Hacker News Reader

Comment by addaon | Hacker News Reader

The joys of HN is finding experts in just about anything, so...

Assuming that NASA is correct about this leak being associated with an interior or exterior weld, what can we learn from the rate of growth of the leak? I assume (with no domain knowledge, seeking enlightenment!) that a leak like this is big enough to reflect a fracture in the weld, and not merely air sneaking past a tiny flaw; and I assume its growth means that the fracture itself is growing, under the combination of load (atmospheric pressure on one side) and cycling (thermal, etc). Is a fracture like this likely to continue to grow in a linear-ish domain, or is there a point where propagation goes non-linear assuming the loads stay constant-ish? Is this even a sane question to ask?

View on HN

Obviously it can't be linear forever, if you assume the crack is along a circumference then eventually there will only be a tiny piece of metal still intact between the two halves, and one would expect it to fail far before then.

These kinds of failures tend to be linear until they are not, and the failure mode is catastrophic. If I were in charge I would have already shut off that module on the risk/benefit analysis of catastrophic loss of the station vs. not being able to evacuate as quickly in a less catastrophic emergency. Maybe they are worried they won't be able to get the hatch open again if they can't equalize the pressure again?

For historical reference, consider Aloha Airlines Flight 243[1].

https://en.wikipedia.org/wiki/Aloha_Airlines_Flight_243

> Maybe they are worried they won't be able to get the hatch open again if they can't equalize the pressure again?

Well note that they currently keep the hatch to the docking port closed unless they're using it, so that presumably mitigates some of the risk (assuming the leak is confined to that location). Not using the whole Zvezda module itself would be more or less impossible from my understanding.

If they decided to stop using that dock port at all then I think that complicates planning and makes everything overall riskier, so while it wouldn't be a disaster to do there's still risk you have to weigh.

Disclaimer: I don't know how space modules are welded or what forces act on a space module.

Welds fail for multiple reasons - heating too fast, cooling too fast, material defects, human error, equipment issues, oxidation. The best human welder will still have some variance in their weld even if they lay a perfect bead.

Once a weld begins to fail, under normal conditions (vibration, stress, etc), it will slowly get worse, but that can also mean deformation around the weld or even sometimes shearing right next to your weld. The article doesn't mention that the growth has been linear only that it is increasing. I expect that the leak is increasing over time in a non-linear fashion as force is applied to the weak area and metal continues to separate and/or shear. Still, 3.7 pounds of atmosphere a day is a very slow leak (relatively), a few bike tires a day.

My back of the napkin math tells me a .2 mm puncture would do this. This must be a very tiny crack.

I'd just paint over all the welds and be happy until it unseals itself again.

From the article, it started leaking years ago, and almost doubled between Feb and Apr of this year, so it is definitely not growing linearly.

> Is a fracture like this likely to continue to grow in a linear-ish domain, or is there a point where propagation goes non-linear assuming the loads stay constant-ish? Is this even a sane question to ask?

It depends on so much that we don't know from the article.

Non-expert, pure speculator here: I cannot help but think of the Titan submersible and its minor flaw in the hull that slowly became more and more of an issue with each dive and pressurization cycle. I wonder if the 'untenable' leak rate that isn't agreed upon is the rate at which an accelerated and catastrophic runaway event occurs?

There was no "minor flaw" in the Titan submersible.

It was a very bad design[1], using substandard material[2] in new and untested ways[3] with poor assembly practices[4]. It was fatally flawed from the start.

1 - Asked about the carbon composite used in Titan’s experimental design, Cameron said, “It’s completely inappropriate for a vessel that sees external pressure.” He went on to say that carbon fiber is very helpful when used for applications subject to internal pressure, like scuba tanks. But, he said, “for something that’s seeing external pressure, all of the advantages of composite material go away and all the disadvantages come into play.”

2 - Another possible "shortcut": https://www.latestly.com/world/titan-submersibles-hull-was-a...

"Expired" bare carbon fiber might be ok; expired prepreg is likely not ok. And I still can't get my head around the apparent mixing of prepreg and wet winding resins in the same laminates.

That article says Boeing has no record of ever selling oceangate composite material. So either he got it for free (Dumpster diving?) or it came from someone else. Maybe some guy in a parking lot: My boss told me to deliver this stuff to Boeing but they wouldn't take it because its expired, I'll sell it to you cheap.

3 - "Thickness, he says, was estimated using micromechanics, and then verified with finite element analysis (FEA).”. >Wonderful. That sort of maybe can work for tension stresses, but can be highly unreliable for compression loading

4 - https://www.youtube.com/watch?v=4PUTbK5AqY8 : >For such a critical joint I see some questionable prep work >in the video. >-Bare hands touching the bond surface. >-Solvent wiping used instead of caustic degreasers. >-Poor wiping technique >-no surface profile/grit blast immediately prior to fixing.

All sources are from commenters on: https://www.eng-tips.com/viewthread.cfm?qid=508005

It's important to remember that the pressures involved are very different. The ISS pressure vessels only have to hold 1atm in, and not several atmospheres out.

To reinforce this point, it is around 375 atmospheres at the titanic. The deep sea is a crazy place.

Equally important is that the ISS is under tensile load (the cylinder is trying to explode), while a submersible is under compressive load (the cylinder is trying to implode).

It's just 1 atm. You can't even say it's trying seriously.

Reply on news.ycombinator.com