[1.12.*] Real Active Radiators v1.1

[Starwaster]

Real Active Radiators is primarily intended for use with Real Fuels and Realism Overhaul but otherwise behaves like the stock Active Radiators with improved behavior.

•  Can always cool parts regardless of ambient temperature. (unlike real radiators, stock KSP radiators will not pull heat if the ambient temperature is hotter than the radiator's temperature)
•  If the cooled part is colder than the radiator's temperature then an added heat pump cost (in ElectricCharge) will be incurred.
• Fixed bug where stock radiator energy transference is divided by the current timewarp scale. During normal x1 warp this resulted in energy being scaled 50x. Increased timewarp scaling reduces energy scaling so that radiators become less and less effective the faster you timewarp. (timewarp 1 = fixedTimeDelta = 0.02)

Known issue: Because of the scaling issue, the VAB reports radiator cooling as 50x more effective than it really is. However, radiator cooling is consistent now and 10 kW of cooling is 10 kW of cooling no matter how fast you're warping. The info in the VAB will be corrected next update.

About Real Fuels: Real Fuels now has multilayered insulation which can be installed by players on a per tank basis in the VAB. Always install some because active cooling works best when used with passive insulation. Also note that unlike Heat Pump, energy usage is more consistent regardless of the tank being cooled. A tank with liquid hydrogen costs the same amount of electricity as a tank with liquid oxygen. HOWEVER: You get less cooling on the LH2 tank than on the LOX tank. The amount of cooling per watt of power depends on the ratio of the cold temperature to the radiator temperature.

DOWNLOAD:

This link always has the latest update: 

https://github.com/Starwaster/RealActiveRadiators/releases/latest

Special thanks to @Bornholio and Schnobs for directing me to the documents that helped me figure out how refrigeration  efficiencies work and numerous conversations on the subject. (not sure who Schnobs is on the forums... who is he?)

Edited August 1, 2022 by Starwaster
Updated for KSP 1.12

[Starwaster]

Former users of the Heat Pump mod should download this file to allow your craft to load if they were using HP radiator parts

https://github.com/Starwaster/RealActiveRadiators/raw/master/GameData/RealActiveRadiators/DeprecatedHeatPumpParts.cfg

Just install that in the RealActiveRadiators folder

If you were using the zzz radiator (slightly triangular, cool looking) I overlooked that one. I'll get that one in here.

Edited May 24, 2018 by Starwaster

[Wallygator]

lovely. just pure unadulterated lovely...

[51ngular1ty]

Does this then eliminate or work around the stock thermal issue with massive core heat during extreme timewarp?  Meaning I can do a max time warp on my isru without worrying that everything will explode?

[Starwaster]

1 hour ago, 51ngular1ty said:

Does this then eliminate or work around the stock thermal issue with massive core heat during extreme timewarp?  Meaning I can do a max time warp on my isru without worrying that everything will explode?

Probably not because I didn't touch anything that dealt with core heat. In fact, as best as I can tell, radiators don't actually 'remove' core heat so much as the core pushes the heat to the radiator. ModuleCoreHeat looks at the list of active radiators (RAR counts because it's actually an extension of ModuleActiveRadiator) and adds heat to the radiator (real heat to the part itself) while removing core heat from itself.

Is this stock ISRU parts you're talking about or ISRU from some mod? I ask because I've been working on the Real ISRU configs and have warped at the maximum rate possible without anything exploding...

[51ngular1ty]

36 minutes ago, Starwaster said:

Probably not because I didn't touch anything that dealt with core heat. In fact, as best as I can tell, radiators don't actually 'remove' core heat so much as the core pushes the heat to the radiator. ModuleCoreHeat looks at the list of active radiators (RAR counts because it's actually an extension of ModuleActiveRadiator) and adds heat to the radiator (real heat to the part itself) while removing core heat from itself.

Is this stock ISRU parts you're talking about or ISRU from some mod? I ask because I've been working on the Real ISRU configs and have warped at the maximum rate possible without anything exploding...

Modded ISRU  powered by Nerteas nuclear reactors plus and a few other mods that use core heat.  From what I understand the behavior has something to do with how the game handles heat during high time warp essentially distributing it to all parts on a vessel.  I just know that if I have a lot of things that use core heat I get RUDs at high time warp.

[Starwaster]

1 hour ago, 51ngular1ty said:

Modded ISRU  powered by Nerteas nuclear reactors plus and a few other mods that use core heat.  From what I understand the behavior has something to do with how the game handles heat during high time warp essentially distributing it to all parts on a vessel.  I just know that if I have a lot of things that use core heat I get RUDs at high time warp.

It sounds like a configuration issue in one of the parts that uses core heat.

Heat at high timewarp doesn't exist in the conventional sense and it sort of gets 'distributed' but not the way it does at low timewarp or no timewarp. Instead what happens is that it goes into analytic mode where it assigns a global temperature to the entire vessel primarily based on its radiative properties. Adding flux determines what the vessel wide temperature will be. For parts to be exploding, something has to be seriously wrong in the core heat configs though if any of the mods that you use are using the analytic interface in their code then it's also possible to screw things up that way too. 

If I knew what parts are causing your explosions then I might have a better idea as to how it will behave with this mod and if it's something I could mitigate in any way. (though my gut tells me that I probably can't; as I mentioned, it's the core modules that are responsible for finding radiator parts and pushing heat out to them, though they limit themselves by what the radiator says it can accept)

[siimav]

Really appreciate the work you have put into this mod.

Is it actually possbile to reduce LH2 boiloff to 0 using radiators? I did a bit of testing by putting a tank with 500 tons of LH2 and LOX into orbit and by just maxing out the MLI layers, I was able to warp at 10000x and lose only 0.1 liters of LH2 per sec. By activating 2 of the largest radiators, the boiloff dropped to 0.05l/s. Do these numbers look about right to you? I think the power requirements of radiators are way too insane to be of any use in RP-1 though. There's no reasonable way to supply 90kW of electricity in that mod and considering the marginally reduced boiloff, the radiators are definitely not worth it. It's possible that the power requirements are only unbalanced in RP-1 and it's fine in RO and stock.

[Starwaster]

2 hours ago, siimav said:

Really appreciate the work you have put into this mod.

Is it actually possbile to reduce LH2 boiloff to 0 using radiators? I did a bit of testing by putting a tank with 500 tons of LH2 and LOX into orbit and by just maxing out the MLI layers, I was able to warp at 10000x and lose only 0.1 liters of LH2 per sec. By activating 2 of the largest radiators, the boiloff dropped to 0.05l/s. Do these numbers look about right to you? I think the power requirements of radiators are way too insane to be of any use in RP-1 though. There's no reasonable way to supply 90kW of electricity in that mod and considering the marginally reduced boiloff, the radiators are definitely not worth it. It's possible that the power requirements are only unbalanced in RP-1 and it's fine in RO and stock.

Yes it is absolutely possible to reduce boiloff to zero.

The power consumption is a bit off and I'm working on that right now.

However, even with the current version you should only see that kind of consumption if the temperature hasn't dropped to normal levels.... It would help if you further had your tank shielded in a fairing before getting to space and turning the radiators on. Or maybe have some radiators on already (the fixed static ones)

Edit: That reply was a little rushed; I had to leave the keyboard suddenly. You really want to keep as much heat from getting in there before getting into space. Some things that help with that are

• Putting the tanks in fairings (if possible; it may not always be practical)
• Make sure that there are umbilicals/clamps attached as long as possible. It's coded to keep the tank temperatures at (or slightly below) their boiloff until they are released.
• Use ServiceModule tank type. (note that it is extremely unrealistic to use this on a tank as large as the one you cited so it's up to you as to whether to do this or not)

Some things you can do as far as configuration 

• Edit your RealSettings.cfg and add QvCoefficient to RFSETTINGS. It's default value is 3.65 and affects how much heat gets through MLI when in a vacuum. Reducing it makes MLI more effective in atmosphere. Setting it to 0 would remove the convective term from the MLI equation entirely. (3.65 is a guesstimated value; it might be too high or even too low and honestly should be calculated on a per-planet basis instead of being a static value)
• In the RealActiveRadiators folder, edit RealActiveRadiators.cfg and edit or delete the cryoCoolerEfficiency section. You will note that temperatures of 20.15 (boiling point of LH2) has an assigned efficiency of 20% - change that if you want the cooling system to be more effective at that temperature, but honestly 20% is probably pretty generous. Deleting the entire section (cryoCoolerEfficiency, brackets and everything in between) will make it default to 70% across the board. Even at 100% efficiency though, it's always going to be less efficient moving heat from cold temperatures to hotter temperatures. The greater the difference the more power it takes to remove each watt of heat.

 

 

Edited May 26, 2018 by Starwaster

[Starwaster]

@siimav Just to make sure it didn't get lost in that wall of text, most of that electrical cost you're seeing is because the tank is already too hot so that kicks things into overdrive a bit. Once it hits its target temperature, the heat being drawn will be equal to the skin-internal heat leakage and I calibrated against a radiator temperature of 300 which at 20% efficiency comes out to about 70 watts of electricity per watt of heat removed. (GRC came up with about 114 W per watt removed so you're getting a discount and you PROBABLY wont see your radiators get that hot anyway)

And the next update will throttle the cooling down a little so it doesn't murder your power distribution. IF I can get the damned throttle code working right.

Edited May 27, 2018 by Starwaster

[siimav]

Did one more test with a 30 ton nuclear stage. It was cheated into orbit and aligned with the sun so that most of the tank is covered by the antenna. I also activated the infinite electricity cheat just to test how low the boiloff would get. After 15 days in orbit, the boiloff was still not reduced to 0. Yes, 0.02l/s at 10000x timewarp is actually almost negligible boiloff. But the thing is that I'm seeing pretty much same numbers with the radiators deactivated. The 100 layer MLI tanks are just too good at reducing boiloff by themselves.

[Starwaster]

20 minutes ago, siimav said:

Did one more test with a 30 ton nuclear stage. It was cheated into orbit and aligned with the sun so that most of the tank is covered by the antenna. I also activated the infinite electricity cheat just to test how low the boiloff would get. After 15 days in orbit, the boiloff was still not reduced to 0. Yes, 0.02l/s at 10000x timewarp is actually almost negligible boiloff. But the thing is that I'm seeing pretty much same numbers with the radiators deactivated. The 100 layer MLI tanks are just too good at reducing boiloff by themselves.

That's interesting, your tank internal temperature is 20.2 - I'm going to assume that it's rounding up from 20.15.

And there's no outgoing internal heat flux and there probably should be because RF sets the target temp to 99% of the boiling point so that it drops it just below. 

Go into your RealFuels folder and edit MFSSettings.cfg file and look for radiatorMinTempMult. It should be set to 0.99

If it's not there then add it.

 

[siimav]

9 minutes ago, Starwaster said:

That's interesting, your tank internal temperature is 20.2 - I'm going to assume that it's rounding up from 20.15.

And there's no outgoing internal heat flux and there probably should be because RF sets the target temp to 99% of the boiling point so that it drops it just below. 

Go into your RealFuels folder and edit MFSSettings.cfg file and look for radiatorMinTempMult. It should be set to 0.99

If it's not there then add it.

 

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?

[Starwaster]

New update with better throttling when power is low. It does assume that all radiators on the craft are getting used so unused radiators can throw the throttling off.

https://github.com/Starwaster/RealActiveRadiators/releases/latest

Might be the last update for 1.3.1 and then it'll be on to KSP 1.4

@siimav that's correct and I'm not sure then why it's not cooling your tank down more than it is. That radiator is drawing a lot of heat from somewhere but it's not from the tank you have your LH2 in......

[siimav]

@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.

[Starwaster]

17 minutes ago, siimav said:

@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.

I must have pressed publish too quickly. It's there now

[Bornholio]

Testing Tomorrow. Looks great. My snoring cat agrees also. Between this and MLI happy nukes for everyone.

[siimav]

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?

Edited June 3, 2018 by siimav

[Starwaster]

19 hours ago, siimav said:

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've noticed this too with certain parts. It is probably a Deadly Reentry thing. DR scales the max temperatures down to realistic levels. Radiators want to keep things at or below 25% of their. For instance the attachment point parts get scaled down to 1000 K. 25% of 1000K is 250K - which is easily exceeded while landed anywhere on Kerbin's equatorial zone. I was testing a Real ISRU lander for power consumption when I noticed that a significant amount of power was being drained that couldn't be accounted for by cryotanks alone.

I'm not sure what the solution will be yet. Either the default target temp needs to be bumped up (which would probably be done by MM patch rather than at runtime) or I need to restrict cooling to parts with RF tanks only.

The part field in question btw is radiatorMax which defaults to 0.25

[siimav]

4 hours ago, Starwaster said:

I've noticed this too with certain parts. It is probably a Deadly Reentry thing. DR scales the max temperatures down to realistic levels. Radiators want to keep things at or below 25% of their. For instance the attachment point parts get scaled down to 1000 K. 25% of 1000K is 250K - which is easily exceeded while landed anywhere on Kerbin's equatorial zone. I was testing a Real ISRU lander for power consumption when I noticed that a significant amount of power was being drained that couldn't be accounted for by cryotanks alone.

I'm not sure what the solution will be yet. Either the default target temp needs to be bumped up (which would probably be done by MM patch rather than at runtime) or I need to restrict cooling to parts with RF tanks only.

The part field in question btw is radiatorMax which defaults to 0.25

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.

[Starwaster]

6 hours ago, siimav said:

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.

Actually ammonia's boiling point is higher at 237.65

[siimav]

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?

[Starwaster]

1 hour ago, siimav said:

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?

Replace the file RealActiveRadiators.cfg with this one https://raw.githubusercontent.com/Starwaster/RealActiveRadiators/master/GameData/RealActiveRadiators/RealActiveRadiators.cfg

The problem is almost certainly caused by my having scaled down the maxEnergyTransfer. If the numbers are looked at though, I have a feeling that even at 1/50th of its original value that it SHOULD Have been enough to cool those parts and that the problem is due to Squad's implementation of ModuleCoreHeat and how it scales heat transfer. I've seen a similar problem with a reactor I wrote a config for constantly shutting down even though the radiators have a cooling capacity of 50kW and the reactor only needs 13kW.

As I mentioned earlier, it's not the radiator code that removes core heat, the core code itself moves its own heat out to the radiator based on what the radiator says it can handle.

Also, FYI (I forgot this earlier) it's ok if the throttling debug field says 100%. It is always at 100% unless there is insufficient power for the heat pump. It's not a percentage of full power, it's a percentage of the amount that was requested.

[siimav]

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.

Spoiler

 

[Starwaster]

@siimav That's usually set to 0 by RO, are you just using RP1 without a full RO install? 

I'll just rip that section out of there; it's got to do with the stock radiator code and it's not really  needed now. In the meantime if you patch it and set it to 0 like RO does then that would fix it.