Building the Falcon 9 aft thermal shield for reentry
From Why Pure Software Is Uninvestable: The SpaceX & YC Way to Build Hardware on Deeptech Decoded, with Jamie Gull.
Transcript
Actually that's the thing that if you do the math of even your cost as an employee to the company, you burn my heart for nothing because you're wasting my time too. Every day I don't get the piece into my hands.

You burn your hours, your managers hours, your VP's hours, and your procurement f like approval hours. Like every it's just
And then also everything cascading to the production too, because it's not like a software I can freeze that part of the module and then you can go ahead with the lunch and then do the whole thing. Actually to get back to the whole thing, can you take us back to the times that you were manually tying GoPros to the floating

Yeah.
debris? For some context, you worked on Falcon 9's reentry systems, right?

Yeah, so I did
There is this fine line between, you know, this fast iterating, but at the same time, you don't want to get drawn into the data collection simulation. And then you also want to break it down small enough that you can iterate fast. In hardware, especially, it's much more challenging than software, obviously.

Yeah. This that it is a great story. It was like so I did the the aft thermal shield on the bottom of the rocket, which enabled the re-entry and landing. So it was the first one. Very exciting project for me because nobody had done it before. and w the we had some unlike most thermal shields which are like very fixed, I wanted it to be incredibly strong. ours had to blow off in the case of an engine Because it needed the they needed to depressurize the engine bay. Cause if you don't, it cascades into other engine bays. So if you lose an engine, and this is like a another another SpaceX philosophy, is like it's okay to lose an engine. back in F9 days, I believe it was you could lose two or something and still complete the mission. Starship now can le lose like, I don't know, three, four, or five.
of how many.

I want to say booster is like 30-ish engines, 27 maybe. Yeah, so like that's not usually done or wasn't that common. And so we had some really weird requirements about having to like let it exp blow out, which makes it really hard to make it strong for re-entry. And we were having trouble on re-entry with the panels coming off. and then the engine bay would be exposed and you know, we had a few rockets blow that way on re-entry, which was okay because of his experiment at that point.


