SpaceX’s Starship V3—still a work in progress—mostly successful on first flight
- Thread starter JournalBot
- Start date
Yes, but F9 isn't doing a full-Kerbal suicide burn at max throttle, where you have to guess the instant to start the burn with extreme precision. There is some slack that they eat up by running at reduced throttle for a portion of the burn such that they can zero altitude and velocity simultaneously while having the ability to adjust a bit. In other words, it's a negative feedback system and if their prediction for landing altitude is a positive distance above the barge, they thrust a little bit less hard, and if it's predicted to land some distance below the barge, they throttle up to stop sooner. It's not as comfortable as having an engine that can hover, but the difference isn't nearly as significant as it sounds, and it's arguably much more important to have a wide throttle range so there's greater control authority. That and having engine redundancy so failure to light isn't isn't a "smoke on the water" situation.
Of your three options, I'd be most willing to bet on three engine suicide burn, even if the crew blacks out for a little bit from the acceleration. Computers are pretty good at making rapid control responses. But can an RVac even run at sea-level without destructive levels of flow separation? That and the lack of gimbal means they wouldn't have sufficient attitude control.
Upvote
7
(7
/
0)
Yes, this thing is absolutely a battleship and obviously highly resilient. Still, just somehow making it in one piece to the ground isn't enough. Like, in this very flight back then it was 4 km off the nominal landing position. In a real flight it would have been a total loss either way since it couldn't have been caught this way.
Also all this isn't just about getting the ship back. This will be just the very beginning. It's about getting it back in a condition they can reuse it. And just because it held together structurally nearly empty during EDL doesn't necessarily mean they could launch it again with 1600 tonnes of propellants on board. It will still have to be structurally sound and they will have to assess this very thoroughly after they manage to catch the first one and have a closer look. I think it's highly probable that they will again need more mitigations to make it truly reusable.
Finally getting a ship back on the chopsticks is only half of the challenge and they don't even know yet how big of a challenge they still have left to deal with then.
I mean, SH/SS already is (and actually already was last year) a very capable heavy lift launcher with a reusable first stage. If it ever will be a capable, affordable and reliable fully reusable launcher still isn't clear by far.
Upvote
-10
(1
/
-11)
We have some degree of evidence pointing in that direction. I'm thinking specifically of (IIRC) IFT-10 and IFT-11, both of which lost specific Superheavy engines during launch, but had those same engines re-light (apparently successfully) during re-entry and/or landing burns.
(I would have to go back and re-watch those videos to be more specific)
So, at minimum, we know that an engine shutdown does not automatically deadline that engine for the rest of the flight. The FCS apparently has the intelligence and discretion to determine that an engine it previously shut down is in good enough condition to try lighting it again.
So it seems reasonable to think that, if it doesn't already, Starship FCS will in future include a decision tree to weigh "Scotty, I need more power or we're all dead!" against "The engines aren't designed to take that much abuse, Captain!" when the situation calls for it.
(...and now I can't shake the mental image of the Starship FCS internal voting mechanism as a VR of the NCC-1701 bridge, complete with high-fidelity personality models of the main characters....)
Upvote
7
(7
/
0)
A production flight would have been mission complete. LoV on landing is SpaceX's problem, not the customer's. Some degree of attrition is to be expected with reusable vehicles, as opposed to the 100% loss rate for disposables. The ability to sacrifice the vehicle to assure mission success is one of the advantages of reusability. That situation changes when flying crew but until then hull losses are just the cost of doing business.
There are no indications of Starship suffering from structural problems. Per your own post Starship is 'highly resilient'. Short of a heat shield burn-through or a catch failure there is no reason to expect a Ship to not be structurally sound after landing.
Catching Ship is the critical point. Refurbishment cost is yet to be determined and will likely need improvement but if they can turn around for less than the cost of new then they're in the game.
Starship is capable and affordable as-is. Out-competing the most cost efficient rocket ever flown will take a substantial amount of work but they're a lot closer than you give them credit for.
Upvote
14
(14
/
0)
It's highly resilient for EDL with about 150 tonnes of dry mass plus 35 tonnes of landing propellants. This is not an indication though for it still being structurally sound enough after all the heating and bending loads to be launched again with 150 tonnes of dry mass, 35 tonnes of landing propellants, 100 tonnes of payload and 1600 tonnes of propellants on top of that. These are two very different things by a full order of magnitude.
Rocket stages usually work with a structural safety factor of 1.2 to 1.5 and the ship can look perfectly fine during and after EDL and then still end up after all that with a safety factor of <1 for the next launch with a full load of 1600 tonnes of propellants. It looking perfectly fine after landing does not necessarily mean it will be able to be launched again.
And I'm not saying it's impossible, I'm just saying that they still will have to carefully analyze all this after getting their hands on the first returned ships and they may well need further structural and heat shield mitigations once they do that. The mass of which again will cut into the deliverable payload 1:1.
Add to that the fact that the Raptor 3 engines in their first flight had a failure rate of 3% during the booster burn and it's totally fair to say that this is not a mature fully reusable launcher at this moment and nobody knows yet if this version will ever be this. They may very well have to go to v4 before they arrive at their "fully reusable with 100 tonnes of payload" target.
Reusing second stages is HARD.
Upvote
-7
(0
/
-7)
If a Ship suffers structural damage during EDL then it wasn't within its designed operating margins. If it was within those margins then the hull is good to fly again. Per definition for a reusable vehicle.
The steel they are using has well understood thermal and structural properties. Standard practice in the steel industry is to laboratory test a sample from each individual batch to ensure that it's within tolerance.
Ship's structural and thermal margins have been extensively tested. The 2020 test campaign gathered real-world data to inform later design and testing procedures. SN15 was recovered and examined to confirm structural loads during the kick-flip and landing. Each new design is subjected to thermal and stress loads during ground testing. The vehicles are instrumented to measure actual mechanical and thermal loads in flight. Partial heat shield failures have been simulated. That data isn't published but it is collected and analyzed. They aren't flying blind.
Recovering an orbital Ship is a necessary step to finalize the design, but they don't need a final design to put it into operation.
Starship has unprecedented engine-out capability. That's a necessity for a highly clustered design and a benefit of reusability - in extremis the landing fuel can be sacrificed to complete the mission. A 3% failure rate isn't great but it's still within reported tolerance. Of concern is the failed boost-back: That's an actual problem.
Nobody is claiming that Starship is fully mature. The point is that it doesn't need to be.
Upvote
12
(12
/
0)
No, absolutely not. You can't be serious about that? "Returnable" isn't the same as "reusable".
Just because the ship can survive EDL to a soft landing with less than 200 tonnes of mass doesn't necessarily mean it can then be launched again with this mass and 1600 tonnes of propellants and 100 tonnes of payload on top of that. There's an order of magnitude of structural load differences between both. If you disagree here I don't know why we're still discussing this.
Upvote
-12
(0
/
-12)
Starship is designed to be reusable. That's a core feature of the architecture and the entire point of this test campaign. Where are you getting 'returnable' from?
Upvote
7
(7
/
0)
It's an iterative design process. I was speaking of what has been tested yet, not what is is meant to do one day. Up to now it's just returnable and in how far THIS version already will be reusable without adding more dry mass (and with this less payload) isn't known yet.
I really don't know what we're arguing about. All I was saying was that I think it's entirely possible that they will have to stretch the tanks to get to 100 tonnes of payload to LEO fully reusable. And now you're saying v3 is already reusable just because it survives EDL. We don't even know if it can carry 100 tonnes of payload to LEO as it is, since the last flight carried at best 40 tonnes of payload.
I've been far too optimistic about Starship in the past (even if not as optimistic as some others) and I'm just trying to be realistic now instead of taking design targets as more than just a wishlist.
In 2022 Starship already was "designed" to lift 100 tonnes to LEO and be fully reusable and we all know how this turned out. I wouldn't have expected back then that even after 12 flights they wouldn't go to orbit and didn't catch or reuse a single ship. And just like v1 didn't reach the design targets v3 also may not. That's all.
Say, when do you expect the first time Starship will be reused and deliver 100 tonnes of payload to LEO and again return after that? I'd say not this year and probably not even next year. I think they may catch their first ship later this year or early next year, then find that they have to modify it to be reusable, this will add more mass and with this diminish the payload and they will go to a v3.5 version with stretched tanks or straight to v4 to arrive at enough payload to bother with for an operational, mass-produced fully reusable launcher. What's your take for an operational fully reusable version with 100 tonnes of payload?
Upvote
-10
(1
/
-11)
It will be this year.
Flight 12 demonstrated the Pez dispenser delivering 100 tons of cargo into a trajectory that, for all practical purposes, has the same energy as an orbital trajectory, but is intentionally designed not to leave junk in space for multiple orbits. Nobody wants to track a bunch of mass simulators as they orbit the planet for the next N years. You are getting downvoted hard because you continue to insist "they HaveN't MaDe it to OrB1TTTT1!!!111!!". When the reality is that they would have (undesirably) been in orbit had they let Starship's engines continue burning for an extra 3 seconds. The engines had plenty of propellant in their tanks and they deliberately shut the engines down a few seconds early to PREVENT the craft from obtaining a full orbital trajectory that clears the atmosphere on the other side of the planet. Heck, they were in an orbit that clears the surface; it just (deliberately) doesn't clear the thick part of the atmosphere. And that was their flight plan. Had they screwed up and overshot into a long-lived orbital trajectory, it would have been a serious mishap investigation.
My prediction is that by end of year they will be launching real production Starlink birds into production orbits, and will be semi-routinely recovering the booster, probably with a non-trivial recovery abort rate due to continued tinkering with the design. And there's a chance they'll attempt a Starship recovery this year, but they also might not, due to that recovery process having much higher stakes. Booster recovery only places the launch tower at risk, and only that if the engines relight post-reentry and shift the booster's impact point from the ocean back towards the tower. Starship's recovery puts at risk their entire flight path from the East of the Boca Chica landing site, and they commit to that flight path risk the moment they begin to reenter the atmosphere. If the engines fail to relight, they probably still have an inertial course that impacts in the gulf, so they can probably manage the launch tower risk about the same as with the booster, but that risk to civilian populations, that's unavoidable, so they need to make damned sure that they are confident they won't break up on reentry. That's why they are investing in satellite-camera based inspection of heat tiles. Any Starship with an even potentially compromised heat shield is going to be unceremoniously dumped into the deep ocean somewhere. To minimize civilian risk.
It's a good plan. The reason they are burning through so many test flights is that they can do about FOURTY test flights for the cost of a single SLS launch (variable cost! not even including amortized R&D expenses!). So for the price of NASA's single uncrewed demonstration flight, SpaceX can burn through dozens of boosters and ships as they practice speeding up their manufacturing line. Heck, there's a bunch of them that they manufactured just for the practice and then scrapped because they already had a newer one for the next flight test.
Last edited:
Upvote
7
(7
/
0)
Did you even read what I wrote? I'm not talking about any of this. I'm not saying "they HaveN't MaDe it to OrB1TTTT1!!!111!!" at all. I'm also not saying they won't be able to get a ship back (they did this several times already) or even catch it (although they haven't done this yet).
I'm just saying that being able to deliver 40 tonnes (this is what they did in flight 12, not 100 tonnes) to orbit (or almost orbit, doesn't matter) and getting the ship through to a soft water landing is not the same as being able to reuse it to again to launch 100 tonnes to orbit, not by far.
The ship can perfectly make through EDL and still not be able to be launched again with 1600 tonnes of propellants on board. They have to actually catch and analyze a ship to assess if it still is structurally sound enough after a reentry to be launched again. And this may (will, IMHO) need further changes to manage this.
And there's even much more to this than just the structures. Like, from the coloring of the steel hull with the last flight it's obvious that the naked steel reached temperatures of about 400° C. The steel may be fine with this, but this will have quite thoroughly baked everything in the ship, like the PEZ dispenser with its motors, chains and belts. They will have to add insulation or other mitigations to make all of this really and reliably reusable. And there will be uncounted other details they will have to care for before being able to launch a ship again. That's iterative development, but it takes its time, step for step. They are not done with this yet.
Being able to land the ship does not automatically mean you can just stack it onto a booster, fill it with 1600 tonnes of propellants and fly it again it as you seem to think. Getting the ship back is a required prerequisite to reuse it, but not a sufficient one.
But you said they will reuse a ship this year to launch 100 tonnes to orbit. If they will do this I will be very happy, but I doubt that a lot.
I think they may be able to catch a ship this year, but even this will be tight. The next flight won't even go to orbit (so it will not be able to make it back to the launch site), so they can at the earliest go to orbit in flight 14. Even if they then go right to a catch attempt with their first orbital flight they would have to relaunch the same ship with flight 15 to make it this year. And they could do this only if they don't do any Artemis-related flights apart from that and if absolutely everything perfectly works from now on. Currently they can get a ship ready maybe every two months, so three more flights already hits the end of the year.
You're either totally misunderstanding what I'm actually saying or this is a typical "he's right but I don't like what he's saying" thing.
What I'm saying is just that progress will be slower than some people think or hope, just as progress in the last 12 flights was slower than some people expected.
But if you really think they will reuse ships this year to deliver 100 tonnes to orbit you again won't read what I wrote.
PS: Not that it would matter, but could please someone explain to me why they're downvoting what I wrote? I mean, what I wrote is just reasonable. SpaceX even took about a year to launch their first landed Falcon 9 booster a second time. With Super Heavy they took four months to reuse it (which was quite an improvement certainly). But do people here really expect that SpaceX does this within months or weeks with a Starship second stage once they manage to catch it later this year or early next year (which I fully expect within this timeframe)?
Getting the ship safely back to Boca Chica will only be the first half of full and operational reusability. Are you people really that thin-skinned when someone says that all of this is hard to do and won't happen without lots of further iterative development and assessments? I really don't understand how people can be that impatient. You people are the very material that haters are made from: You expect miracles and then will be sooner or later be disappointed because what you get is just solid iterative engineering that takes its time and in the end will be not really a miracle.
It really seems to me as if there are just haters and fanbois here anymore. The haters say "Starship just explodes every time and will never work!" and the fanbois say "Everything will work perfectly the next launch!" and anyone who's just reasonable and factual about things will be hated by both sides just the same. What a situation to be in.
Last edited:
Upvote
-11
(1
/
-12)
"...PS: Not that it would matter, but could please someone explain to me why they're..."
I'm not.
You're just a (well-informed) natural worrier, I think.
Generally, that's an important - if unpopular - role.
But none the people here are actually making crucial decisions about the progress of Starship. (I assume.)
So the (rash?) enthusiasm is harmless.
And, you did volunteer to be the party-pooper, Eeyore!
Just my inexpert thoughts about the vibes here.
I'm here to learn a little, mostly.
Including from you.
And, everyone else - you could all lighten up, too!
Argue, - yes please.
But lay off the anonymous votes?
Upvote
-4
(1
/
-5)
It is an iterative process but reflight isn't a 'one day' goal. It's something they've been preparing for at every step since day 1. They've already reflown a booster. The only reason they haven't reflown a Ship yet is that they aren't allowed to overfly land for a catch. Without that restriction they could've caught and reflown a v1 Ship. As it stands the FAA are waiting for EDL to mature before they will allow a catch attempt.
They don't need 100t to refly. If anything it's the other way around: A successful reflight will tell them where they can save weight and those savings will go directly to payload.
Ship is instrumented with full telemetry. They know how hot it gets inside and where any hotspots occur. That's critical for heatshield development. Any problems in that direction would have been discovered and addressed in v1. If anything new cropped up in v3 then they know about it and are already working on a solution.
Any tank stretch is going to be informed by known stress loads. They have enough data to model it accurately. At this point it's bog standard structrual engineering.
Booster 12 was the first catch and Booster 14 the first to refly. Based on that I expect the third or fourth caught Ship to be the first to refly.
I expect Starship to be operational (launching payloads) by the first catch attempt, first reflight at the latest. I expect full payload capacity to be achieved after reuse is worked out. There might be a v3.5 before the final v4 stretch.
Assuming that the v3 test campaign goes well, I expect the first Ship catch attempt within six months and the first reflight within six months of that. Full payload is a parallel track and may take longer. I expect that improved reliability will contribute as much to payload improvements through reduced reserve fuel* as will be gained from either weight savings or engine performance improvements.
* I don't believe that 40t is v3's actual payload capacity. I think they are carrying no more than 2/3 of maximum so they have ample reserve performance for problems such as the engine failures during the most recent flight.
Upvote
11
(11
/
0)
I downvote you because you are complaining about SpaceX not doing something that they should not do, that they would be irresponsible if they did do, and which they easily could do. It's literally almost the most stupid complaint you could have. And you just keep at it, like if you have absolutely no clue how stupid the complaint is.
Upvote
3
(4
/
-1)
I still don't understand the problem. If they are successfully launching 100 tonnes of Starlinks to orbit, and are landing the Starship, but are then retiring the ship after each flight, then, what, they'll still have the most cost-effective heavy lift platform ever built?
Second stage reusability is hard. This we agree on. If it takes them a few tries to get it right, so what? They spent a lot of tries getting F9 landings to work. It seems like you are upset that they didn't meet your prior optimistic predictions, and are now trying to over-correct with extreme pessimism that even if they recover Starship it will somehow not be reusable, and then saying that in a way that makes it sound like you think they haven't done their homework and don't have a credible plan to ever make it work.
Upvote
10
(10
/
0)
I think most of what you're writing is reasonable. I'd really only disagree with this bit, and that's because I'm a wildly optimistic fanboi. If things go well, I think SpaceX may be able to reach a monthly cadence, and have up to six flights this year.
Upvote
0
(0
/
0)
Um, what do you think they're complaining about, or think that SpaceX should do differently? I thought uhuznaa was merely saying that there's a gap between recovery and reuse, and spanning that gap might take more time and adjustments than people expect.
Upvote
3
(3
/
0)
I suspect the complaints and downvotes are because he is making strong statements of opinion as if they are mostly incontrovertible fact, and that we are all being naive not to see it. That is a common enough occurrence around these parts, but in this case it is made worse by claiming that only once a ship is caught will the real work of reuse will begin.
Such work has been on-going since the beginning and huge visible improvements have already been made. Flaps no longer burn through, tiles no longer fall off, steel no longer gets visible heat discoloration, etc... but none of that matters because apparently there will be massive internal structural damage for some reason only known to him.
Upvote
8
(8
/
0)
Of course it's just my opinion, what else? And I'm not pretending there WILL be "massive internal structural damage", I'm just saying that they don't know in which condition the ship is after EDL apart from it surviving EDL and the flaps and heat shield looking good now.
But the ship surviving down to the landing with about 200 tonnes of mass does not necessarily mean that it will also be able to survive being launched again with a mass of about 1900 tonnes (dry mass plus landing propellants plus payload plus 1600 tonnes of propellants) as some here seem to think. They will have to very carefully assess the ship after the first catch. It would be a miracle if the ship wouldn't need more modifications then.
What irks me is the totally unrealistic expectations here. Like insisting in SpaceX being able to reuse their first ship this year and launch 100 tonnes to orbit with it. I'm still following this closely on NSF and elsewhere and the consensus right now is a launch every two months at best looking at the progress with building new stages currently. That's still incredibly fast but not fast enough to fly often enough for rapid tests and apply the lessons learned by this within half a year.
Some people here are basically programming themselves to be disappointed later. I've learned to not expect miracles and instant progress during the last 12 flights. Progress will be slow and there will be regressions here and there and maybe even v4 will be needed for 100 tonnes of payload to orbit fully reusable. v3 will only be able to do that if they could go from 40 tonnes payload to 100 tonnes with no further changes (like reducing dry mass by a lot and/or adding more propellants) and not need to add any dry mass to make the ship reusable instead of just returnable.
But after all we don't really need to argue about all this, we can just wait and see.
Upvote
-2
(1
/
-3)
Yeah exactly, and I'm repeating this over and over because it seems some people haven't really read what I wrote or didn't understand it.
Upvote
-1
(1
/
-2)
Yes, of course. As I said, Starship already is a highly capable heavy lift launcher, no doubt. With an expended second stage they could already have launched 100 tonnes of payload to orbit last year. That's a whole lot more than nothing.
The thing is just is that some people here expect it to be a fully reusable launcher with 100 tonnes of payload this year. And it won't. This will take its time just as it took its time to arrive at where it is now. And I have no problem with that, I just like to discuss which problems will still need to be solved to arrive at that.
Upvote
-2
(0
/
-2)
That doesn’t sound like opinion and that’s the problem. You don’t know that it won’t, you just need to be pessimistic and are finding reasons to be so, as if e.g. SpaceX doesn’t have cameras inside Ship showing the state of the internals just before crashing into the ocean.
Upvote
5
(5
/
0)
It is? I don’t think it has proven its capabilities yet. However, I’m willing to be corrected on this.
Have they launched with a 100t dummy mass yet?
Upvote
-3
(0
/
-3)
Performance and reliability are still short, and Ship still has some boxes to check, but SuperHeavy has demonstrated its full flight profile: launch, stage, boost back, catch, refurb, relaunch. Ship still needs orbit circularization, catch, refurb, and relaunch, but those are gated by reliability requirements.
No, but it has launched with ~40t plus a similar mass of landing fuel. If you expend the second stage then that fuel can be replaced 1:1 with payload. I don't think breaks 100t but it's close enough for partial reuse to be viable if they can sort out the reliability issues.
Upvote
5
(5
/
0)
I imagine if you strip off all the extra bits for recovery (HS, fins, etc.) you are likely over 100t.
Upvote
2
(2
/
0)
And this opinion of yours is based on what, exactly? Rectal extraction?
Upvote
2
(4
/
-2)
'zackly.
SpaceX has a whole lot of instrumentation in the SS/SH prototypes, dozens of cameras, hundreds if not thousands of sensors, and enormous amounts of telemetry (temperatures, pressures, voltages, accelerations, power use, control surface positions, ...). They've (presumably...) been studying that telemetry and addressing/fixing issues that they've discovered that would presumably interfere with reuse. Given that they went from v1 to v2 to v3, presumably they've adjusted their designs and manufacturing enough that it made engineering sense to collect all the changes into batches. There may be, and almost certainly will be, additional changes to improve reusability. But this isn't SpaceX's first rodeo, and if the telemetry has led to questions, they've almost certainly responded with design changes and/or sensor/telemetry changes to resolve them.
Conversely, however, this rocket is new in a lot of ways that no rocket has ever been. It's "old tech" steel and high-tech navigation and control. They've got a lot of problems to solve, and appear to have workable solutions for all the big ones. Better ones are still significant options in many cases, optimizing propellant settling and transfer are still essentially new practices and likely amenable to significant improvements. So regardless of how the development path shakes out, they will undoubtedly be optimizing hardware and procedures for another several years. I don't think SpaceX, or most people here, think that it will be fully operational and complete, sprung from Zeus' forehead fully formed and perfect, anytime in the next two years.
Upvote
9
(9
/
0)
The opinion that structural load differences between landing with 190 tonnes off mass and launching again with nearly 2000 tonnes of mass are far enough apart that surviving the former after reentry does not mean it will necessarily again survive the launch doesn't need much explanation.
Mind you, I'm not saying that it won't be able to be launched again as is. I'm just saying that they don't know yet. While some people here say that if the ship survives reentry with 200 tonnes of mass this means it will also be able to be launched again with 2000 tonnes of mass. This is absolutely not a given.
But I'm repeating myself again and if you didn't understand what I was saying (or don't want to understand it) this won't help. So I'll better stop wasting my time.
Upvote
-9
(0
/
-9)
Damage in flight is a known quantity. Several Ships have in fact landed with significant damage that would prevent a reflight. This much is not in dispute.
This is the dispute: SpaceX know what damage is occurring. They've known since the first reentry attempt. Every Ship is instrumented to detect damage as well as off-nominal conditions that might cause damage. Every test includes mitigations for problems discovered on previous flights. They don't need to wait for a catch to find out what's going wrong, nor do they need to wait until then to fix it. Catching would speed up the process but it isn't necessary. That work is already in progress.
Upvote
8
(8
/
0)