[StarDucks]

What is the standard process for failure analysis for a company like SpaceX and how do they likely integrate the lessons into the next iteration?

Do companies like this have some form of high level "failure procedure" they go though?

[101404]

They are rapidly iterating through SN versions. This kind of thing is expected to happen.

This one was SN1. SN2 already has "better welds" (Musk tweet from some days ago). They built a welding system that better fits their needs and should produce much stronger welds. So this explosion just shows that that was probably a good idea.

See more rockets blow up over the next 20 iterations. Plan is SN1 to SN20.

[refulgentis]

It wasn't supposed to explode. The first one wasn't supposed to explode either.

[101404]

It wasn't the plan, but it was always a possibility. That's why they changed the welding equipment some days ago, because the current welds aren't good enough.

At a US Air Force event yesterday, Musk said that currently they are focusing primarily on production equipment and methods, not so much on the actual rocket.

[refulgentis]

Ah Elon, you always have an excuse

[nemosaltat]

I worked there briefly after leaving the Navy. I worked in nuclear power on submarines, both of which have incredibly high standards for QA. During my short time at SpaceX, I was fairly shocked by their QA process and ultimately their failure analysis “procedure.” In June 2015, they hadn’t had a catastrophic failure in something like 8 years. We all went in on Sunday (happened to be Elon’s birthday) to watch the launch from the Hawthorne campus. Coming from the Navy, I was expecting weeks or even months of “stand-down” and review. Ultimately, the failure was quickly attributed to a failed rod end bearing and if IIRC we are back to business as usually by Thursday. It was that response (or lack thereof) that prompted me to seek alternate employment. I wish SpaceX all the best, but can’t imagine being there as they move closer to manned flight.