Also, as a general principle, an orbit is completely determined by the position and velocity at a single point in said orbit. If you’re trying to determine if your craft has reached orbit, a single radar fix is sufficient.
Any sign of when they are going to perform this test? Is it the next test after the SN15 refly or at some distant point in the future?
Distant point in the future. Where distant means probably 2022 sometime.
I don’t think they’ll go for it until they’ve done a bunch more Starship hops (at increasing altitude) as well as a few Super Heavy (booster) hops. SpaceX seems to be limited by the Raptor production rate, and even with no payload I’d guess they’d have to throw away at least a dozen Raptors for the orbital test. SpaceX is trying to cost-reduce them as much as possible, but I doubt they’re anywhere close to their $250k/each target.
It certainly won’t happen before SN20; I believe Musk has said that’ll be the first with the necessary features for orbital flight. And they’ll want to hop that a few times as well.
So overall I think we’ll see quite a few hop-style tests before the orbital test. It’s a little surprising that they seem to have no plan whatsoever of recovery–though maybe they think they can pull it from the ocean if it does a soft landing. Or maybe they don’t want to wait for their ocean platforms to be finished, or that their costs will be so low that they can just sacrifice one. Hard to say right now.
28 Raptors in the booster, six on Starship. If they are splashing both of them, that’s 34 Raptors going bye-bye. Or are you thinking they will fly with fewer because of no payload?
At some point, they need to test SuperHeavy with all 28 engines, because acoustic energy, vibration and fuel distribution are likely to be the show-stoppers if there are any. They might get away with just the three engines on Starship at the very beginning, accepting the inefficiency of non-vacuum engines for the first tests, assuming Starship is light enough.
We should prepare for the cadence of Starship flights to slow down. As they start working on orbital flights and adding operational hardware like the clamshell for payloads, payload adapters, heat shields, vacuum engines, space refueling hardware and the like, each Starship is going to take more time and money to build, and SpaceX will have to do more ground testing because of the cost of loss of vehicle.
By the time we get to on-orbit refueling tests, each flight will be much more expensive and complex, and they won’t be firing those tests off every couple of weeks.
We still have a long way to go.
That’s right. With no payload, I’d guess they need under half the number of engines, and I don’t think they’ll expend more than they need.
You’re right that they’ll need to get to the full amount eventually, but the early tests don’t necessarily need to be orbital. And the orbital tests don’t need to test every other piece.
I’m not sure we should expect tests to slow down. Once they start landing the stages, I’d expect a massive increase in flight rate.
I predicted 2022 above for the first orbital test, but I’m wondering now if they won’t try to go for something much sooner. Musk said this not long ago:
If they’re really expecting a bunch of failures, they need to start iterating as early as possible. If they can get booster landings dialed in, then it’s not that big a demand on Raptor production (or other costs).
Elon Musk sais something and you believe him? OK, fair enough. I don’t. But you sure will enjoy this article. I did too, but in a different way as you are going to (assuming you read it).
Musk fanboys need to take his statements with a much bigger grain of salt than they do. At the same time, the anti-Musk types seem to willfully misunderstand that almost everything he says is subject to change and dosed with a huge amount of optimism (“aspirational goals”). If both types just dialed down how seriously they take his tweets, etc. by like 50%, they’d be much more relaxed. Take that in mind when I repost his tweets.
The whole Starship program has in fact been making unreasonably rapid progress even if it’s probably still not as rapid as Musk hoped. It’s always hard to make predictions about rapidly iterating development programs, but at the very least it’s clear they haven’t stalled or anything. Every new test flight has either uncovered a novel problem or retired some risk. They’re still making rapid forward progress.
Thanks for the link to the article; I read it the other day but probably others will enjoy it as well. Note that the author is Robert Zubrin, who has been a longstanding supporter of Mars exploration and at this point is fully onboard with Starship. At one point he was advocating for a mini-Starship, though based on the article, maybe he’s given up on it.
It’s tough to underestimate the potential of Starship even if it misses its cost targets by a factor of 10. I’m really excited about its lunar potential, even if Mars is much further off than Musk hopes.
Holy shit. Mach 25 belly flop?
Yep. And that’s the minimum velocity it’ll ever reenter at. From the Moon or Mars, it’s higher.
It’s one of the main reasons why second-stage reusability is so hard. Re-entry from orbit demands extreme heat protection, which is heavy. The Shuttle cheated by discarding its fuel tank. But for Starship, the whole thing is coming back. If you’ve noticed the black patches on various Starship flights, those are tests for the heat protection tiles. The orbital vehicles will have one side entirely covered in tiles. I’m not sure what the mass is, but I’m sure it’s not insignificant.
NASA has been very picky about making sure there are no possible protrusions or gaps in heat shields. SpaceX originally wanted to return Dragon on land, and had a plan for landing legs that would deploy through the heat shield. NASA nixed the plan.
So for me that biggest potential issue is the moving fins, which have gaps and hinges that must be protected by a heat shield. Also, the flaps have to survive the heat, and will have to have hsat shields on the bottom. I would be worried about things like flutter, warping of hinges, burn-through around the hinge gaps, all that kind of stuff.
Capsules are much easier, because they self-stabilize during re-entry. So long as the heat shield is intact, there’s very little to go wring. Starship may need its flaps deployed rather than folded back to maintain stability, which would make protecting them from heat much more difficult,
None of that should be a show-stopper, but as Musk says, it may take a few flights burning up or losing control before they figure out how to do this, and maybe a few revisions of the hardware.
Certifying this to carry people will be a real challenge, given the many variables on each flight, The good news there is that a fully reusable system with a high cadence can prove out safety simply through repetition. Mudk has said there will be ‘hundreds’ of unmanned flights before one of these ever carries a person.
One big difference is that Starship is made from stainless steel, while the Shuttle was made from aluminum. Some low level of hot gases getting through the heat shield should be more survivable. In fact one Shuttle mission was almost lost due to damaged heat tiles, except that by sheer luck there happened to be a steel reinforcing plate beneath one tile that had been completely lost in a critical area.
I agree that it’s going to be a while before we see manned flights, and that NASA may never be happy with certifying Starship for them (to/from Earth, that is). But NASA can afford the alternatives here, whether Crew Dragon or Orion. And Musk’s Mars dreams don’t necessarily require NASA.
Decided to run some orbital flight numbers here. My intuition was about right, though this doesn’t mean this is the path SpaceX will choose.
Say that Starship dry mass is 120 t and Super Heavy dry mass is 200 t. Also say that MECO happens at 2400 m/s and S2 does another 6600 m/s.
If S2 only gets 3 standard Raptors (not vacuum models), it’ll probably get around 3550 m/s exit velocity. That means the wet mass needs to be 770 tons. The Raptors have ~225 t force each, which gives a little under 1 g initial acceleration–totally fine for a second stage.
So S1 dry+payload needs to be 970 t, and if we assume 3400 m/s exit velocity (rough average between sea level and vacuum performance), then we get 1965 t wet mass. S1 obviously does need >1 g thrust, say 1.3x, which gives us 12 engines.
So 15 engines total for no payload. Actually I didn’t reserve anything for the boostback, so maybe add a few more, though obviously this also depends very much on the actual masses instead of my estimates, plus I didn’t optimize the staging velocity or anything like that. Still, I think it’s in the ballpark.
Yeah, I agree it’s do-able. And they may do something like that for the first couple of flights. But the biggest risks of the whole stack involve making 28 Raptors work together, and the structural limits of such a giant rocket. Launching it easier with fewer engines and no payload is fine for aerodynamic testing of Starship, but very quickly they are going to have to load up the whole system with full fuel and all three engines.
Look at Strahip itself. It’s flying with three engines and limited fuel, because all they care about right now is figuring out the belly flop and landing, and in that situation the conditions are nearly the same as return from orbit - only using 1-3 engines for the burns and having only landing fuel aboard.
But the minute they get that phase nailed and try for full orbital tests, a real test of the system will require all the engines and full fuel and a dummy payload. That’s where it’s going to get expensive. I’m guessing SN15-20 are coming in at somewhere around $5 million each. The Raptors are probably still costing close to a million bucks, I’d guess.
What would a full test of the system cost? Let’s say the next batch of Raptors come in at $500,000 each - still ridiculously cheap compared to other rocket engines of tgat size. 34 Raptors alone is then $17 million. The Booster is gigantic. Call that $5 million, and another $5 million for the Starship body without clamshell and other flight hardware. Then maybe half a million for fuel?
We are encroaching on $30 million for full orbital tests. Maybe half that for early ones with fewer engines and fuel. I could be off by a lot sincevthis is just a WAG, but if I am I would guess it’s on the low side.
Building a Raptor for $250,000 sounds very difficult. A GE90 turbofan costs $27 million dollars, and therefore likely has a manufacturing cost of around $10 million. Hell, a Pratt and Whitney PT-6 costs about a million bucks, and it’s a very small and low-powered engine, comparatively. I know turbofans are not rocket engines, but still… There are a lot of parts on a Raptor made of very expensive materials. Those turbo pumps can not be eady to make.
All that goes to show that they need to get reusability going as quickly as possible 
. I don’t agree though that the engine count of the full stack is a big concern. The Falcon Heavy has 27 engines in close proximity and it’s never had an issue. Sure, the Raptors are more powerful, but to me it feels like they’ve nailed the analysis of many-engined vehicles.
It’s reentry that worries me, because of the so many unknowns. You mentioned several of the reasons earlier, but there are many others, like the fact that the Starship hex tiles are a different material from the Dragon PICA-X, plus SpaceX hasn’t seemed to have really figured out the mounting yet (they’ve fallen off or broken on basically every test flight so far).
They need to retire some of that risk as quickly as they can, and learn as much as possible about the rest. That means test flights as early as possible. In the meantime, they’ll nail down the booster landings, which I expect to be fairly easy, and will reduce the cost of future orbital flight tests.
It’s possible that somewhere in there they’ll make some money back from commercial payloads.
It is pretty amazing that SpaceX can build their engines for so cheap. But arguably, a rocket engine is simpler and easier than a modern turbojet. Even with massive reusability, they don’t need nearly the same lifespan. And in some ways they’re less sensitive to efficiency than engines for commercial jets. Tom Mueller said this on the subject:
Even airliners are kind of expensive for our cost trades. From Elon, I’d always get, you know, how much does it cost to make a Model S, a Tesla? Start from there, and work my way up to the rocket.
Like, here’s a conversation I had maybe about five years ago on the Merlin 1D when we first developed it. He asked me; he said, “How much do you think it costs to make a Model S?" And I’m like “I don’t know; 50 thousand dollars?” He said “No, about 30 thousand dollars.” That’s the marginal cost for that car.
And he said, “How much does that car weigh?” And I said, “About 5 thousand pounds.” And how much does a Merlin engine weigh? I go, “About a thousand pounds?” So, he’s like, “So why the heck does it cost, you know, some fraction of a million dollars to make a Merlin engine?”
And I mean, he has a good point. And the material you’re using isn’t aluminum, it’s not stamped, so I’ll give you a factor of five. So it’s equivalent to a five-thousand-pound rocket engine. So why’s it 20 times the cost? So that’s the way we look at it and the way we think at SpaceX trying to get the amortization cost of the rocket down. Once you start reusing it, the real big cost becomes the amortization cost of the rocket, the operational costs, and the fuel costs, which is basically the same model as the airliners.
…
Like I said, we don’t use space— we avoid space vendors like the plague. When we started developing the Merlin engine, you know, I needed valves; I needed liquid oxygen and kerosene valves that had to work. So I went to some of the vendors that supplied these valves and I said, “Hey, can you give me a good price on your existing product?” And no, they couldn’t. So I said, “Can you design a much lower-cost one?” So they came back; and you know, if it takes two weeks or a month to give you a quote, you already got the wrong vendor. If it takes them that long to just give you a price, how long does it take them to build the actual part?So they come back with a quote of hundreds of thousands of dollars for their part, and you know, it’s going to take eighteen months to develop it. And I say, “No, I need it in like three months.” And so they kind of laugh at you. And so we ended up designing our own components; you know, pre-valves, main valves. We’d already developed the injector, the combustion chamber; the main parts of the engine. We were hoping we could just go buy some of this other stuff from existing suppliers, and no, the cost was just— the cost and schedule weren’t close for us. So we— Anybody that provides you know, space hardware to government contractors is just not at the level we want to be at. So, that’s how we get the cost of the hardware down; also, we had to have control of our own test site. So we developed a test site in Texas and did our own testing because a huge cost of developing and testing rockets is the test site cost. So we needed to get that under control.
Not super specific, but it does seem that vertical integration helps them a lot. Aerospace suppliers are just absurdly expensive, so they build as much as possible in-house.
NASASpaceFlight seems to think the orbital test will get a full set of Raptors:
Super Heavy will have a full set of Raptors; the number has likely changed since the 2019 Starship update when Super Heavy was to have 28 engines. The outer ring of engines will not have Thrust Vector Control while the inner ones will. Starship is set to utilize six Raptors, three sea-level and three vacuum-optimized.
As such, this test flight alone will require a large stock of Raptor engines.
While Raptors arriving into Boca Chica are in the SN50 and SN60 range, SpaceX’s test site in McGregor is hosting engines in the SN70-80 range, while it is understood that production in Hawthorne is working on engines past SN150 — with a current shipping rate of several Raptors per week.
I didn’t realize that their production rate had gotten so high. Could well be that their price per engine has already seen significant drops. A Raptor every day or so is impressive for such a sophisticated engine.
SpaceX can let out a mild sigh of relief:
(5) Savings.–The Administrator shall not, in order to comply with the obligations referred to in paragraph (1), modify, terminate or rescind any selection decisions or awards made under the human landing system program that were announced prior to the date of enactment of this division.
It’s still not set in stone, but SpaceX does seem to have some allies in the senate. It seems very unlikely that NASA would revoke the contract at this point.
There’s been various bits of excitement related to the HLS decision. Maria Cantwell, a Senator from Washington, is attempting to amend the Endless Frontier Act, to give Blue Origin another chance at a contract (by adding $10B to the NASA budget).
What’s interesting, and promising for SpaceX, is that as far as I can tell exactly no one is suggesting that the SpaceX win be taken away or somehow be rebid. That includes Blue Origin itself, which has said:
So they don’t want to start the process over; they just want NASA to get more money so they get the #2 win by default (Dynetics might not like that).
Bernie obviously doesn’t have many kind things to say about Musk, but even in his critique he’s focused almost entirely on Bezos:
The news coverage has likewise been almost entirely anti-Bezos:
So I don’t think SpaceX has anything to worry about here. No one’s denying that they got a clean win (and the SN15 landing surely didn’t hurt). The only question is whether there’s a #2.
I think the Cantwell bill requires NASA to pick a second candidate, but doesn’t actually have the budgetary power to increase NASA’s budget. So it’s an unfunded mandate of the sort that has crippled NASA programs in the past.
I don’t understand how you plan a moon program that has two wildly different landers in the mix. Whether there is a Starship or a Blue Origin lander .waiting at the moon would have serious implications for the entire program. For one thing, you’d either have to ship a complete Blue Origin lander for each mission, or somehow deliver a new descent stage and integrate it with the ascent stage in LLO. On the other hand, reusing Starship requires a lot of fuel to be sent there.
It seems to me to be adding a lot of complexity to plan for both.
For the moment, but again I don’t think it affects SpaceX. For the time being, NASA can say to Blue Origin that they will have to make do with $1/yr in funding, which may increase if and when NASA gets more funds. SpaceX, which already won their bid, will get the funding they asked for.
Well, the minimum capabilities are the same, in principle. The mission details are up to the companies to figure out. As long as the landers can carry X people and Y tons of cargo, etc., NASA doesn’t have to think too hard about it.
Of course, NASA will likely very much want to use the extra Starship capability. Maybe they will just decide to fill it with bulk cargo like food/water/solar panels/etc. that would be useful in an emergency and for a future base, but isn’t strictly necessary for the mission.
I guess the difference in astronaut training might be a factor. Surely they’ll have to have completely separate training programs for the two vehicles. Which would ultimately mean that only one group gets to be the lucky first back on the moon.
I continue to not worry much about Nelson. This article came out a few days ago:
Nelson gives no indication of weakening their commercial-oriented strategy. He got a question from some (idiot) congressperson about whether Artemis should emulate Apollo, and responded:
Nelson gently pushed back on this as well. “In the Apollo program, we got to the Moon with American corporations,” Nelson said. “They did all the work. NASA supervised it. NASA had a reason to supervise it, because NASA’s responsibility is to make sure it is safe. We’re just continuing that in a different way.”
It’s a very diplomatic answer, with a clear message: no, NASA isn’t giving up on commercial space. From NASA’s perspective, it’s just a different funding model. Nelson also reminded Congress of the old “no bucks, no Buck Rogers”:
But Nelson was having none of this. “The Congress appropriated $850 million,” Nelson told Babin. “And so you can only get so many pounds of potatoes out of a five-pound sack. If you all are generous… then we’re going to try to rev it up.”
I’m not seeing even a whiff of signaling that Nelson plans on changing the current strategy, or even that he will pull his punches when it comes to Congress applying pressure. Of course, it remains to be seen if Congress will supply enough funding for a second vehicle.