Storing Cryogenics in Zero-G

Maintain your cool

Storing cryogenic fuel in zero g presents some unique engineering challenges for the Artemis Project.

By design, the temperature of the cryogenic tanks will be right at the boiling point of the liquid fuel or oxydizer. Most spacecraft maintain this temperature by venting vapor overboard as the liquid boils. To keep the mass of the fuel tanks down, we don't run the fuel at high pressure -- no more than 3 atmospheres will do the job.

Constant evaporation of the fuel carries away heat that gets into the tank. Insulation does not prevent heat from getting through, but it does slow down the rate at which heat transfers to the liquid fuel. If you look up the heat of vaporization of oxygen and hydrogen, and estimate the amount of heat that makes it past the insulation, you can estimate the amount of fuel we will use for cooling during the mission.

The phase change from liquid to gas absorbs a tremendous amount of heat compared to simply heating the liquid. Compare the heat of vaporization to the specific heat of the liquid. Specific heat is how much energy it takes to raise the temperature of one unit of liquid by one degree; heat of vaporization is how much energy it takes to evaporate that same amount of liquid.

Don't blow it out your vent

Now, here's the problem. In an unaccelerated state, the cryogenic liquid would tend to evaporate wherever it comes into contact with an object hotter than its condensation point. Liquid near the vent nozzle could flash into vapor, pushing liquid propellant overboard. This would waste a lot of fuel because the liquid is so much more dense than the vapor.

What Her brother did

The Apollo S-IVB stage used propulsive venting, just a few micro-g's, to keep the fuel settled in the tanks so that they didn't have liquid fuel near the vent nozzle. Propulsive venting simply means that we vent the vapor through a little rocket nozzle pointed to provide a tiny bit of thrust in the direction we want to accelerate.

Propulsive venting worked fine for the Apollo missions because they were done with the S-IVB stage as soon as they completed translunar injection. After TLI, they used room-temperature hypergolic fuels. The Apollo service module had tanks of cryogenic fuel for generating electrical power, so it also had propulsive vents to maintain a tiny acceleration on the vehicle.

Sister has a different opinion

In the design for the Artemis Project's Reference Mission vehicles, we use cryogenic fuels for every stage of the flight except the Ascent Stage. That means we will have the Lunar Transfer Vehicle orbiting the moon for several days with cryogenic fuel in its tanks.

More problems at the gas station

Cryogenics also becomes a problem if we want to swap out empty tanks for full ones in Earth orbit, or store tanks of cryogenic fuels at a space station. Keeping a constant acceleration on a space station would eventually add up to quite a trajectory change. The problem gets compounded by the fact that the acceleration experienced by the fuel in the tanks will change as tanks are swapped between the station and the moonship -- we'll have the stuff sloshing around in the tanks as we move them.

Just to add to the fun for the space station, we have the additional complexity of acceleration due to atmospheric drag in low Earth orbit. Drag can add another 40 microgravities to the normal state of operation. Drag acceleration might be sufficient to keep the cryos settled, but we will have to reboost the station every six months or so, to keep it from reentering Earth's atmosphere. Acceleration during reboost will be in the opposite direction from the drag acceleration, so our tank vents will be on the wrong end of the tanks during that maneuver unless we rotate the whole station every time we want to reboost.

So, handling cryogenics in orbit is going to be a complex operation where we have to be constantly mindful of the orientation of the fuel tanks, the acceleration vectors, and where the tank vents are located. It might even be more trouble than it's worth; but we win a great prize if we can figure out how to do it.

Rocket scientists do it on impulse

The specific impulse of H₂-O₂ engines is much greater than common room-temperature hypergolics (about 460 seconds for H₂-O₂ compared to about 260 seconds for MMHD-N₂O₄). For many lunar missions, the difference adds up to a huge difference in the payload we have to launch from Earth.

Consider, for example, a mission where we send 10,000 lbs from low Earth orbit to the surface of the moon. In this example we'll assume it's all one vehicle, no staging and no expendable tanks. So, we have a series of rocket engine burns that like this:

        Translunar injection      10,000 ft/sec
        Lunar orbit insertion      2,800 ft/sec
        Descent orbit initiation     200 ft/sec
        Landing                    7,000 ft/sec

                           Total  20,000 ft/sec

Those numbers look like they're rounded off, but they're good to three decimal places; so this is a reasonably accurate example. Now, using the inverse rocket equation ...

   mo = mf*exp(delta_V/(g*Isp))

... and g = 32.174 ft/sec/sec, we get:

        Propellants            H2-O2  MMHD-N2O4
        Isp         sec          460        260  specific impulse
        Mf          lbs       10,000     10,000  final mass (on moon)
        delta_V     ft/sec    20,000     20,000  change in velocity
        Mo          lbs       38,626    109,227  initial mass (in LEO)
        Delta_M     lbs       28,626     99,227  propellants burned
        Mass Ratio   -          3.86      10.92  (Mo/Mf)

Cold shoulder, but a cheap date

See the problem? If we use cryogenic propellants, about 25% of the mass we put into low Earth orbit lands on the moon as useful payload. If we use hypergolic propellants, we get less than 10% or our original mass as useful payload. If launch to LEO is the major cost driver, an MMHD-N₂O₄ rocket will cost us 2.5 times as much as an H₂-O₂ rocket, for the same mass of payload delivered to the moon. So we want to use H₂-O₂ unless there just isn't a feasible solution to the boil-over problem.

Putting the proper spin on it

There's still hope. If we can't make propulsive venting work, we might be able to use a centrifugal separator. We just let the liquid flash into vapor, and blow liquid propellant into the tank vent. However, instead of letting it dump overboard, we run the vented fluids (liquid and gas) into a centrifugal separator. We recover the liquid and pump it back into the tank, while the gas vents to space. That keeps the tank cool while minimizing the amount of propellants we lose during the mission.

If we can make the centrifugal separator work, we could put this system on every large cryo tank we have to handle. That makes the whole system work, including storing cryogenics at an orbiting facility and swapping out fuel tanks for the moonships.

There might also be an option of recovering the vented gas at the space station, so that we can recompress it and cool it down again. We would have to dump the excess heat into space with radiators. The mass of the radiators might make this infeasible for the moonships, but we're less worried about the total mass of an orbiting fuel dump.

So although handling cryogenic fuels presents some problems, we at least have an idea for an engineering solution. The next step is to proceed with a conceptual design for the centrifugal separator, pumps and plumbing to return the liquid to the tank, and the whole system for handling fuel tanks on orbit.