I think a smaller N number of more dives would yield better results. Down, up, down up...repeat until failure. In parallel do it with another hull at the same

I think a smaller N number of more dives would yield better results. Down, up, down up...repeat until failure. In parallel do it with another hull at the same time. Or maybe more than one. Compare. Testing to failure is a nice option because it finds certain types of unwelcome surprises. But it's not really complete, either -- there are probably a lot of ways that a submersible can be compromised, and there's no way to be sure what caused this particular failure.