SpaceX Starship / Super Heavy Discussion Thread

Miscellaneous and Personal Stuff I Must Share
222

I was thinking in terms of direction, more than just speed; that’s why I specified vectors. Relative to the original vector, the first stage is deflected downward, the upper stage is directed downward.

I’d have to think about what the resultant speeds are. There’s no actual force pushing the two stages apart.

I’m not talking about thrust, I’m talking about the timing of the separation relative to how much the rocket has rotated.

The rocket is travelling, pointing along its velocity vector. At some point, it starts rotating (pitching up, in aircraft parlance). When it separates, the two pieces are going to start travelling in different directions. The new directions depend on when the separation happens.

How precise does that need to be, and how precise is the timing and operation of the clamps that hold the stages together? Suppose the clamps release a second too late, and the rocket has rotated a little farther than planned; the two stages will go in slightly different directions than planned. How much leeway is there in the pitch angle when the clamps release to still achieve the desired orbit?

223

Even more confusing: NASA Space Flight, which actually has nothing to do with NASA at all (they just happen to spend a lot of time covering NASA). Basically a bunch of enthusiasts.

I think they try to go by NSF now, like KFC did to indicate that they are unrelated to chicken products. Anyway, they run this page:
https://www.nasaspaceflight.com/

224

I think there’s an xkcd for this one…

Consider the centers of mass for the booster, Starship, and the total system. Induce a rotation just before stage separation, and the velocity vectors of the individual CoMs will have a small offset relative to the total CoM. Release the clamps, and the stages retain these small deltas, but there is no longer a force that is slowly rotating them (the clamp force is the origin of the acceleration that is changing the velocity vectors). So the stages separate with the relative velocity they last had.

I know, but thrust is what’s important. If the craft ends up with a slightly different velocity than it wanted, it’s not a big deal–the control system will easily correct for that. The only issue is that if it spends a lot of time with the wrong vector, it will waste propellant and not make orbit.

I think it could easily be 10-20 degrees off in the clamp release and still be just fine. The velocity will be off by just a few meters/second at most. Thrust vectoring can correct for that easily and cheaply (in delta V terms).

225

I remembered from the Everyday Astronaut interview (#2) with EM that the goal was to get into orbit without blowing up, then he dialed back and said if the booster did it’s job without the ship blowing up he would count that as progress, then he said “to be frank if it takes off without blowing up the stand, that would be a victory…[because]…it’s harder for us to make a Stage 0 than to make a booster or a ship.”

So my question is, with all that damage, did Stage 0, in essence, “blow up”?

226

There’s probably more damage than they’d have liked, but no, Stage 0 didn’t blow up. The launch tower is still there, as is the tank farm and all of the support infrastructure. The OLM itself appears to be largely intact aside from having a giant hole below it.

LSLGuy joked about this, but I’m not even sure he’s wrong–the crater gave them a headstart on the excavation they’ll need to build a more robust flame diverter system.

227

Did Musk talk a lot about the good possibility of the ship blowing up before the launch happened, or did he move the goalposts after the fact?

228

It was all well before the launch. Most specifically, in the Twitter Spaces thing he did a day or so before the first attempt on Monday. In fact he spent almost the whole time dodging questions about later stages of flight and subsequent missions, and emphasized over and over that they just want to clear the pad and that everyone should lower their expectations.

The SpaceX announcers said exactly the same thing before Monday’s flight attempt.

229

Some quotes from that event (which was on Sunday):
https://www.reddit.com/r/SpaceXLounge/comments/12ot0qi/starship_launch_discussion_on_twitter_spaces/

If we don’t destroy the launchpad, it’s a success

probably tomorrow will not be successful, if by success one means getting to orbit

230

Another video.

I’ve seen some speculation here and there that with all the concrete being kicked up the rocket may have been hit. That this much concrete flying is not typical (I’m not an expert in this by any means).

231

How far does it rotate before the separation? Assume it’s at 0° before the rotation starts. You want the upper stage going up and forward, and the first stage going down and back. Seems to me you’d want about 315° of rotation before you release the clamps. Is that in the ballpark?

Incoming!

232

I don’t think so–I think it’s only going to be 10-20 degrees. And they may start at a “negative” angle to split the difference. For the separation itself, what matters is the rate of rotation, not the angle. So it just needs enough time to accelerate and then separate when there is enough relative velocity.

They’ll want to separate when Starship is pointing more or less in the same direction as before. 10-20 degrees off is fine. 45 is probably too much.

The booster will be going the wrong direction–it needs to rotate back by ~90 degrees to perform the boostback. But it’s lighter weight and has more margin. So it should be the one to perform the big rotation, not Starship (which needs every ounce of performance it can get).

233

Thank you.

234

Sure thing. And as a general note for new launches: this was well within “normal” levels of success for brand-new launch vehicles, even totally conventional ones. “Failed at stage separation” is the bread and butter of new launch vehicle failures (which SpaceX themselves are no stranger to, as one of the Falcon 1 failures was due to an unusual stage separation problem). So for Starship to do this well while being a vastly more ambitious vehicle than most is doing pretty well.

Generally, this is because boosters can get a reasonable degree of ground testing in before flying. You can do a static fire of the engines and test a bunch of other stuff. But for stage separation onward, you have very limited ability to test on the ground. You do your best to get it right, then hope for the best.

235

They need some sort of…I don’t know…unexciting company to drill some holes.

236

Ok, this is all massively speculative, but here goes.

I think that most of the problems can really be traced to the kicked-up debris. They severely damaged a few engines to start with, shutting them down, and damaged a few others (or surrounding infrastructure, like fuel lines) so that they shut down later in flight.

This caused, among other things, underperformance of the boost stage. By itself, this didn’t cause the failure. But it meant that at the nominal stage separation time, it was nowhere high or fast enough. It actually still had plenty of fuel, since the damaged engines weren’t pulling anything. And it still had enough thrust to accelerate.

The announcers called out the upcoming stage separation, since that’s when it should have happened, but it was still far off from the vehicle’s perspective. Gorsnak’s video shows what looks like the hydraulic units failing, which would have prevented engine gimbaling, or at least rendered it too weak to be effective. The vehicle may have even fallen back on differential throttling, which wasn’t enough.

So what looked like the stage separation maneuver was (coincidentally) actually the loss of vehicle control. It probably would have kept going and maybe had a successful separation if it weren’t for that, though with the failed engines it probably still wouldn’t have made orbit.

Instead, the booster kept firing, since it had plenty of fuel left, and was trying its damnedest to regain control even though it was hopeless. Mission control let it go for a while since it was presumably still within its corridor, and they may as well get as much telemetry as possible, but eventually they had to press the button (or it was automatic).

Why did the hydraulic units fail? Again, probably damage from the debris. Possibly it was something unrelated, but I think it’s a reasonable guess.

New vehicles ditch the hydraulics completely for an electric gimbaling system. That should make them more robust by eliminating some points of failure, but of course with that much damage who knows.

237

Maybe they will need to build that flame trench after all.

238

I think that’s pretty much inevitable at this point.

My guess is that they’ve been avoiding it because it’s a massive inconvenience. They currently have this platform that they can park under the stand and raise up to vehicle level, so as to perform engine maintenance and other things without having to move the whole vehicle back to the construction bay. That’s likely to be impossible with the trench.

A giant hole also makes more mundane work more challenging. With the current design, workers can mill around at ground level without any special measures (unless there’s a crane lift happening or some other above-ground work). But with a flame trench, workers can’t get anywhere close without significant safety measures to prevent them from falling in. Unless you build a guardrail or something–but how does that survive?

Nothing insurmountable here, but it just makes everything less convenient than it is today, which makes it slower and more expensive. They probably don’t have a choice, though.

239

That doesn’t make sense to me.

To use an extreme example, suppose the rocket starts pitching up, and when it’s exactly 90° from the direction vector, the clamps open and the stages separate. The upper stage will be propelled backwards, and the first stage forwards (relative to their common center of gravity). Seems like that would be the opposite of what you want.

What am I missing?

240

Well, the upper stage then flips forward and resumes its boost to orbit, while the lower stage begins boostback.

241

Right, but the direction doesn’t actually matter that much. A few meters per second is negligible compared to the total delta-V the vehicles are capable of. What matters for separation is that the velocity vectors point away from each other. And that’s going to be the case regardless of the angle, but go up with an increasing rate of rotation.

Of course, you wouldn’t intentionally do this, which is why I suspect the angle relative to the overall velocity will be fairly small, like 10-20 degrees. Starship stays out in front and the booster can do its flip from behind. But there should be significant tolerance for error here. Again, we’re only talking a few meters per second compared to an overall velocity of multiple kilometers per second.