Thermal Control and atmosphere

[Jarin]

Okay, as I use SSTO spaceplanes more and more, this is becoming more of an issue. My current series of light shuttle fly a shallow enough ascent profile that their PE is around 40km by the time their AP hits 70. This means I spend a LOT of time in the upper reaches of the atmosphere. So here's the issue, heat sinks appear to basically not work at all as long as there's even the slightest hint of air around you. Compare screenshots below of just below and just above the 70km line.

Spoiler

The heat bar on the right is my cockpit, incidentally. If I'd taken the shot closer to 60km, you'd have seen a very nice cherry-red outline glow as it neared overheat, despite the TCS right behind it. Is there any way around this? Am I missing something about KSP thermal mechanics? There's other bodies with atmosphere, why are thermal control systems almost completely inoperable here? How are you supposed to cool things like mining/refining on Eve or Laythe?

[diomedea]

Heat is transferred through conduction, convection and radiation, both in reality and in KSP.

Radiation occurs from any hot body, the transferred heat Q proportional to Stefan-Boltzmann constant, the emissivity of the material, the area of radiative surface, and the temperature at the fourth power.

Convection is when a part is immersed in a fluid (like atmosphere), the transferred heat Q proportional to the heat transfer coefficient of the material, the surface, and the temperature difference with the fluid.

Conduction occurs by contact with other parts at different temperature.

Radiators work by having a large surface and high emissivity (and also by pumping heat so they're always very hot), but still may not be enough to control heat build-up if conduction/convection is high.

Now, please have a look at the AeroGUI panel with the pic here (http://imgur.com/ktMhQct), taken just below kerbin's atmosphere limit. Due to the high speed of the vessel, there's a serious difference from the static ambient temperature at that altitude (212.32°K) and the measured external temperature (5968.07°K). Thanks to the very low density of air at that altitude, convective flux isn't too high (5.48 kW with the cockpit) or the vessel would easily burn; but still you should notice it's a positive flux (while instead, radiative flux is negative, a net loss of heat; but at the same time, conductive flux from adjacent parts is very high). Just above the atmosphere limit, in KSP the air density is made = 0 and therefore the external temperature drops suddenly at the temperature of void (4°K). So you have that marked difference by being just above or below the atmosphere limit.

KSP thermal mechanics work, at least just as designed. As anybody else, you may need to build and launch vessels considering the above, so fitting enough of radiators and heat shields, but also avoiding to stay too long at very high speed in atmosphere.

[Jarin]

@diomedeaSo basically, just like SOI, atmosphere is an abstracted hard border. I'm having trouble actually forming any practical advice out of the science above, though. Help?

1. Does "positive flux" in the atmosphere mean that the radiator is doing nothing?
2. How does that affect my previous question about heat management for offworld mining and refining on Eve/Laythe? I've managed to explode refining-rig components on minmus before, let alone someplace that removes all my heat-management ability.

[diomedea]

@Jarin: yes, quite close to how KSP treats SOI, does with atmospheres too. Real-world accuracy was sacrificed to have a simplified model, much easier to compute (and to play with).

Please enable the visibility of thermal data with parts action menus (Mod-F12 to open the console, then Physics tab, then Thermal, then "Display Thermal Data in Action Menus", that will show at the bottom of the parts action menus (opened right-clicking on parts) data as the radiation, conduction and convection fluxes.

1. Radiators actually work, even when the net thermal flux is positive. For parts connected with radiators, conduction flux should be negative anytime the radiators pull more heat from the part, than the heat that part is receiving from others. For radiators themselves, the radiative flux would show how much they are radiating outside.

2. Mining actually creates a build-up of heat from the drills and converters. You have need of radiators actively pulling heat from those parts. Please know there's a sweet-spot of a specific temperature with converters, when their efficiency is the highest, so you only need to deplete heat as needed to not go beyond that temperature.

Hope the above is enough help, but please ask if you need more, if not me others will answer.

 

[Jarin]

That helps, thanks @diomedea. 

Separate heat question. I know Thermal Control Systems (the deployables) draw heat from an entire craft, but how far away do the non-deployable heat dispersers draw from? Do they have to be attached directly to a heat generating part? Or one step away, I assume, since our primary heat generators (drills, refineries, nukes) don't have surface-attach... but is it further than that?

[bewing]

One step for fixed radiators only -- the parent part plus anything attached to that. And the TCS systems do work in the atmosphere -- your pic clearly shows it working at a 32.3% load, which is a lot better than nothing.

[foamyesque]

1 hour ago, bewing said:

One step for fixed radiators only -- the parent part plus anything attached to that. And the TCS systems do work in the atmosphere -- your pic clearly shows it working at a 32.3% load, which is a lot better than nothing.

 

I've heard elsewhere it's two, but either way what I do is put a 2x2 plate in front of my ISRU, then stick a pair of large fixed radiators and three drills onto that via offset & rotate tools. Works great IME, and you can put the whole package in a -50 cargo bay to be carted about by planes.

[Magzimum]

In real life, there is a large difference in the design of cooling equipment for atmosphere or vacuum. Is this the same in KSP? I.e. is the radiator good for space, and the thermal control system good for atmospheric use? Or are both equally good for both situations?

Take for example the Radiator Panel (large) and the Thermal Control System (small):

-	Radiator panel (large)	Thermal control system (small)
Cost	450	450
Mass	0.05t	0.05t
Elec.	0.025 ⚡/s	0.025 ⚡/s
Diss. heat	10,000	2,500

If presented like this, then the only trade-off is the retractability of the TCS vs. the much higher heat dissipation of the radiator.

Note: I hope that these numbers are still correct - I pulled them out of the wiki (which isn't always accurate), not the game itself to which I have no access at this moment.

 

 

[diomedea]

@Magzimum: those numbers are still correct; also other values in the config for those parts are the same.

There is no intended difference due to the presence of atmosphere on the efficiency of those parts. The different sizes should allow for different vessel designs. Of course I would never keep a TCS open while moving in atmosphere (those have the windResistance stated at 2.5 in their config, the part will break with enough dynamic pressure), so have to thank the OP for running that experiment.

However, in the design considerations, the orientation of the radiating panels has to be included. Panels radiate heat outside, but also are very efficient at catching heat should they face a heat source (also in real world). As the TCS designs all have rotating panels (that will always face away from Sun), their efficiency will generally be higher than fixed radiators (unless the case radiators are kept facing away by keeping a vessel direction, or having them in shadow).

[Magzimum]

58 minutes ago, diomedea said:

@Magzimum: those numbers are still correct; also other values in the config for those parts are the same.

There is no intended difference due to the presence of atmosphere on the efficiency of those parts. The different sizes should allow for different vessel designs. Of course I would never keep a TCS open while moving in atmosphere (those have the windResistance stated at 2.5 in their config, the part will break with enough dynamic pressure), so have to thank the OP for running that experiment.

However, in the design considerations, the orientation of the radiating panels has to be included. Panels radiate heat outside, but also are very efficient at catching heat should they face a heat source (also in real world). As the TCS designs all have rotating panels (that will always face away from Sun), their efficiency will generally be higher than fixed radiators (unless the case radiators are kept facing away by keeping a vessel direction, or having them in shadow).

That's a really interesting comment - I did not know that the heat model was so sophisticated that it includes actual orientation of the part. I had never realized that radiators that face the sun don't work so well because they also absorb heat.

Can radiators also heat up each other? (Normally, if you'd place two radiators in a parallel setting, with the radiating surfaces facing each other, they would not work).

Can radiators heat up the ground below them? (Normally, if a radiator faces the ground, it would evenually heat up the ground, and then stop working unless the ground is able to conduct that heat away - and heat conductivity of soil is quite poor for dry ground).

For me, I understand heat transport phenomena quite well, but I'm mostly struggling to figure out what is in the game, and what it not.

[diomedea]

I know of no mechanic in KSP for radiation flux to be intercepted by other parts or a body surface. In real world that is certainly true, the solid angle another object has in relation to a radiator (so, figure a sphere centered on the radiator, with the other object projected on the sphere surface) would define the relative amount of radiation captured.

It shows you have good knowledge about thermal phenomena. However let me one consideration from what you wrote. Radiators in KSP work by irradiating energy, in a similar way to how radiators work on a space vessel. Radiators we commonly know on Earth also irradiate a bit, but mainly work by convection.

When a radiator is at the same temperature of the air around it, convection ceases to work (and that perfectly matches what you wrote).

Instead, radiation works with temperature at the fourth power (and surface of radiator, and Stefan-Boltzmann constant) and emissivity of the material. Because of difference in emissivity, is common to have a radiator at a lower temperature (but with emissivity = 1, as a black body) losing heat to another facing object with low emissivity (e.g. silver, emissivity = 0.02). Some devices in real world use this phenomenon to actually pull heat from one warm surface to another at even higher temperature. Heat pulling is simulated with TCS in KSP.

[AeroGav]

@Jarin I have more or less stopped using radiators on my craft.

There is possibly a use for them when for example coming back from Mun or Minmus with multiple aerobraking passes, you can get rid of the heat from the last pass before starting a new one.

In your search for a more efficient ascent they are probably creating more drag/performance loss than you are gaining.   

What I find helps

1. Use inline cockpits, and place them as far back as practical

2. Wings are very good at radiating heat, larger the better.   Fuel tanks make good heat sinks.  If possible, attach big S wings to the cockpit then use the offset tool to move them where needed.  The game still treats them as connected.

3. Engine nozzles are also good, but obviously nukes are a source of heat themselves.   Attaching a Rapier directly to the back of a mk1 cockpit is good.  With a mk2 this doesn't work as you need some kind of adapter or bicoupler in between to not have massive drag.

4. If you got engines on sponsons/nacelles either side of main fuselage,  then make the root part of each sponson/nacelle an engine pre-cooler, and attach these either side of the cockpit.    Pre-coolers are good radiators. Again, use the offset tool to slide them to where they are actually needed.

.

 

One other thing... lift.        A high lift design keeps you above 40km once speed exceeds mach 6.   This reduces heat a lot.

If you look at this video of me launching a mk1 with nukes and no oxidizer

I've linked to the bit where i'm getting into my speedrun.    You can see that at 1390 m/s and 16.5km, heat bars first appear.   I climb a bit, the heat bars drop off at 23.8km even though we accelerated to 1550m/s.   Unfortunately we've lost too much engine power going this high, so i head down again.   At 1615 m/s and 22km the heat bars come back with a vengeance, so I decide to bring the speedrun to an end.    At this point, you'll notice my angle of attack has been very low in the speedrun, less than 2.5 degrees to deliberately stop the plane climbing.   Once I light the nukes, I wind on a load of pitch trim to get us up to 5 or so where lift/drag ratio is best,  and we really start to zoom upstairs.  The heat bars soon disappear again.

[foamyesque]

AFAICT precoolers are just an ordinary fuel tank with an intake add, and have no special thermal properties.

 

It's also true that there's some very weird behaviour with TCS as you move in and out of atmosphere; if you turn on a thermal overlay (or KER), you can see that as soon as you leave the atmosphere the TCS skin temperature will shoot way up. It does not radiate as efficiently as it should in atmosphere because it isn't as hot as it should be.

[crackerbal]

euhm am i alone to use radiators only in case of any specific part create heat and need to been cooled ( ex. isru )

i see no reason to put them as coolers on ssto : 1. drag 2. weight 3. they are heating by their own in atmo

it's always cool to have some Tcs once in space, but why in atmo flight ? it's like put a heatshield on the top of a rocket

[bewing]

If you place them correctly, they will suck some heat out of the nose on reentry. It can make the difference between success and failure, if you are on the edge of failing.

[AeroGav]

On 18/11/2016 at 7:08 PM, Jarin said:

Okay, as I use SSTO spaceplanes more and more, this is becoming more of an issue. My current series of light shuttle fly a shallow enough ascent profile that their PE is around 40km by the time their AP hits 70. This means I spend a LOT of time in the upper reaches of the atmosphere. So here's the issue, heat sinks appear to basically not work at all as long as there's even the slightest hint of air around you. Compare screenshots below of just below and just above the 70km line.

  Hide contents

The heat bar on the right is my cockpit, incidentally. If I'd taken the shot closer to 60km, you'd have seen a very nice cherry-red outline glow as it neared overheat, despite the TCS right behind it. Is there any way around this? Am I missing something about KSP thermal mechanics? There's other bodies with atmosphere, why are thermal control systems almost completely inoperable here? How are you supposed to cool things like mining/refining on Eve or Laythe?

BTW is this ascent profile getting close to what you were doing?  The aircraft almost stops climbing at 46km (vertical speed falls to about 35m/s) before starting up again as it reaches orbital velocity.   PE ends up right behind our current position.

 

[Jarin]

On 11/23/2016 at 9:24 PM, AeroGav said:

BTW is this ascent profile getting close to what you were doing?

I only do one run, no lobbing and diving (though after seeing 1650m/s on airbreathers alone, I might change that), but here's what I was doing: Low-altitude acceleration to somewhere between 600 and 1000m/s, then shallow climb, hitting 1300m/s before I pass 20km. From there, I flatten out, adjusting my pitch by watching "Time to AP" in KER, and keeping it moving in front of me. Typically kick in the orbital engines around 1450m/s (I just wait until the jets stop giving me acceleration), and continue just barely pushing the time-to-AP, which has me flying fairly close to prograde horizontal all the way out of the atmosphere. Kill engines and coast once AP hits around 110, which is around 40-50km; then coast. I'll regularly have a positive-alt PE before I cut the engines.

Edited November 27, 2016 by Jarin

[ForScience6686]

If you're using radiators to get you out of the atmosphere you're doing something wrong.   that is not there intended purpose because they won't help much.   pushing the craft through the air generates the heat around the craft,  so what good would it do to put your radiator in all that heat?   

Instead,  go for more vertical speed to get you up and out of the thicker air faster.   or in the case of re entry,  increase your drag coefficient.

I only use radiators for nukes and drill/refining,  and nothing else as they are dead weight.  I either adjust my craft or ascent/descent  profile.

[fourfa]

2 hours ago, ForScience6686 said:

I only use radiators for nukes and drill/refining,  and nothing else as they are dead weight.  I either adjust my craft or ascent/descent  profile.

QFT