second question:
Can one identify being in this state?
It seems problematic - Dunning and Kruger wrote that being in the bottom quadrant means you can't identify that you are in the bottom quadrant.
But I think there are some symptoms that one can learn to recognize.
Number 1: These people are idiots.
If you find yourself with this opinion, you are almost assuredly in the bottom quadrant.
Pretty much any engineering organization will have done a trade study, also known as a "trade-off study" or simple "trades".
(read)
Here's on that I found that looks at different methods of getting from the earth to the moon.
Trades take the professionals a lot of time and effort to do well and the decisions aren't always clear cut.
If you don't understand the alternatives and you think the decision is easy, you are in the bottom quadrant.
If you are wanting to give your "hot take" on a situation, you are in the bottom quadrant.
A related sign is "they should just..."
I was going to call this "simplistic solutions", but "they should just" is so often used that I think it's a better label.
Here are some examples:
(read)
The "just" part is the indication of a simplistic statement about what is almost always a very complex issue. There are generally very obvious reasons that "they should just" solutions don't work.
There is the related "why don't the just?", which at least adds in some uncertainty but is generally the same attitude.
Number 3: Certainty...
Here's a question for you. Is liquid methane a better fuel than liquid hydrogen? There's a very straightforward answer to this question....
It depends...
It depends on a bunch of different things. What is the scenario, what are the goals, what constraints do you have, is cost more important than performance or vice versa, etc.
Experts are experts because they understand that the world is not simple, there are always tradeoffs and they are often complex. That is the whole point of doing trade studies.
If you are very certain about the answer to a question like this, you are likely in the bottom quadrant.
Number 4: Dismissal of Expertise
There are experts on pretty much every topic.
Not only are there books on concrete repair, there are numerous research papers on thermally damaged concrete. Some people literally spend their whole career on this topic.
That is not to say that experts are always right. Just that you need high-level understanding of a topic and the details of the expert analysis to determine if they are right.
I asked on twitter how long people thought it would take to repair the damage at Boca Chica, and these are the responses I got. About 90% of people expressed an opinion, but I'm not clear how anybody came up with a number.
And yes, surveys like this are fun and I respond to some of them myself, but they rarely generate useful predictions.
And our final question, what can one do to be better informed?
First, read the Wikipedia article on a topic. The space-related ones are generally quite good and the vast majority of their information is correct.
This is the entry level to being somewhat informed, and I'm surprised at the number of people who do not take this basic step.
Follow the references and links.
If you really want to understand a topic, you're going to need to read original sources. You will usually find some links from the Wikipedia articles.
#3: Do deeper research
Google scholar is your friend. This will *mostly* point you to more reliable information, though I'll note that you will find a lot of articles and studies that are written by *advocates*, so they may be one-sided. But they are interesting because they are in more depth, though that also makes them much harder to read in general.
You will find that some articles are available in PDF, but many are behind a paywall.
Sci-hub is a site of dubious legality that allows you to access many papers that would normally be behind a paywall. Because of the legal issues, it tends to move around a lot so you'll need to search for it. Try using "where is sci hub" as your search, and if you don't see this page, you're probably on a site you don't want to be on.
If you have a university affiliation, you may be able to access articles that way, and your local library may also provide some access.
If you find a paper that's useful, I highly recommend downloading the pdf and saving it. Not only do papers sometimes go away, it's maddening when you know there's a paper out there but you are unable to summon a query that finds it.
For anything *technical* from NASA, the NASA technical reports server is a good place to start, though prepare to be a bit frustrated by the search functionality.
There's also the NASA image and video library which is either maddeningly incomplete or very poorly indexed; I am regularly frustrated by my inability to find NASA images that I know exist.
Note that the NASA centers maintain their own sites that may also have technical information.
If you are interested in speculative information, there is no better site that Winchell Chung's atomic rockets site.
#6 Don't watch videos...
Strange advice from somebody who has a youtube channel on space, but the percentage of bad space videos on youtube is pretty high.
My real advice is to be choosy - you know the things that help you identify when somebody is in the bottom quadrant of a specific topic, and keep those in mind.
I'm hoping that the information I've presented will help you understand what it's like to be in the bottom quadrant and how to get out of it, so that you suck less at space.
If you enjoyed this video, please write your hot take on it...
There's a widespread belief that NASA is the right US organization to guide human spaceflight - that they are "the pros" at this and everybody should defer to them.
But are they?
This video looks at how reliability is managed in aerospace and NASA's history in that area.
We'll start by defining two terms.
Loss of Crew - or LOC - is any scenario that kills the crew members during the mission.
To pick a lesser-known example, in 1971 the Soyuz 11 crew was killed during reentry when an air vent opened and depressurized their capsule.
Loss of Mission - or LOM - is any scenario that prevents the successful completion of the mission but does not kill any crew members.
In 2018, the launch vehicle on the Soyuz MS-10 flight to the international space station malfunctioned, but the abort system worked correctly and they landed safely.
Here's an relevant tweet.
Kathy Lueders ("leaders") (head of commercial crew for NASA) says that the LOC number of SpaceX's Crew Dragon is 1 out of 276.
One out of 276. What does that number mean?
The implication is that we can fly 276 missions before anybody dies.
But that's not what it really means, for two reasons.
The first reason is because of how probability works, and the second is because of how those numbers are generated
We will start with a journey to the land of math, a land that is populated with many strange and confusing symbols.
We will travel to Probabilistan, and we will start with...
Backgammon...
For our purposes, you need to know two facts about backgammon:
Doubles are very useful, as they give you four moves instead of two
Double sixes are highly prized
We're going to look at the probability of getting double sixes.
We know that the chance of rolling a six is one in 6, so the chance of rolling double sixes is one in 36, or about 0.028 (2.8%)
How many times do I need to roll to get double sixes? 1 in 36, so 36, right?
Any backgammon player can tell you that it's not 36 times. You might roll them the very first roll of the game, or you might roll 100 times without getting them.
Here's a graph from a simulation that rolled dice until it got double sixes and tracked how many throws it took. It did this 10,000 times. The green line shows 36 throws.
What can we figure out from this chart?
First, ten percent of the time we will thrown double sixes in the first 4 rolls. That's very lucky if we are playing backgammon, very unlucky if we are flying a spacecraft.
Second, if we throw the dice 25 times there is a 50/50 chance that we get double sixes. I think the 50/50 point is probably the most useful one to think about from a risk perspective; if our risk is 1 in 36, we have a 50% chance of a LOC in 25 flights.
As a rule of thumb, take the denominator (36 in this case) and multiply it by two thirds, and you'll get pretty close to the 50/50 number.
Third, the 90% point is at 82 throws. We have a 10% chance of rolling or flying 82 times or more before the happy or unhappy event occurs.
Note also that this smooth curve comes from doing this 10,000 times. We generally don't have that amount of data.
There are two places these numbers come from.
Empirical numbers come from measurements in the real world.
Analytical numbers come from performing analysis of the underlying system.
We'll start with some empirical numbers from the aviation industry
Aviation doesn't use the term "loss of crew", but we can come up with equivalent numbers.
The Boeing 737 was launched in 1967, and it had 50 fatal events in about 58 million flights. That gives us a measured LOC of one in 1.6 million.
In 1997, Boeing launched the 737-600 series, and it had 12 events in 100 million flights, for a LOC of about one in 14 million.
That's an order of magnitude improvement in about 30 years, and it came from improvements in design but also because of improvements in operations, maintenance, and flight crew training.
How good are those estimates? Look at the numbers in red. Both planes had a number of incidents and that makes it more likely that the numbers we calculate are good. As backgammon players know, while the outliers do happen, over time everything balances out towards the 50/50 spot.
It may be useful to ask a different question - what range of values gives a 95 percent chance of the range containing the real number
This is calculated using a confidence interval, where our real number is 95% likely to be between the two limits. Note, however, there's still a 5% chance that it could be outside those limits.
If we apply that calculation to the 737-600, we find that there's a 95% chance that the real number is between 1 in 4.7 million and 1 in 16.1 million.
Let's look at two more planes...
In 2016, Boeing launched the 737 Max. It had two fatal events in the first 650,000 flights, for a measured LOC of one in about 325,000. Much, much worse than the 737-600 series.
But the most interesting example is the Concorde. It first flew in 1969, and flew for 31 years with a perfect safety record for 88,000 flights. Then it crashed on takeoff in 2000, and that gives it a measured LOC of one in 88,000.
Let's add in the 95% confidence intervals. The 737-MAX has a range of 1 in 87,000 to 1 in 2.6 million. It's unlikely that it is close to the safety of the 737-600.
The concorde has a range from 1 in 16,000 all the way up to 1 in 3.4 million. It could be a very risky plane, or it could be safer than the early 737. We simply do not have enough data.
This demonstrates a significant limitation of the empirical method; it doesn't work well with small numbers of events.
What if events are very rare or it's a new vehicle?
You can use a technique called probabilistic Risk Assessment, which is designed to evaluate risk with complex technology.
To greatly oversimplify the process, risk is computed using two quantities:
The severity of possible adverse consequences
And
The probability of occurrence of each consequence.
This is often done with fault trees analysis.
This is the fault tree diagram for a rocket engine on an ejection seat. The top node shows the overall failure, and the three nodes underneath show situations that would lead to the overall failure:
The rocket engine is not connected to the ejection seat
The rocket engine explodes
The rocket engine does not work
Underneath each of those are sub-scenarios, and then, sometimes, sub-sub scenarios, with the tree continuing many levels deep
Each of those sub-scenarios has a specific probability estimated for it, and then those probabilities flow up to provide an estimated overall probability - in this case, 1 in about 77,000.
Each probability comes from an analysis of how likely a specific issue is.
What are the limitations of this approach?
First, it's complex, expensive, and slow - and it needs to be kept up-to-date as designs change.
Second, the results depend on the expertise and the imagination of the analyst; scenarios that the analyst does not think of do not exist in the analysis.
Third, it is hard to capture operational issues. Here are a few examples of the kind of issues that are difficult to foresee:
In 2015, the second stage on SpaceX's CRS-6 mission exploded because struts from a supplier did not meet specifications.
In 2013, a Russian Proton rocket exploded because three sensors were installed in an improper orientation.
In 2019, a Boeing Starliner abort test parachute failed because it was not connected to the parachute harness.
Because of these limitations, PRA provides a lower bound for the risk; the actual risk may be higher.
Now let's move to some examples at NASA.
Early in Apollo, NASA commissioned General Electric to do a probabilistic risk assessment for the apollo program.
The number NASA got back from GE said that there was a 10% or less chance of success. NASA was worried about the reaction such a number would create if widely known, so NASA kept the number secret.
And they decided to avoid PRA analysis in the future.
We therefore have no updated PRA for the actual Apollo missions, but we can maybe estimate it.
Apollo had 11 missions with humans on it.
In 1967, a fire broke out in the Apollo 1 capsule during a test and killed three astronauts.
In 1970, Apollo 13 blew up an oxygen tank in the service module, which very nearly became a LOC incident.
That gives us a LOC measurement of 1 in 11, maybe 2 in 11 if you count Apollo 13.
What did NASA learn about reliability from Apollo?
Here's a quote from Will Willoughby, head of reliability and safety for Apollo
Statistics don't count for anything... they have no place in engineering anywhere.
The Apollo approach rolled right into the shuttle program
With their disdain for statistical methods like PRA, NASA instead chose failure modes and effects analysis.
FMEA looks at three things:
The severity of the effect on the customer
The likelihood of occurrence
The ability to detect the failure
These are all ranked on a 1-10 scale, and then multiplied together to produce a risk priority number.
Here is a table that describes how to come up with those three numbers.
Let's look at a couple of examples:
We have a minor defect that shows up occasionally and we can not detect it. That gives us an RPN of 2 times 4 times 10, or 80
Or
We have a defect that is immediately hazardous, shows up occasionally, but we can always detect it. That gives an RPN of 10 times 4 times 1, or 40
Wait, what? A minor problem that is very hard for us to detect rates *higher* than a major issue that we can always detect? That might rank a problem with the shuttle toilet higher than a failure in the maneuvering thrusters that would prevent the crew from returning
This is one of the issues with FMEA. Some organizations instead use a risk ranking table that ignores the ability to detect number.
There's another issue because FMEA is done by people, and most people are human. As humans, we like the middle of ranges - the 4-7 range in these rankings. They are comfortable. The big numbers are less comfortable and likely require more effort to justify, so the ratings we give are going to be pushed more towards the middle of the scale. That tends to mush things together.
The biggest issue, however, is one you've probably already noticed. FMEA provides no risk estimate.
How did that work out for shuttle?
Well, at the beginning of the space shuttle program, NASA was quoting an LOC number of 1 in 10,000 to 1 in 100,000. Getting close to airplane numbers. To put that in perspective, the 1 in 10,000 number means that if you flew shuttle weekly for 133 years, you would have a 50/50 chance of losing a crew.
Then Challenger happened on the 25th flight of shuttle
Richard Feynman, the well-known and well-respected physicist and Nobel Laureate was part of the investigation commission, and he did a deep investigation into the reliability of shuttle and wrote a detailed analysis. Here's a link if you are interested.
As part of his analysis, Feynman polled NASA employees on what they thought the LOC number was. Management's number was 1 in 10,000.
Engineers estimated 1 in 50 to 1 in 200.
Feynman finished his analysis with the following: (read)
I would be remiss in my responsibilities if I mentioned Richard Feynman and didn't mention two very interesting and enjoyable books. His later books are also good.
After challenger, NASA started doing limited PRA analysis, but they didn't complete a full PRA analysis until 17 years later, in 2003, the same year of the Columbia Disaster.
That analysis was updated as the program progressed. Interestingly, they didn't just evaluate the current PRA but updated the PRA for the whole program lifetime, and that gives us a very interesting diagram.
*****
https://ntrs.nasa.gov/api/citations/20110004917/downloads/20110004917.pdf
We'll start with the early years.
The first obvious question is "where is challenger?" It shows up right here, with a risk of 1 in 10.
The estimate for the whole early period is a 1 in 10 probability of LOC. Note that there was almost a LOC event on STS-1; I'll put a link to my video in the upper corner.
After challenger, the risk only decreased from 1 in 10 to 1 in 17, despite the widespread impression that it was a much safer vehicle.
This middle section is concerning; that jump from 1 in 38 to 1 in 21 is an increase of 80%, due purely to a new way of applying foam to the external tank.
A bit later we see Columbia, in a section labelled at 1 in 47.
And finally - right at the end - we end up with the 1 in 90 figure, but it really only applies at the end.
The empirical value - two failures in 135 flights - is about 1 in 68. Given the numbers on the chart, it's tempting to suggest that NASA got lucky, especially given the other close calls in the program.
*****
https://ntrs.nasa.gov/api/citations/20110004917/downloads/20110004917.pdf
https://www.thespacereview.com/article/3785/1
As shuttle was winding down, NASA was doing a lot of work on the upcoming Constellation program. The astronaut office noted that it would be nice if the next launch system was less likely to kill astronauts, and proposed an LOC goal of 1 in 1000.
The Ares 1 rocket was the crew delivery vehicle in Constellation, and it was built using a Shuttle solid rocket motor as the first stage, a single RS-25 space shuttle main engine in the second stage, and then an apollo-ish Orion capsule with escape system on top.
NASA presented their reliability numbers in the final study on Constellation, and estimated the LOC number at 1 in 2021.
Note that this number is 22 times better than the shuttle numbers.
Here's one of my favorite quotes from the Constellation reliability study.
Predictions of future reliability must go beyond the raw data of history and capture the improvements in design, production and management that tend to make the next flight more reliable than the average of the previous flights.
This basically says that we should ignore the historical numbers - shuttle numbers, I assume - because we've learned so much since then. Despite the fact that shuttle showed that this assertion was very much not true - the shuttle did not get more reliable because of the flights before Challenger or Columbia, and the 1 in 90 number is at the end of shuttle, reflecting a more mature program than constellation.
When the Ares I solid rocket design was announced, the crew survival office went nuts - at least, "went nuts in a very NASA engineer way"
Here are some excerpts from their report.
The whole constellation program was cancelled, but that's a story for a different video..
After constellation, nasa was working on the space launch system - or SLS - and continuing work on the Orion capsule.
Detailed information on PRA efforts for SLS and Orion were hard to locate.
I found this table from 2013 that looks at the ascent reliability. If you take those numbers and calculate an aggregate reliability, it comes out to 0.993429, or 1 in 152. Note that this number is just for ascent.
The Aerospace Safety Advisor Panel Annual report for 2014 published this table, which looks at the LOC for a trip around the moon, similar to what the upcoming Artemis 1 mission will do.
The panel wrote "The ASAP was pleased to note that the NASA exploration systems development division has now established LOC probability thresholds for its programs. The panel was less pleased that these thresholds were not significantly safer than the actual historic performance of the space shuttle.
*****
https://foia.msfc.nasa.gov/sites/foia.msfc.nasa.gov/files/FOIA%20Docs/42/SLS-RPT-077_SLSP-Reliability-Allocation-Report.pdf
See table 3-1, multiply the reliability of each item (including two boosters and 4 engines), and you get 0.993429, or 1 in 152
2014 NASP report: https://oiir.hq.nasa.gov/asap/documents/2014_ASAP_Annual_Report.pdf
Section 3
In 2011, the commercial crew program came along, and Boeing and SpaceX were awarded contracts. As part of the contract, the contractors needed to meet a LOC of 1 in 270 for the entire mission including ascent, a 9 month stay on station, and reentry.
This has proven to be quite challenging but SpaceX has been validated to have an LOC of 1 in 276, meeting the requirement.
Starliner has not yet met it's development milestones and therefore it does not have a public LOC.
*****
https://web.archive.org/web/20140113160636/http://www.airspacemag.com/space-exploration/Certified-Safe.html
I knew a lot of NASA history, but I was surprised when I looked at the details.
NASA does not have an enviable safety record.
Apollo was somewhere around 1 in 11 for safety, and the shuttle was 2 in 135. Overall, NASA lost 17 humans.
The Russian Soyuz is around 2 in 156, which looks comparable to shuttle, but both losses were early in the program. Since 1971, they have been 0 in 146. Overall, they lost 4 humans.
NASA ignored the best technique for risk estimation for years.
SLS is purportedly only as safe as shuttle, 1 in 75
On the other hand, commercial crew has a threshold of 3 times safer than shuttle at 1 in 270. SpaceX has delivered that with Dragon, and Boeing is currently under certification of that Starliner.
Why is NASA the expert at this? My point is not to denigrate the skills or dedication of NASA employees, but as an organization the s have not been impressive.
If you enjoyed this video, please tell your cat
Interestingly, an issue cropped up with Ares I.
The thrust on space shuttle rocket boosters oscillates at about 11 Hz. The Space shuttle has two boosters where the oscillations of the two boosters will tend to cancel out the vibration rather than reinforce it.
The shuttle stack weights about 880 tons at SRB burnout while the Ares I stack weighs about 260 tons at SRB burnout.
That means that the vibrations due to thrust variation will be around 3.4 times worse with Ares I compared to shuttle even without the smoothing effect of two boosters.
Here's a modeling of the Ares I ascent. It shows near the end of the solid rocket booster the oscillation will be an 11 Hz vibration of 5-7 g.
There have been studies that looked at the effect of vibration, and once you get to 0.5 g, you see significant effects. Above 3.7 g is likely to cause health problems.
Ares I is expected to be 5-7 Gs. This is not only bad enough to break astronauts, it's bad enough to break equipment as well.
Before we move on from constellation, I do want to share one chart with an important point.
This is for a lunar launch mission. Note that the booster launch only provides 3% of the risk in the mission; the lunar operations are much more likely to lead to LOC events than the launch.
The point here is a risk that might be high on a mission to ISS might be minimal on a lunar mission.