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

[PeteW]

What you just described is known as "Helium Hammer" and was observed, with near catastrophic effects in early spaceflight projects dealing with cooling systems using liquid helium.
 

The helium used in F9 is not a liquid though. Would you get this same effect through simply heating then re-cooling gaseous helium?

True, but any nitrogen or oxygen impurities in the helium would be transitioning between supercritical fluid and liquid at some point between the GSE tanks and the COPV. That could be an additional complication, especially if the inlet to the COPV is at the bottom were liquid impurities would collect.

Anyway, as others have pointed out the thermoacoustic effect can still occur in just a gas phase.

[Space Ghost 1962]

These effects mentioned above are much closer IMHO to this anomaly than past discussion posts.

[rsdavis9]

These effects mentioned above are much closer IMHO to this anomaly than past discussion posts.

2nd it.
This does seem to fit what we know and we can see of the anomaly.

[wannamoonbase]

These effects mentioned above are much closer IMHO to this anomaly than past discussion posts.

2nd it.
This does seem to fit what we know and we can see of the anomaly.

3rd it.

[LouScheffer]

These effects mentioned above are much closer IMHO to this anomaly than past discussion posts.

Also fits well with Gwynne's comment:

We’re running a lot of tests at our test facility in Texas and we’re learning an awful lot.

IMO, this comment implies some tricky interaction.  Although tests might show that COPVs, or sub-cooled LOX, have unidentified properties, it's hard to see how this could be "an awful lot".   Interactions, however, could be very tricky and easily deserve the "awful lot" comment.

Furthermore, problems like this are not unheard of.  Ed Kyle, in L2, reported that the shuttle Centaur screamed when being pressurized (italics his), which presumably must be some sort of resonance.

So overall, I agree this line of investigation looks promising.

[iamlucky13]

One example of acoustic resonance coupling to mechanical resonance leading to high-cycle failure was in the Flow Control Valves of the Shuttle Orbiter where the poppet in the valve fractured (chunk of metal broke off the metal sealing face) resulting in its inability to seal fully closed.  Given that this valve controlled the flow of gaseous hydgrogen from the SSME back to the LH2 tank to provide ullage pressure, this was a big deal at the time.  There are lots of NSF articles on this.

I remember this well. I ultimately gave up trying to rebook my reservations repeatedly to follow the launch date across multiple changes and just had a nice quiet vacation to KSC.

But I missed the one and only chance I would have had to see a shuttle launch as a result.

If I remember right, there is a relief to prevent excess ullage pressure, but they were concerned about valve fragments damaging the gaseous H2 line and presenting a fire hazard to the vehicle. They installed doubler plates on elbows in the plumbing to mitigate against that, and did impact tests with high speed simulated fragments to ensure they wouldn't penetrate the line.

https://www.nasaspaceflight.com/2009/02/sts-119-targeting-march-12-pending-prcb-and-frr-meetings/

[Long EZ]

The obvious solution to fix the problem of a Helium COPV rupturing inside of the LOX tank is to remove it from the LOX tank. This would increase the volume of LOX and reduce the loads on the tank and its struts. But you say that these Helium tanks need to be cooled to hold enough Helium. So the question becomes how do you cool them. One way would be to wrap a tube around it that has liquid Nitrogen flow through it. Another would be to blow cold Nitrogen gas over the COPVs. You would need some form of insulation around the COPVs so that could be a composite shell with a layer of foam. If you make the shell strong enough if could dampen the effects of a rupturing COPV. With proper check valves you could design the system to have enough He with one tank leaking or ruptured. The other question would be where you would put the Helium COPVs. I would suggest the engine compartment of the second stage. It would seem that there is a lot of space there for the second stage. For the first stage I would doubt that there would be enough space in the engine compartment, so maybe the solution for that stage would be to use an autogenous  pressurization system for the LOX tank and He for the fuel tank. Thinking about the Apollo 13 problem, the solution was to have nothing combustible in LOX tanks.

[ejb749]

The obvious solution to fix the problem of a Helium COPV rupturing inside of the LOX tank is to remove it from the LOX tank. This would increase the volume of LOX and reduce the loads on the tank and its struts. But you say that these Helium tanks need to be cooled to hold enough Helium. So the question becomes how do you cool them. One way would be to wrap a tube around it that has liquid Nitrogen flow through it. Another would be to blow cold Nitrogen gas over the COPVs. You would need some form of insulation around the COPVs so that could be a composite shell with a layer of foam. If you make the shell strong enough if could dampen the effects of a rupturing COPV. With proper check valves you could design the system to have enough He with one tank leaking or ruptured. The other question would be where you would put the Helium COPVs. I would suggest the engine compartment of the second stage. It would seem that there is a lot of space there for the second stage. For the first stage I would doubt that there would be enough space in the engine compartment, so maybe the solution for that stage would be to use an autogenous  pressurization system for the LOX tank and He for the fuel tank. Thinking about the Apollo 13 problem, the solution was to have nothing combustible in LOX tanks.

And for the first stage?

[JasonAW3]

The obvious solution to fix the problem of a Helium COPV rupturing inside of the LOX tank is to remove it from the LOX tank. This would increase the volume of LOX and reduce the loads on the tank and its struts. But you say that these Helium tanks need to be cooled to hold enough Helium. So the question becomes how do you cool them. One way would be to wrap a tube around it that has liquid Nitrogen flow through it. Another would be to blow cold Nitrogen gas over the COPVs. You would need some form of insulation around the COPVs so that could be a composite shell with a layer of foam. If you make the shell strong enough if could dampen the effects of a rupturing COPV. With proper check valves you could design the system to have enough He with one tank leaking or ruptured. The other question would be where you would put the Helium COPVs. I would suggest the engine compartment of the second stage. It would seem that there is a lot of space there for the second stage. For the first stage I would doubt that there would be enough space in the engine compartment, so maybe the solution for that stage would be to use an autogenous  pressurization system for the LOX tank and He for the fuel tank. Thinking about the Apollo 13 problem, the solution was to have nothing combustible in LOX tanks.

The solution you're suggesting would add a significant amount of mass to the stage.

[spacenut]

Why not move them to the chilled kerosene tank like the first stage?

[Herb Schaltegger]

Why not move them to the chilled kerosene tank like the first stage?

Because the chilled kerosene isn't nearly cold enough to allow the same cold helium storage as putting them in the LOX tank. They'd need bigger, more massive helium tanks, and consequently displace more volume (fuel), resulting in that much lower overall performance.

Look, the key here isn't to somehow make the stage capable of surviving a pressurization system failure, or somehow avoid putting the COPV's in the LOX tank ... it's to prevent whatever error (manufacturing, operational) that resulted in the over-pressure.

[MP99]

The obvious solution to fix the problem of a Helium COPV rupturing inside of the LOX tank is to remove it from the LOX tank. This would increase the volume of LOX and reduce the loads on the tank and its struts. But you say that these Helium tanks need to be cooled to hold enough Helium. So the question becomes how do you cool them.
...
The other question would be where you would put the Helium COPVs. I would suggest the engine compartment of the second stage. It would seem that there is a lot of space there for the second stage. For the first stage I would doubt that there would be enough space in the engine compartment, so maybe the solution for that stage would be to use an autogenous  pressurization system for the LOX tank and He for the fuel tank.

The gas from the COPVs is heated in a heat exchanger (on the Merlin GGs, I believe) which reduces the mass required for pressurization.

If you had to re-engineer that (Herb's input excepted), ISTM that liquid helium bottles in the engine compartment could be a feasible alternative. This would go through uprated Merlin heat exchangers to feed the same gasses to the tanks.

These would be much smaller than the gas bottles (liquid is much denser), and not have the extreme pressures, but would need more mass for insulation.

Gas pressure would be provided by GSE while on the pad, including bringing the tanks up to launch pressure.

This would be a substantial change, and they'd need to also survive reentry OK.

Cheers, Martin

[stempus]

These effects mentioned above are much closer IMHO to this anomaly than past discussion posts.

2nd it.
This does seem to fit what we know and we can see of the anomaly.

3rd it.

Next one. Why we had to wait so many pages for this to came up...?

[Stranger]

The gas from the COPVs is heated in a heat exchanger (on the Merlin GGs, I believe) which reduces the mass required for pressurization.

If you had to re-engineer that (Herb's input excepted), ISTM that liquid helium bottles in the engine compartment could be a feasible alternative. This would go through uprated Merlin heat exchangers to feed the same gasses to the tanks.

These would be much smaller than the gas bottles (liquid is much denser), and not have the extreme pressures, but would need more mass for insulation.

Gas pressure would be provided by GSE while on the pad, including bringing the tanks up to launch pressure.

This would be a substantial change, and they'd need to also survive reentry OK.

Cheers, Martin

But then it starts to lag Falcon-9 for at least a year.
And Musk likes to apply the principle of the constant small changes to the rocket. It apparently stopped working

[MP99]

Agreed - it would push back RTF quite a long way.

Cheers, Martin

[Stranger]

Agreed - it would push back RTF quite a long way.

Cheers, Martin

yes

[guckyfan]

If, a big if, they come to the conclusion that there will always be some risk with the present setup, they could switch from aluminium liner to stainless steel liner. No need to rework the whole Falcon 9.

[Stranger]

If, a big if, they come to the conclusion that there will always be some risk with the present setup, they could switch from aluminium liner to stainless steel liner. No need to rework the whole Falcon 9.

then it is necessary to change liner all have made the Falcon 9 and carry out additional testing. This is also a long time. 2-3 months before RTF, imho. Not November and not 2016

[Jim]

If, a big if, they come to the conclusion that there will always be some risk with the present setup, they could switch from aluminium liner to stainless steel liner. No need to rework the whole Falcon 9.

then it is necessary to change liner all have made the Falcon 9 and carry out additional testing. This is also a long time. 2-3 months before RTF, imho. Not November and not 2016

Likely don't have a pad to launch from

[eweilow]

If, a big if, they come to the conclusion that there will always be some risk with the present setup, they could switch from aluminium liner to stainless steel liner. No need to rework the whole Falcon 9.

then it is necessary to change liner all have made the Falcon 9 and carry out additional testing. This is also a long time. 2-3 months before RTF, imho. Not November and not 2016

Likely don't have a pad to launch from

There's still Vandenberg for Iridium missions