Hacker News Reader

Comment by simonpure | Hacker News Reader

At this stage of the investigation, preliminary review of the data and debris suggests that a large breach in the cryogenic helium system of the second stage liquid oxygen tank took place.

...

Pending the results of the investigation, we anticipate returning to flight as early as the November timeframe.

View on HN

Note that the second-stage helium system was also the source of the previous explosion of a SpaceX mission to the International Space Station (which blew up during launch).

https://en.wikipedia.org/wiki/SpaceX_CRS-7#Launch_failure

But:

> Through the fault tree and data review process, we have exonerated any connection with last year’s CRS-7 mishap.

I found that puzzling too... doesn't this mean it was the same component, though perhaps failing in a different way?

Sounds like that pressurization system needs to be taken out back and shot.

Frankly, that's how you increase your failure rate. If a system fails, you figure out why, figure out any other failure cases that may have been missed, and you fix them.

You don't ditch something that largely works, and replace it with something brand new, for which you have zero data. That's how you lose rockets, cargo, and lives.

Same thing applies to software: you encounter an error, you analyze why it happened. Then ask: what other places could have this error? Can I eliminate this entire class of error by design or other method or do I need to fix them all individually?

You organically end up with a more robust system that way. Don't throw out the baby with the bathwater.

And then there is the "Did you turn it off and on again" school of thought. I've had to fight against that so hard before when the correct issue is inducing something to drop core, or fix a leaky program using valgrind.

Assuming they were correct about the first incident, it was caused by a faulty strut that was supposed to hold the helium COPV in place, but failed. If they've already ruled out a strut failure in this case, it seems pretty safe to say that they are unrelated.

Though it must be said that NASA apparently never signed off on the strut explanation. So conceivably it was never the strut, in which case it could be the same issue. Then again, I assume the people working on this are much smarter than me and have much more information, so they probably considered and rejected that too.

The same system. There would be multiple components of the helium pressurization system that have failure modes causing tank overpressure.

The previous failure was with a strut holding the tank. That's not really the same 'component', depending on one's specific definition.

Perhaps taken out back and blown up would be more apropos?

Only in a loose sense. The CRS-7 failure was a strut failing which held the tank. This time, the tank itself is suspected to have failed. These are probably completely unrelated.

They might turn out to be related in an unexpected way. For example, they might have redesigned the strut in such a way as to cause excessive stress on the surface of the tank near the attachment point, leading to rupture.

These systems are probably more complex than the cruddiest Enterprise software that any poor soul has ever had to maintain, so I guess anything is possible.

George Carlin would have a field day with that wording.

Reply on news.ycombinator.com