The reasoning behind SLS sort of made sense 10 years ago. We didn't know that Starship, New Glenn and Falcon Heavy were coming down the pipe. We decided that the US & NASA needed super-heavy lift capability. Previous attempts to design a heavy lift vehicle were lost to political regime change and congress in-fighting. Previous attempts to design a single-stage-to-orbit vehicle were cancelled due to its virtual impossibility.
So NASA & Congress designed a program that was unkillable. It greased too many wheels and lined too many pockets to make it easy to kill. It was also designed ultra-conservatively using mostly existing designs so there was no technical risk.
So now when we complain that it's a pork-filled boondoggle that's impossible to kill, that was the plan, and there was a certain logic behind it. It makes no sense in a world with Falcon Heavy, Starship & New Glenn, but who would have predicted that with confidence in 2010?
The Constellation Program [1] that preceded SLS had two vehicles - Ares I for ISS crew and supplies after Shuttle was retired, and Ares V for occupying the moon. The missions gave some justification for needing vehicles. Their design was primarily a means to preserve existing contracts and jobs. Additional wheel greasing made it happen.
That stuff about re-using parts to reduce risk is just what people say because it sounds good, but isn't inherently true. Kind of like how "drug delivery" and "machine learning application" are go-to pseudo-justifications for many research proposals.
Constellation was eventually canceled when ISS resupply went to the Commercial Crew Program [2]. Ares V was stripped down and reimagined as the mission-less SLS. SLS is not really intended to do much more than exist for the time being.
There are a lot of factors that will go into canceling SLS - infighting between NASA centers, preserving jobs and contracts with Old Space companies, even more wheel greasing. I'm interested to see how it plays out.
I was going to mention the issue with reuse of old designs “being conservative” but you got there first. Mandated reuse of some component of a prior vehicle is, by definition, an extra constraint on the new design. Extra constraints make design harder, not necessarily safer, and the use of old stuff in the new vehicle to some extent kills the “flight proven-ness” of the old thing.
(Generally speaking. Maybe some of this is less of a factor or more depending on the system... point is you can’t say up front that old hardware is a good idea)
Thank you for elaborating on that point. From some of the other comments, it seems that it's maybe a contentious issue.
There's a lot of pressure to include hand-wavy references to "heritage hardware" in the risk assessments for various reasons, but it's not well-substantiated where it exists.
I’d imagine it’s partly due to the view of the public that NASA’s launch tech has regressed since the Apollo program. The Saturn V had crew escape, could deliver 110 tons to leo, and was cheaper per launch than the shuttle. The F1 engine is still considered superior to the shuttle main engines. Pitching new tech when you haven’t matched capabilities from 60 years ago is hard.
Combine this with incumbents looking to keep new entrants out of the market and heritage, and reuse become the operative words
Shuttle was canceled because its bay wasn't needed, it was overly expensive, and its service record was pretty accurately tracking the projected 1 shuttle loss for every ~50 flights. There wasn't a strong commitment to the missions at the time, so making something completely new was out of the question. But there was a lot politically and strategically riding on Shuttle, so they had to find some way of keeping some lights on. Those points can be elaborated, but the decision really wasn't much more complicated than that.
The F1 and SSME are apples and oranges. They have separate uses and achievements for their times. I haven't heard of either being widely considered objectively superior to the other.
At what stage was a shuttle loss per 50 flights projected? Can you expand on this, do you have a link? Was that due to the boosters or some other design constraints? Would a smaller shuttle have fared better in terms of reliability?
They started PRA of Shuttle after Challenger [1]. I got that number from the SLS PRA and Reliability people, and might have read it in a report too, but I don't remember many more details about it. There's a lot that goes into PRA [2].
Shuttle size was dictated by intended cargo and maneuvers, so it's not really possible to say how changing size would affect it since you'd have completely different design constraints and objectives.
The F1 isn't superior to the SSME. It doesn't really excel at anything (other than being very big maybe.)
From another article on that website:
"The RS-25 is still considered to be about the best engine ever made with a fairly high thrust to weight ratio and unmatched efficiency." [1]
IIRC the F1 still holds a few thrust records (thrust per one engine maybe ?) but is certainly much less efficient than modern engines from Spacex/Ble Origin & Russian construction bureaus (much worse ISP & TVR).
Too big, too expensive, too low flight rate. These all problems contribute to each other.
Also, large solid rockets are very troublesome. They can not be fueled on the pad, so they are heavy and dangerous from casting to stacking to launch. There will never be a high launch rate reusable rocket with solids.
Big was the whole point, people decided we needed the capability to send large masses beyond Earth orbit in a single launch. Anything smaller would mean "giving up" on human spaceflight to the moon or Mars. Doing it in a single launch is silly, but that thinking was widespread at the time.
And a low flight rate was also a given. Delta IV Heavy only files about once every 18 months, there's insignificant demand for anything even bigger. That was why NASA decided they had to do this themselves, because it would be insane for any commercial company to build a rocket without a market.
And expensive was also part of the plan, you can't spread enough pork to make it unkillable if it isn't expensive.
Big, expensive, low flight rate. Those weren't problems, they were requirements.
I think you misunderstood what I said. I understand that NASA thought it needed the vehicle. However I'm saying it did not in fact need it for any missions. (The idea for docking in earth orbit at the start of a mission is not new.)
Very different thing, and you actually demonstrate those points.
Everything starts with the misplaced thought that it needs to be so big. Not actually required.
Probably correct thought that it needs to be NASA-made since it wouldn't make sense commercially, but this is a consequence of misplaced thought 1.
Expensive to spend pork. Again, not needed for the mission, if something more sensible was chosen in the first place.
There could have been a capsule flying on Atlas V for many years already. The ESAS study just had all kind of really bad assumptions that made Ares happen instead.
Yes, bad assumptions led to a bad result. Some of those bad assumptions made sense 10 years ago. Some of them didn't. I was far too generous to Congress by only talking about the former and not mentioning the latter. Multi-launch for large beyond-Earth missions made much more sense 10 years ago than a big rocket. Heck, multi-launch probably made more sense in the 60's than Apollo, if it wouldn't have been for that 1969 hard deadline.
Multi-launch was considered for the Apollo program, but was ruled out because it had higher technical risk than just scaling up the rocket and making the payload extra light. If we didn't have the hard 1969 deadline we would likely have gone with the multi-launch option back then.
It's an interesting alternate reality to imagine: we might have had a more interesting moon program that had the ability to naturally scale to larger and larger missions, instead of repeating basically the same thing until the public gets bored and cancels it.
Yeah, first orbital docking was done 1966 on Gemini 8. The architecture for Apollo was decided much earlier already (it did include the lunar orbit rendezvous and docking which was indeed risky).
We now have 54 years of repeated low earth orbit dockings so maybe the architecture designs could start taking that into account.
Even SpaceX is planning on doing orbital refueling.
There were plans for more interesting moon missions under Apollo Applications Program, cancelled because of budget cuts. (and what was left was botched as Skylab)
Big was political. It was a requirement added because SLS could provide it, not because it was needed. It was there to stop the order-of-magnitude cheaper per pound smaller launchers from being proposed instead.
Part of all this was strong arm twisting from the politicians involved to stop all work on in-space propellant transfer and propellant depots (because with them, you might as well launch lots of smaller cheaper launchers to carry up the propellant.)
Right. I will go further: if SLS is the best that could be done, then human space activities themselves would make no sense. SLS is a solution to no real problem beyond "how do we deliver $$$ to the appropriate pork consumers?"
If you believe humanity has a future in space, then you must believe it's possible to do much better than SLS. So do that, not waste time on a pointless and hideously expensive diversion.
> It was also designed ultra-conservatively using mostly existing designs so there was no technical risk.
Politicians can't design a rocket and cannot get what is "ultra-conservative". Rockets are not Lego blocks that you can just mix and match. They're explicitly designed a certain way to support a certain load profile. Even if the parts of the SLS _look_ like they have low risk, they do not and that's why the vehicle has taken so long to develop.
One example, if the shuttle boosters were could actually have been used directly they wouldn't have needed to extend the number of segments from 4 to 5, redesign the liner between the segments, and completely redesign the nozzle. It's a new booster.
Second example, even though the shuttle external tank looks like it's being used, in previous iterations the shuttle external tank did not support any axial loads, it was simply held on to the bottom of the Orbiter. Now the external tank has to withstand the entire axial load of the rocket so it's basically redesigned from scratch.
Yes, it's still rocket engineering. It's never trivial.
What would have been a more conservative design? Reusing the Apollo design was considered and discarded because so much knowledge and tooling had been lost that it was considered to be less conservative than the SLS design.
I love the whole "lost Saturn V engineering" trope. It sounds so good despite being bullshit. Rocket dune and NASA did an extensive knowledge retention project for the F-1 and J-2 engines. NASA also maintains several copies of extant F-1 and F-1A engines. If we approached Rocketdyne tomorrow with a billion dollars they could restart F-1A production. The J-2X is in fact the second stage engine for the SLS stack.
The SSME was selected for the SLS (and Ares V before it) because it had twice the burn hours of the F-1 and the SRBs existed and were a known quantity. The SLS (and Ares V) can reuse more of the Shuttle's infrastructure than an F-1 based design. The SRBs allow for use of smaller first stage tankage and the less powerful SSMEs which allow for reusing Shuttle launchpads which replaced the Saturn launchpads at LC-39.
While the SLS is moving into boondoggle territory the Ares V and SLS designs made economic sense to reuse Shuttle designs and infrastructure. It didn't have anything to do with the myth of lost engineering knowledge.
Perhaps the problem was deciding that the U.S. needed to have a super-heavy lift capability in the first place. A lot of the U.S. space program seems to be rather aimless, doing things because they seem cool rather than practical. We shouldn't really be surprised when situations like this occur.
I strongly disagree with this opinion. The US space program is incredibly practical, particularly in the last 3 decades.
* An unbelievable amount of scientific data about biology and microgravity comes from the ISS. To say this is “aimless” only is true if you believe humans will always live on Earth and only Earth.
* A wealth of fundamental knowledge about the universe comes from spaced based telescopes. These are helping us understand the laws of nature, some of the most raw R&D.
* There are significant military applications to our space program. These range from GPS, surveillance, communications, and potentially weapons. There are some government space launches that we know very little about their mission because it is highly classified. The Space Shuttle had a use case for rapid troop deployment. There is significant military applications to the engineering knowledge of rockets. The difference between a rocket and a missile is mostly the application.
* Our exploration of our solar system provides possible future avenue for economic development. There is more fresh water, raw minerals, and energy inside our solar system but outside of Earth than there is water, mineral wealth, and energy on Earth. Certain things we consume are non-renewable, and we will need to find sources for them outside of Earth.
* It gives hope, inspires young people to become engineers and scientists, and gives us a feeling of awe as we go through the process of understanding our existence in the context of the cosmos.
EDIT: This isn’t to say there is major bloat inside the space program, and that NASA is not in need of a significant culture change.
I mostly agree, but what in the world do you mean by "The Space Shuttle had a use case for rapid troop deployment"? I can't picture that being anything but the worst way imaginable to get like 2 or 3 guys somewhere.
The key word here is likely rapid. Doing a suborbital hop is much faster than going through atmosphere, because you get to accelerate to much higher velocities. Even supersonic planes take forever to get anywhere, while a suborbital hop would let you get to any place on the planet in under 2 hours.
There never was a space shuttle sitting around ready to launch on-demand though. A 2 hour travel time is irrelevant if it takes weeks to prep for the mission.
The space shuttle could also only land at a handful of runways on the planet.
I’m having a hard time finding the source, but a mission requirement of the Space Shuttle was something along the lines of “Deploy a dozen marines to Moscow in less than 1 hour.”
Yeah, that's just mythology. The concept doesn't work out.
The shuttle did have a military requirement: the air force wanted to be able to launch into an orbit over the south pole, then up across Russia from the south, to return and land after that single orbit. That pretty much forced the use of wings that could glide, vs capsule style re-entry, and opened the path to most of the shuttle's design flaws.
As for why the air force wanted this, the presumption is surveillance, but I don't think they've ever officially answered, other than denying that they ever considered a bomber mission for the shuttle.
IIRC the requirement was to be able to snatch a Soviet satellite & land wit it in less then one orbit. Here it is demonstrated in KSP (with a SSTO, not shuttle though): https://youtu.be/pFCQEgnKwdY
On the other hand Soviets were reportedly scared to death of shuttle starting on a "normal" mission only to use it's wings and heat shield to dip into the atmosphere and change it's inclination, so that the new orbit passes directly over Moscow, from an unexpected angle. Then it would drop a 20 ton worth of RVs with nukes in a decapitation strike without any warning.
It is estimated that this is one of the reasons why the Energie & Buran program were started - they also wanted part of the "fun".
Maybe somebody wrote a requirement about that once, but it doesn't sound very practical. I know they were dreaming up a lot of fanciful things when they were setting the requirements though.
I think the actual shuttle had a life support capability for like 7 people. Not sure of the actual flight crew requirement, there's probably room for like 3-4 other guys. It can land on a runway, but it's gotta be really long, and there's only one shot, and it'll be really obvious and has no ability to evade defenses. No chance of it landing in Moscow unless the Russians let it. And then you'd be in the middle of a huge military base or something, and what are 4 guys going to do except be killed or captured immediately.
Maybe they could get there in an hour. From a specific launch pad in Florida, and if the shuttle is all set up to launch, which takes months. And it'll be blindingly obvious to everyone in the world that something just launched heading straight at Moscow.
You could put a crew module into the payload bay. Also the Shuttle's life support systems were designed for multi-day missions so they may have been able to handle a two hours flight with more people.
> There are significant military applications to our space program.
It's interesting looking at the military and space, since the military seems to be more objective focused than NASA. Early on (1960's), the military was planning on a manned station (MOL), but ended up going with automated satellites in the end because they made more sense. Likewise the Space Force believes it might eventually send up astronauts, but doesn't think that will happen for decades. They're interested in sending people up when it's useful to do so, not just sending them up for the sake of sending them up.
Just to play devil's advocate for a bit: "space fans" excuse the wastefulness and obvious corruption of NASA.
"Microgravity research" is entirely pointless. A total waste of time. It's an excuse to stay on the ISS, that is all. There would have to be a thousand problems solved with fundamental human biology for us to live in microgravity, some of which are at the genetic level. Alternatively, and much more simply, we could use rotating space stations to simulate gravity. That requires just a few dozen engineering problems to be solved, none of which are fundamental. Admitting that the problem was "solved" in the 1960s would remove the excuse to keep sending people to the ISS, cut NASA's budget, and retrench a bunch of bureaucrats. So of course, NASA management is carefully ignoring the simple solution so that they can keep receiving funding for solving the difficult problems.
"Space telescopes" are fantastic! We need more of them! But NASA keeps blowing their budget on the aforementioned waste of time microgravity research they're doing on the ISS instead of the much more scientifically useful telescopes. Many have been cancelled or delayed. Remind me, where's the James Webb space telescope again? Oh yeah... on the ground, that's where.
"Military applications" are not a good thing. "Every gun that is made, every warship launched, every rocket fired signifies, in the final sense, a theft from those who hunger and are not fed, those who are cold and are not clothed."
"Resources in space" is commonly trotted out, but first consider that your argument for the militarisation of space is directly contrary to an argument that space has resources. Military spending, as the quote above by Eisenhower points out, is wasting resources that could have gone towards productive uses. But even ignoring that, the problem with resources in space is the ludicrous cost of moving them to where we want to use them, which is here. Even if, say, we get costs down to $50 per pound of returned raw material, there's only a handful of metals that would make that worth the bother, none of which are abundant (or concentrated!) in sources outside the Earth.
My point is that all of these apologetics for space, or more accurately, for NASA's current approach are false.
We can spin space stations, there's no need for microgravity research.
There's no territory in space worth fighting over. No oil wells in the desert we need to bomb. Space doesn't need to be militarised.
There are no places outside the Earth where liquid water had concentrated minerals into worthwhile deposits. A trillion tons of gold is worthless if it's evenly spread out in a billion-trillion tons of rocky asteroids.
Nobody really wants to live in space. If you think that you do, I challenge you to move to a place much nicer than any place in the solar system for a year, like Bouvet Island. It's a barren island just off the coast of Antarctica. Live there for a year, come back and tell me you like it, or shut up about how it's the inevitable future the human race is yearning for.
PS: Meanwhile, SpaceX is doing all of the truly useful things that we need in space. We need a global satellite-based Internet service that isn't slow and expensive. That's astonishingly useful, right now, and isn't something NASA could ever deliver despite a much higher budget than SpaceX.
> We can spin space stations, there's no need for microgravity research.
My understanding is that the problem with 'spin gravity' is that unless you are talking about enormous truly enormous space stations, then the human vestibular system becomes a major issue.
Plus, then there's the issue with ensuring that the station remains balanced, and the additional material strength requirements.
I don't think "spinning space stations" is quite as simple a solution.
A "simple workaround is to have two masses connected by cables, that spin around a common barycenter. That way you can get good fake g at human safe RPM. And it reduces the balancing issues significantly.
While traversing between the two parts might be tricky, the craft could be divided into manned and unmanned part connected by cables, to reduce "traffic" between the two to a minimum.
"On Sept. 14, 1966, the Gemini XI spacecraft is tethered to an Agena target vehicle. Gemini XI command pilot Charles "Pete" Conrad and pilot Dick Gordon are maneuvering their craft to keep the tether taut between the two. By firing their side thrusters to slowly rotate the combined spacecraft, they were able to use centrifugal force to generate about 0.00015 g of artificial gravity."
Why this is not being used now ? I guess because it was not seen as a priority for Apollo and followup (US) missions as they usually took just two weeks at a maximum. Health effects of microgravity have been also understood in more details & how they can be overcome for the short flight durations.
With the ISS & missions lasting half a year (and longer for some crews) artificial gravity would indeed be more useful, but hard to pull off for a big permanent installation like the ISS,
that needs to maneuver to accept spacecraft while also providing reasonable microgravity environment for experiments.
> The Space Shuttle had a use case for rapid troop deployment.
No it didn't. The DoD didn't even want it before it even flew.
Very little of the things you talk about has much to do with the NASA Human Directorate and that is the many part of NASA that people dislike, and this is not just from the outside, even inside NASA this is well known.
> using mostly existing designs so there was no technical risk.
If only that was so. Using a bunch of old tech in a new way does not actually lead to great success. The re-qualification of the RS-25 alone took years and cost 100s of millions. And that is without producing new ones.
SLS is the last major gasp of a military industrial complex that has completely taken over congress. The fact that we're paying $146M for each engine on the SLS (takes 4 to launch) and SpaceX is promising an entire launch for Starship at <100M is obscene. (Oh, and SLS ends up at the bottom of the ocean...Starship you can just refuel).
While I don't agree with all thing Musk, he's revolutionized the global spacelaunch industry by reducing costs 10x. Now, totally dominant in an industry, he's making that rocket obsolete in favor of a better one (Spaceship).
If you think it's the last major gasp, you're gonna find yourself awful disappointed. One does not simply kill a jobs program. There'll be SLS, and then some other 2 trillion dollar airplane, then some satellite system, and again and again forever. Nobody votes to kill these projects because of the jobs they maintain. There will always be more
Find a way to do it keeping the same contractors and adding a couple in some states with senators and representatives who are simultaneously members of the majority party and vulnerable, then maybe you've got a bill, though it might still get filibustered. NASA is a political work of art
There's also a strategic capabilities argument that never goes away.
I'm anticipating a significant investment in synthetic biology facilities in the next few years. Ideally these would be 99% automated, remotely operated via secure dedicated networks, and maintained by small on-site skeleton crews. They'd be geographically distributed, with a majority in 'red states', but not because of any jobs they might create.
I'm not understanding what you're suggesting. What is a "synthetic biology facility" (research center? Manufacturing plant?), who is paying for them, why are there no jobs, why are they in red states, and since when did synbio become a low labor field? If its private sector investing in synbio manufacturing, they'll do that wherever the tax credits are. If it's the government, they aren't going to be doing any manufacturing, and research needs people.
I have no clue what GP is proposing, but "strategic capability" projects tend to be in red states because those are the low-population states where you can spread things out enough to be somewhat resistant to nuclear attacks. Being far from the coast is preferable for similar reasons (more time to intercept missile attacks).
Some of the low population states have had certain facilities allocated specifically because the facilities would be high priority target for ICBMs, thus making them "nuclear bomb sponges".
Yes, if you put your facility in the middle of San Francisco you give the enemy two targets for one bomb. Better make them work for it (and have a slightly higher chance that either your facility or part of San Francisco survives)
The problem with synthetic bio is that it's going to be pretty hard to drag synbio people away from boston, and I don't think the NIH is leaving bethesda any time soon.
I guess my comment is pretty scattered rereading it now - I guess that's because as you mentioned we don't know what exactly was being proposed - synthetic bio is wide enough it could refer to a million different wildly divergent things.
The MIC will find a way to make a flying car/moon rocket that fires Hellfires and a gatling railgun. Let's call it a multi-program total cost of $40 terabucks that could've given every American a nice home (household assets are $130T but this counts both impoverished and wealthy persons) and UBI; but no: self-serving, subsidized wealth transfer to the Cheney's and the Erik Prince's comes first.
The plans for Starship are haisy, not scrutinised and it has not been delivered. So far design has changed massively, it was scaled down a lot, and it's stil a vehicle looking for a purpose.
I have significant doubts that we will see it launch at that price-point withing the next 10 years.
All the best luck to spaceX, but you can't base the entire national space programm on something that flaky.
SLS solidly gets you to the moon, and you van make real plans on it. If it turns out to be redundant, that's ok
I remember when people said that about Falcon Heavy and people were defending SLS by saying 'Falcon Heavy is not real, SLS is real and exists'. 2017 was a simpler time.
You don't need SLS for a moon program, Starship or not. Its a major waste of money and only needed because the whole architecture was designed to need it. It makes neither economic, nor infrastructural sense considering the rest of the infrastructure and American space industry.
Starship will very likely fly, even at 100x over its target price, its a bargain. Putting in 30+ billion into a project, even if we assume complete 100% success from now on, will not succeed in the mission 'going to the moon and staying'. At a same-time invest basically 0 in a system that could literally revolutionize human space travel, is beyond nonsensical.
I would bet a fair amount of money on Starship transporting humans before SLS. SLS Core stage just spent almost a year preparing for a single static fire test, and has already fallen behind again.
> I have significant doubts that we will see it launch at that price-point within the next 10 years.
But what if it is only half as expensive per ton to low-Earth orbit as Falcon-9/Heavy? And if it can only deliver 100 tons at a time, not 150?
That's still going to dominate the launch market for some time.
If they eventually prove it to be safe for human passengers and can demonstrate in-orbit refueling? It then replaces SLS, and is still a fraction of the cost.
While I agree that Starship is very much an early-stage project it has a clear commercial utilization: Starlink.
Assuming demand for Starlink is strong and SpaceX will be able to scale it up like crazy since Starship will give them far more LEO capability than anyone else.
Starlink doesn't need Starship. It works right now and the current SpaceX vehicles are sufficient for what it does. Maybe you could talk about Starship as a means to increase the total receiving area, number of satellites, and thus total bandwidth, users, etc. Maybe it could reduce the necessary size of the terrestrial receiver.
But... this is both speculative and marginal. If Starlink needs more hardware, they can ramp up Falcon 9 rocket use, which is already partially reusable, and will continue to increase reuse. Starship could drive down launch costs, but it can't reduce the cost of the satellites themselves. You'll hit a floor where it could reduce costs, but not remotely enough to justify it.
Starship is on a completely different playing field. It strives for a VERY large payload with a VERY large fraction of reuse. This makes no sense unless actual people are riding on it at some point. The demand for orbital transport is not enough otherwise. It only makes business sense by assuming some future activities will happen which will bring in a massive amount of funding. This is beyond conventional business risk, this is a leap of faith.
> Starship could drive down launch costs, but it can't reduce the cost of the satellites themselves. You'll hit a floor where it could reduce costs, but not remotely enough to justify it.
You're assuming that satellites are so expensive that the difference between Falcon 9 and Starship is not relevant but we don't know this. Satellite hardware is usually very expensive but SpaceX is building them internally as a series product so they might be able to drive the marginal cost/unit much lower.
It's safe to assume constellation bandwidth (and potential revenue) scales linearly with total payload mass, meaning it is proportional to ($satellite_cost_per_keg + $launch_cost_per_kg). Unless satellite cost is much higher than launch cost there are benefits from switching to a cheaper launcher.
Since Falcon 9 design is mostly frozen and still requires throwing away an upper stage for every launch it has a price floor of its own. I honestly wouldn't be surprised if launch is already more expensive than satellites.
I will concede all those points except for the last one.
> Starlink, the NASA Moon program, commercial sat buissness, SpaceX Mars Plans, SpaceX Space tourism and so on. A cheaper vehicle always has more usage.
Go back to basic college Econ. There is a supply curve and a demand curve. Reusable rockets allows providing the same commodity (orbital transportation) at a lower price.
The problem that everyone seems to subtly know but doesn't like to say aloud is that the demand curve is highly inelastic.
It is completely possible that cheaper rockets result in more usage, but less total revenue. This happens with lots of things. Think about food production, weather is bad, rice harvest is very low during a year in a nation that mostly eats rice. We have historical examples that total value of the crop goes up, despite less being available. That's highly inelastic demand for you. If supply of orbital launch goes up, then total revenue can easy decrease.
In this case, the only way for total revenue to increase a great deal is for genuine innovation to happen. Yes, space tourism, Mars plans, etc. would fit the bill. However, all of these are bets, and I don't think they're particularly good ones. The best bet would be for US government itself to realize what is happening and double-down on the military and scientific windfall they can get from it. I don't feel like this will happen on its own, and the general public isn't engaged enough. I worry that SpaceX's success could usher in its own demise, and we lose out on the opportunity of a generation.
With Falcon 9 you always throw away the second stage, even if you reuse the rest - that adds up. Also even though 60 says per launch and the many hundred in orbit are a monumental achievement already, it pales in comparison with their plans. Also the stats have finite lifespan, so with many thousand in orbit even a weekly 60 sat launch migh not be enough to replenish & add more sats over time.
When did that happen? Originally, and I thought still today, it was quite literally the reverse - it was Starlink that was created to fund Starship, and Starship has a well-defined purpose: getting to Mars.
Starship launch price should only be a couple millions (see the article).
<100M is _technically_ right, but very far off. Since Starship will be fully reusable, it's upfront building investment (the ~100M price-tag you are referring to), spread over the lifetime of the vehicle, plus staff/maintenance/fuel for launches.
Falcon 9 was already a +10x reduction (~~1.5B per launch -> ~70M, likely cheaper for reused boosters), and Starship will be another +10x cheaper than F9. This means Starship will be >100x cheaper than competitors (excluding small-sat rockets like the Electron)
Where is that 10x reduction and the ~1.5B per launch from for the Falcon 9 comparison? I mean Falcon 9 certainly was cheaper than the competition even without reuse, but I'm quite sure the difference wasn't as large as you're stating there.
The Falcon 9 did disrupt the industry for sure, but you don't need a 10x price reduction for that.
Of course, a large part of the cost of those engines is that they're designed to be reusable (a difficult task with hydrogen/oxygen engines due to thermal shocks and hydrogen embrittlement). This made sense when they were being built for the Space Shuttle. It no longer makes sense now that they are using the same engines for an expendable first stage.
Which makes the plans to throw them away with each launch even more silly. They are cobbling together jobs programs from space shuttle components with no real strategy except "worked before" and "we can charge a lot of money for this".
Reusability doesn't make engines costly. All liquid rocket engines are reusable. They have to be, so they can be tested before being put on the vehicle, and so during development you can keep reusing test articles.
The Merlin engines on the Falcon 9 are reusable (and on the first stage, they are reused). They are reported to cost SpaceX just $400K per engine to manufacture.
>All liquid rocket engines are reusable.
As always there are exceptions - rs-68 has an ablative nozzle that is rated for a single flight, so you need to test without that I guess.
Other engines could be similarly rated for jinutes of runtime by default for maximum performance. SpaceX aldo repeatedly ststed Merlin and Raptor were designed for cheap and fregvent reuse.
Yes, Merlin was designed to be cheap. That's why it's cheap. Being reusable was not an impediment to it being cheap. Being reusable is not an excuse for being expensive.
A cheap rocket engine needs a cheap development process, which means you want your test articles to last a long time. Far from being an obstacle to cheapness, reusability is a requirement for an affordable engine development program.
Let's go way back to Usenet and listen to Henry Spencer on this subject.
"To reiterate my original point: almost any regeneratively-cooled liquid rocket engine is reusable, even if it was built for expendable vehicles. The design requirement for the F-1 was 20 starts and 2250s of firing; as part of the test program, six of them accumulated over 5000s each."
The Merlin engines don't run hydrogen, though. Because of hydrogen's propensity to embrittle metals it comes into contact with over time, it's much harder to design a hydrogen engine for long runtimes (vs. say, a one-minute static fire plus a eight-minute flight).
Any hydrocarbon engine will have hydrogen in the thrust chamber, and in the gas generator. They all run a bit fuel rich to maximize Isp, and that will mean the hot partially burned gas has some molecular hydrogen in it.
What makes hydrogen engines more expensive is the low density of LH2, which greatly increases the pumping power needed to bring the propellant up to pressure for injection into the thrust chamber.
Reusability is not binary, saying 'all liquid rocket engines are reusable' is pointless. It's like saying surely a Formula 1 car is equivalent to a Tacoma in terms of 'reusability', because both can be restarted. Spoiler: they're not.
His material is SpaceX heavy but far from exclusive, and he works very hard to recognize and compensate for any biases he might have when doing deep dives and analysis.
There's quite a lot of technical material in many of his videos, presented in an organized, pleasant and approachable fashion.
It’s crazy to see this here. I knew Tim in college; it’s crazy because he bought this old cosmonaut suit as a prop for his photography and it just kind of took off on random corners of the Internet and then next thing I know he’s doing this stuff with NASA and now he’s on the front page of HN.
the channel is one of my top recommendations, but I think another thing that should be pointed out is he posts an article version of each of his videos. I really really wish this was more common.
Who would have known the SpaceX experiment would work? There is some case to be made that US Gov was overly conservative and not willing to innovate, but I suspect that's safer than using taxpayer money on a SpaceX-like iniative that suffers as many failures as SpaceX did initially. Can you imagine the political fallout? We don't have an appetite for that from the Gov any more, not like we did in decades past.
NASA and the US Gov made the decision to fill a strategic gap the only way they knew how (and the way that had worked in the past). Now, other options are emerging, but not available and certainly not certified.
While it might not have been known that SpaceX would work, it's been known since the 1960s that cost optimized launchers could be much cheaper than what was being done.
Arthur Schnitt's MCD design philosophy dates from the late 1960s.
One of the tragedies of the Shuttle program was that it was pursued instead of this. Simply cost optimizing and evolving the Saturn 1B could have led to something a lot like the Falcon 9 (perhaps not with the reusability, at least at first), years before SpaceX.
Personally, It smacked me in the face as obviously going to eat the rocket industry's lunch in 2013 with Falcon 9 v1.1. Cost optimised boosters while testing landing tech on paying commercial flights was a no brainer.
Only reason SpaceX exists is because of the US Gov commercial programs. They wanted to create a more robust private space industry and they did it. Tax payer money was spent on SpaceX as the US Gov was the biggest customer for a long time. This is what governments do, they create markets. It was intentional and it worked just like it did countless times in the past.
There are space companies other than SpaceX which participated in the same contracts and were nowhere nearly as successful.
* Boeing did not yet deliver commercial crew.
* Antares rocket did not expand beyond ISS delivery.
* Blue Origin is not yet flying, despite immense funding.
* Virgin Galactic did not yet fly any customers.
The other great space launch success in the last decade was RocketLab but they're far too small to be relevant to SLS.
I didn't think the early SpaceX rockets did much that was technologically novel.
The real reason why everyone expected the commercial crew program to turn into big contractor grift is because they saw the same story play out 100 times before and 100 of those times the big contractors found a way to capture the money and send the project into development hell. The 101st time was different.
Long duration projects face the risk of looking silly when something radically changes the landscape. SLS took the safe route back to mostly expendable launchers from a "reusable" system that was the result of extensive feature creep and left the agency with some psychological scars (and some extremely cool and inspiring museum pieces).
As the author says, some things are impossible until they aren't. Starship may still hit a development wall, but, at this point, it's doubtful. If it works, it changes the landscape completely.
While I see a lot of discussion about the contents of the article (politics, engineering, etc...) I want to comment on the article itself. Very well done. A delightful read during lunch. Bravo!
It's not really fair to put the N1 in the category of successful giant rockets that have flown. Yes it flew, it exploded every time. No third stage of an N1 ever made it to low earth orbit.
So ironic Saturn-V is drawn as carrying 140 tons, while Apollo CSM is 30 tons and Lunar Module is 15 tons, yet Space Shuttle is drawn carrying just 27 tons. Definitely a case of apples and oranges.
The chart is tons to LEO. Saturn V wasn't designed for LEO, but it certainly could have dropped that much there. (Skylab used a pretty fair fraction of that capability.)
When you're comparing launchers you have to pick some baseline target orbit. LEO is a good lowest common denominator, but isn't quite fair as launch systems optimized for higher energy orbits (like the Saturn V) lose some of their advantage. But a higher orbit is even less fair to systems designed for LEO; choose anything else, like GEO or lunar insertion, STS would have a big fat 0.
Yeah, I know, but the point is Saturn V is shown as carrying to LEO 140 tons - and 140 tons was actual translunar payload (Apollo CSM + LEM, ~45 tons) plus dry 3rd stage (~15 tons) plus unspent fuel in the 3rd stage (about 2/3 of initial load of 105 tons of fuel). At the same time Space Shuttle doesn't show the empty mass (a lot) and unspent fuel (very little) of Space Shuttle "second stage", which is Orbiter, and mentions only payload in the bay of that Orbiter.
We could arbitrarily say Energia brings 0 tons on orbit - because it's, strictly speaking, a suborbital rocket, the payload needs to add some ~100 m/s of velocity to get to orbit, this was done to avoid getting empty stage to orbit (so no littering, no necessary maneuver to deorbit). Or, alternatively, we could say Energia brings 170 (!) tons to LEO - indeed, if one pays no attention to factual suborbital speed and adds mass of empty 2nd stage of Energia, which is 78-86 tons (http://buran.ru/htm/rocket.htm), one gets this number.
So it's important to carefully compare similar things. 140 tons for Saturn V is too different from 27 tons of Space Shuttle by method, not by result.
Thank you for a well reasoned argument. I would love some formalism of actual useful lofted weight, vs apparent weight lofted to some stated orbit. Getting 27 tonnes to LEO but leaving 50 tonnes in unstable orbit is not 77 tonnes of useful loft, to me (fictive example)
They are listed with numbers about payload to (quite low) LEO. Apollo CSM and the Lunar Module were launched into a much higher-energy trajectory. The Saturn V could have launched 140 tons to LEO, it just never did. IIRC the most massive thing it ever launched was ~90 tons of Skylab.
> They are listed with numbers about payload to (quite low) LEO.
Then one should include the mass of Shuttle Orbiter. That the orbiter is used as a second stage... Saturn-V lists mass of third stage alright.
> The Saturn V could have launched 140 tons to LEO, it just never did.
No, it never could. It's a popular illusion. The only way to reach 140 was to count parts of Saturn V reaching orbit as payload. When you count payload only, that number falls. It's like counting the mass of Sputnik 1 adding the weight of the second stage of R-7. Or, in other words, a satellite with mass 140 tons couldn't be brought to LEO by Saturn V.
To get Saturn V realistic number for LEO in similar sense to Shuttle's 27 tons (did Shuttle ever got 27 tons to orbit? Could it even be launched safely with 27 tons payload, given that it could only land with about 15 tons in the bay?), you should look at ~77 tons of Skylab. There is at least a hint that LEO payload is not all which gets to LEO.
I think you're confusing mass of the payload with something else. Mass of the payload to LEO is just how much mass is brought to LEO. According to "Alternatives for Future U.S. Space-Launch Capabilities" [0], the Saturn V did in fact bring 140 tons to LEO...
I think people are confusing the mass which appears on orbit with payload. Payload is (as meant when the word "payload" was created) rather the goal; in case of Saturn V it was Apollo stack which needed to be launched to the Moon. Similarly, in Space Shuttle case payload is something which is in the shuttle trunk; that's why typically a shuttle flies. Illustration for why 140 tons isn't the payload of Saturn V is that Saturn V wasn't able, or designed, to launch arbitrary 140 tons to LEO, and Skylab is a implicit confirmation of that.
The definition of payload to LEO in case of higher-than-LEO missions can include booster stage and fuel, but then we have another strange case of Shuttle with IUS - the payload could fly higher, and was rather small, yet we don't consider Orbiter's mass as part of LEO payload in this case. Since, as another consideration, we usually consider Saturn V as a 3-staged rocket, not a 2-staged (as flew with Skylab) with a space booster stage which is used for LEO insertion as well as for subsequent translunar injection, it would be logical not to include 3rd Saturn's stage into LEO payload. Then Saturn V LEO payload becomes too small to be meaningful (TLI requires a lot of fuel).
It would be better - here - to consider Space Shuttle LEO payload to include Orbiter's mass - that justifies large liftoff weight. As for what's Saturn V LEO payload, this question is better be met with clarifications.
"The payload of 140 metric tons is derived from weight data pro-vided in Richard W. Orloff, Apollo by the Numbers: A Statistical Reference, NASA SP-2000-4029 (National Aeronautics and Space Administration, updated September 27, 2005), available at http://history.nasa.gov/SP-4029/SP-4029.htm. In that reference, 140 mt is the weight of the Apollo 17 command-and-service modules, the lunar module, the spacecraft/lunar module adapter, the instrument unit, and the S-IVB stage (the third stage of the Saturn V), including the fuel remaining in that stage needed to propel the Apollo command-and-service modules and lunar module from low earth orbit to the moon."
So this number includes the S-IVB stage and unspent fuel. Which ought to raise the question of the definition of the payload - we don't usually consider launcher's parts among payload mass.
> Mass of the payload to LEO is just how much mass is brought to LEO.
No, it's usually not that simple - Space Shuttle here doesn't include Orbiter's mass (I think it's a mistake), Sputnik payload didn't include R-7 second stage (I think it's right), Energia doesn't have 0 LEO payload even though it releases the payload below LEO (I think this definition is also more meaningful).
Related to this, I found this article [1] interesting, from Arstechnica in September: Charlie Bolden [former NASA administrator] says the quiet part out loud: SLS rocket will go away
Pah, the feeling when your shallow post gets all the upvotes but the link to a Covid-risk calculator of a renowned research institute that you submitted earlier gets none. Priorities, folks, pri-or-i-ties... ><
I find the point that SpaceX has no plans beyond the next step hard to believe. I could imagine they perhaps have no firmly officially approved plans like a full waterfall project, but surely there must be already a lot of trade studies on the next steps, and at least some rough plans.
Or maybe the even have approved plans, but with the understanding that they could change quickly as new data comes in.
Otherwise a project like StarShip is likely not feasible. For example for a Mars project you must at least have some idea that long term live support is feasible with the chosen size. They likely have that idea based on data from ISS, but I bet there were some preliminary studies on this at least.
> surely there must be already a lot of trade studies on the next steps, and at least some rough plans.
Starship was always designed to refuel in Earth orbit, land on Mars, refuel on Mars and fly back to Earth in a single stage. This kind of architecture has never been attempted before and it has a lot of long-term implications for the design, for example the same reentry maneuver must work on two planets and orbital refueling is mandatory.
All other rocket designs consider "planetary landing" as something for the payload to worry about.
You are thinking of ULA’s Vulcan rocket, and there are indications that their engine reuse plan has been entirely abandoned. Aerojet has modified the Shuttle’s RS-25 engine design to make it explicitly non-reusable (in the hopes of reducing costs).
Huh, it looks like I might be. I could have sworn this was a block 1B or block 2 feature, that they were going to eject the engine unit and refurbish it, but looks like I must be mistaking it for the Vulcan. That's a shame they might have abandoned their engine re-use plan.
Tim Dodd is an amazing and pure soul, bless his heart. His fair, thorough, and well-researched explanations of this topic, and others, including his other recent video on just Starship vs Falcon 9 and Heavy, are real eye-openers.
And, as others here have said, yeah, hindsight is 20/20, given that the SLS will almost certainly be a boondoggle, and barring any major failures on South Padre or something... inconceivable... its utility will likely be obviated before it even launches.
But, even as an avid newspace fan for the past 10+ years, I still have to say, it wasn't terribly obvious Starship would launch on "Elon Time"... or "Elon Time 2.0.", even only two years ago when Elon unveiled "Dear Moon", he hadn't even announced he'd be building Starship out of stainless steel. Yeah, and he builds them like water towers are built, in ring sections, not advanced composites. I shudder to think back even 10 years...
Like, since it's pretty much fully reusable, yes, both stages, it could cost less than a million dollars to launch 150 metric tons to low earth orbit, which is pretty handy, because they're using it to launch a network of thousands of communications satellites in ridiculously low orbits that'll probably decay in only a few years and it's looking like that's actually gonna work really well. Yeah, way better than the one time Motorola tried it.
Oh, and if you wanted to get 150 metric tons to GTO, you just need a single orbital refueling launch. The moon, mars, just a couple more orbital pitstops.
Oh, and it uses a rocket engine design that's never even gotten off the test stand the handful times it was looked into before.
Yeah, hindsight's a... cough
Oh, did I mention there's a global pandemic? And President Trump got voted out of office. And Apple replaced all their Intel chips with the one they have in their iPads and smart phones, and it wound up being way better. Yeah, Intel's in a pretty tough spot, actually. Oh, and try to find as many Pokemon cards as you can, some people have paid their mortgages off that way.
Yeah... Time travel's gonna be hard to explain. Especially since nobody has DVD burners anymore.
From the article lead:
>NASA has certified SpaceX’s Falcon 9 to launch the agency’s most important science missions, giving the agency new options that could result in lower costs.
NASA sees reducing launch cost as a top scientific priority. The mission of Starship is to do the same.
The title made me think the post was something to do with starship.rs, comparing it with a newer, potentially even better shell prompt that I had yet to hear about. I feel weirdly disappointed.
So NASA & Congress designed a program that was unkillable. It greased too many wheels and lined too many pockets to make it easy to kill. It was also designed ultra-conservatively using mostly existing designs so there was no technical risk.
So now when we complain that it's a pork-filled boondoggle that's impossible to kill, that was the plan, and there was a certain logic behind it. It makes no sense in a world with Falcon Heavy, Starship & New Glenn, but who would have predicted that with confidence in 2010?