Cooling Method Experiment

[beowulf2010]

As I mentioned in @Gurgel's Compact Chiller thread, I wanted to test out the differences between 3 different cooling methods. The results have both been predictable and surprising.

Initial Setup: Using Debug mode, I created 6 identical sets of rooms to test how much cooling gets transferred between a typical 4 Wheezewort Hydrogen cooling room and 20,000kg of water. Attached is the completed build before inserting the water into the target rooms. Please note that for whatever reason, the bottom left door keeps resetting to closed, it should be set to open before running the experiment if you want to use my parameters. 

The mess of pipes under the 6 rooms is there to make sure all 6 coolant rooms get filled at the same rate due to identical length pipes and I didn't want to go back to straighten it up.

After inserting whatever liquid you want to cool, all that is needed is to hook the Manual Generator up to the rest of the power grid with (preferably) conductive wire.

I've only run this for 20 cycles so far (have a lot to do today so I can't sit here and get data every 10 cycles like I'd like) but the results are already showing.

Rooms 1/2 (Left hand pair) are using passive cooling through Tungsten Tiles and Wolframite doors. Room 1 uses 2 doors, room 2 uses only 1 door.

Rooms 3/4 (Center pair) are using Hydrogen filled radiator piping. Room 3 actively pumps the gas while room 4 fills a pipe bridge driven recirculating radiator through both rooms.

Rooms 5/6 (Right pair) are using liquid filled radiator pipes. Like the central pair, room 5 is the active pumping while room 6 uses bridge driven re-circulation.

Results:

Cycle 10: 99.1, 99.1, 98.6, 98.9, 97.0, 96.8

Cycle 20: 96.8, 96.7, 95.6, 96.6, 94.7, 94.8

Initial Conclusions:

1) The best way of cooling (between these three choices) is using liquid filled radiators. I think most of us already knew this from the higher volume that liquid pipes carry (10kg versus 1kg) and the better heat capacity of most fluids versus gas (4.179 Water and 6 Polluter Water versus the 2 of Hydrogen).

2) Active pumping versus re-circulation: Not sure if I can actually make a conclusion on this yet as liquid shows similar results, but gas active pumping did result in a full degree better performance. This being said, 20 cycles is a small time frame and we're looking at 3.4 versus 4.4 degree temperature drop.

3) Passive is the worst in terms of raw performance but is best for micro temperature management as cooling/heating gets shut off immediately upon the doors opening. There does not seem to be any meaningful difference between 1 or 2 doors in the vacuum space. Possible factors for this might be leaving extra metal tiles or the fact that I put the water below the metal tiles. I don't know if 800kg of water interacts with metal tiles above them the same way a full 1,000kg tile of water does.

4) The liquid pumping rooms almost ruined my initial experiment run due to how much more heat gets put into the cooling room. By the time the temperatures stabilized at 95C around cycle 13, the Wheezeworts had gotten up to 85C.  Using the more efficient polluted water coolant, more volume of water to cool or even a slightly higher temperature could have stifled the Wheezeworts ruining the experiment.

5) There are reasons to use 3 cooling methods. Passive is exactly that but take space. Hydrogen cooling takes more time but has a wider temperature range to work with. Liquid cooling is more powerful but can more easily freeze or boil if safeguards are not put in place to prevent it from over heating/cooling.

I'll let it run for a couple hours tonight and post again with those results, though I expect to see more of the same.

Attached is both the save file ready for the experiment to start and the experiment after 20 cycles in case anyone wants to look at what I did.

Comment are definitely welcome, especially if I screwed something up in the build or have misunderstood a critical part of ONI physics.

Coolant Test - Before Elements.sav

Coolant Test - After 20 cycles.sav

[BionicSandwich]

you forgot about the AETN. that thing is a beast when it comes to cooling.

[beowulf2010]

1 hour ago, BionicSandwich said:

you forgot about the AETN. that thing is a beast when it comes to cooling.

Nah. Everything that is true for a small scale 4 Wheezewort room will hold true for the beastly AETN. Only real difference is the minor hydrogen usage. 

[Neotuck]

4 hours ago, beowulf2010 said:

Only real difference is the minor hydrogen usage. 

and the temp cap

wheezewarts stop cooling when gas reaches -60C

AETN stops at -174C

[Coolthulhu]

And the part where people don't misspell AETN to sound like a name of a disgusting disease.

[Kabrute]

3 minutes ago, Coolthulhu said:

And the part where people don't misspell AETN to sound like a name of a disgusting disease.

would you please elaborate

[wachunga]

Your "passive" cooling setup is very sub optimal. Tile/tile heat interactions are limited by the lower conductivity material. The hydrogen to tungsten interaction is limited by the hydrogen, the tungsten/wolframite by the wolframite, and the tungsten/water by the water. You get around this restriction by using temperature shift plates between the two materials. A shift plate is just a 3x3 "building" that can be built pretty much anywhere. Buildings multiply the conductivities of themselves and the tiles they interact with. If you really want to test what "passive" is capable of then you need a setup like this.

6 tiles of hydrogen interact with the plate (preferably diamond) at a boosted rate because the conductivity of the plate helps. That plate also interacts with 3 tiles of tungsten, again at a boosted rate. So you effectively link 2 tiles of hydrogen to 1 tile of tungsten with a conductivity boost, in addition to the "normal" interaction. Versus just the "normal" interaction of 1 tile of hydrogen to 1 tile of tungsten limited to hydrogen's conductivity. Do note that plates DO NOT directly interact with other plates. I only mention this because it's a very common misconception, people build these long chains of plates that are poor design.

Try this test again with an optimal "passive" setup and see what you get.

[beowulf2010]

30 minutes ago, wachunga said:

Try this test again with an optimal "passive" setup and see what you get.

Yeah, you're right that Tempshift Plates should make things way better. I chose not to in this case for reasons I'm not sure I even remember. I can edit and rerun tonight. 

[wachunga]

Choosing not to use diamond for example is perfectly reasonable as diamond isn't exactly trivial to get. Just wanted to make sure everyone is on the same page about how to setup a door controlled heat exchanger.

[beowulf2010]

Yeah, no problem. Like I said, I can't even remember why I didn't use Tempshift Plates onto and out of the metal tiles. Cooking up some Taco meat so it'll be a while before I can edit stuff. 

[beowulf2010]

Welp, added Diamond Tempshift Plates and it did improve the door cooling. But liquid radiators are still better.

After the first 10 cycles, door cooling was a half degree better than gas radiators, but liquid was another degree and a half better than the doors.

I can keep this running for a while today but I think we know what I wanted to double check.  Liquid radiators > metal tile/door/metal tile > gas radiators.