siimav

What capsules/cockpits are are you testing this with? If the parts have an entry cost of 1 then no training is required before being able to assign crews to them.

Yes, it does. With cryogenic fuels, it's always wise to orient the vessel in a way that minimizes the tank surface area that the sun is shining on. Heat conduction is also implemented so keeping the distance between sunlit areas and tanks also helps.

The HeatPump mod was redone and is now called Real Active Radiators. It also has a separate GH repo and forum thread. The new version works a lot better with high timewarp but be aware that the EC consumption is a lot higher (and more realistic) than before.

I would suggest grabbing the ScanSat mod and generating your own maps of Mars. Specifically the slope map allows pinpointing the flattest areas on planetary bodies.

The radiator part was from USI: link. I'm building a hydrolox mining base in RO with USI + RealISRU and thought that adding one of those USI heatpumps is a good idea because drills quickly saturate the cooling capacity of regular radiators.

I found a couple of more issues while using this mod. First, it doesn't really play nice when you add radiator parts from another mod onto the same vessel. Found out it's because of part.FindModuleImplementing<ModuleActiveRadiator>() does not check whether the returned item is actually of the correct type. Another issue that randomly happened to me was that the throttle went totally insane on high time warp. I was seeing crazy high negative throttle values, like -700%. This seems be caused by refrigerationCost randomly getting negative values. I think I have managed to get both of them fixed on my own fork. So would it help you if I created these fixes as pull requests on GitHub?

Yes, I do have the full RO suite of mods installed. Did a search against my gamedata folder and RO does indeed try to set overcoolFactor to 0.0186367. What makes it extra strange, is that there are no other mods that set it to 0.25 either except the stock part configs from Squad. Anyway, I also ripped out the offending stock code and can now get LH2 boiloff to absolute 0. Thank you for taking your time to explain how KSP's internal radiator mechanics work.

Thanks, the nuclear reactors are now working fine after overwriting this config file. I did some more debugging about the LH2 boiloff issue I previously mentioned. It's clear that cryogenic cooling is not applied to my tanks that contain LH2. After adding some logging code, I found out that it's because [tank temperature] < [radiator temperature] * overcoolFactor. From the screenshot: 293.5 * 0.25 = 73.4 while the tank temperature is at 21K. Not sure what the overcoolFactor is supposed to control though.

One more issue I have now run into when testing the Nertea's Near future electrical, is that this mod causes the stock radiators to become unable to cool the nuclear reactors. I'm not sure what stat it is that causes this behavior. If tweaking the stock radiators is necessary for cryogenic cooling to work, maybe this mod should leave stock radiators as-is and create of copy of them that have the changes needed for cryo cooling?

I think running cryogenic cooling only on parts where (part.maxTemp * part.radiatorMax < 110) also works. This should include only parts that contain cryogenic fuels because methane has the highest boiling point of the lot (109.15 K). Based on my testing, it looks to be working fine with RealFuels and cooling is only applied to parts that actually need it. Power requirements are also not so insane any more.

Hey, @Starwaster, I have a space station is orbit that contains no cryogenic fuels but the cryo radiators are running at 100% throttle. Any ideas what might be causing it or how to diagnose the issue?

I have to agree when we are talking about KSP 1.2.2/RP-0. In KSP 1.3.1/RP-1 they introduced a tag system to most parts that is used for adding a pricing multiplier to rollout costs. In case of turbopump engines, that modifier is 4 whereas solids only have a multiplier of 1.1. So even though a SRB part may cost 2 times as much as an liquid engine + tank, the final cost will actually be slightly cheaper when you also factor in the rollout costs. Unfortunately those multipliers are not documented anywhere and you have to poke around in the config files to find the actual multipliers for each tag.

The tank types that were in 1.2.2 are no longer used. There's now different levels (1 to 4) of default tanks and service module tanks that are unlocked by researching materials science tech tree nodes. Cryogenic tanks were not available in RP-1 for a long time. Couple of weeks ago an update to RealFuels was released that added a MLI slider that you can now use to control isolation levels on your tanks. Adding part configurations to TREE-parts.cfg should actually work to make these available in RP-1. Are you sure you assigned the parts to some tech node and added pricing for it?

@Starwaster Did a test by changing the radiatorMinTempMult setting to 0.9. I'm still seeing a internal temp on 20.2 and boiloff is also at 0.02. So it looks like the radiators do no provide sufficient cooling or they are not cooling the tank at all. BTW, the latest 1.01 release does no include dll file.

Checked that settings and it's correctly set as 0.99. If I set it to for example 0.9 does it mean that it would try to cool the tank to 0.9*20.2K?