Tank of pressurized hot hydrogen, poor man's NERVA?

[Pds314]

So, after a long series of random tangents starting with how to improve the propellant efficiency of a water bottle rocket, I've imagined a nightmarish very interesting pressure rocket.

Grab a tank made of tungsten or some other durable high-temperature material.

Insulate it with several inches of vacuumed out carbon aerogel.

Fill it full of hydrogen compressed to 850 atmospheres, such that it is nearly liquid.

Raise the temperature near the melting point of your tank. (In my head, I was thinking 3400 C). This should raise the pressure to around a Gigapascal. This could be done via an electric element, also made of tungsten.

Open a tiny valve in the bottom to open the gates of hell activate the rocket.

Enjoy an ISP of up to 684 seconds (more if the hydrogen dissociates) without ever touching a nuclear reactor.

It would bleed off heat slowly, but since carbon aerogel weighs essentially nothing and is unbelievably good at insulating things, the losses would be negligible.

There is a problem. The thing is a giant thermobaric bomb. 22.5 MegaJoules released per kg of H2 in the sudden pressure explosion that would occur. 141 more MJ the instant that H2 runs into O2.

And also, how thick does white-hot tungsten have to be to hold back 10000 atmospheres?

Edited November 17, 2014 by Pds314

[LaytheAerospace]

There is a problem. The thing is a giant thermobaric bomb. 22.5 MegaJoules released per kg of H2 in the sudden pressure explosion that would occur. 141 more MJ the instant that H2 runs into O2.

The bigger problem is the pressure vessel. We've known for a long time that you can get useful propulsion from compressed gas without combustion. Kids have been playing with pump bottle rockets for decades. The problem is storing a useful amount of gas in a pressure vessel that's still light enough to be propelled by the compressed gas. I suspect that there's probably a point where making the pressure vessel stronger to hold more gas results in less useful thrust, and that point is probably quite low compared to something like a traditional rocket.

There's also the problem of the rapidly decreasing strength of metals combined with the increasing pressure as you heat the pressure vessel up. Even with a huge vessel, you're likely to blow it out before getting anywhere near the melting point.

[Pds314]

The bigger problem is the pressure vessel. We've known for a long time that you can get useful propulsion from compressed gas without combustion. Kids have been playing with pump bottle rockets for decades. The problem is storing a useful amount of gas in a pressure vessel that's still light enough to be propelled by the compressed gas. I suspect that there's probably a point where making the pressure vessel stronger to hold more gas results in less useful thrust, and that point is probably quite low compared to something like a traditional rocket.

There's also the problem of the rapidly decreasing strength of metals combined with the increasing pressure as you heat the pressure vessel up. Even with a huge vessel, you're likely to blow it out before getting anywhere near the melting point.

True...

Also, the fact that liquid H2 has a density of 71 kg/m^3 doesn't help.

And blowing things up would be BAD.

Something with a similar amount of H2 to the total mass of a Saturn V would make a 14-kt explosion, and then within a fraction of a second, consume all of the O2 within a kilometer, making a 91 kt explosion.

I suppose that does technically mean I am trying to design a pressure vessel to contain a 14-kt explosion, and use it in a controlled way, doesn't it?

[LaytheAerospace]

True...

Also, the fact that liquid H2 has a density of 71 kg/m^3 doesn't help.

And blowing things up would be BAD.

Something with a similar amount of H2 to the total mass of a Saturn V would make a 14-kt explosion, and then within a fraction of a second, consume all of the O2 within a kilometer, making a 91 kt explosion.

I suppose that does technically mean I am trying to design a pressure vessel to contain a 14-kt explosion, and use it in a controlled way, doesn't it?

Something that would work better, I think, might be nanotubes. Tungsten nanotubes are theoretically possible, and should be much better at resisting heat and radiation than carbon nanotubes. Take a thin tungsten tank, wrap it in a woven mesh of tungsten nanotubes, then heat it up. You won't get anywhere near the melting point, but I bet it would hold a hell of a lot of pressure, and be very light to boot.

Now...where'd I put my tungsten nanotube loom?

[Kryten]

Theoretically possible according to who? Under what conditions is tungsten supposed to be able form covalent bonds with itself?

[xenomorph555]

Something with a similar amount of H2 to the total mass of a Saturn V would make a 14-kt explosion, and then within a fraction of a second, consume all of the O2 within a kilometer, making a 91 kt explosion.

This sounds epically awesome!

I guess this basically happened on N-1 flight 2.

[GregA]

So, after a long series of random tangents starting with how to improve the propellant efficiency of a water bottle rocket, I've imagined a nightmarish very interesting pressure rocket.

Grab a tank made of tungsten or some other durable high-temperature material.

Insulate it with several inches of vacuumed out carbon aerogel.

Fill it full of hydrogen compressed to 850 atmospheres, such that it is nearly liquid.

Raise the temperature near the melting point of your tank. (In my head, I was thinking 3400 C). This should raise the pressure to around a Gigapascal. This could be done via an electric element, also made of tungsten.

Open a tiny valve in the bottom to open the gates of hell activate the rocket.

Enjoy an ISP of up to 684 seconds (more if the hydrogen dissociates) without ever touching a nuclear reactor.

It would bleed off heat slowly, but since carbon aerogel weighs essentially nothing and is unbelievably good at insulating things, the losses would be negligible.

There is a problem. The thing is a giant thermobaric bomb. 22.5 MegaJoules released per kg of H2 in the sudden pressure explosion that would occur. 141 more MJ the instant that H2 runs into O2.

And also, how thick does white-hot tungsten have to be to hold back 10000 atmospheres?

NASA is one step ahead of you. They are proposing meta stable hydrogen rockets, with twice the isp of Nerva

http://www.nasa.gov/sites/default/files/files/Silvera_FinalReport.pdf

[LaytheAerospace]

Theoretically possible according to who? Under what conditions is tungsten supposed to be able form covalent bonds with itself?

I should have been more specific. Tungsten disulfide. Read about them first a few years back in the context of space elevators. The article was talking about how carbon nanotubes, despite being very strong, aren't well suited to a hostile radiation environment, like you find in space. Inorganic nanotubes, particularly tungsten disulfide, were given as a potential alternative that could withstand the environment without degrading over time, like carbon would.

I looked again, and couldn't find the article. Saw lots of unrelated information about the tungsten disulfide nanotubes, but nothing specifically related to its potential use as a ribbon in a space elevator, or how it stands up to radiation.

Edited November 18, 2014 by LaytheAerospace

[Idobox]

Both ISP and thrust will go down relatively quickly, since thrust is directly linked to pressure (unless you open the valve wider and wider to compensate), and the temperature should go down with the pressure too (it's essentially an adiabatic decompression).

[trekkie_]

as you release pressure, the pressure reduces. the only way to counteract this is to have a tank that changes in volume to maintain pressure (which would require energy, and energy equals weight), or to somehow trap pressure in nanotubes and release it through some catalyst that dissolved the nanotubes in a controlled chain reaction and expels the pressure. but the mass of the nanotubes themselves may pose a problem to the energy you get out of it, and there really isn't any mastered material that can handle the pressures we would need.

at this point, other propulsion methods look better. space travel will ultimately need to focus on refuelable sources of propellant, or become propellant-less entirely.

[magnemoe]

Why not heat the hydrogen on the way out instead of heating the entire tank. This also let you use liquid hydrogen.

You could use an huge mirror and heat an tungsten block the hydrogen flow trough with sunlight like an solar oven, optional use laser from base.

Upside is that you don't need the heavy reactor, downside is that you need the voluminous and complicated mirror system.

Would work best on small probes in deep space.

If you have an huge power source on-board vasmir would be better.

[gooddog15]

Why not heat the hydrogen on the way out instead of heating the entire tank. This also let you use liquid hydrogen.

You could use an huge mirror and heat an tungsten block the hydrogen flow trough with sunlight like an solar oven, optional use laser from base.

Upside is that you don't need the heavy reactor, downside is that you need the voluminous and complicated mirror system.

Would work best on small probes in deep space.

If you have an huge power source on-board vasmir would be better.

Or you could have a reactor to power the ship, and use its waste heat to heat up the tank.

also, I would imagine that you could use electric space heaters if they can attain the same temperatures.

Edited November 19, 2014 by gooddog15

[rodion]

Tank of pressurized hot hydrogen

Yes, this sounds simply delightful. What could go wrong?

[Idobox]

Or you could have a reactor to power the ship, and use its waste heat to heat up the tank.

also, I would imagine that you could use electric space heaters if they can attain the same temperatures.

That would be exactly a NERVA. The OP wanted to get something similar, but without needing a reactor.

[Bill Phil]

That would be exactly a NERVA. The OP wanted to get something similar, but without needing a reactor.

Technically no. A NERVA is a NTR developed* by NERVA. A NTR is what you're thinking about. Having a reactor heating up the tank itself isn't like NERVA, but rather a different concept.

*key word here. They never really built any flight models as far as I know

[magnemoe]

Or you could have a reactor to power the ship, and use its waste heat to heat up the tank.

also, I would imagine that you could use electric space heaters if they can attain the same temperatures.

Hating the tank is an bad idea in any setting, you want to use liquid hydrogen to decrease tank weight. Having 1 ton of liquid hydrogen would be far lighter than a ton under pressure. You heat the hydrogen in a small chamber and use this as reaction mass.

Benefit of vasmir is that its 5-10 time better isp than nerva as you don't only heat hydrogen you also accelerate the ions like in a ion engine.

It however require lots of electrical power, nerva or solar thermal only need heat and you let the reaction gas cool the engine so you don't need lots of cooling systems.

[NFUN]

Hating the tank is an bad idea in any setting, you want to use liquid hydrogen to decrease tank weight. Having 1 ton of liquid hydrogen would be far lighter than a ton under pressure. You heat the hydrogen in a small chamber and use this as reaction mass.

Benefit of vasmir is that its 5-10 time better isp than nerva as you don't only heat hydrogen you also accelerate the ions like in a ion engine.

It however require lots of electrical power, nerva or solar thermal only need heat and you let the reaction gas cool the engine so you don't need lots of cooling systems.

Uh.......?

[Kryten]

Uh.......?

Tank mass is lower because it's a lot easier to contain. Same reason ISS gets oxygen supplied as water then electrolyses it; liquid tanks, even cryogenic ones, are far lighter than pressurised gas tanks.

[magnemoe]

Tank mass is lower because it's a lot easier to contain. Same reason ISS gets oxygen supplied as water then electrolyses it; liquid tanks, even cryogenic ones, are far lighter than pressurised gas tanks.

Yes, i was obviously thinking total ship weight not the weight of one ton hydrogen

Pressure tanks are heavy and don't hold much gas.

[Armchair Rocket Scientist]

Technically no. A NERVA is a NTR developed* by NERVA. A NTR is what you're thinking about. Having a reactor heating up the tank itself isn't like NERVA, but rather a different concept.

*key word here. They never really built any flight models as far as I know

It's a terrible concept. If you must use electric heating elements, then pass the hydrogen through a tungsten pipe that's been heated to thousands of kelvins or something, and have the benefit of your spacecraft not being under thousands of atmospheres of pressure. Failing that, build a Poodle Thruster. It's basically an NTR, but uses the waste heat from RTGs instead a full nuclear reactor. It should get about the same ISP as an NTR, but at far less thrust.

[Bill Phil]

It's a terrible concept. If you must use electric heating elements, then pass the hydrogen through a tungsten pipe that's been heated to thousands of kelvins or something, and have the benefit of your spacecraft not being under thousands of atmospheres of pressure. Failing that, build a Poodle Thruster. It's basically an NTR, but uses the waste heat from RTGs instead a full nuclear reactor. It should get about the same ISP as an NTR, but at far less thrust.

I never said it was good, just that it's not a NERVA. Heck, a modern NTR wouldn't be a NERVA and would probably have an ISP advantage of 50 seconds or so.

[Wallygator]

For some reason after reading this thread I started visualising a solar powered toaster stapled to the rear end of the Hindenburg. With George C Scott somewhere in the middle...

Recalling my basic chemistry and Physics, I don't see this working very well.

[K^2]

Somebody should actually compute the thickness of the tank walls in that original proposal, just for giggles. You can have all the I_SP you want, but it won't do you any good if your mass ratio is practically unity.

[architeuthis]

Somebody should actually compute the thickness of the tank walls in that original proposal, just for giggles. You can have all the I_SP you want, but it won't do you any good if your mass ratio is practically unity.

Ok.

Realistic treatment of failure theories for thick walled pressure vessels isn't really something they teach you during mechanical engineering undergraduate education. Practically speaking you should use FEA for a problem like this. But the yield strength for most polycrystaline metals takes an epic nosedive once you get above 0.6 of their melting temperature. Material strength also depends on things like strain rate, and the specifics of the material's annealing process. Anyway, here is my guess at it:

Assuming the tank is perfectly insulated, there is no heat transfer via radiation at the boundary, conduction is steady state and everything is in thermal equilibrium, and the ambient pressure is a vacuum. There is no loading on the pressure vessel apart from the internal pressure. No stress concentration factors such as notches or geometry changes. No factor of safety. I know that some of these factors interact in complicated ways, so I'm naively going to go with the basic distortion energy theory (I'm assuming that tungsten behaves ductile-ly at these really high temperature for low strain rates). Suffice it to say, take this calculation with a bit of salt:

Trying be a little realistic I'll say that this is wrought tungsten at 2200C... This is going to reduce your isp by about 20% unfortunately.

Anyway, here we go

Y_strength=68.95E6; %Yield stress for tungsten at 2200C [Pa]

p_i=850*101.3E3; %internal pressure [Pa]

dia=8.412; %assuming the pressure vessel has the same diameter as the Space Shuttle External Tank [m]

r_o=dia/2;

%sigma_t=principle stress tangential(hoop)

%sigma_l=principle stress longitudinal

%plane stress condition holds

syms r_i sigma_t sigma_l

[sigma_t, sigma_l, r_i] = solve(sigma_t==r_i^2*p_i/(r_o^2-r_i^2)*2, sigma_l==r_i^2*p_i/(r_o^2-r_i^2), Y_strength==sqrt(sigma_t^2-sigma_t*sigma_l+sigma_l^2), sigma_t, sigma_l, r_i,'Real', true);

thickness=vpa(r_o-r_i)

ans =

2.3649408607072881546752561420743

2.36 m thick... yeah, I'm guessing your mass ratio isn't going to be super awesome with tanks like that.

Edited November 28, 2014 by architeuthis

[rodion]

Ok.

Realistic treatment of failure theories for...

great analysis, but i think you may do well from exploring the benefits of LaTeX's math markup.