KSP 1.2.2 - Radiators do not work correctly in atmospheres

[foamyesque]

Just now, sal_vager said:

None of that actually matters though, as I said the radiators stop cooling parts when the outside air is hotter than the part, that's why you're seeing the effect that you are.

 

18 hours ago, foamyesque said:

That's physically stupid behaviour, then. It should only disengage if the radiator is being heated beyond its operational temperatures.

 

I think we've been round this before? The fact that that is what is happening doesn't address the point that it should not be happening. All you're doing right now is telling me that it's intentionally wrong.

[sal_vager]

Just now, foamyesque said:

I think we've been round this before? The fact that that is what is happening doesn't address the point that it should not be happening. All you're doing right now is telling me that it's intentionally wrong.

How do you know that radiators would work at all under these conditions? There has to be a reason why the Shuttle didn't use its radiators until it was safely in space, and the Blackbird didn't use radiators at all.

If radiators were viable why were they not used?

[foamyesque]

Just now, sal_vager said:

How do you know that radiators would work at all under these conditions? There has to be a reason why the Shuttle didn't use its radiators until it was safely in space, and the Blackbird didn't use radiators at all.

If radiators were viable why were they not used?

The Shuttle was still in atmosphere; so is the ISS. They both have functional radiators, no? And the Blackbird was, in fact, black. Y'know why it was black? To increase its radiant cooling. Moreover the cockpit had, in fact and of course, a cooling system, because otherwise you'd have a roasted pilot. The similarities to why I want one are remarkable.

[sal_vager]

2 minutes ago, foamyesque said:

The Shuttle was still in atmosphere; so is the ISS. They both have functional radiators, no? And the Blackbird was, in fact, black. Y'know why it was black? To increase its radiant cooling. Moreover the cockpit had, in fact and of course, a cooling system, because otherwise you'd have a roasted pilot. The similarities to why I want one are remarkable.

Barely in atmosphere, but as I said already Kerbins atmospheres isn't like Earths, but that doesn't answer my question and I'm genuinely curious why the Blackbird and Shuttle didn't have radiators on the exterior of the craft, the Shuttle kept them safely within the bay, the Blackbirds paintjob was also for night time camoflage, but even though it helped with radiating from the airframes skin it also wasn't an actively fed heat exchanger.

The subject of heat, especially at such high speeds where the air around you is being compressed, is a complex one, and it'd be wrong to make radiators that can magically work in conditions where they could not in real life.

I'm looking up this subject to see why they should/should not work, as far as I can tell they wouldn't do very much as they'd be saturated long before, so your cockpit would still start to overheat as the heat wouldn't be removed from the cockpit faster than it is dumped into the cockpit by surrounding conditions, and the radiators ability to lose its heat will diminish as the air thins, and radiation becomes the primary means of heat loss.

[Starwaster]

23 minutes ago, foamyesque said:

 

 

I think we've been round this before? The fact that that is what is happening doesn't address the point that it should not be happening. All you're doing right now is telling me that it's intentionally wrong.

If it were not this way then radiators would be so OP as to render heat shields unnecessary. That's fact not speculation. They used to behave that way before and had to be changed

I dont agree with the solution because external conditions should only affect how radiators shed heat not whether it can be moved around internally. If that were true irl then refrigerators, which move heat from a cooler part to a hotter one, would not be possible. They do that even if external temps are hotter than internal. 

[sal_vager]

1 hour ago, Starwaster said:

If it were not this way then radiators would be so OP as to render heat shields unnecessary. That's fact not speculation. They used to behave that way before and had to be changed

I dont agree with the solution because external conditions should only affect how radiators shed heat not whether it can be moved around internally. If that were true irl then refrigerators, which move heat from a cooler part to a hotter one, would not be possible. They do that even if external temps are hotter than internal. 

Actually that's not what I'm finding, this effect was in the game at least as far back as 1.0.5, I can install 1.0.0 and try it too if you like, but from what I see they have always done this, it's not new.

They can be made to work constantly, it's a game, they can be made to spawn an endless stream of Kerbals when they get hot if we really wanted, but the question is if they were used in real life would they really do what you are wanting them to do, or is there some reason why using radiators like this to cool a spaceplace would be a waste of time?

If it turns out that they should not be expected to cool your craft there's no sense having them do so, except for just gameplay reasons, but people keep telling me they want realism.

[foamyesque]

4 minutes ago, sal_vager said:

I'm looking up this subject to see why they should/should not work, as far as I can tell they wouldn't do very much as they'd be saturated long before, so your cockpit would still start to overheat as the heat wouldn't be removed from the cockpit faster than it is dumped into the cockpit by surrounding conditions, and the radiators ability to lose its heat will diminish as the air thins, and radiation becomes the primary means of heat loss.

It's a question of thermal balance and radiator operating temps.

So, assume a radiator in thermal equilibrium. By definition, in thermal equilibrium, incoming thermal energy must match outgoing thermal energy. This means, then, that Conductive In + Convective In + Radiant In = Conductive Out + Convective Out + Radiant Out. Conductivity In - Conductivity Out is the domain of heat pumps; it will generally be positive. Radiant In - Radiant Out depends on the balance of solar incident flux and radiator temperatures, but it'll be strongly negative for any temperatures we're interested in in Kerbin orbit. The question is then Convection In - Convection Out. This depends on the temperature differences between the radiator and the surrounding atmosphere and is the only one that is directly affected by the external temperature. However it is also affected by the external atmospheric density (as it depends on the air to actually carry heat away). So there's two competing factors at play, but at nil convection flux -- which the tooltips tell me is what KSP is calculating to be the case at high altitude, and which is correct -- that drops out.

Part skin temperatures in KSP are governed, primarily, by the balance between net convective flux and net radiant flux (excepting flux to the interior, but if the interior is thermally soaked to effectively isothermal as in the screenshots that's negligible). What that means is that you can tell whether a radiator would be able to actively cool a ship, at a glance, by seeing if its skin temperature is below that of it's headroom temperature. If it is, it has unused cooling capacity.

KSP's radiators are already magic, too. It's difficult to find a working fluid that'll handle 1800K+ and it's also very, very difficult to spread it finely enough that it can cool enough to absorb the full amount of heat from whatever it's trying to cool before being passed back to it. Radiators flown in space are, AFAIK, exclusively low-temperature ones designed to operate at around 280-300K. High temperature concept ones hit perhaps 900K, and are meant for cooling high-temperature things (e.g. a nuclear engine or nuclear reactors). Meanwhile, KSP's radiators are treated as black bodies with extraordinarily powerful heat pumps that, themselves, generate no heat, which also goes some way towards answering your question about "what about RL".

(Also, given the Blackbird's cockpit was actively cooled, one could then regard the Blackbird's skin as one giant radiator-- after all, heat's gotta go somewhere.)

 

 

 

[sal_vager]

Well from what I can see @foamyesque it's a deliberate effect, as @Starwaster says it's coded into the radiators, I couldn't tell you why though, he's probably right that this is a gameplay choice, otherwise we'd see stegosaurus spaceplanes with radiators all the way down their backs, hardly authentic.

Looks like the Blackbird just never got hot enough to need active cooling of anything other than the cockpit, it was able to handle the heat so additional cooling wasn't used as it would add mass and complexity.

[foamyesque]

Just now, sal_vager said:

Well from what I can see @foamyesque it's a deliberate effect, as @Starwaster says it's coded into the radiators, I couldn't tell you why though, he's probably right that this is a gameplay choice, otherwise we'd see stegosaurus spaceplanes with radiators all the way down their backs, hardly authentic.

Heh. As I said, I think that's because KSP radiators are ridiculously overpowered for what they're supposed to do. If they had significantly reduced headrooms that'd be solved, but that'd require some reworking of the thermal properties of the ISRU components (possibly needed in any case given some of the glitches people say exist in those).

[sal_vager]

8 minutes ago, foamyesque said:

Heh. As I said, I think that's because KSP radiators are ridiculously overpowered for what they're supposed to do. If they had significantly reduced headrooms that'd be solved, but that'd require some reworking of the thermal properties of the ISRU components (possibly needed in any case given some of the glitches people say exist in those).

Well the headroom is a cfg value, you're welcome to try tweaking it but I don't know if core heat will just overpower everything.

ISRU expects you to be on the ground most of the time by the way, not flying through the air with an external temperature of 6000k, or in low Sun orbit.

[foamyesque]

Just now, sal_vager said:

Well the headroom is a cfg value, you're welcome to try tweaking it but I don't know if core heat will just overpower everything.

 

Yeah. I would fiddle with them, but if radiators just sort of arbitrarily turn themselves off there's really not much point, and altering that is not within my own capabilities.

[sal_vager]

1 minute ago, foamyesque said:

 

Yeah. I would fiddle with them, but if radiators just sort of arbitrarily turn themselves off there's really not much point, and altering that is not within my own capabilities.

Well that's just code, but I can't change that without good reason, but if there were already better balanced values for the cfg so they wouldn't be OP it might help.

Won't stop radiator spam of spaceplanes though.

[foamyesque]

32 minutes ago, sal_vager said:

Well that's just code, but I can't change that without good reason, but if there were already better balanced values for the cfg so they wouldn't be OP it might help.

Won't stop radiator spam of spaceplanes though.

 

Sure it will. If they have a max-headroom of 1200K, say, and cease to be active radiators past that (and, perhaps, have the heat pumps operate on a declining curve of some sort?), they'd provide no more cooling than a wing or opened cargo bay would for the not only the peak heat loads but a very significant chunk to either side. This'd be both a physically consistent solution and should allow them to prevent accumulated heat from causing problems once the incoming heat drops. This is particularly relevant for parts which have differing internal/skin temperature tolerances, but it's also true if you wanted to, e.g., cool RAPIERs or nuclear engines in flight.

 

A heavy duty rebalancing of the thermal system might see more parts with split tolerances, much lower internal tolerances for things like pods and non-structural parts, integral heat pumps for cockpits, a precooler that actually does something, tweaks to skin thermal capacities, etc, to make heat management more complex than "add a couple radiators to an ISRU", but that's a much bigger project.

 

Edited December 11, 2016 by foamyesque

[sal_vager]

3 minutes ago, foamyesque said:

 

Sure it will. If they have a max-headroom of 1200K, say, and cease to be active radiators past that (and, perhaps, have the heat pumps operate on a declining curve of some sort?), they'd provide no more cooling than a wing or opened cargo bay would for the not only the peak heat loads but a very significant chunk to either side. This'd be both a physically consistent solution and should allow them to prevent accumulated heat from causing problems once the incoming heat drops. This is particularly relevant for parts which have differing internal/skin temperature tolerances.

But doesn't adding more radiators increase the effective radiative surface of the part, providing 1200k of cooling per radiator on top of the parts own surface area?

[Starwaster]

As you say,  

2 minutes ago, foamyesque said:

 

Sure it will. If they have a max-headroom of 1200K, say, and cease to be active radiators past that (and, perhaps, have the heat pumps operate on a declining curve of some sort?), they'd provide no more cooling than a wing or opened cargo bay would for the not only the peak heat loads but a very significant chunk to either side. This'd be both a physically consistent solution and should allow them to prevent accumulated heat from causing problems once the incoming heat drops. This is particularly relevant for parts which have differing internal/skin temperature tolerances.

 

A heavy duty rebalancing of the thermal system might see more parts with split tolerances, integral heat pumps for cockpits, a precooler that actually does something, tweaks to skin thermal capacities, etc, to make heat management more complex than "add a couple radiators to an ISRU", but that's a much bigger project.

 

It would also help if they failed at a much lower temperature. Their current failure point is 2700 K. That's close to (but not quite at) Apollo's heat shield surface temperatures. Which is fine for an ablative shield that's getting reduced to char and shedding its heat through ablation. To keep radiators from being so OP they should have lower maxTemp and they should lose the thermalMassModifier as well. The combination turns them into heat sinks and not just radiators.

Really, an across the board rebalancing of maxTemps needs to be done. Maybe not to the same level of Deadly Reentry or Realism Overhaul. Baby steps after all, but maxTemps really need to come down a bit for a lot of parts.

[foamyesque]

2 minutes ago, sal_vager said:

But doesn't adding more radiators increase the effective radiative surface of the part, providing 1200k of cooling per radiator on top of the parts own surface area?

Sure, but with no heat pumps engaged if the headroom's hit or exceeded, it'd be just the same as increasing your radiation area by adding a structural sheet (or opening cargo bays, or whatever). I suppose you could try placing them in locations with low skin temperatures and get them to run a bit longer, but the fixed panels have a limited range and the TCS ones are fragile.

[sal_vager]

If by fail you mean explode they should do that at 2500, they all have the same max temperature..

maxTemp = 2500

Also maxTemps were raised due to player feedback.

3 minutes ago, foamyesque said:

Sure, but with no heat pumps engaged if the headroom's hit or exceeded, it'd be just the same as increasing your radiation area by adding a structural sheet (or opening cargo bays, or whatever). I suppose you could try placing them in locations with low skin temperatures and get them to run a bit longer, but the fixed panels have a limited range and the TCS ones are fragile.

From my testing today I see that they do seem to reduce the heat they take from the part once they are saturated.

[foamyesque]

5 minutes ago, sal_vager said:

From my testing today I see that they do seem to reduce the heat they take from the part once they are saturated.

I know; you've said so in this thread before too, I believe. My point is that if you retain that behaviour, which is a correct one, and eliminate the arbitrary external-temperature-turn-off, which is not, and you then reduce the headroom, you should wind up with radiators that don't need a special case, still work as radiators in the other circumstances you'd want them to, and aren't spammable in the way you're concerned about.

[sal_vager]

3 minutes ago, foamyesque said:

I know; you've said so in this thread before too, I believe. My point is that if you retain that behaviour, which is a correct one, and eliminate the arbitrary external-temperature-turn-off, which is not, and you then reduce the headroom, you should wind up with radiators that don't need a special case, still work as radiators in the other circumstances you'd want them to, and aren't spammable in the way you're concerned about.

You'd need more for ISRU, but I don't see why that would be a problem.

Would you be willing to write this up as a feedback issue on the tracker? I can then more easily bring it to the devs attention.

[foamyesque]

Just now, sal_vager said:

You'd need more for ISRU, but I don't see why that would be a problem.

Would you be willing to write this up as a feedback issue on the tracker? I can then more easily bring it to the devs attention.

 

I would be, but where is it? I see the one for the 1.2 pre-release...

[sal_vager]

21 minutes ago, foamyesque said:

 

I would be, but where is it? I see the one for the 1.2 pre-release...

Here's the main public tracker for 1.2.2

[foamyesque]

16 minutes ago, sal_vager said:

Here's the main public tracker for 1.2.2

Hm, looks like I need a separate account. Wheeee.....

[sal_vager]

1 minute ago, foamyesque said:

Hm, looks like I need a separate account. Wheeee.....

Yeah, I assumed you had one actually.

Also I should apologise to you and starwaster for muddling this thread up so much, if I'd just looked at the code at the beginning I'd have known what was going on.

[foamyesque]

36 minutes ago, sal_vager said:

Yeah, I assumed you had one actually.

Also I should apologise to you and starwaster for muddling this thread up so much, if I'd just looked at the code at the beginning I'd have known what was going on.

Eh, no worries. Tracker issue is up: http://bugs.kerbalspaceprogram.com/issues/13412

[sal_vager]

On 10/12/2016 at 1:33 AM, LatiMacciato said:

possibly that is still connected tho..

...tried a stock game (only hyperedit), drill is not cooling, radiator works on Kerbin ground to cool perfectly, stops working while attached on an asteroid.
There MUST be something wrong with the cooling stuff inside KSP at all.

 

Could you upload that save somewhere for me please @LatiMacciato.