Thermal Control Systems and Radiator Panels

[NewtSoup]

@Geschosskopf all modern reactors need cooling even when scrammed, that's why they need cooling towers.  So I don't see nuclear engines needing  TCS as implausible.  I make no assumptions of the underlying code at all which is why I went with a simple on/of state and have the cooling rate proportional to fuel flow ( throttle setting ).  

Of course you can run nuclear engines happily without cooling in KSP so it doesn't really matter either way.

 

[Geschosskopf]

1 hour ago, NewtSoup said:

@Geschosskopf all modern reactors need cooling even when scrammed, that's why they need cooling towers.  So I don't see nuclear engines needing  TCS as implausible. 

Oh, I totally agree that there's still heat to deal with for a while even after SCRAM, and that while this is above the material limits of the engine/reactor, it has to be dealt with somehow.  It's just a question of how you deal with it, fuel or radiators?  This question in turn is function of other design parameters, such as the rate at which the reactor heats and cools, and the degree to which you can provide the cooling with fuel at non-optimum temperatures before it starts being too wasteful.  In some cases, radiators might be better, in others fuel.

[Brainlord Mesomorph]

OK I've been doing some of my own tests:

I've been concerned that there is some bad advice in this thread, specifically that the fixed radiators "don't do anything" and that nukes "don't need cooling."

Experiment 1:  I build a test rig, a probe core, fuel, and 4 nukes pointing forward, and 4 more pointing opposite them. A TWR of zero. I lob this thing into orbit, turn on the engines, in 10 minutes it blows up from overheating. Specifically it was the quad adapters that went first, but the nukes, the fuel tanks, even the probe were all red lining on heat. If it hadn't been the adapters something would have gone.

Experiment 2: same test rig, but I added 8 large radiator panels (4 at each end). This one ran for 23 minutes before it blew up. Again the quad adapters, but this time they were neck and neck with the nukes, I really thought the nukes were going to blow first.

Observations:
Nukes do need cooling (esp. at low TWR).  Fixed radiators do indeeed do "something".

Theories:
It looks to me like the fixed radiators were designed (by the devs) to mitigate the minor heat issues that you will get using nukes at low TWR. (and IMHO if you're not using nukes at low TWR then you;re not taking full advantage of nukes) while the deploy-able TCS units are for the major heat issues you get w/ ISRU.

[NewtSoup]

3 minutes ago, Brainlord Mesomorph said:

Observations:
Nukes do need cooling (esp. at low TWR).  Fixed radiators do indeeed do "something".

Good to know!.  I've not much experience of Nukes so far and the ship I built under guidance used no cooling at all and I've seen ships on youtube with a pair of nukes use no cooling.  I knew the Fixed Radiators would cool the parts they are attached
to.  I wonder if your test rig would still explode if instead of a quad adapter which places the nukes close together, uses a gantry with fixing points so the engines are spread out.  Even LfOx engines can explode when close together and run continuously.

[Brainlord Mesomorph]

If four engines (any heat sources) are attached to any part it would be the same regardless of proximity,  (in this sim, AFAIK)  but if there are more parts between them THAT would help spread it out.

8 minutes ago, NewtSoup said:

I've seen ships on youtube with a pair of nukes use no cooling.

its all a question of TWR and length of burn.

Also if you're doing heat test in the atmosphere, that's not the same as heat tests done in vacuum. 

[NewtSoup]

The temperature is climbing rapidly

 

Engines blew at 6 minutes, which is a conceivable for a long burn.

How do you think TWR affects the overheating?

Going to take off the Forward Facing engines and see how long an overheat takes.

TWR in a vacuum here is ~9 according to KER in the VAB with 4 rear facing engines.  Which is not low.  In space though TWR is actually 1.9 in LKO but also climbing as my altitude climbs.  Interesting!

Internal temp is climbing just as quickly as before with 4 engines.  In fact it seems to be climbing quicker!

 

Edited February 17, 2018 by NewtSoup

[NewtSoup]

The TWR is obviously climbing as I leave Kerbin's SOI. 

[Frozen_Heart]

There was one version where nukes overheated so easily to the point of barely being usable. The two suggestions in response were add radiators or turn down the heat they produced.

 

The devs did both. Added radiators and dialed down the heat produced making said radiators pointless.

[Brainlord Mesomorph]

20 minutes ago, NewtSoup said:

How do you think TWR affects the overheating?

(not think, know) LFO or nuke, all engine exhaust is hot, if the ship is accelerating swiftly (i.e the ship is light,  a high TWR)  exhaust blows out of the engine cone swiftly and is nice and cool. But; if the ship is very heavy (low TWR) the same thrust STAYS in the engine cone pushing against the other (hot) exhaust that hasn't lefty the cone yet, heat builds up. KSP models this correctly. Look at the color of any engine cone after a long burn; if was a heavy load it will be white hot.

 

1 minute ago, Frozen_Heart said:

The devs did both. Added radiators and dialed down the heat produced making said radiators pointless.

I disagree with "pointless" And my recent experiments prove it 

[Brainlord Mesomorph]

1 hour ago, NewtSoup said:

Engines blew at 6 minutes, which is a conceivable for a long burn.

you think 6 minutes is a long burn? LOL.

20, 30 minute burns are common for me. I used to have hour- long burns before I learned about Pe kicking,

 

EDIT: wait, your rig blew up in 6 minutes? mine took 10. That would mean that the quad adapter was BETTER at absorbing and transferring heat than the girder segment you used.

So not only does proximity not matter (KSP does not model radiant heat) but part selection matters.

Edited February 17, 2018 by Brainlord Mesomorph

[NewtSoup]

It’s  all good information. 

Yes, for me 6 minutes is a long burn the furthest I’ve been was Duna, once, in a twin ion engine craft.  I cut the burn into 6 minute chunks at kerbin PE. Please forgive my inexperience

Edited February 17, 2018 by NewtSoup

[NewtSoup]

Craft exploded at ~12 minutes

Probably due to the lack of exhaust gas collecting around it then.

8 hours ago, Brainlord Mesomorph said:

(not think, know) LFO or nuke, all engine exhaust is hot, if the ship is accelerating swiftly (i.e the ship is light,  a high TWR)  exhaust blows out of the engine cone swiftly and is nice and cool. But; if the ship is very heavy (low TWR) the same thrust STAYS in the engine cone pushing against the other (hot) exhaust that hasn't lefty the cone yet, heat builds up. KSP models this correctly. Look at the color of any engine cone after a long burn; if was a heavy load it will be white hot.

 

 

Bear in mind I've just run my tests with no cooling at all.  I am testing to see if Nukes will fly with no cooling.  The answer is yes but watch the heat.  Probably advisable to use some TCS if you can spare the mass.

Edited February 17, 2018 by NewtSoup

[NewtSoup]

Added 4 small TCS tot he gantries.  Craft Exploded at 8m03s MET. 

Fascinating!. 

Adding the cooling system apparently accelerated the tendency to overheat.

Added some Medium TCS to the gantry - the ship is still going to explode.

When the engines internal temperature reaches 877K the TCS then kick in and once their cooling reaches 99% the engines then go back to heating up.

This also happened with the small TCS so I wonder why the overall time to overheat was shorter than with no cooling at all.

[NewtSoup]

Currently at the 23 minute mark.  Internal temperature is 1666 k and climbing slowly.  Current velocity is ~18650 m/s
 

Ok.. The craft may not explode.  The internal temperature has fallen to 1623 Kelvin.

My bad.  I had throttled down by accident switching desktops.  The craft will eventually explode.

Jeb should be boiled in the bag by now.  Lander can internal temp is 458k ( holding steady )  good thing Kerbals don't pass out with high temperatures.

Edited February 18, 2018 by NewtSoup

[NewtSoup]

At the current rate of dt the ship will explode at ~54 minute MET.  I'm bored and pulling the plug... Time to add 4 small TCS to the 4 Medium ones.

Jeb also reports he's lost more than 78% of his body weight and can we please splash down near one of the poles.

[NewtSoup]

Ok.. I stuck with it. MET was 1hr 2min 40 seconds when the first engine exploded, after that the TCS cooled the remaining 3 sufficiently that there was no more rapid deconstruction.  Though the G-Forces from the resultant spin made Jeb reconsider the Kedgeree he had for breakfast.  Rice grains look so much like incoming asteroids when stuck to the console!.

Really I consider 1 hour of full burn on Nukes "enough".. however I'm going to add 2 small TCS to see if I can mitigate the heat production from the NERVS completely.

[Xordus]

It's pretty dumb that such an important part of the ISS IRL is totally negated in KSP.  Lame...  I always wondered what the point of them was, I just never used them due to lack of necessity. 

It would be nice if there was some use for parts that already exist in the game in various sizes and strengths.....  Seems like they could make this part of 1.4.  It can't be intentional to have made all of these parts that have no use at all..

[NewtSoup]

Running the new test with 2 small TCS added to the lander can I'm at the 54 minute mark the temperature in the lander can is MUCH lower at a balmy 294k

My other observation is that the rate of increase in internal temperature decreases as you approach max T

Damit I messed up again by throttlign down while switching desktops.  Ctrl Alt Arrow key switches desktop on my machine.  Ctrl reduces throttle.

The results will be skewed slightly

 

 

Going to have to run this test again as the throttle down has skewed too much. 

But not now.. it's 03:35 and I have wine to drink before bed!  ( also known as liquid fuel )

Edited February 18, 2018 by NewtSoup

[Archgeek]

15 hours ago, Brainlord Mesomorph said:

(not think, know) LFO or nuke, all engine exhaust is hot, if the ship is accelerating swiftly (i.e the ship is light,  a high TWR)  exhaust blows out of the engine cone swiftly and is nice and cool. But; if the ship is very heavy (low TWR) the same thrust STAYS in the engine cone pushing against the other (hot) exhaust that hasn't lefty the cone yet, heat builds up. KSP models this correctly. Look at the color of any engine cone after a long burn; if was a heavy load it will be white hot.

 

I disagree with "pointless" And my recent experiments prove it 

Erm, TWR isn't that important, save for needing longer burns.  Exhaust gases definitely aren't involved...that 800s Isp means they're ripping out the nozzle at 7848m/s, and no where near your craft's delicate bits for any amount of time.  For a heavy, low TWR ship, you're going to burn longer, letting the engines get hotter and make everything else get hotter.

7 hours ago, NewtSoup said:

Craft exploded at ~12 minutes

Probably due to the lack of exhaust gas collecting around it then.

[...]

Bear in mind I've just run my tests with no cooling at all.  I am testing to see if Nukes will fly with no cooling.  The answer is yes but watch the heat.  Probably advisable to use some TCS if you can spare the mass.

A decent time, that.  You'd be surprised what kind of passive cooling you can pull off, especially if there's a sizable heat sink to keep it from flowing to sensitive components.

 

Incidentally, I've completed the LV-N portion of those tests!: 

 

 

 

[NewtSoup]

Re-Running my above test with the same ship (4 medium TCS and 2 small ) at 2370k the increment in temperature is just 0.5k/s or thereabouts

[Geschosskopf]

7 hours ago, Archgeek said:

Erm, TWR isn't that important, save for needing longer burns.  Exhaust gases definitely aren't involved...that 800s Isp means they're ripping out the nozzle at 7848m/s, and no where near your craft's delicate bits for any amount of time.  For a heavy, low TWR ship, you're going to burn longer, letting the engines get hotter and make everything else get hotter.

Yup, that's it right there.  The engine produces X amount of heat per second, so the more seconds the engine runs, the more total heat you have in the ship.  There really aren't any exhaust gases in the game at all.  Fuel simply vanishes and the mathematical results of the thrust and heat generation functions are applied to the ship, while graphics of fire and smoke appear at the engine bell.

There's a small volume behind the bell in which other parts have can heat applied to them, as if they were actually in the exhaust stream.  However, this "stream" is very short and narrow, and neither it nor the graphical flames radiate heat at all.  Thus, we can mount radial engines so their fires pass only inches away right down the length of the central stack, and the central stack receives no heat from this.

 

7 hours ago, Archgeek said:

A decent time, that.  You'd be surprised what kind of passive cooling you can pull off, especially if there's a sizable heat sink to keep it from flowing to sensitive components.

This is what complicates the whole issue.  The longer the burn lasts, the more time the accumulated heat has to percolate through all the ship's parts.  The thermal properties and physical arrangement of these parts then determine how much heat gets to which other parts in how much time.  Thus, even with the same engine running for the same amount of time, whether or not you actually need radiators (and what type, and where to mount them) depends on the design of the rest of the ship.  This is why people see different outcomes in their games, resulting in much confusion.

This brings up the subject of your recent experiments.  They're very interesting---thanks for doing them.  However, the test rig's configuration bears no resemblance to any ship a player would actually use in the game, and is in fact designed specifically NOT to have any built-in heat sinks.  Thus, it gives a good idea of the relative effectiveness of various parts as radiators, and lets you know how much heat LV-Ns produce, but it doesn't tell a player how many of what type of radiator he might need on his real ship.

This percolation through the rest of the ship also somewhat ties back into the TWR question.  Let us say you've got a ship you want to send from Kerbin to Jool.  The basic design of the ship is constant---the same payload, the same fuel tanks, and you're going to be using LV-Ns.  And, of course, the dV required to get from Kerbin to Jool is the same, too.  The question is, from a heat safety POV, is it better to use fewer engines (and thus low TWR and a longer burn) or more engines (high TWR, shorter burn)?

My own experience, which I haven't rigorously tested, is that high TWR causes less problems with heat.  Although the rate of heat production during the burn is higher because you've got more heat sources, heat production stops sooner, so there's less time for heat to move around the ship while heat is still being produced.  Thus, the fuel tanks (which are full to start with so are great heat sinks) absorb the heat before they get saturated, then gradually feed it into the rest of the ship while simultaneously radiating some to space.  The result is not much heat ever making it up front where the delicate things are.  OTOH, with low TWR, the burn can last long enough that the fuel tank can become thermally saturated, at which point it ceases to provide insulation to the rest of the ship, meaning more heat eventually reaching the squishy bits.

Does this match your observations?

[NewtSoup]

$&%$%^$££~@! after 2 hours of testing I messed up my test again by accidentally throttling down.  I am going to have to change the throttle controls to something else.

That said my current design is not going to explode.  Ever. With 4 medium TCS and 2 small TCS the increase in heat levels off around 2395 Kelvin.

Given that most craft will have a much greater thermal mass than my skimpy gantry craft I'd say 1 Medium TCS per Nuke engine will always be sufficient.

[Archgeek]

5 hours ago, Geschosskopf said:

This brings up the subject of your recent experiments.  They're very interesting---thanks for doing them.  However, the test rig's configuration bears no resemblance to any ship a player would actually use in the game, and is in fact designed specifically NOT to have any built-in heat sinks.  Thus, it gives a good idea of the relative effectiveness of various parts as radiators, and lets you know how much heat LV-Ns produce, but it doesn't tell a player how many of what type of radiator he might need on his real ship.

This percolation through the rest of the ship also somewhat ties back into the TWR question.  Let us say you've got a ship you want to send from Kerbin to Jool.  The basic design of the ship is constant---the same payload, the same fuel tanks, and you're going to be using LV-Ns.  And, of course, the dV required to get from Kerbin to Jool is the same, too.  The question is, from a heat safety POV, is it better to use fewer engines (and thus low TWR and a longer burn) or more engines (high TWR, shorter burn)?

My own experience, which I haven't rigorously tested, is that high TWR causes less problems with heat.  Although the rate of heat production during the burn is higher because you've got more heat sources, heat production stops sooner, so there's less time for heat to move around the ship while heat is still being produced.  Thus, the fuel tanks (which are full to start with so are great heat sinks) absorb the heat before they get saturated, then gradually feed it into the rest of the ship while simultaneously radiating some to space.  The result is not much heat ever making it up front where the delicate things are.  OTOH, with low TWR, the burn can last long enough that the fuel tank can become thermally saturated, at which point it ceases to provide insulation to the rest of the ship, meaning more heat eventually reaching the squishy bits.

Does this match your observations?

Indeed, the test rig is designed to provide an upper limit.  If someone has a ship going somewhere, they know the dV, and probably have their TWR, from which they can get their burn time.  They compare the number of engines they're using and the burn time in question with the experiment, and find they need an amount of radiative capacity somewhere less than the radiators used in the experiment that came closest to their calculated burn time.

I'd agree that, ignoring radiators, or pushing radiators to their limits, a higher TWR is safer -- high burn times let the nukes get really hot, which increases their ability to cool themselves and thus the rest of the ship, but with the internal flux number having doubled, it's not enough to make up for the low-grade heat having time to work its way around the ship and build up in dangerous places.  However, when radiators aren't overpowered, the case gets more interesting -- what's superior, more engines at 3 tonnes ea. for lower burn time, or say a large radiator panel three parts up the tree from each engine, letting the engine and the tank its attached to get hot to produce more rad flux, but pulling out heat that gets any further?  Or perhaps a medium TCS or two anywhere you like, which will  prioritize just about cryogenically cooling sensitive parts followed by making sure nothing explodes (until its overwhelmed, if you don't have enough combined passive and active radiation capacity across the ship)?  Or if you're feeling cheeky, popping a couple of Big-S spaceplane tails on the tanks the nukes are attached to, letting them get as hot as they can to passively cool enough for an 11min+ burn?

It seems to me the use case is to allow lower-TWR designs to not explode, saving the mass of more engines and thus giving similarly-designed craft more range.

 

[MaverickSawyer]

Reviving this thread for a slightly different tack...

 

I want to make a sundiver probe, but I'm not sure I can keep the radiators adequately shaded, as I'm trying to keep the craft small:

Mk1 pod for scale

I've used the Draco heat shield from the Probes Plus pack as a shadow shield due to its high temperature rating (4500K!), but stripped off the ablative as previous missions revealed that the ablative wouldn't last very long anyways.  The small surface mount radiators are nominally rated for 50kW each, with a max of 274kW each, and I opted to use RTGs (50% scale, c/o Tweakscale) so I don't have to have solar panels sticking out into the sunlight to soak up heat. However, if I can swig it, I'd prefer to add on a pair of deployable radiators to better reject the heat I'm going to build up on this mission.

Couple of questions...

-Do high-temperature shadow shields do anything to mitigate the heat absorbed by vessels down close to the star? If so, is it based off of a point source, or off the actual model of the star's photosphere?

-If the above is true in either regard, how effective are radiators that far down the gravity well? What's the effective heat rejection rate, in percent of normal, basically.

-If the photosphere model mentioned above applies, is there any effective means of protecting a radiator as you close in on the Sun?

[NewtSoup]

Revive away - have you considered using fuel cells as a means of powering TCS and Radiators?  What is the practicality of using a captive asteroid as shield / fuel along with other means?  Just throwing ideas out here.  Cold you in fact create a sphere segment with TCS on the shadowed sides and the afore mentioned asteroid or just huge fuel tanks?

I think you should do lots of testing with sun divers in various formats to see which ones best survive