Saturday, November 18, 2023

Starship's IFT-2 Went Almost as Well as it Could Have

My perspective on the test flight is that it didn't make every milestone I was watching for or cared about, but it did pretty darned well.  

To begin with, the pad improvements seemed to prevent anything bad from happening - in particular, no chunks of concrete flew up into the engines destroying some and causing a fuel leak that eventually would kill the vehicle.  

All 33 engines ignited and ran for the entire required amount of time.  Hot staging worked; the engines were cut in groups until only three remained running, Starship 25's six Raptor engines lit; first the outer ring of three vacuum raptors, and then the inner three sea level Raptors lit as Booster 9 was dropping away.

Booster 9's destruction as it dropped away strikes me as unimportant - in the short term it's a "don't care."  My example is the Falcon 9 as they were working on getting it to land successfully.  The way I understand their cost model is that once the paying customer's payload was on the way, the booster was garbage so they were going to experiment on the piece of garbage as if it was something they found on the side of a road.  If they want to just throw a booster away, they'd do what every other launch company in the world does; drop it and "fuggedaboutit."  They were going to throw this one away but they were going to play with it first.  They started to play with it and it turned out less than ideally.  Don't care.  

Look at that array of 33 engines, all of them lit and firing, all of their exhaust trails looking to be the same color.  In other views you can see the Mach diamonds in the exhaust better than this shot.  John Kraus photo from X.

As is often the case, Scott Manley on YouTube posts a video in which he has taken a preliminary dive into the behavior of the booster.  His conclusion is that the booster's engines were interrupted by the flip and burn back maneuver the booster was doing.  Picture a long, skinny tank being rotated end over end; at some point, inertia can have the fuel at the top of the tank, and unable to flow into the engines.  It helps to know that often, with stages that coast between burns while on orbit, they use thrusters of some sort, or some other power to start the stage moving.  The inertia gets the fuel to the bottom of the tank, where it needs to be.  The tanks on the Super Heavy booster still have tons of LOX and Liquid CH4 in them and the abrupt rotations might well have forced engines to run without one or both fluids.  

The moment of staging - at the bottom of the booster, you can see a triangle caused by the three engines left running.  The bright spot at the top of the booster/bottom of the ship is the hot staging taking place.  The "psychedelic color" cloud is the exhaust plume around the two.  Image credit: SpaceX.

Unfortunately, Manley doesn't have as much to say about the Starship's failure.  Starship made it into space; in videos that he goes through the data coming down from Ship25 freezes and stays at a pair of values for a while.  The speed is 24,124 km/hr or just under 15,000 mph while LEO orbital velocity is generally given in the vicinity of 17,000 mph.  Altitude was 148 km, or 91.96 miles.  Blue Origin and Virgin Galactic sell seats on tourist rides to space that don't go that high or that fast.  Starship seems to have been around 15% short of orbital velocity, but this was billed as a suborbital flight so they'd better be short of orbital velocity.  I don't know what their targets were.

Starship seems to have also undergone a spontaneous explosion.  Excuse me, it underwent a RUD, to borrow Elon's old joke everyone has been using today.  I'll start using my own redneck version of that, IDBU - It Done Blowed Up - pronounced id-boo.  Scott shows a few seconds of evidence in his video that imply something went wrong that caused Starship to consume excess LOX. 

I view the Starship anomaly more seriously than the booster not being reusable.  Of course, they weren't going to reuse this ship after this flight; it was going to belly flop in to the Pacific north of Hawaii, but I'm automatically suspicious because while there were test flights of Starships before, none of them went this high.  Is the problem something related to flying in this range?  Doesn't seem possible.

I've already seen mentions of the next test, IFT 3, being NET February '24.  Three months instead of seven?  I'm not sure the FAA could evacuate their buildings in three months if they were burning down. 


  1. The flight met major milestones, so not disappointed. Now they need to learn what went wrong with Starship. Better to learn on a non-critical flight so you can hopefully avoid the same problem in the future. Space-X has a history of learning form the "OOPS" moments. They are willing to take more risk, and have a faster learning curve because of it. That's why they are launching more flight per year than their combined competition.

  2. The OLT worked. Hot staging worked. New actuators worked. Starship made it above 100km. And both survived MaxQ. Now it's just the fiddly little bits.

    Good one, SpaceX!

  3. Screen tanks would prevent propellant starvation until the tanks were actually empty. But that does cost extra.

  4. Starship clearly had an oxygen anomaly. What is different from all the other, successful flights? The hot staging. I suspect something in the engine bay was damaged and took a few minutes to become critical. I wouldn't argue with SpaceX engineers, but the design of the HSR isn't open enough (seat of the pants guesstimate, it's only ~40% open). Take a look at the Russian version to see what I mean. There's going to be a lot of shockwave plasma bouncing around in there and resonance is a bitch.

    Actually, I lied. I'm an engineer, thus I will argue with anybody's engineers. It's in the genetic profile of people drawn to engineering to be opinionated.

    1. Just to add: remember all the discussions about launching Starships with the engines so close to the ground? Many were amazed that there was no apparent damage from the practice. Well, they are much, much closer to the top of the booster during staging than they ever were to the ground.

      Just watch. The next iteration of the HSR will be a few feet higher. And the QD will be modified yet again.

    2. I've been reading Eric Berger's book about SpaceX and this morning read about the second test flight, where they almost made orbit except for overlooking the 11th biggest risk and had turbulence issues in the second stage tank. It seemed the moral of the story is that it takes three successful test flights to iron out everything. Flight two was markedly better than flight one, but flight two failed from a relatively minor risk - or a bigger risk that was underrated.

      The parallels to IFT1 and IFT2 leapt off the screen at me.

      The alternative to making the interstage a little taller is to wait a little longer before the inner three Raptors get turned on - in which case, why do hot staging at all?

  5. I would go with the acronym BURG: Blowed Up Real Good.

  6. As the day has gone by, I've run into a couple of important additions to the picture.

    Probably the biggest is that by mid-afternoon yesterday, Elon Musk was saying they had driven out to the launch pad and there's no damage worth mentioning. Tweeted here.

    The other one is more related to this subtopic, the interstage. In this SpaceX video you can watch the three sea-level Raptors carefully and see that at the moment you can see all three, 11 seconds on the clock, they're all pointed as far away from center as they get. By the 14 second mark they're all pointed on axis, to maximize thrust. This means that to minimize damage from the sea level raptors blasting the top of the dome, they point them as far away from the centerline as they can go.