Simple Liquid Hydrogen Setup

[Jumpp]

This is a simple Liquid Hydrogen producer. Once booted up, it can take input hydrogen at 50C and produce output hydrogen at well below vaporization (-258ish) at nearly 300kg/cycle. (It's doing it as fast as the single pumps can operate, so it ought to be a neat 300 kg/cycle, but my actual observed measurements came in at 296/cycle over several cycles.)

The key problem with producing LH2 is that super-coolant freezes at -266.2, and therefore can't be safely passed through an aquatuner at temps below -252.2. However, the condensation point of Hydrogen is -252.2. So if you use a temp sensor and shutoff to filter by temperature such that coolant below -252.2 doesn't go through the tuner, the system tends to settle into an undesirable stable equilibrium just below -252.2 such that the coolant has little cooling power but the tuner can't help.

So this cooler works by taking parcels of coolant in the awkward range from -252C to -256C and warming them slightly by passing them briefly through a warmer 'pre-cooling' chamber.

Input hydrogen (at 50C in my tests) is first pumped into a pre-cooler chamber (the upper half of that double vessel) and from there it's pumped to the main cooling chamber.

The coolant, on leaving the main cooling chamber, takes one of three paths:

- Parcels above -252C go to the aquatuner and then to the main cooling chamber.
- Parcels below -256C go back into the main cooling chamber.
- Parcels between -252C and -256C go through the pre-cooling chamber, where they're warmed slightly by being passed through a single length of non-insulated tungsten pipe, then back into the main cooling chamber.

The liquid pipe thermo sensor on the left is set to below -256C, and the one on the right is set to above -252C.

There's no automation on the input pumps. They just pump as fast as the high-pressure vents will permit.

The output pump runs only when there's at least 500kg of hydrogen in the tile, at a temp below -257. (In practice, the liquid level sensor is almost always the bottleneck. At 500g/s in at 50C, the liquid in the cooling chamber rests at around -260.

Some notes:

I ran my tests with 370kg of super-coolant in the system.

To pause the entire system, just disable the liquid reservoir that's above the tuner vessel. Pretty soon all the coolant in the system will collect there.

The liquid reservoirs are essential. They buffer the output on every line out of the temp-sorter unit such that there's no possibility of a jam.

It's important that the radiant line in the main cooling chamber passes first through the accumulated LH2. If you bring that line in above the liquid level, the LH2 tends to settle at temperature uncomfortably near to (or just above) the vaporization point.

The system is built with a minimum of exotic materials. All the insulated tile is ceramic. All the metal is gold, except for that one tungsten pipe in the pre-cooling chamber which could probably be any metal. Most of the insulated pipe is ceramic. The only exotic materials in the system are the super-coolant and the insulated output pipe, which is insulation.

The system requires no vacuum. I ran my tests in a breathable O2 atmosphere. The only vacuum I used was for the reservoir to hold the output, but that's just a measurement tool. In prod you'd just run the output line direct to the fuel tanks.

The tuner vessel here is cooled with a loop powered by six wheezes. That turns out to be not enough. I don't know the correct number. (A cocktail-napkin calculation suggests that it might require 31 wheezes to run at full speed for any length of time. Without that ludicrous amount of cooling, you'd need to throttle the system somehow. A liquid valve on the line out of the tuner vessel, just before it passes behind the center reservoir, would do.)

The temp-sorter unit is fail-safe: If there were a jam (not possible without over 5 tons of super-coolant in the system) or if the system loses power (which it could), the only failure is that some coolant that ought to go to the tuner instead goes to the pre-cooling chamber.

 

[Lifegrow]

I ran a Liquid O2/Liquid H build for over 400 cycles, and I found that providing you use a buffer reservoir you can easily automate it with just a pipe thermo sensor set to above -254 degrees C. 

As with all liquid O2/H builds, the real key is to keep a small body of the liquid in the room at all times for stability - that way you're essentially balancing your coolant temperature with the end product. It also aids in avoiding pipe breakage when you begin pumping the liquid to it's destination (a simple hydro sensor set to above 500kg for example) as it can absorb the heat generated by the pump itself.

Heres the first iteration of what I ultimately ended up using for ~400 cycles

Spoiler

 

 

[Carnis]

I had 0 issues with a bottomfed simple loop supercoolant with:

If > -252, aquatuner.

If < -252 cooling chamber.

My equilibrium ended up at 256-257 on The bottom Part of The chamber, I did not see any problematic lh2 evaporation. Seems a simpler system can work equally.

[horgerj]

The hysteresis properties of the phase changes are useful for liquid hydrogen - you can set the thermo sensor to allow super coolant into the aquatuner above -255 °C. This puts the super coolant coming out at -269 °C, which is enough to condense incoming hydrogen instantly, but still safely over the *actual* freezing point of the coolant (-266.2 °C- 5 °C = -271.2 °C). I've been running this for a looooong time, and haven't frozen anything yet!

[Jumpp]

Thanks very much, all. I'm glad to learn this doesn't have to be so complicated.