[1.12.x] Heat Control - More radiators! (August 22, 2021)

[Captain Sierra]

After sitting back for a while and letting things happen, I come back to read Streetwind's and Mikegarrison's comments, only to feel quite vindicated, and maybe just a little smug.

In seriousness, Mike is right. The simple answer is to regress to the 0.90 version of reactor heating. If the goal is radiators everywhere as you stated, could you just tweak that system to do globalize the reactor heat output and radiator heat rejection instead of checking each reactor and only parts attached to it? Instead just sum the heat rejection the reactors require, sum the heat rejection of all radiators in the moment (to account for retracted/deployed) and base the efficiency/power output calculations off that.

Alternatively stick with the 1.0.2 system since, as most of us in the thread agree, stock f'd you over. No point bending backwards for something that just destroyed your entire balancing paradigm. /rant

[Fraz86]

That way you guarantee that the conformal set can cool the reactor in all cases... that do no involve going closer to the sun, because that's going to break the balance.

Have you actually tested this issue? I had the same concern, but when testing this patch, I HyperEdited my spacecraft into orbit of Moho, and the reactor steady state temperature only increased 2-3K.

You could also make reactors be fully isolated at all times. Make them just not participate in exchanging heat with the vessel. That way it's easier to balance for the conformal set, because it always guarantees conformal conditions.

I really think this is the way to go (well, not necessarily fully isolated, but very low heatConductivity). It's the only way, within this system, to achieve robust reactor/radiator relationships. I would argue that it doesn't entirely defeat the purpose of the heat simulation, as players will still need to consider issues of radiator saturation if they have any other heat-producing parts on the spacecraft.

Fraz made some comments before about ignoring sunk costs, but I think he's not really doing that himself. There are a lot of things we don't model, like ionizing radiation, nuclear waste dumps, Kerbal "no-nukes" protestors trying to block the launches, etc. Why are we trying to model reactor heating the way we are doing?

What sunk costs am I holding on to? Reactor heat itself? As my recent patch demonstrates, it's actually really easy to balance reactor heat if we take Nertea's radiators out of the picture.

---

Nertea,

For you consideration, I have outlined one potential approach to working through the reactor heat/radiator issues:

Step 1 - Update NFE for 1.04, without conformal radiators. This involves (i) tuning reactor properties to be well-matched with stock radiators (as I did in my example patch), and (ii) solving the ≥1000x time warp exploding ship issue (e.g., by disallowing ≥1000x time warp with active reactor, or turning off reactor heat production at ≥1000x time warp).

Step 2 - Update HeatControl for 1.04, as a parts-only pack (which would now include conformal radiators, and should probably exclude heat pipes and exchangers). This would involve tuning the conformal radiators to work as matching sets with their corresponding reactors, which I still believe is possible, though it will be messy. Doing so will probably require tuning the emissivity values of each radiator individually to achieve the desired performance.

Step 3 - If/when motivated to do so, develop an optional "HeatControl+" add-on that basically restores your pre-1.03 radiator behavior and localized heat management, including heat pipes and exchangers.

Edited July 2, 2015 by Fraz86

[Nertea]

I wrote a long post but the forum ate it.

Summary: I really need to work out what's going on about over 100x warp. There may be actually some more issues than we think. I'll see what I can do later (next week probably) . Feel free to play around with more solutions in the meantime.

Thanks for all your ideas and work guys.

Edited July 2, 2015 by Nertea

[Streetwind]

Timewarp and heat don't really seem to get along all that well, yeah. In 1.0.4 even less so than before. A lot of people report massive issues when touching an atmosphere under timewarp, or using physics warp during reentry. And I noticed that the steady-state point for a reactor-radiator pairing is different at 50x warp than it is at normal time.

I'm currently sort of leaning in Fraz's direction about insulating reactors, mostly because glowing hot crew cabins break my immersion. Alternatively the insulator parts from Heat Control can help with that. Some people may not be bothered as much as I am, so maybe solving the issue with additional parts is acceptable. Also returns a little bit of purpose to Heat Control as a mod.

For pure part balancing, fully isolating reactors isn't required. It all depends on how much you want to push the conformal radiator thing. From how I see it, even with full reactor isolation you'll have to run with completely inconsistent numbers between the three sizes to make it work with the existing radiators.

Which leaves one more option: new radiators. If you made new models specifically targeting the required surface amounts, you could make conformal sets that work consistently. Of course, that means extra work for you... =/

[Fraz86]

I have a new & improved version of my NFE patch. This version brings back conformal radiators, tweaking their emissivities such that a "matching set" will achieve thermal equilibrium just below the reactor's nominal temp. Relative to a stock radiators of equivalent mass, conformal radiators are now twice as expensive and 5-10K warmer at steady state, in order to balance their much lower surface area. Conformal radiators' maxEnergyTransfer was adjusted to match stock (per unit mass). This update also adjusts reactor costs to offset the decreased cost of radiators.

Basically, I believe this patch addresses the issues with the current NFE test release, aside from the time warp problem.

@PART[*]:HAS[@MODULE[FissionReactor]]:Final
{
	%heatConductivity = 0.00001
	%skinInternalConductionMult = 0.00001
}
@PART[reactor-0625]:Final
{
	@description ^= :$: The manufacturer recommends 2x Thermal Control System (small) for heat dissipation.:
	@mass -= 0.02 //offsets +0.02 change in recommended radiator mass (4x 0.02 vs 2x 0.05)
	@cost += 970 //offsets -970 change in recommended radiator cost (4x 467.5 vs 2x 450)
	@MODULE[FissionReactor]
	{
		@NominalTemperature = 800 //steady state ~780K with recommended radiators
	}
}
@PART[reactor-125]:Final
{
	@description ^= :$: The manufacturer recommends 2x Thermal Control System (medium) for heat dissipation.:
	@mass -= 0.02 //offsets +0.02 change in recommended radiator mass (6x 0.08 vs 2x 0.25)
	@cost += 6975 //offsets -6975 change in recommended radiator cost (6x 1912.5 vs 2x 2250)
	//steady state ~770K with recommended radiators; no need to modify nominal temp of 800K
}
@PART[reactor-25]:Final
{
	@description ^= :$: The manufacturer recommends 2x Thermal Control System (large) or 8x Thermal Control System (medium) for heat dissipation.:
    //no change in recommended radiator mass (4x 0.5 vs 2x 1.0 vs 8x 0.25)
	@cost += 27900 //offsets -27900 change in recommended radiator cost (4x 11475 vs 2x 9000 vs 8x 2250)
	@MODULE[FissionReactor]
	{
		@NominalTemperature = 800 //steady state ~770-780K with recommended radiators
	}
}
@PART[reactor-25-2]:Final
{
	@description ^= :$: The manufacturer recommends 2x Thermal Control System (large) or 8x Thermal Control System (medium) for heat dissipation.:
	@cost += 27900 //offsets decrease in radiator costs
	@MODULE[FissionReactor]
	{
		@NominalTemperature = 850 //steady state ~825K with recommended radiators
	}
}
@PART[reactor-375]:Final
{
	@description ^= :$: The manufacturer recommends 6x Thermal Control System (medium) for heat dissipation.:
	@cost += 20925 //offsets decrease in radiator costs
	@MODULE[FissionReactor]
	{
		@NominalTemperature = 825 //steady state ~800K with recommended radiators
	}
}
@PART[radiator-conformal-1]:Final
{
	@mass = 0.025
	@cost = 450 //twice as costly as stock radiators for given performance, in exchange for higher performance for given surface area
	@emissiveConstant = 2.4 //steady state ~785K with M-EXP
	@MODULE[ModuleActiveRadiator]
	{
		@maxEnergyTransfer = 1250 //follows the "mass x 50000" algorithm of stock radiators
	}
}
@PART[radiator-conformal-2]:Final
{
	@mass = 0.085
	@cost = 1530 //twice as costly as stock radiators for given performance, in exchange for higher performance for given surface area
	@emissiveConstant = 5.2 //steady state ~785K with MX-4
	@MODULE[ModuleActiveRadiator]
	{
		@maxEnergyTransfer = 4250 //follows the "mass x 50000" algorithm of stock radiators
	}
}
@PART[radiator-conformal-3]:Final
{
	@cost = 9000 //twice as costly as stock radiators for given performance, in exchange for higher performance for given surface area
	@emissiveConstant = 2.25 //steady state ~785K with MX-1
	@MODULE[ModuleActiveRadiator]
	{
		@maxEnergyTransfer = 25000 //follows the "mass x 50000" algorithm of stock radiators
	}
}

[Psycho_zs]

Conformal radiators could be generalized to diameters. GR-1 could be reduced in width (it clips through neighbors' corners when mounted in fours on 2.5m).

From the other end, reactors could be slightly remodelled to replace radiator sockets with more generic rings where radiators could be placed anywhere. This way it will be easier to tune radiators, including conformal to reactors by their quantity.

Considering stock radiator wide bases, HC radiators could also swap pointy plugs for a wider bases.

[Streetwind]

Re: proposed changes:

Not sure it's worth folding the radiator cost into the reactors, because the reactors have always been on the extremely expensive end already, and a bit less cost might do them good. The important stuff is the mass, which you seem to have taken care of. Are the radiators competitive/useful to have in scenarios other than conformal sets? (Can't test right now, am at work)

With a config like that we have to say goodbye to beautiful emissives though

(Also, argh! Must look away... from inconsistent... numbers... )

[includao]

Forgive me if I'm being stupid, but wouldn't it be good to follow what Interstellar extended mod did with WasteHeat generation or maybe integrate these reactors with their reactors? It seems Interstellar Extended mod is working fine with the current heat changes.

[Streetwind]

Isn't Interstellar running on a completely custom heat system? I know it used to, but I must admit I haven't looked at it in a while.

The point is, Near Future and Heat Control intentionally do not run a custom simulation. They are built with the intent of supporting the stock system. There are only some issues because the definition of "the stock system" keeps changing

[includao]

They use a custom resource called WasteHeat that is captured by a Generator with a certain x% efficiency to generate electricity, so they avoid the stock system. Maybe it is not realistic, but at least it is functional so they've avoided these problems.

[Captain Sierra]

They use a custom resource called WasteHeat that is captured by a Generator with a certain x% efficiency to generate electricity, so they avoid the stock system. Maybe it is not realistic, but at least it is functional so they've avoided these problems.

Part of what makes this mod (or at least used to, its arguable this aspect has been lost due to stock's choice of direction) is that not only is it useful for cooling reactors, its also viable for dealing with LV-Ns and other heat-producing components and better managing your vessel's heat buildup. Converting to a second system like KSPI removes that element, however its not off the stable right now. Mikegarrison suggested reverting to the 0.90 version of the system, which defines a custom heat system (because a stock one didnt exist) that was heavily arbitrary. Heat is handwaved as a few numbers the mod tracks (not unlike the black magic of stock) but unlike KSPI, its not a resource.

TLDR: before stock heating, the reactors did run on a custom arbitrary system. They were converted over for a variety of reasons (which I do not remember all of so Nertea will have to comment) without too much difficulty. The addition of stock radiators which lack a key feature (refrigeration) means we're now in a major balancing bind converting to stock again.

[Fraz86]

Are the radiators competitive/useful to have in scenarios other than conformal sets?

Yes, if you need smaller radiators. Performance per unit mass is almost exactly the same for conformal vs stock radiators, with just a tiny advantage to stock that won't usually matter in-game. Conformal radiators are twice as expensive as stock per unit mass (though actually a bit less expensive than pre-1.03), in exchange for much smaller surface area & volume.

(Also, argh! Must look away... from inconsistent... numbers... )

Indeed. However, I believe this is the lesser evil of the available options. And really, I'm mostly just compensating for the inconsistency introduced by KSP's inaccurate surface area calculations. This should actually result in better consistency from a gameplay perspective.

[Nertea]

I'll have a look at your changes tonight Fraz.

Have you tried playing with radiatorMax at all?

[Fraz86]

Have you tried playing with radiatorMax at all?

If my understanding of radiatorMax is correct, I think the default value (0.25) should be fine. No real advantage to "activating" radiators at a lower temperature. I suppose we could use a slightly higher value (e.g., 0.35 like LV-Ns) so reactors don't needlessly activate radiators while operating at low outputs, but this would only matter in very specific situations.

Edited July 2, 2015 by Fraz86

[Psycho_zs]

KSPI uses completely different model. I would advice against it.

Both old HC and stock radiators operating within the same thermal framework. The only difference introduced in 1.0.3 is skin. How radiators manipulate the heat is in the governance of radiators themselves. This can be another argument for preserving old HC mechanics. In a way it does not drift so far away from stock, it does not even have to be a revamp.

Throwing another idea for old HC:

The advantage of old HC mechanics is, again, heat pump. Consider taking this mechanics a bit closer to reality by making heat transfer dependent on temperature difference between radiator and parent part (like in real fridge between hot and cold ends of the circuit). This would allow tweaking radiator performance by varying maximum temp difference it's pump can achieve. So the area of radiator would not be the only vital characteristic.

[Fraz86]

This would allow tweaking radiator performance by varying maximum temp difference it's pump can achieve. So the area of radiator would not be the only vital characteristic.

My patch's tweaking of radiator emissivity is basically intended to simulate this mechanic, in an abstracted manner. That's how it's possible to balance conformal radiators with stock, despite their very different surface areas.

Nertea, for the record, even with good balance, I still recommend moving conformal radiators to a separate parts pack, so that players can choose not to complicate their radiator options if they prefer to stick with stock.

Edited July 2, 2015 by Fraz86

[Nertea]

Conformal radiators could be generalized to diameters. GR-1 could be reduced in width (it clips through neighbors' corners when mounted in fours on 2.5m).

From the other end, reactors could be slightly remodelled to replace radiator sockets with more generic rings where radiators could be placed anywhere. This way it will be easier to tune radiators, including conformal to reactors by their quantity.

Considering stock radiator wide bases, HC radiators could also swap pointy plugs for a wider bases.

Ah, no. Well not right now. I really like my models and crap, and don't have any desire to do a redo on them, not for a while.

Forgive me if I'm being stupid, but wouldn't it be good to follow what Interstellar extended mod did with WasteHeat generation or maybe integrate these reactors with their reactors? It seems Interstellar Extended mod is working fine with the current heat changes.

Nah. Interstellar always used its own heat system which is really not coming here.

KSPI uses completely different model. I would advice against it.

Both old HC and stock radiators operating within the same thermal framework. The only difference introduced in 1.0.3 is skin. How radiators manipulate the heat is in the governance of radiators themselves. This can be another argument for preserving old HC mechanics. In a way it does not drift so far away from stock, it does not even have to be a revamp.

Throwing another idea for old HC:

The advantage of old HC mechanics is, again, heat pump. Consider taking this mechanics a bit closer to reality by making heat transfer dependent on temperature difference between radiator and parent part (like in real fridge between hot and cold ends of the circuit). This would allow tweaking radiator performance by varying maximum temp difference it's pump can achieve. So the area of radiator would not be the only vital characteristic.

The thing is, this all requires more code, and requires what is essentially a parts mod to change stock gameplay. Which I don't want to do . Generally anything that I add should not modify stock parts except in a particularly beneficial or optional way.

[Nertea]

New test release

Removes conformal radiators, tweaks reactor temperatures and masses a bit. Also stops reactor heat production at > 100X time warp. See if that helps (I have a worry here).

Is there some effective way to quantify radiation capacity of stock radiators at optimal temperatures? It would be good to place a descriptive field on them (for HC) for better intercomparison between parts. I can add a part description field easily enough.

[Wolfox]

Is there some effective way to quantify radiation capacity of stock radiators at optimal temperatures? It would be good to place a descriptive field on them (for HC) for better intercomparison between parts. I can add a part description field easily enough.

I doubt you didn´t read (or didn´t know) this yet, but just in case, THIS might help :-)

I am not fully aware how exactly heat transfer between parts works, but if it does how I think it does, the ammount of heat should be calculable from skin size and radiation effectiveness of each radiator. If this presumption is right, then only distinctive factor should be effective radiation surface (assuming that yours and stock have the same heat dissipation per square meter properties). I believe there is no connection between other properties than part temperature and ammount of heat radiated by radiator connected to it. Correct me if I am wrong please.

PS: Shouldn´t the radiator be painted black? (for highest radiation factor)

EDIT: However this does not solve comparison between heat generated by your generators and needed radiator power, like your NF EC mod does (and it does it very well).

Edited July 3, 2015 by Wolfox

[Nertea]

Oh trust me, I know .

It's less the theoretical BB radiation and more the effective heat removal ability. They don't need to be painted black either. A common misconception - it's perfectly possible to have high (>0.95) emissivities with bright coloured materials, as we're generally talking about emissivity in the thermal region of the spectrum while we see in the visible spectrum. Those only overlap significantly at very high temperatures.

[Wolfox]

Oh trust me, I know .

It's less the theoretical BB radiation and more the effective heat removal ability. They don't need to be painted black either. A common misconception - it's perfectly possible to have high (>0.95) emissivities with bright coloured materials, as we're generally talking about emissivity in the thermal region of the spectrum while we see in the visible spectrum. Those only overlap significantly at very high temperatures.

Then you might want to quantify radiator effectivity as (very crude, I know) capability to remove amount of heat per unit of time. That way you could recalculate both stock and mod radiators for easy comparison, and also to determine heat produced by NF EC generators (to preserve easy calculation).

Edited July 3, 2015 by Wolfox

[Fraz86]

New test release

Removes conformal radiators, tweaks reactor temperatures and masses a bit. Also stops reactor heat production at > 100X time warp. See if that helps (I have a worry here).

Looks good! Here are my initial observations:

• The >100X time warp temperature jump no longer occurs in my tests. Now the temperature drops dramatically at >100X time warp. This is technically exploitable - you could build a ship with no radiators and lots of batteries, start the reactor at 0% output, go to 1000X time warp, adjust output to 100%, fill batteries, decrease output back to 0%, turn off time warp, then use the stored electricity to power engines or whatever. Thankfully, it's a bit difficult to imagine when this would really be useful, considering that - if you're going to 1000X time warp to fill batteries, you might as well just use solar panels or RTGs.
  
• It looks like my wishlist of proposed changes was almost entirely implemented, so I don't have much else to add. The only exceptions are MX-4 mass (which remains at 1.22 instead of 1.2) and reactor costs. I agree with Steetwind's assertion that, given the expensiveness of reactors, it seems reasonable to leave them alone. The only additional suggestion I can think of is, I suppose it wouldn't hurt to increase radiatorMax of reactors to 0.35 like LV-Ns (though, as I said before, this will only matter in very specific circumstances).
  

Is there some effective way to quantify radiation capacity of stock radiators at optimal temperatures? It would be good to place a descriptive field on them (for HC) for better intercomparison between parts. I can add a part description field easily enough.

Given that heat radiation varies significantly depending on the target temperature (e.g., ~775 for M-EXP, MX-4, & MX-1, ~800 for MX-2, ~ 825 for MX-L, and potentially much higher for nuclear, ion, and plasma engines), I worry that including specific heat radiation capacity information will end up being more confusing than helpful to the average player. Perhaps we could simply describe radiation capacity in terms of an arbitrary unit where 1.0 is defined as the capacity of the Thermal Control System (small). Thus, the TCS (med) is 5, TCS (large) is 20, GR-EXP is 0.5, GR-4 is 1 ²/₃, and GR-1 is 10. Multiply everything by 6 if you want only whole numbers.

Edited July 4, 2015 by Fraz86

[smjjames]

Oh trust me, I know .

It's less the theoretical BB radiation and more the effective heat removal ability. They don't need to be painted black either. A common misconception - it's perfectly possible to have high (>0.95) emissivities with bright coloured materials, as we're generally talking about emissivity in the thermal region of the spectrum while we see in the visible spectrum. Those only overlap significantly at very high temperatures.

Are you merging HC back into NFE or wheres the latest test version of HC?

[SSMI]

I must be doing something wrong because this mod isn't making much sense to me. I am running 1.0.4 with the released version of this mod. I also have Hot Spot (https://kerbalstuff.com/mod/937/Hot%20Spot) installed.

I am trying to run an LV-N "Nerv" Motor. I have a RBMK-1000 heat exchanger above it. I then have 8xXR-1100 attached to that. Above that there are various hydrogen tanks from near future technologies (1xHM-M-10 and 4xHB-30). As I run the engine it stays perfectly cool which makes sense to me. The radiators however overheat and explode much faster than I would expect. Beyond that they also don't radiate like it seems they should. The hydrogen tanks radiate (Thermal Rate ®) more energy than they do hot. They never get above about 10kW. The tanks get to something like 100kW.

It is even more confusing because in VAB it says they will get 1100kW.

What is going on here?

[Nertea]

I must be doing something wrong because this mod isn't making much sense to me. I am running 1.0.4 with the released version of this mod. I also have Hot Spot (https://kerbalstuff.com/mod/937/Hot%20Spot) installed.

I am trying to run an LV-N "Nerv" Motor. I have a RBMK-1000 heat exchanger above it. I then have 8xXR-1100 attached to that. Above that there are various hydrogen tanks from near future technologies (1xHM-M-10 and 4xHB-30). As I run the engine it stays perfectly cool which makes sense to me. The radiators however overheat and explode much faster than I would expect. Beyond that they also don't radiate like it seems they should. The hydrogen tanks radiate (Thermal Rate ®) more energy than they do hot. They never get above about 10kW. The tanks get to something like 100kW.

It is even more confusing because in VAB it says they will get 1100kW.

What is going on here?

Well, if it says 1.02 on the can and you're using 1.04, I'd wager that it has something to do with things.

Looks good! Here are my initial observations:

• The >100X time warp temperature jump no longer occurs in my tests. Now the temperature drops dramatically at >100X time warp. This is technically exploitable - you could build a ship with no radiators and lots of batteries, start the reactor at 0% output, go to 1000X time warp, adjust output to 100%, fill batteries, decrease output back to 0%, turn off time warp, then use the stored electricity to power engines or whatever. Thankfully, it's a bit difficult to imagine when this would really be useful, considering that - if you're going to 1000X time warp to fill batteries, you might as well just use solar panels or RTGs.
  
• It looks like my wishlist of proposed changes was almost entirely implemented, so I don't have much else to add. The only exceptions are MX-4 mass (which remains at 1.22 instead of 1.2) and reactor costs. I agree with Steetwind's assertion that, given the expensiveness of reactors, it seems reasonable to leave them alone. The only additional suggestion I can think of is, I suppose it wouldn't hurt to increase radiatorMax of reactors to 0.35 like LV-Ns (though, as I said before, this will only matter in very specific circumstances).
  

Given that heat radiation varies significantly depending on the target temperature (e.g., ~775 for M-EXP, MX-4, & MX-1, ~800 for MX-2, ~ 825 for MX-L, and potentially much higher for nuclear, ion, and plasma engines), I worry that including specific heat radiation capacity information will end up being more confusing than helpful to the average player. Perhaps we could simply describe radiation capacity in terms of an arbitrary unit where 1.0 is defined as the capacity of the Thermal Control System (small). Thus, the TCS (med) is 5, TCS (large) is 20, GR-EXP is 0.5, GR-4 is 1 ²/₃, and GR-1 is 10. Multiply everything by 6 if you want only whole numbers.

Alright, that's good that the 100x warp stuff works. I was occasionally seeing some weird issues in testing. Let me know if you see any more temperature jumps.

And, we'll probably need some kind of rating type thing instead of a number. I think it's necessary to have some kind of metric, when there are many radiators that can be chosen.

Are you merging HC back into NFE or wheres the latest test version of HC?

There is no test version of HC, that will come after NFE is locked off.