SpaceX’s Starship spacecraft exploded on Thursday minutes after lifting off from Texas, dooming an attempt to deploy mock satellites in the second consecutive failure this year for Elon Musk’s Mars rocket program.
Several videos on social media showed fiery debris streaking through the dusk skies near south Florida and the Bahamas after Starship’s breakup in space, which occurred shortly after it began to spin uncontrollably with its engines cut off, a SpaceX livestream of the mission showed.
The failure comes just more than a month after the company’s seventh Starship flight also ended in an explosive failure. The back-to-back mishaps occurred in early mission phases that SpaceX has easily surpassed previously, indicating serious setbacks for a program Musk has sought to speed up this year.
Ugh, I don’t actually hate the hardware rich testing strategy SpaceX uses, but you’re supposed to learn from your failures before launching the next rocket.
This looked like almost exactly the same issue - fire in the skirt area that catastrophically destroyed the engines.
Pretty sure the other fire was on the inside due to an engine leak. This fire/leak was on the outside.
Both around the enginee, but different spots.
The thing they changed was cutting holes in the area where the fire was last time to “ventilate” it. I think it’s the same fire just blowing out the holes and melting something different.
They were also going to dump extra nitrogen or something into the area to prevent the fire from happening, so that would have had to also fail.
Either way if its the same problem in some way or another hopefully they can fix it before raptor 3 (which should solve a lot of leaks as there’s less tubing). I think raptor 3 is over a year out.
I’d kinda expect a little longer delay as well between the next launch if it was the same issue. If your fix failed, it might be more complicated than you thought.
They are learning, but the versions launched are never the current version at launch time, they are all older models. As changes are made those changes go into the next one being built, not the ones currently under construction already, unless it’s an easy retrofit, like when they were changing the tile layouts.
And even those retrofitted changes may not be adequate depending on the cause. When dealing with lost vehicles you also have limited forensic evidence to use to determine root causes. Cameras don’t see everything sensors don’t log every aspect. And fixes you thought were adequate, may not actually be enough. That’s why they iterate and try again in the real world, instead of just simulating everything and hoping it works at launch for 10x the cost. These issues sometimes taking multiple tries to fix are an expected parts of the process.
Exact same symptom. Who knows what actually started it either time.