The Flimsiest Cryogenic Fuel Tank: LH2 in a Bag

[TheRadicalModerate]

Fair warning:  This may be a stupid idea.

A year or two back, in one of the periodic "hydrogen vs. methane" food-fights that we have, I suggested the idea of a "flimsy" LH2 tank for translunar or interplanetary missions.  The basic idea is that the vast majority of propellant is consumed within a few minutes of starting a departure burn.  If you can find a way of making a really, really thin-walled tank, with no insulation and a very low ullage pressure (and therefore very low hoop stress), then the biggest problem with LH2, which is its lousy structural mass fraction, can be significantly mitigated.

At the time, I was thinking of very thin-walled metal tanks that could be pressurized just enough to be launched empty from Earth (very low mass, very high volume), and then filled at a depot.  But it's recently occurred to me that I wasn't thinking big (or dumb) enough.

Could we put cryogenic propellant in a bag?

Two possibilities:

1) A bag that would expand as it's filled with LH2, then remain flexible throughout the cryo fill and engine draw-down process.  This would be preferable, because no autogenous ullage gas would need to be added during draw-down.  The bag would simply collapse--without rupturing!--as it emptied.  But the bag would almost certainly have to be made of unobtanium.

2) A bag that would be flexible to fill with low-pressure ullage gas, after which it would be filled.  As it emptied, ullage gas would keep it inflated until all LH2 was gone.  This is still really challenging, but might be possible.

Let's fool around with a 500t-capacity spherical bag.  For LH2 (density=71kg/m³), that would require a volume of 7040m³, for a diameter of 23.8m.  For a rocket with an acceleration of 3m/s², that's a hydrostatic pressure of only 5.1kPa at the bottom.  However, the bag will also need some ullage pressure, or the rate of boiling will be difficult to manage, even for a fast "fill-'n'-go" conops.  Looks to me like LH2 subcooled down to about 13K will remain a saturated liquid down to about 0.25bar.

A great thing about such bags is that you could launch tens of them at a time and deploy them as needed.  If even a flimsy tank is rigid it takes up a whole launch in volume.  You might be able to load it partially with LH2, but the hydrostatic pressure at max acceleration will 10x-20x greater than it would be on interplanetary departure, requiring needless tank mass just to survive launch.  With the bags, you just fold 'em up, inflate/fill them in space, and discard them when they're empty.

There are some big caveats, of course:

1) I'm basically positing a bag made of unobtanium.  However, DuPont claims that Kapton has been used in applications all the way down to 4K, although I doubt any of those applications involve any significant stress.  However, I'd think that things would get a lot easier with a bag that was pre-inflated before LH2 loading began.  And it doesn't have to be perfect:

a) It can boil like crazy (see #2 below).
b) It can leak quite a bit.
c) It can even be supported by external stringers if necessary.

Mostly, it just needs not to rupture.

2) This simply doesn't work unless there's a very carefully engineered LH2 depot in orbit, with very fast fill, and the departing vehicle can go from being topped off to initiating its departure burn in a matter of minutes.  That doesn't require any new science, but it certainly requires some innovative ops.

3) This only works if all of the prop in the bag is immediately used.  If the vehicle needs any delta-v on arrival at its destination, that has to be stored in a more traditional tank.  However, because the total mass of the vehicle will be dramatically less by then, the tanks can be small.

4) With such low pressures, the LH2 will require a multi-stage turbopump to avoid cavitation.  That'll be a very low T/W engine.  But you're saving so much tank mass that even with the extra pump mass, you should be far ahead of the game.

5) Lots of boil-off and possibly even leakage is inevitable.  You size your bag to allow for that.  As a result, there's an implicit assumption built into this that LH2 at the depot is pretty cheap.

6) I've concentrated on LH2 here, because the most likely use for this kind of system would be for a nuke.  However, I'd think that exactly the same kind of bag could be used for methane and LOX.  The biggest problem is that, as density goes up, the hydrostatic head also rises, and the bag needs to be stronger.  It might easily be necessary to put LOX into a traditional tank.  Still, if the LH2 portion is super-light, you'd be way ahead.

I know almost nothing about materials, and this idea doesn't make sense if the materials are unlikely to exist in the not-too-distant future.  But the problem with LH2 has always been that it's leaky and fluffy, making the dry mass needed for storing it for flight difficult and heavy.  If you can find a way to ignore its leakiness and fluffiness by just getting rid of it really quickly, that would be a game-changer.

Silly cartoon attached.

[redneck]

A couple of things occurred to me.  Use two bags with the inner one mostly hydrogen tight with minimal strength    Outer one is stronger but less cryogenic capable with insulating/pressure gas between.   Have to cycle the Warmer insulating  gas to keep it from freezing out or warm the leaking hydrogen from the main for the burn time.  Then pump down the ullage for later use. 

Second thought was put the cargo on top and thrust on bottom of tank.  Weight of cargo under thrust compresses the propellant bladder. 

[edzieba]

A few thoughts:
- A bag is just a a balloon stage without taking advantage of pressure stabilisation.
- There are multiple origami designs that could be adapted to form collapsing tanks, including but not limited to the classic cylindrical bellows. That means as long as your tank wall can flex it can expand and contract without needing to stretch (i.e. you can use stainless foil rather than unobtanium)
- If you're going for subcooled LH2, tolerate boiloff, and want to eschew pressurisation, then the Gordian option is not a bag, but a cup. You need to fill it under acceleration and may need a thin foil over the top to provide mechanical pressurisation to prevent sublimation, but as long as the engine runs and the walls are shielded from the sun (or in the outer solar system) it will stay liquid from head pressure.
- A balloon tank takes advantage of pressure rather than eschewing it, but means the tank walls have still some minimum strength required due to fluid head. As well as stringers or thicker tank walls, that could be achieved by overwrapping with separable bands: as the burn progresses, shed the band to shed mass without compromising the inner envelope or needing to stage.

[laszlo]

Instead of a pump, pull the tank through rollers (like an old-fashioned washing machine wringer) to squeeze the contents towards the exit.

[Twark_Main]

I wonder if anything can be done with a rigidizable metal layer, similar to Echo 2.

https://en.wikipedia.org/wiki/Project_Echo

Echo 2's skin was rigidizable, unlike that of Echo 1. Therefore, the balloon was capable of maintaining its shape without a constant internal pressure; a long-term supply of inflation gas was not needed, and it could easily survive strikes from micrometeoroids. The balloon was constructed from a 9 μm (0.00035 in)-thick mylar film sandwiched between and bonded with two layers of 4.5 μm (0.00018 in)-thick aluminum foil.[10] It was inflated to a pressure that caused the metal layers of the laminate to plastically deform slightly, while the polymer was still in the elastic range. This resulted in a rigid and very smooth spherical shell.

I imagine thin Kapton would have a non-trivial H2 permeability, so a metal layer may be called for.

[lkm]

Inflatable fuel tanks are already a thing. In as much as people have thought of them before.

Apparently Von Braun thought of them for his Mars Project, but they are used in the Scorpion design.

https://www.researchgate.net/publication/339201364_Scorpion_a_Design_Study_for_a_General_Purpose_Space_Transportation_System

[TheRadicalModerate]

I wonder if anything can be done with a rigidizable metal layer, similar to Echo 2.

https://en.wikipedia.org/wiki/Project_Echo

Echo 2's skin was rigidizable, unlike that of Echo 1. Therefore, the balloon was capable of maintaining its shape without a constant internal pressure; a long-term supply of inflation gas was not needed, and it could easily survive strikes from micrometeoroids. The balloon was constructed from a 9 μm (0.00035 in)-thick mylar film sandwiched between and bonded with two layers of 4.5 μm (0.00018 in)-thick aluminum foil.[10] It was inflated to a pressure that caused the metal layers of the laminate to plastically deform slightly, while the polymer was still in the elastic range. This resulted in a rigid and very smooth spherical shell.

I imagine thin Kapton would have a non-trivial H2 permeability, so a metal layer may be called for.

A non-trivial H2 permeability would be fine, though, unless a substantial amount of H2 is lost in a few minutes.

[chopsticks]

I've been thinking about a similar idea for Starship refilling. A bag inside the tanker with pressurized gas collapsing the bag forcing the propellant into the receiving ship.

[TheRadicalModerate]

I've been thinking about a similar idea for Starship refilling. A bag inside the tanker with pressurized gas collapsing the bag forcing the propellant into the receiving ship.

That does have the advantage of not needing continuous ullage acceleration for a couple of hours, but it's a lot more complicated than using an electric pump with the two ullage spaces connected to equalize their pressures.  That can reduce the time needed for ullage acceleration to the point where prop consumption is very small.

[TheRadicalModerate]

- There are multiple origami designs that could be adapted to form collapsing tanks, including but not limited to the classic cylindrical bellows. That means as long as your tank wall can flex it can expand and contract without needing to stretch (i.e. you can use stainless foil rather than unobtanium)

That's a good idea.  I wonder what happens to the sharp angles in a bellows under even modest pressure, though.

I also wonder if you could get by with vapor-deposition on Kapton.  It might not be a super-cheap tank to manufacture, though, especially after you figured out the tooling to create the bellows folds without something bad happening.

- If you're going for subcooled LH2, tolerate boiloff, and want to eschew pressurisation, then the Gordian option is not a bag, but a cup. You need to fill it under acceleration and may need a thin foil over the top to provide mechanical pressurisation to prevent sublimation, but as long as the engine runs and the walls are shielded from the sun (or in the outer solar system) it will stay liquid from head pressure.

This is sort of the "zero ullage pressure" limiting case, which gets back to the whole question of how many stages you need on your turbopumps.  And, needless to say, it has some kinda gnarly contingencies if you lose ullage thrust for even a moment.
 

- A balloon tank takes advantage of pressure rather than eschewing it, but means the tank walls have still some minimum strength required due to fluid head. As well as stringers or thicker tank walls, that could be achieved by overwrapping with separable bands: as the burn progresses, shed the band to shed mass without compromising the inner envelope or needing to stage.

That's a lot of separation events, especially for a crewed interplanetary mission.

What's the limiting case for composite overwrapping?  If you inflate a Kapton (or other notyetobtanium miracle of chemical engineering) bag on the ground, then spin it to wind the sparsest possible amount of overwrap onto it, do you think it would collapse or fold in a reliable enough way to re-inflate on-orbit?

[Asteroza]

Ahem

https://forum.nasaspaceflight.com/index.php?topic=53390.msg2208675#msg2208675

[TheRadicalModerate]

Ahem

https://forum.nasaspaceflight.com/index.php?topic=53390.msg2208675#msg2208675

Oops. Missed that one.  Sorry.

Some comments/questions:

1) With the kind of conops I was thinking about, you wouldn't need multiple cycles of crumple and expansion.  One cycle would do it.  If you wanted the bag to crumple as prop was removed, you'd need one and a half.

2) I noticed that they were doing all their testing at LN2 temperatures.  Does something odd happen if you go down into the subcooled LH2 regime?

3) I'm having a failure of imagination of what the uncrumpled form of this looks like.  Is it a cylinder, or something else?

4) While this takes care of the crumple, it's not clear what happens in the presence of even modest hoop stress.  Again, this becomes a question of just how many stages you need on your turbopumps to avoid cavitation.

[Asteroza]

Ahem

https://forum.nasaspaceflight.com/index.php?topic=53390.msg2208675#msg2208675

Oops. Missed that one.  Sorry.

Some comments/questions:

1) With the kind of conops I was thinking about, you wouldn't need multiple cycles of crumple and expansion.  One cycle would do it.  If you wanted the bag to crumple as prop was removed, you'd need one and a half.

2) I noticed that they were doing all their testing at LN2 temperatures.  Does something odd happen if you go down into the subcooled LH2 regime?

3) I'm having a failure of imagination of what the uncrumpled form of this looks like.  Is it a cylinder, or something else?

4) While this takes care of the crumple, it's not clear what happens in the presence of even modest hoop stress.  Again, this becomes a question of just how many stages you need on your turbopumps to avoid cavitation.

Can't blame you, site search is always wonky.

1. But you need to test, so in the need you need multiple cycles
2. I'm assuming LH2 by definition is problematic due to the extreme temperature, plus the increased cost of performing such experiments. But the lead author of that paper says he's working on a LH2 version
https://hydrogen.wsu.edu/kjell-westra-2/
3. something like the following pictures suggests a cylinder
4. If a cylinder, then hoop stress might not be a problem per se, though the crease is a known weakpoint

[redneck]

Another thought is that slush hydrogen has even less vapor pressure than LH2 and might facilitate the bag concept. Beyond that, since this is leaving a dedicated facility in microgravity, could the H2 be actually frozen with just enough 'bag' to slow sublimation?

Could the 'bag' be contained or supported  inside a structure with uses either en route or at the destination? Dome with other uses at the bottom and girders with other uses supporting the sides?

 'Bag' as a low pressure hangar at destination? As in 5 psi CO2 contained at Mars such that workers have less restricting suits for semi-EVA work? Or on moon,
asteroid, free space etc...

Possibly aluminum bag that becomes solid fuel portion of hybrid engine when slowing down at destination?

[TheRadicalModerate]

Another thought is that slush hydrogen has even less vapor pressure than LH2 and might facilitate the bag concept. Beyond that, since this is leaving a dedicated facility in microgravity, could the H2 be actually frozen with just enough 'bag' to slow sublimation?

Could the 'bag' be contained or supported  inside a structure with uses either en route or at the destination? Dome with other uses at the bottom and girders with other uses supporting the sides?

 'Bag' as a low pressure hangar at destination? As in 5 psi CO2 contained at Mars such that workers have less restricting suits for semi-EVA work? Or on moon,
asteroid, free space etc...

Possibly aluminum bag that becomes solid fuel portion of hybrid engine when slowing down at destination?

If you support the bag inside a pressurizable structure, you might as well just put the prop in that structure and have done with it.  The whole idea is to make the dry mass of the container so low (because it's just a bag) that you get to take full advantage of the high Isp of hydrogen (either as burning hydrolox, or heated by a nuke) without incurring the high dry mass of the great big traditional tank.

The advantage of a collapsible bag is that you can send a bunch of the 'em, folded up, on a single launch.  You still have to make a bunch of LH2 in orbit somehow, but that's a different problem.  (My guess is that you launch water, electrolyze it using solar energy on the depot, and save the LOX for some other mission--unless of course it's a hydrolox mission, when you'll use most but not all of the LOX you electrolyze.)

[Asteroza]

Another thought is that slush hydrogen has even less vapor pressure than LH2 and might facilitate the bag concept. Beyond that, since this is leaving a dedicated facility in microgravity, could the H2 be actually frozen with just enough 'bag' to slow sublimation?

Could the 'bag' be contained or supported  inside a structure with uses either en route or at the destination? Dome with other uses at the bottom and girders with other uses supporting the sides?

 'Bag' as a low pressure hangar at destination? As in 5 psi CO2 contained at Mars such that workers have less restricting suits for semi-EVA work? Or on moon,
asteroid, free space etc...

Possibly aluminum bag that becomes solid fuel portion of hybrid engine when slowing down at destination?

If you support the bag inside a pressurizable structure, you might as well just put the prop in that structure and have done with it.  The whole idea is to make the dry mass of the container so low (because it's just a bag) that you get to take full advantage of the high Isp of hydrogen (either as burning hydrolox, or heated by a nuke) without incurring the high dry mass of the great big traditional tank.

The advantage of a collapsible bag is that you can send a bunch of the 'em, folded up, on a single launch.  You still have to make a bunch of LH2 in orbit somehow, but that's a different problem.  (My guess is that you launch water, electrolyze it using solar energy on the depot, and save the LOX for some other mission--unless of course it's a hydrolox mission, when you'll use most but not all of the LOX you electrolyze.)

If you are facing ground launch volume restrictions, sending a small water tank + electrolyzer + crycooler + tensegrity or inflatable frame + sunshield over frame +collapsible bags may work out, if you are building up a transport craft in orbit. Bag design may even favor puller style vehicles, aka Valkyrie, where you have a hammerhead propulsion block, and a primarily tensile trailing structure supporting cargo and hanging propellant bags.

Designing a hanging collapsible bag for propellant is an interesting design problem. Ullage pressure tends to inflate it, so how to handle the feed lines both when collapsed for transport and stretched when in use is tricky.

[redneck]

"""If you support the bag inside a pressurizable structure, you might as well just put the prop in that structure and have done with it.  The whole idea is to make the dry mass of the container so low (because it's just a bag) that you get to take full advantage of the high Isp of hydrogen (either as burning hydrolox, or heated by a nuke) without incurring the high dry mass of the great big traditional tank.

The advantage of a collapsible bag is that you can send a bunch of the 'em, folded up, on a single launch.  You still have to make a bunch of LH2 in orbit somehow, but that's a different problem.  (My guess is that you launch water, electrolyze it using solar energy on the depot, and save the LOX for some other mission--unless of course it's a hydrolox mission, when you'll use most but not all of the LOX you electrolyze.)"""


Not arguing for more dedicated structure. Suggesting that items that may have a use at the destination may be used to help structurally. Wasn't thinking about a pressurizable structure, just a bit of support from items that are eventually payload. 

[Asteroza]

"""If you support the bag inside a pressurizable structure, you might as well just put the prop in that structure and have done with it.  The whole idea is to make the dry mass of the container so low (because it's just a bag) that you get to take full advantage of the high Isp of hydrogen (either as burning hydrolox, or heated by a nuke) without incurring the high dry mass of the great big traditional tank.

The advantage of a collapsible bag is that you can send a bunch of the 'em, folded up, on a single launch.  You still have to make a bunch of LH2 in orbit somehow, but that's a different problem.  (My guess is that you launch water, electrolyze it using solar energy on the depot, and save the LOX for some other mission--unless of course it's a hydrolox mission, when you'll use most but not all of the LOX you electrolyze.)"""


Not arguing for more dedicated structure. Suggesting that items that may have a use at the destination may be used to help structurally. Wasn't thinking about a pressurizable structure, just a bit of support from items that are eventually payload.

Were you think of truss elements that could be reassembled for other uses?

[TheRadicalModerate]

Not arguing for more dedicated structure. Suggesting that items that may have a use at the destination may be used to help structurally. Wasn't thinking about a pressurizable structure, just a bit of support from items that are eventually payload.

Were you think of truss elements that could be reassembled for other uses?

I'm prejudiced toward thinking about this as a nuke.  Odds are that a nuke dumps all of its payload, which makes arrangements for its own arrival at wherever it's going.  That makes it unlikely that anything gets reused other than the nuke, which wants to be as light as possible, since it probably has to do a pure braking burn if it's to be captured.

Second most likely is a hydrolox architecture that has to aerocapture the bejeezus out of itself, and you're not recycling any odds and ends from that, either.

There are obviously other architectures, but let's keep it to the irreducible minimum:  How light can we get away with for a departure burn?

[redneck]

Not arguing for more dedicated structure. Suggesting that items that may have a use at the destination may be used to help structurally. Wasn't thinking about a pressurizable structure, just a bit of support from items that are eventually payload.

Were you think of truss elements that could be reassembled for other uses?

I'm prejudiced toward thinking about this as a nuke.  Odds are that a nuke dumps all of its payload, which makes arrangements for its own arrival at wherever it's going.  That makes it unlikely that anything gets reused other than the nuke, which wants to be as light as possible, since it probably has to do a pure braking burn if it's to be captured.

Second most likely is a hydrolox architecture that has to aerocapture the bejeezus out of itself, and you're not recycling any odds and ends from that, either.

There are obviously other architectures, but let's keep it to the irreducible minimum:  How light can we get away with for a departure burn?

It seems to me that the lightest would be the pure tension bag with puller vehicle as suggested by Asteroza with the H2 as slush or even frozen. Tiny inducer/transfer pump at the bottom of the bag. 0.5% of H2 mass for tankbag??