I’m not too familiar with him. Does he have a good reputation as a journalist? There’s not much on his wiki page.
He’s probably right, but it is hard to characterize something that cost $3B to make exploding as a success. He claims they can afford to make many more, but … can they really? SpaceX is a private company so we’re not privy to its finances but I didn’t think that they had that many billions of dollars. I know they value the company at ~$200B but they certainly have vastly less cash. Wasn’t Musk saying last year that the company was going bankrupt (EDIT: it was in 2021 not last year)?
That was the van from the NSF crew (one of the several streaming channels covering the launch). It was a platform for their camera equipment. They regularly put cameras well within the damage zone to get a better shot, knowing that they might be destroyed (it’s happened a bunch of times before). It’s all remotely controlled; no human is anywhere close to it.
Uh, where did you get that $3B number? It’s complete nonsense. It could plausibly be the total Starship development cost so far (though that’s unknown), but the cost of this particular test is nowhere close to that; probably under $200M.
I’m pretty sure the staging never began. For whatever reason, be that deliberate control or due to malfunction of the staging system.
During the later stages of normal flight the vid cut to a split-screen view with one half inside the interstage shroud where we see the 2nd stage engines above and the top of the first stage below. The two segments never move relatively, nor do we get any outside light intrusion in there until well after the stack is behaving badly. And then they cut to a purely ground view of the already-gyrating stack shortly before the destruct event.
That’s not to say that some aerodynamic anomaly didn’t occur. Maybe some of the first stage grid fins fell off or went stupid, etc. But if all the yaw/pitch we saw, even the first bit, was loss of control, then a much better bet is as you say a thrust asymmetry and/or hydraulic problem.
So, here’s a collection of interesting tidbits. It begins with the first few seconds of launch. I noticed what appeared to be debris arcing away from the launch site. Check out 0:18-0:20 and watch just to the right of the rocket.
There is very significant damage to the pad:
There’s a ‘before’ photo in the comments for comparison. No flame diverter seems like a bad plan in retrospect. There are also various photos of parts of the base with debris strewn about, and the tank farm now includes several tanks with significant damage.
Then there’s a “small” fireball near the base of the rocket about 30s into flight. (Actually quite large because the scale of this thing is ridiculous.) Speculation that it’s one of the hydraulic pump units.
Speculation: lack of flame diverters on the pad resulted in debris being kicked up into the rocket, causing damage that resulted in the loss of hydraulics. No hydraulics = no gimbal control = spinning out of control.
I was willing to assume this. I was just surprised that an intentional camera position included a vehicle. Once upon a time it would have just been emplaced (preferably hardened) remote cameras. Leaving a vehicle there feels careless, or at least hasty and wasteful.
Right. $3B is a reasonable estimate for that. But it also includes:
The complete design and test work for the Raptor 1 and 2 engines, including vacuum variants, which are pretty much the most sophisticated engines in existence.
A factory to produce said engines at an extreme rate: more than one a day
The design of the Superheavy and Starship vehicles themselves
Again, a factory to produce those vehicles at a high rate.
A complete launch facility in a location that was pretty much a swamp.
A series of pure test articles and flights (Starhopper et al)
A series of semi-complete vehicles and test flights (SN8 through SN15)
A bunch of other prototypes (recall that today’s flight used Booster 7 and Ship 24–those aren’t random numbers; they really did have all those previous prototypes built in various stages of completion)
A few more complete vehicles, either finished or nearly so.
A bunch of other stuff I’m not thinking of.
So today’s test is just a tiny fraction of the full budget. It’s hard to even really put a number on it since it reflects a small portion of the output of the production system they built, which can build vehicles at a high rate whether they’re needed or not, but in any case the figure is small.
I think the debris was rather more energetic than they expected, but ultimately it was a calculated risk. Even with the amount of it, they got kinda unlucky there. They’ll probably make the money back via their channel, in any case. And better to take out the back window than the camera tower on top (which probably cost more than the van).
Which is a fairly egregious failure of ground systems engineering. You’re not really allowed to go “oops, bigger blast zone than expected, sorry for the debris damage.”
I understand agile processes allow for “go fast, break stuff, fail forward” but operations safety has to be an exception.
The Orbital Launch Mount (OLM), i.e., the milkstool thing, did its job. This is a hugely complicated structure, as complicated as the vehicle itself, and had serious room for failure. For instance, there are 20 arms around the ring that serve as launch clamps and provide services for the engines. These had to retract perfectly on launch. If one failed to do so, it would catch the edge of the booster and probably rip that portion off. Badness ensues.
The autogenous pressurization system worked. No indication of losing pressure due to ullage collapse or otherwise.
While the engine failures are obviously not desirable, clearly the system was highly resilient to them happening. No cascading failures like with the N1 rocket. Starship lost a few immediately, and a few more over time, and nothing too untoward happened.
Structurally, it looked great. Survived max-Q, and remarkably, even held together when the whole thing was spinning. I might even suggest the thing is overbuilt.
The bad:
Well, it didn’t make it through stage separation. It would have been nice to see that. But SpaceX has been setting expectations low.
As shown in the pics above, the rocket left a hell of a crater below the OLM. They really need to solve this problem. Whether that means a flame diverter or something else, I don’t know, but a big layer of concrete is clearly not sufficing.
Raptor reliability needs improvement. Though this might relate to the debris kicked up from under the OLM. Can’t blame the engines if they failed from to a 10 kg chunk of concrete hitting them.
Overall, the massive cheer the SpaceX employees let out after the flight termination system engaged should tell you everything out whether they considered it a big success or not. Sure, more would be better, but they proved out a huge portion of the newly developed technologies. This is an excellent first flight of a system that does so many new things.
ETA: @gnoitall a few posts up; thread’s moving fast.
I’m not sure “safety” is the right word when were talking about damage to equipment vs damage to people. At least not in the admittedly rough and ready world of experimental space launch. In my industry, kicking rocks through truck windshields is definitely under the rubric of “safety”.
It may well be sensible to give these issues mere napkin-level treatment rather than the dozen man-years of engineering analysis NASA would have done. And still one of the early shuttle launches destroyed a flame trench and damned near destroyed the vehicle just leaving the pad.
These may, repeat may, be situations where the quality and quantity of the actual data to calculate from is so vague or poor that any honest calc ends up with the error bars swamping the result. At which point you flag the risk as “uncontrolled” and make a cost / schedule risk assessment to armor up to an arbitrary big number taking X time and Y dollars or armor down for much less than X and much less than Y, and fix the resulting damage, if any.
And meanwhile instrument the shit out of the situation so now you can do calcs that yield meaningful results, not just WAGs to 12 decimal places.
The NSF guys had special permission to be there. I’m sure they signed a waiver saying that they accepted any damage that happened. They wanted to be within the blast zone for better footage, hoped it wouldn’t be too bad, and kinda lost that bet. It would only be a problem if the debris field was actually way larger than even SpaceX’s worst models, but there’s no evidence that was the case.
And yeah, they do need to fix this before Starship is launched regularly, but for now what matters is whether there was ever any risk to humans or property damage outside of expectations.
I wonder whether a part of the ground damage solution is simply a taller OLM. Although that word “simply” conceals a LOT of complexity. The good news is that if the do decide to build a flame trench, Starship has already given them a nice head start on digging the hole.
I was thinking about that, too. It’s certainly possible, but damn, that’s getting to be pretty high up. Imagine a 400 ft vehicle on top of a 400 ft mount–and then riding a manlift to do some maintenance on the tippy-top.
