SpaceX Falcon 9 - AMOS-6 - (Pad Failure) - DISCUSSION THREAD (2)

[te_atl]

So heres a question I got asked, and maybe someone knows or can come up with a good answer.   

So we know that SpaceX was just in the process of but not finished with the O2 loading phase of the second stage, and the tanks were not pressurized for firing.  This per their statements.  That being the case, why was there such a high pressure of helium being supplied at this time?   If they start with pressure inside the lines/COPVs equalized to the same pressure of the fluid outside, then the lines/COPV's rupturing from over pressurization would be impossible.  Pressurization to quarter or half flight specifications should easily be handled by the lines and COPV's early on.   As the O2 tank gets more and more filled, then the HE pressure could be gradually increased, on a pace to reach launch pressures in the COPVs at or about the same time as LOX fill goes to replenish mode.     A line or tank that is failing under pressure will tend to show increasing signs of distress as pressure is gradually increased, and will likely start a small leak that will grow as pressure increases until it fails.  This should be detectable.   Rapid pressurization of a line or tank and held at that pressure can result in sudden failure if it has a fault. 

Now its possible that this is very similar to what they actually do, but I haven't seen that anywhere, so I couldn't answer.  Any ideas?

[Jim]

So heres a question I got asked, and maybe someone knows or can come up with a good answer.   

So we know that SpaceX was just in the process of but not finished with the O2 loading phase of the second stage, and the tanks were not pressurized for firing.  This per their statements.  That being the case, why was there such a high pressure of helium being supplied at this time?   I

To load the COPV's.  As the temperature drops in the COPV's, more He can be loaded.

[CameronD]

This isn't the first time they've loaded these rockets, so I'd expect them to have their procedures down tight.  Besides if anything went wrong procedurally, I'd think they'd know right away - not still be looking for the problem.

No, something failed here.. but what that was, it seems they don't yet know.
 

[Fred Bonyea]

Audio - With reference to the initial screech moments before the first fireball and understanding this was from a distant mic, it does sound like - (and this is NOT something you want to experience), a pressure burst disk popping into a tubular vent stack.

[Rocket Science]

Audio - With reference to the initial screech moments before the first fireball and understanding this was from a distant mic, it does sound like - (and this is NOT something you want to experience), a pressure burst disk popping into a tubular vent stack.

Welcome to the forum! I have listened to that audio several times and have been down to the Cape often over the past 40 plus years and I'm beginning to think that sound may have been possibly a bird and sounded familiar. We have a few locals on here and perhaps some ornithologists can confirm it or not... It could have been mentioned as a bird in thread 1 as I went "cross-eyed" reading it...

[vandersons]

The sounds have been done to death in the previous speculation thread (and possibly also in this one) and it was established that the noises came from the vicinity of the camera.

Reasoning being that the high pitch of the sounds before the explosions excludes them from coming from the rocket without those same sounds being deafeningly obvious to everyone at the Cape at that time. High frequency sounds have a harder time traveling long distances. The camera was positioned approximately 4km from the pad.

[rsdavis9]

The sounds have been done to death in the previous speculation thread (and possibly also in this one) and it was established that the noises came from the vicinity of the camera.

Reasoning being that the high pitch of the sounds before the explosions excludes them from coming from the rocket without those same sounds being deafeningly obvious to everyone at the Cape at that time. High frequency sounds have a harder time traveling long distances. The camera was positioned approximately 4km from the pad.

Except Elon commented on them. We don't know if that was from the uslaunch video or some video he has from the pad.

[Fred Bonyea]

Welcome to the forum!

Thank you, I have lirked for a long time - especially on this thread.

The sounds have been done to death in the previous speculation thread (and possibly also in this one) and it was established that the noises came from the vicinity of the camera.

Reasoning being that the high pitch of the sounds before the explosions excludes them from coming from the rocket without those same sounds being deafeningly obvious to everyone at the Cape at that time. High frequency sounds have a harder time traveling long distances. The camera was positioned approximately 4km from the pad.

Except Elon commented on them. We don't know if that was from the uslaunch video or some video he has from the pad.

If it is a bird, it is a good imitation of a very high pressure (>800 psi) release. I suppose a Cape bird could imitate a lot of pad sounds, but the timing on this one is consistent with a stress-fractured aluminum helium tank bleeding through a composite overwrap. Four kms is not a difficult distance for a high-pitched factory whistle - especially if downwind. 

It is interesting that they have narrowed the investigation down to the COPV area. One has to wonder if a supporting brace was really the root cause of failure last year. If the top of the COPV bottle decided to behave like a nozzle, mechanical support failure would have been a secondary, rather than primary failure; and one that could occur in any of the stages with the embedded tanks. 

In any case and despite the SpaceX optimism, it is unlikely that we will see a quick RTF - Customers bought-off on that the first time.

[Jim]

1.  If it is a bird, it is a good imitation of a very high pressure (>800 psi) release. I suppose a Cape bird could imitate a lot of pad sounds, but the timing on this one is consistent with a stress-fractured aluminum helium tank bleeding through a composite overwrap. Four kms is not a difficult distance for a high-pitched factory whistle - especially if downwind. 

2. It is interesting that they have narrowed the investigation down to the COPV area. One has to wonder if a supporting brace was really the root cause of failure last year.

3.  If the top of the COPV bottle decided to behave like a nozzle, . 

1.How does stress-fractured aluminum helium tank bleeding through a composite overwrap that is in another tank, much less immersed in LOX make such sound?

2.  Where does it say that they have "narrowed the investigation down to the COPV area"?

3.  They would have seen a different signature.

[robert_d]

Are we allowed ANY comment on improbable first causes in this thread? That is the word Mr. Musk used himself.
Maybe something like someone listing two and then everyone votes off one. Then a third would be listed and compared to the survivor. Single Surviving "crazy" theory allowed to stay here until evidence conclusively refutes it?

[Mader Levap]

Are we allowed ANY comment on improbable first causes in this thread?

Nope. There is wacky thread for that.

[FinalFrontier]

Went over the videos again. Seems like first sign of fireball is near the lower part of the second stage, roughly just below where the common bulkhead between LOX and RP1 tanks are.

This is speculative, but it seems to me while the rupture started in with the LOX tank itself rupturing the fire or ignition occurred around this bulkhead after the rupture, which would make sense considering RP1 and LOX would have mixed.

Musk mentioned a "soft bang" heard in audio prior to the rupture, a few seconds before. Estimates range from 8 to 3 seconds before. An accurate estimate would be around 5 seconds before but exact time can be found from video analysis.

A guess would be that the initial "soft bang" was the common bulkhead being punctured or otherwise partially failing as a result of over-pressure in the LOX tank, which caused the fuel and oxidizer to mix, ignition follows a few seconds later. Ignition source could be anything really, the point being you had an ignitable mixture at this point. 


Between this and SpaceX's statements so far, it is seeming less likely to me that a COPV burst explosively, if one did fail it had to have still been a rapid failure, but not one that totally burst the vessel or the entire second stage would have blown apart perhaps with a more limited fire as the pressure would have dispersed the elements of the stage quicker. This suggests two things, one is that there was definitely a helium system failure and that is the root cause, and two is that it's more likely the helium plumbing itself failed. While a strut was the cause, helium plumbing rupture was the actual event which terminated CRS7. It is therefore possible a strut failed again, but I think that is highly unlikely or SpaceX would have already detected it.

So what this tends to suggest is that some portion of the helium plumbing failed energetically enough to puncture/fail the common bulkhead 5~ seconds before explosion allowing fuel and oxidizer to mix, but perhaps not energetically enough to burst the entire stage right away.  This would explain why they said "they have eliminated all obvious possibilities" because this is not a part of the helium system you would think would fail like this, especially under conditions in which we were not at flight pressure, or more likely were at roughly 1/2 of flight pressure.


Following the logic here a bit, to me this re-enforces the idea of a manufacturing; procedure; or testing+checkout mistake being the issue. I cannot see very many other reasons why the helium system/plumbing would fail at lower than to 1/2 of flight pressure in such a way.

FWIW

[chrisking0997]

so SpaceX would not know that the bulkhead failed?  that 5 seconds is well before the 93 milliseconds in question...not sure how that theory jives with what they have told us

[Rocket Science]

Welcome to the forum!

Thank you, I have lirked for a long time - especially on this thread.

The sounds have been done to death in the previous speculation thread (and possibly also in this one) and it was established that the noises came from the vicinity of the camera.

Reasoning being that the high pitch of the sounds before the explosions excludes them from coming from the rocket without those same sounds being deafeningly obvious to everyone at the Cape at that time. High frequency sounds have a harder time traveling long distances. The camera was positioned approximately 4km from the pad.

Except Elon commented on them. We don't know if that was from the uslaunch video or some video he has from the pad.

If it is a bird, it is a good imitation of a very high pressure (>800 psi) release. I suppose a Cape bird could imitate a lot of pad sounds, but the timing on this one is consistent with a stress-fractured aluminum helium tank bleeding through a composite overwrap. Four kms is not a difficult distance for a high-pitched factory whistle - especially if downwind. 

It is interesting that they have narrowed the investigation down to the COPV area. One has to wonder if a supporting brace was really the root cause of failure last year. If the top of the COPV bottle decided to behave like a nozzle, mechanical support failure would have been a secondary, rather than primary failure; and one that could occur in any of the stages with the embedded tanks. 

In any case and despite the SpaceX optimism, it is unlikely that we will see a quick RTF - Customers bought-off on that the first time.

Listen at 39 secs. for example...

[Moskit]

so SpaceX would not know that the bulkhead failed?  that 5 seconds is well before the 93 milliseconds in question...not sure how that theory jives with what they have told us

In such case they would likely know it failed.
They might not know however what caused this failure or rather what caused the fire (source of heat/ignition), if we are to take literally previous SpaceX announcements.

[te_atl]

To load the COPV's.  As the temperature drops in the COPV's, more He can be loaded.

Correct me if I'm wrong on what you are saying.
1) The colder the HE is in COPV, the more HE can be placed in it before achieving flight time pressure.   
2) HE heats as it expands into the COPV from the feed lines, raising the temperature of the HE in the COPV.
3) Prefilling the COPV's to flight temperature at the beginning heats the HE and COPV as high as the HE expansion will take it at first.
4) The COPV is immersed in subcooled LOX as the LOX loads, which cools the COPV and its HE down to the LOX temperature. 

But that being said, immersion in the LOX will trend towards equalizing the temperature of the COPV and the HE it contains regardless to the LOX temperature of whether its at flight pressure, half flight pressure, or less.   

However, HE can be chilled far below LOX temperatures.  Why isn't it loaded at a gradual increase of pressure at lower temperatures?   Obviously you don't want your COPV's or feed lines dropping below around 55K, or you'll end up with slush in the LOX tank.  But loading it enough colder that expansion into the COPV doesn't raise the HE temperature above the target LOX temperature.   The LOX then wouldn't act to chill the COPV and HE, but rather to maintain it at the desired temperature.

Again, not knowing exactly how they load it, I don't know that they don't already do this.

[rsdavis9]

Again, not knowing exactly how they load it, I don't know that they don't already do this.

I suggested some kind of thermal shock of hot copv and cold lox.
I was told spacex would have thought of this and this would have been a obvious failure scenario.

[Jim]

1) The colder the HE is in COPV, the more HE can be placed in it before achieving flight time pressure.   

No, I am not describing how they load it.  Just that at flight pressure, more He mass can be put in the COPV at LOX temp vs room temp.  I don't know the exact loading script as it pertains to time, pressure and temp

[iamlucky13]

Again, not knowing exactly how they load it, I don't know that they don't already do this.

I suggested some kind of thermal shock of hot copv and cold lox.
I was told spacex would have thought of this and this would have been a obvious failure scenario.

Don't take it personally. The rapid temperature change and issues with some composites forming cracks as a result is a known issue. Digging through NASA's technical reports server a couple weeks ago, for example, I found papers discussing the matter that were 20+ years old. It's not obvious to a lay person, but it is well-discussed within the spaceflight community.

Of course, obvious failure scenarios are not out of the question. Just not generally the leading candidates.

[Fred Bonyea]

1.How does stress-fractured aluminum helium tank bleeding through a composite overwrap that is in another tank, much less immersed in LOX make such sound?

I described the sound as being similar to the rapid release of very high pressure through a fluted burst disk - I wouldn't expect the sound of such an event to survive for 4 km, so I am looking for something that might.

I have heard composite bottles burst many times, but all but one of the bursts (including Mr. Musk's bottle) were filled with water, so there was little expanding gas.

But in one test, the composite bottle was not properly filled with water. The quick depressurization from ~ 900 psi was deafening two bays away, and literally raised the roof. I don't know how much fidelity there is in the audio, but a diminishing screech only a few ms long is exactly what I would expect to see/hear in a very rapid decompression due to a filament wound bottle failure, regardless of the medium.

2.  Where does it say that they have "narrowed the investigation down to the COPV area"?

http://www.spacex.com/news/2016/09/01/anomaly-updates

"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. All plausible causes are being tracked in an extensive fault tree and carefully investigated. Through the fault tree and data review process, we have exonerated any connection with last year’s CRS-7 mishap."

"Suggests" doesn't exactly mean 'narrowed'; but it does suggest this is where they are 'focusing'; which is a synonym for "narrows".

3.  They would have seen a different signature.

Not being privy to any of the SpaceX musings, I can only paint a skeptics loop around the first investigation. It was in SpaceX's best financial interests to return to flight asap; and nothing could have been much more simple than throwing in an extra brace or two.

Others have pointed out that the temperature expansion coefficients for aluminum are very different than those of composites, and we know rocket science is largely the mitigation of material strain and stress. Sometimes it takes failure in a component after many successes to realize how thin these margins are.  (We found out with the Space Shuttle that it can be a fatal failure to ask a chief engineer to put on his management hat when assessing a possible failure mode.)