Hacker News new | ask | show | jobs
by nomel 732 days ago
No it's not.

> While Starliner is docked, all the manifolds are closed per normal mission operations preventing helium loss from the tanks,

> “We can handle this particular leak if that leak rate were to grow even up to 100 times,”
2 comments
fransje26 732 days ago
If your propulsion system is developing leaks as you go, and, for good measure, you are also loosing thrusters in critical flight phases, than yes, this is terrifying.

What you have is a known failing system, with a very high probability of additional failure as soon as it is repressurised and it enters its next stress cycle. Which is normally when these things like to fail.

link
nomel 731 days ago
> What you have is a known failing system

Absolutely not. Engineering in the real world doesn't work like this. You don't design for perfection. You have intentionally defined specs set to what can be intentionally accommodated in your design.

In this case you literally have is a system operating at 1% of the value that can be accommodated. That is not a problem. Something undesirable happening does not necessarily mean the system is failing, or that it's even a practical problem.

link
fransje26 721 days ago
Mmm.... No.

That's absolutely how it works in engineering. You build fail-safe systems when you can, if safe-life systems where you must.

There is no redundancy in the structural integrity of an aircraft wing. Once it falls off, everybody dies.

Similarly, there are little redundancy margins in a spacecraft propulsion system. You will plan for a thruster malfunction, but if you loose your entire control system in flight, or if you develop 5 different leaks in flight, than it's safe to say you have a failing system. At no point in the design phase were any of those failure modes deemed "acceptable".

link
protocolture 732 days ago
Its terrifying that the build quality is so poor due to the issues we have seen.

The helium leak is obviously acceptable, but this thing is meant to return humans to the surface, can we trust that it wont have issues in other areas?

link
nomel 731 days ago
> can we trust that it wont have issues in other areas

We can trust it's not perfect. For this problem, operating at 1/100 the failure point probably shouldn't stir too many emotions. There's very little in your daily life operating at those types of margins, except maybe the ground you walk on.

link