SpaceX Falcon 9 FT - ORBCOMM-2 - Dec. 21, 2015 (Return To Flight) DISCUSSION

[dorkmo]

since we're clumping japan into this large deming thing, i think it is interesting to compare to the the recent airbag fiasco. honda and other car companies trusted takata to provide parts at certain spec. was honda testing these inflators? what percent failed? how many should honda independently test to be confident in takatas work.

[Kabloona]

since we're clumping japan into this large deming thing, i think it is interesting to compare to the the recent airbag fiasco. honda and other car companies trusted takata to provide parts at certain spec. was honda testing these inflators? what percent failed? how many should honda independently test to be confident in takatas work.

The Takata issue involves the aging of ammonium nitrate inflator pellets over a period of many years, in which temperature-induced phase changes and moisture contamination caused cracking of the ammonium nitrate pellets and increased their burn rates. So even 100% testing of the inflators right out of the factory wouldn't have identified the issue. It takes years for the inflators to go bad. Unlike the SpaceX strut which was likely defective right from the factory.

(Incidentally, people have been trying to phase stabilize ammonium nitrate for use as a rocket propellant ingredient for a long time, but it's exactly this weakness--the phase change with temperature--along with an extreme tendency to absorb moisture--that has made ammonium nitrate impractical as a solid propellant oxidizer.)

So since this is really an entirely different issue, could we please stick to SpaceX RTF stuff? I know there's a long downtime here, but as has been suggested above, the history and methods of automotive/industrial quality control deserve their own place elsewhere.

[Herb Schaltegger]

since we're clumping japan into this large deming thing, i think it is interesting to compare to the the recent airbag fiasco. honda and other car companies trusted takata to provide parts at certain spec. was honda testing these inflators? what percent failed? how many should honda independently test to be confident in takatas work.

The Takata issue involves the aging of ammonium nitrate inflator pellets over a period of many years, in which temperature-induced phase changes and moisture contamination caused cracking of the ammonium nitrate pellets and increased their burn rates.

There's actually a bit more to it than just aging inflator pellets, but I can't say more due to an honest-to-goodness NDA and ethical consideration. That said, your general premise (that it's a more complicated issue than "100% testing") is dead on accurate. There are plenty of things you CANNOT test to 100% load as acceptance test methodology, even things that don't involve flammable or explosive materials. Elastics, for instance, or metals that suffer from creep and/or plastic deformation. The ACT of testing impacts the life of the item and/or the system into which is installed or intended to be installed.

[Jim]

What if the strut design was wrong in the first place.  Wrong material, wrong manufacturing technique, etc.  Some technique or quality that allows for variability in the final product.

[sdsds]

As SpaceX returns to being a company actively launching payloads I think they deserve a lot of praise. They are approaching the point where they will declare their first Full Thrust vehicle to be ready for flight. But not only is this a substantially different vehicle from those they have flown before, they are attempting to reach flight readiness along a path different from the paths they took with prior vehicles.

That's because this is their first flight since they have changed their corporate "risk posture." Previously they were comfortable with the level of risk associated with using parts like the strut that failed; now they are not. So even though they've previously brought three launch vehicles to flight readiness (F1, F9 1.0, F9 1.1) this time they're asking not just "are we on the path" but also "does this path truly lead to readiness?"

I wish them good fortune, and hope they will be a shining example of how to "do it right!"

[CameronD]

There are plenty of things you CANNOT test to 100% load as acceptance test methodology, even things that don't involve flammable or explosive materials. Elastics, for instance, or metals that suffer from creep and/or plastic deformation. The ACT of testing impacts the life of the item and/or the system into which is installed or intended to be installed.

That's why you take random samples off the production line and test them instead - and test them again, usually to destruction.  Statistical analysis then tells you, based on the number of samples and your production metrics, just what your chances are of a faulty one getting through and you run with that.  Sure, it's no 100% guarantee, but by including a range of NDT's and material scanning techniques you can get mighty close (like better than 99.5%) - and certainly good enough for commercial aviation.

To answer Jim:  If the test specification is right, you can detect design and tooling issues also.

[Rik ISS-fan]

When the pictures of the TEL for LC-39A showed up, a question arose in my mind. How sure are we the RTF will be from LC-40? The Falcon 9 v1.1FT is a lot longer than the F9 v1.1, the TEL deals with the bending moments caused by the payload on top of the second stage. Because both stages are longer, the arm is longer and thus the moments are larger. Does someone know if a new TEL will be or has been build (or the TEL is modified) for LC-40?
Or is it more likely the RTF will use LC-39A, and LC-40 will have to wait for a new TEL and other modifications (Cooling systems for densified propellents)?   

[cscott]

From the pictures is obvious that the new TEL is nowhere near ready yet.  No base, no plumbing.  If the RTF is happening "soon" it will certainly not use the new TEL.

[cscott]

What if the strut design was wrong in the first place.  Wrong material, wrong manufacturing technique, etc.  Some technique or quality that allows for variability in the final product.

I believe there were posts earlier in this discussion which implied exactly that.  That casting the parts allowed for the possibility of defects, so the new replacement struts would be forged instead to eliminate that failure mode.  Or vice-versa, my memory of the discussion is foggy.

[woods170]

From the pictures is obvious that the new TEL is nowhere near ready yet.  No base, no plumbing.  If the RTF is happening "soon" it will certainly not use the new TEL.

And by extension the RTF will not use LC-39A.
LC-39A will not be hosting any launch until well into 2016.

[abaddon]

The Falcon 9 v1.1FT is a lot longer than the F9 v1.1 [...]  Because both stages are longer

I thought the 1.1FT was a little longer (a small 2nd stage stretch plus a longer interstage) and the 1st stage was the same length.  Certainly not "a lot longer" in my opinion, but do you have numbers for how much longer you think/know it is relative to the 1.1?

[Rik ISS-fan]

And by extension the RTF will not use LC-39A.
LC-39A will not be hosting any launch until well into 2016.

Than we all missed news about modifications to LC-40. For F9FT they need a reinforced TEL and cooling/ Cryogenic systems to densify the LOX and Kerosine. any news on this?

I don't now how much longer the F9 v1.1FT is conpaired to the F9 v1.1. But if I am not mistaken the TEL at Vandenburg became useless because of the stretch. The TEL for LC-40 was already elongated to be compatible with the F9 v1.1, I doubt they could do that again.   

[Lars-J]

And by extension the RTF will not use LC-39A.
LC-39A will not be hosting any launch until well into 2016.

Than we all missed news about modifications to LC-40. For F9FT they need a reinforced TEL and cooling/ Cryogenic systems to densify the LOX and Kerosine. any news on this?

I don't now how much longer the F9 v1.1FT is conpaired to the F9 v1.1. But if I am not mistaken the TEL at Vandenburg became useless because of the stretch. The TEL for LC-40 was already elongated to be compatible with the F9 v1.1, I doubt they could do that again.

Why do you doubt it? The TEL is not monolithic, it can be modded. The main structure appears to have 3 sections, two joints where changes/extensions can be made.

The main issue with the VAFB TEL was for FH, not the F9 FT stretch.

[sublimemarsupial]

And by extension the RTF will not use LC-39A.
LC-39A will not be hosting any launch until well into 2016.

Than we all missed news about modifications to LC-40. For F9FT they need a reinforced TEL and cooling/ Cryogenic systems to densify the LOX and Kerosine. any news on this?

I don't now how much longer the F9 v1.1FT is conpaired to the F9 v1.1. But if I am not mistaken the TEL at Vandenburg became useless because of the stretch. The TEL for LC-40 was already elongated to be compatible with the F9 v1.1, I doubt they could do that again.

Why do you doubt it? The TEL is not monolithic, it can be modded. The main structure appears to have 3 sections, two joints where changes/extensions can be made.

The main issue with the VAFB TEL was for FH, not the F9 FT stretch.

Also the LC-40 TE was a new build for v1.1, not a modification of the v1.0 TE.

[Karloss12]

There's no need to invoke supernatural beings! Yes, all parts can have defects. Most defects are of kinds that can be detected. In general there is no need to test-to-destruction to eliminate defective parts.

Putting this back into context: what SpaceX is doing now, they could have done before.

They're not buying struts from that supplier any more, that's what they're doing differently.

And perhaps more importantly, they're load testing each strut.

Where has it been said that they aren't buying struts from that supplier?

As I remember, the problem was with the bolts in the strap rather than the strap structure itself.

All your questions are answered here:

http://shitelonsays.com/transcript/elon-musk-talks-failed-crs-7-dragon-mission-2015-07-20

 Elon may have used the word "bolt" as shorthand for "special attachment fitting on the end of the strut" because he didn't want to get into too much technical detail in public. Otherwise they'd simply be changing bolts or bolt test protocol instead of changing the strut design and the supplier.

It just looks like the head of a bolt snapped off.  You can't just magically fabricate and test thousands of struts in two weeks.  But you can have a crate of thousands of mass produced bolts delivered the next day for testing.

The reference to changing supplier in the "long term" could mean anything.  It could mean the redesign of the strut could happen in two years after many more of the same strut has flown.  It all sounds like "brush it under carpet speak" for "there is absolutely nothing wrong with the design of the strut or bolt, we just need to test the bolts to pick up any material defects".

26:38 Elon Musk: We have a pretty good sense for, ... . There's one strut sort of holding it in the vertical direction. The strut is oversized to be able to handle far in excess of the projected load. And in this case if failed far, far below, it failed five times below its nominal strength, which is pretty crazy. It appeared to fail at the bolt head, most likely.

24:22 Elon Musk: We have been able to replicate the failure by taking a huge sample, essentially thousands,....

16:43 Marcia Dunn with Associated Press: What is the size of the strut that failed? How long was it? How thick was it?.....

17:00 Elon Musk: I'm not sure the exact dimensions of the strut matter that much, I mean, it's about two feet long and about an inch thick at its thickest point......

22:20 Elon Musk: In this case we don't think that the strut is something we need to make in-house. We are going to be, long term, switching to a different strut design from a different supplier, most likely, but I don't think this is likely to be done in-house.

[vulture4]

I heard that at almost the exact time of the failure some in-house people who had been testing the struts reported they had had some failures. I do not personally see any reason to change suppliers as long as the supplier understands the problem and corrects it.
They could test 10% to failure, which should be about five times the rated load, and 100% to the expected flight load, at least initially. The test is not what ensures the part is made correctly, rather it is ensuring that the manufacturing process is properly controlled.

[cscott]

I'm not sure I believe that rumor.  According to what we heard here, they tested the struts after the failure and could *not* get any to fail.  Only after they had exhausted the other possibilities did they go back and test much much larger number of struts, and then they got *two* to fail, in total.  Your story doesn't line up with that sequence of events.

[Johnnyhinbos]

I'm surprised people are focusing on QA/QC for this strut and yet no discussion on fault tolerance. I mean, the F9 doesn't need nine Merlin 1D engines to get off the pad, rather there's that many for fault tolerance.

Personally I find it amazing that one strut could fail and take out the entire shooting match. I would think if a strut holding a tank failed at 1/5 rated load then in a properly designed fault tolerant system the surrounding strut work should be able to take on the added load.

Just thinking out loud...

[vulture4]

You may well be right, still it was not long after the failure that the rumor spread about the struts. Someone must have considered the possibility pretty quickly. I'm still not clear on how they ruled out a failure in the helium tank itself.

[kdhilliard]

... was not long after the failure that the rumor spread about the struts.

Was that just word-of-mouth, or do you recall the rumor spreading publicly?  While there was a lot of COPV speculation, I don't recall any public mention of struts until Musk's 2015-07-20 conference call.

~Kirk