I've done a lot of manufacturing with all sorts of metals. I am not sure one can blame a vendor for a grain structure problem. This is a testing failure. A failure to identify parts that don't pass tests.
The problem might very well be that the very tests a part must pass will weaken the part to the point that it is not usable or less reliable. In other words, some tests are destructive.
Here's on of many interesting pages that came up by searching for "how to test aluminum for grain structure".
I guess I am saying I hope a vendor isn't blamed for this when the reality of the matter might very well be that testing to 100% certainty is impossible.
But there's so much margin for error on these specific parts. Even if the test only ensures 60% strength, and weakens the part by 50%, that's still strong enough.
A part being mildly out of spec might not be a vendor problem. A part being 20% strength is absolutely a vendor problem.
Why was this designed such that the failure of A SINGLE STRUT would be catastrophic?
I don't like to blame a vendor for what should be an engineering problem, whether this means design or testing.
The problem here --assuming it is as described-- is that someone designed a system with the assumption that none of these struts would fail. And, furthermore, executing on a design where the failure of ONE strut could cause a disaster.
Anyhow, that's what it looks like to me given what's been released.
Why was this designed such that the failure of A SINGLE XXX would be catastrophic?
There are plenty of things in a rocket that can fail that would take the entire vehicle with it. Structural mechanics is pretty well understood and loads are well predictable, so it seems perfectly reasonable to me to design with the assumption that a part with a 10x safety factor will not fail.
If you didn't, your rocket would never get to orbit anyway.
I completely agree on the responsibility for acceptance testing, but _someone_ is responsible for the grain structure of the material, after all that's one of the most important properties of metals aside from their constituent metal content, which is itself chosen for its influence over the microstructure of the alloy.
Yes, agreed. That said, I prefer to function from a mindset where I place fault in engineering first. What I mean by this is, I assume it is my responsibility to verify, as much as possible, that components and assemblies meet the required specifications. In other words, don't just engineer the parts. Take the time to engimeer a "failure is not an option" process as well.
I suspect SpaceX's costs are going to, over time, increase significantly as they continue to learn that playing it loose isn't always possible in that business. They are famous for going for COTS in order to save money.
This is NOT a put-down. I think what they are doing is fantastic. It is obviously redefining aerospace. At the same time I am astounded that critical structural components are not 100% tested. That said, metalurgy isn't my area of expertise, which means my opinion here could be complete nonsense. Perhaps this particular failure mode can only be tested through destructive methods (sectioning?) which means you can never be 100% certain to be flying good metal.
Are SpaceX famous for COTS though? My impression was that they manufacture a ridiculous amount of stuff in-house for ... well pretty much this exact reason.
I'd argue that the likely result here is Elon Musk will continue his campaign of "hell with it, we'll build our own".
I've only been exposed to this in oil & gas, but there is a problem that non-destructive testing of drill stems does not test for all failure modes
Edit: i.e. NDT'd pieces are not immune from failure (but other factors such as age and in-service time/rotations/re-thread/re-collar runs etc. can provide additional context to NDT results and should correlate more or less with actual failure patterns).
In a single-use part, non-destructive testing mightn't be a complete picture of how the part will perform.
The problem might very well be that the very tests a part must pass will weaken the part to the point that it is not usable or less reliable. In other words, some tests are destructive.
Here's on of many interesting pages that came up by searching for "how to test aluminum for grain structure".
https://www.google.com/webhp?sourceid=chrome-instant&ion=1&e...
I guess I am saying I hope a vendor isn't blamed for this when the reality of the matter might very well be that testing to 100% certainty is impossible.