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

[Robotbeat]

There's a LOT of weight bringing a titanium tank all the way to orbit. Not LEO, but GTO. And it may itself have problems.

What about shrink-wrapping the COPV in a fluoropolymyer thermoplastic? Totally seal off the fibers from oxygen intrusion. Shouldn't add much mass.

I imagine it would significantly increase the handling difficulties though.  If you design a solution where no microcracks or scratches can be tolerated, you have to be very, very careful in how they are handled.  I think if they can design a solution that allows for less fragile handling they will.  Not that your proposed solution might not be feasible.

Why not? If there's a scratch and oxygen gets through, then you're just back to where you came from (which almost always works, remember!). The idea is to have a defense-in-depth approach without having to dramatically change a major system (that will happen on later Blocks).

[sdsds]

I'm still struggling a bit to comprehend one bit of the SpaceX statement. In the short term, this entails changing the COPV configuration to allow warmer temperature helium to be loaded.

Is it possible that warmer helium requires physically larger bottles?

[dglow]

I'm still struggling a bit to comprehend one bit of the SpaceX statement. In the short term, this entails changing the COPV configuration to allow warmer temperature helium to be loaded.

Is it possible that warmer helium requires physically larger bottles?

Chris says SpaceX have installed an additional COPV.

[HMXHMX]

I'm still struggling a bit to comprehend one bit of the SpaceX statement. In the short term, this entails changing the COPV configuration to allow warmer temperature helium to be loaded.

Is it possible that warmer helium requires physically larger bottles?

Chris says SpaceX have installed an additional COPV.

Someone will no doubt refresh my memory if I am in error, but I seem to recall the stage used to have four bottles in an earlier configuration, reducing to three about the time SpaceX went to densified LOX.

[Steven Pietrobon]

Agreed but do they have the room in the S2 to mount the bottles externally? Would appear to be a better and safer solution but only if they don't have to stretch the stack any more.

For the second stage, the bottle is quite small and would fit under the RP-1 tank. For the first stage, the bottle is quite a bit larger. It might be able to fit inside the pusher on the first stage LOX tank, inside the second stage engine bell. I had worked out the volumes, but I can't find the post where I posted the information.

Can't liquid hellium bottle be installed inside of LOX tank? It would simplify thermal control..

Yes, that could could be a possibility. I believe the bottle is like a thermos flask with two walls and a vacuum between the walls.

Why would it be lighter?  Granted, moving the helium bottle out of the lox tank would make the lox tank a little smaller and lighter, but then the unpressurized structure of the stage would need to be larger.  Unpressurized volume will tend to be heavier than a similar pressurized volume, because it is not supported by pressure in flight.

We're talking about using liquid Helium, not supercritical Helium under very high pressure. Thus, the Helium bottles would be lighter since they would be much smaller and under low pressure (about 0.2 to 0.3 MPa versus 38 MPa). I believe the SpaceX vehicles are currently performance restrained by its length. That is, a longer vehicle would give better performance due to a greater propellant load, but the increased bending moments would make it too dangerous too fly. Thus, keep the same length, but increase the propellant load. The bottles should be placed where they don't extend the length of the vehicle, as described above.

Carbon fiber is my bread and butter, but in this case (if I could not re-package it outside of the oxygen tank) I'd just go all titanium or aluminum for the helium tank.  Take the weight hit and move on. I realize that is not trivial in the second stage of a rocket, but I'd rather take the performance hit than continue with the risk.  If that isn't tolerable, design around it in the next F9 variant or successor (re-package or go autogenous).

Titanium is dangerous to use in LOX. Under impact or fracture, Titanium can ignite. Aluminium Helium bottles were used in the Saturn V, but they would be quite heavy.

[sdsds]

So warmer helium consuming more volume within the oxygen tank will displace some propellant, preventing the new "short term" F9 configuration from providing as much total impulse as the AMOS-6 configuration?

[meekGee]

Remember that SpaceX is currently designing an entire rocket and spaceship  out of carbon composite, which will hold cold LOX and hot GOX in tanks for extended periods of time, and under very high loads (think of the pressure head at the bottom of the BFR tanks during launch)

They must have some confidence in their ability to do so, and if they can't solve the He issue without switching away from composites, they have a much bigger problem.

I think they figured out the compatibility issues, and this last failure was an edge case that they now know about.

The operational record of the bottles on F9 will complement ground based testing in that respect.

[DaveJes1979]

Titanium has a very excellent *specific strength* (aluminum is pretty good also). using an all-metal tank to replace F9's carbon fiber + metallic liner pressure vessel the weight difference would not be high. One would probably only really feel pain if you were replacing a liner-less composite tank design with an all-metal design.

Metallic tanks are zero risk.

[HMXHMX]

Agreed but do they have the room in the S2 to mount the bottles externally? Would appear to be a better and safer solution but only if they don't have to stretch the stack any more.

For the second stage, the bottle is quite small and would fit under the RP-1 tank. For the first stage, the bottle is quite a bit larger. It might be able to fit inside the pusher on the first stage LOX tank, inside the second stage engine bell. I had worked out the volumes, but I can't find the post where I posted the information.

Can't liquid hellium bottle be installed inside of LOX tank? It would simplify thermal control..

Yes, that could could be a possibility. I believe the bottle is like a thermos flask with two walls and a vacuum between the walls.

Why would it be lighter?  Granted, moving the helium bottle out of the lox tank would make the lox tank a little smaller and lighter, but then the unpressurized structure of the stage would need to be larger.  Unpressurized volume will tend to be heavier than a similar pressurized volume, because it is not supported by pressure in flight.

We're talking about using liquid Helium, not supercritical Helium under very high pressure. Thus, the Helium bottles would be lighter since they would be much smaller and under low pressure (about 0.2 to 0.3 MPa versus 38 MPa). I believe the SpaceX vehicles are currently performance restrained by its length. That is, a longer vehicle would give better performance due to a greater propellant load, but the increased bending moments would make it too dangerous too fly. Thus, keep the same length, but increase the propellant load. The bottles should be placed where they don't extend the length of the vehicle, as described above.

Carbon fiber is my bread and butter, but in this case (if I could not re-package it outside of the oxygen tank) I'd just go all titanium or aluminum for the helium tank.  Take the weight hit and move on. I realize that is not trivial in the second stage of a rocket, but I'd rather take the performance hit than continue with the risk.  If that isn't tolerable, design around it in the next F9 variant or successor (re-package or go autogenous).

Titanium is dangerous to use in LOX. Under impact or fracture, Titanium can ignite. Aluminium Helium bottles were used in the Saturn V, but they would be quite heavy.

The bottles contain just shy of 20K cu inches (and no, I'm not going to do this in metric) so at  SF = 1.5 (too low in my opinion but meets the NASA spec) and a PV/W of 1E6, the weight of the current bottle is about 160 lbm.  The PV/W of a high strength aluminum is about 250K, so the difference (per bottle) is about 470 lbm, or 1880 lbm for the stage.

Not a problem for LEO as the current F9 configuration is way overpowered, but 0.85MT loss is painful for GTO missions.  Perhaps not as painful as losing another multi-hundred million dollar spacecraft, though.

[Comga]

The bottles contain just shy of 20K cu inches (and no, I'm not going to do this in metric) so at  SF = 1.5 (too low in my opinion but meets the NASA spec) and a PV/W of 1E6, the weight of the current bottle is about 160 lbm.  The PV/W of a high strength aluminum is about 250K, so the difference (per bottle) is about 470 lbm, or 1880 lbm for the stage.

Not a problem for LEO as the current F9 configuration is way overpowered, but 0.85MT loss is painful for GTO missions.  Perhaps not as painful as losing another multi-hundred million dollar spacecraft, though.

0.33 m^3. 
It took less time to type "20,000 cubic inches in cubic meters" into Google than it took to type "(and no, I'm not going to do this in metric)"

[Odysseus]

....
We're talking about using liquid Helium, not supercritical Helium under very high pressure. Thus, the Helium bottles would be lighter since they would be much smaller and under low pressure (about 0.2 to 0.3 MPa versus 38 MPa)
.......

Are you sure? I thought one needs the high pressure He (in supercritical state) to pressurize the LOX and to spin up the turbo pumps. "Real" liquid He could not provide that pressure, but only the related vapor pressure.
Cheers

[eriblo]

....
We're talking about using liquid Helium, not supercritical Helium under very high pressure. Thus, the Helium bottles would be lighter since they would be much smaller and under low pressure (about 0.2 to 0.3 MPa versus 38 MPa)
.......

Are you sure? I thought one needs the high pressure He (in supercritical state) to pressurize the LOX and to spin up the turbo pumps. "Real" liquid He could not provide that pressure, but only the related vapor pressure.
Cheers

It's run through heat exchangers in the engine to increase the volume in both cases, with liquid it just takes a bit more heat (and you need a small pump to bring it up to pressure first).

[Odysseus]

I thought we had this information earlier, see. e.g. reply #1146. This is the reason why I was surprised.
Liquid He would mean having it below the critical point. Where should this LHe come from?
Cheers

[envy887]

Remember that SpaceX is currently designing an entire rocket and spaceship  out of carbon composite, which will hold cold LOX and hot GOX in tanks for extended periods of time, and under very high loads (think of the pressure head at the bottom of the BFR tanks during launch)

They must have some confidence in their ability to do so, and if they can't solve the He issue without switching away from composites, they have a much bigger problem.

I think they figured out the compatibility issues, and this last failure was an edge case that they now know about.

The operational record of the bottles on F9 will complement ground based testing in that respect.

The head in the main tanks will only be ~30m on the pad and depleting equally as fast as the g-forces rise. For the density of LOX that's a pressure about 2 orders of magnitude lower than the 40,000 kPa in the He bottles on Falcon 9. Also, there will be nothing colder than LOX around, so SOX formation is quite unlikely. Eliminating the He will solve a lot of potential issues.

[meekGee]

Remember that SpaceX is currently designing an entire rocket and spaceship  out of carbon composite, which will hold cold LOX and hot GOX in tanks for extended periods of time, and under very high loads (think of the pressure head at the bottom of the BFR tanks during launch)

They must have some confidence in their ability to do so, and if they can't solve the He issue without switching away from composites, they have a much bigger problem.

I think they figured out the compatibility issues, and this last failure was an edge case that they now know about.

The operational record of the bottles on F9 will complement ground based testing in that respect.

The head in the main tanks will only be ~30m on the pad and depleting equally as fast as the g-forces rise. For the density of LOX that's a pressure about 2 orders of magnitude lower than the 40,000 kPa in the He bottles on Falcon 9. Also, there will be nothing colder than LOX around, so SOX formation is quite unlikely. Eliminating the He will solve a lot of potential issues.

Oh, lower than COPV type pressures for sure, but the pressure at the bottom spikes with the dynamic loads, so higher than the static g load.

The point is, comments above were about the general idea of putting composites and LOX in contact, and what I'm saying is that SpaceX is intending to do so for an entire two-way mission profile to Mars.

So they need to solve the issue related to this incident, but are not backing off from the very idea if composites in LOX.

Using liquid He and pre-heating? Maybe. That's very cold though.  How do you keep it from boiling in the pipes?  Maybe you don't, and just keep bleeding out vapor (GHe), dewer style.

[Odysseus]

Buckling - just a thought:
Is it a credible option that the buckling of the liner is caused by local pressure on the liner surface? Caused by a higher pressure outside around the COPV, pressing LOX through porosities of the composite material?
The loading of He occurs after LOX loading.
Cheers

[meekGee]

I'll add this though, to the above.  Not knowing that buckling is happening is not good.

I wish we knew more about what kind of buckling.  If something is operating very close to its structural limits and there's bucking on top of that, it will fail structurally, probably every time.

But this hasn't happened, not even in this case.

So I hope there's some more information made available, but sadly not expecting any..

[Johnnyhinbos]

In COPVs the pressure containment mechanism is the composite overwrap. The AL liner is just a container and isn't very thick compared to an all AL pressure vessel. Hence the weight savings. Therefore the AL liner is far more prone to mechanical distortions than an all AL container. I suppose it's even a possibility that there are acceptable distortions ('buckling', if you will) that are permissible from a pressure vessel perspective within some limits, just as there are permissible amounts of bend, divits, line corrosion, etc in AL and steel cylinders - I am a certified PSI visual cylinder inspector, though I don't do SCBA tanks.

So perhaps these buckles were always know about but never were considered a threat until now...

[HMXHMX]

The bottles contain just shy of 20K cu inches (and no, I'm not going to do this in metric) so at  SF = 1.5 (too low in my opinion but meets the NASA spec) and a PV/W of 1E6, the weight of the current bottle is about 160 lbm.  The PV/W of a high strength aluminum is about 250K, so the difference (per bottle) is about 470 lbm, or 1880 lbm for the stage.

Not a problem for LEO as the current F9 configuration is way overpowered, but 0.85MT loss is painful for GTO missions.  Perhaps not as painful as losing another multi-hundred million dollar spacecraft, though.

0.33 m^3. 
It took less time to type "20,000 cubic inches in cubic meters" into Google than it took to type "(and no, I'm not going to do this in metric)"

But I know PV/W in English units by heart; not so much for metric.  ;(

[MarekCyzio]

I found this:
https://shellbuckling.com/presentations/otherTopics/pages/page_180.html

Buckling seems to happen near the neck of COPV.

Edit - more here:
http://tiny.cc/7vj8hy

Edit 2 - shortened link, sorry!