Serious Bug in Convective Heating in The Game

[arkie87]

I posted this in gameplay questions, but did not get an answer...

So, i've noticed that some of my craft appear to explode for no reason due to overheating in 1.0, so i decided to debug a bit.

I think i've found a serious problem:

When i look at the ram intake temperature, it is larger than air stream temperature, and yet, the convective flux is still positive (when it was below air stream temperature, convective flux was negative)-- why?

This obviously violates Newton's law of cooling ( Q=h*A*(T-T_fluid) ). Yes, h might be a complicated function of Reynolds, attitude, Mach etc... but these only effect magnitude, not sign. Thus, if T < T_fluid, T cannot become larger than T_fluid...

To highlight this point, another user, Windy_Skunk, performed an experiment of hypersonic flight at sea level with temperature damage turned off. Here are his results:

Temperature approaches 20,000 K while air temperature, even after all the shock heating, approaches 1505.6 K.

Incidentally, 1505.6 is EXACTLY the same value as the surface velocity... so something is fishy...

[Mr Shifty]

I notice in your first image that ext. temp and surface velocity are exactly the same as well: 943.9. Hmm...

[ZetaX]

Has someone tried to find out if this is a bug in the display (which could have happened easily) or an actual bug in the temperature handling (which would sound a bit weird)¿

[michaelsteele3]

So, i've noticed that some of my craft appear to explode for no reason due to overheating in 1.0

No offense, but are you trolling?

• Wrong Forum
  
• You DID in fact get an answer on your last thread in Gameplay Questions and Tutorials, but you just kept arguing with others on the other thread about it, even going as far as to insulting.
  
• Yes, crafts explode when they overheat. It's a thing.
  
• It's obviously not a huge problem if nobody else seems to notice it. They are trying to help you, and you say stuff like"Let me google that for you:sticktongue:"
  

[arkie87]

• Wrong Forum
  

I like to post things in science forum since people here generally are more willing to discuss physics/science rather than just giving me an answer and get offended when i argue why their answer makes no sense. So i think this is the perfect form. I didnt place it in bug reporting because i think we might be making progress i.e. discovering that velocity = Temp_ext...I might post it in bug forum once i am more confident that there is definitely a bug...

You DID in fact get an answer on your last thread in Gameplay Questions and Tutorials, but you just kept arguing with others on the other thread about it, even going as far as to insulting.

First of all, no, i didnt get a good answer. Everyone just gave an explanation ignoring the equations, like you are here, saying crafts explode when they overheat and hypersonic craft get hot. That is not helpful, which, once again, is why i came here. For some reason, everyone on Gameplay Questions and Tutorials just wants to give answers, and get insulted when you argue with them, instead of defending their positions (like you are doing here).
Second, I didnt mean to be insulting (my use of emoticons were supposed to convey i was being playful). But calling me a troll I think is insulting.

No offense, but are you trolling?

...

Yes, crafts explode when they overheat. It's a thing.

I realize that. That is not the issue we are discussing. This is condescending and unhelpful...

It's obviously not a huge problem if nobody else seems to notice it. They are trying to help you, and you say stuff like"Let me google that for you:sticktongue:"

He admitted he had no idea what newtons law of cooling was, but was participating in the discussion anyway... that is not helpful...

If there is a bug, then it is a HUGE problem...

Just because others havent noticed it, doesnt mean its not a problem. Other, inexperienced players (or players who never played with deadly re-entry) might not be familiar with what can and should cause re-entry problems, so they might have assumed they were doing something wrong...

- - - Updated - - -

Has someone tried to find out if this is a bug in the display (which could have happened easily) or an actual bug in the temperature handling (which would sound a bit weird)¿

Yeah, i dont know. I would love to see the source code, or a Wiki on how KSP handles re-entry heating (convective, radiative, and conductive heat fluxes).

- - - Updated - - -

I notice in your first image that ext. temp and surface velocity are exactly the same as well: 943.9. Hmm...

Good find!

It seems to always be the case above Mach 1-2... Below Mach 1-2, i think temp reads normally...

[PakledHostage]

If there is a bug, then it is a HUGE problem...

Just because others havent noticed it, doesnt mean its not a problem. Other, inexperienced players (or players who never played with deadly re-entry) might not be familiar with what can and should cause re-entry problems, so they might have assumed they were doing something wrong...

Even if it isn't just a display bug as Rathlon suggests below, it probably isn't a good idea to hold your breath while you wait for it to be fixed. I gave up on my re-entry heat mod (the original re-entry heat mod that predated deadly re-entry) when I couldn't get my heat effects (which were based on simplified compressible flow, convective heat transfer and radiative heat transfer models and which were surprisingly consistent with real world re-entry heating data that I could find, given the simplicity of my physics model) to correlate with the visual effects that were added in v0.19 of the game. When I first introduced the mod several months prior to v0.19's release, Moach even went over the code with his brother (HarvesteR) and Moach passed along some suggestions from Harv to me following their conversation. Clearly they know that there are alternatives to the way they coded it, but KSP's developers have said over and over again that they are making a game first; it isn't intended to be a physics simulation.

Edited April 30, 2015 by PakledHostage
Revised post in consideration of Rathlon's good point, below.

[Ralathon]

I think what's happening is that Squad messed up and instead of the external temperature you're seeing part velocity. Thanks to Adiabatic compression the external air is heated by some function of velocity and is convecting heat into the part.

Post it in the bug report forums as a minor display bug and be done with it.

[arkie87]

I think what's happening is that Squad messed up and instead of the external temperature you're seeing part velocity. Thanks to Adiabatic compression the external air is heated by some function of velocity and is convecting heat into the part.

Post it in the bug report forums as a minor display bug and be done with it.

I would agree with you if velocity always equaled temp_ext, but that isnt the case always (for low mach numbers or in space), so i dont think thats the root of the problem (or at least, not the ONLY problem).

- - - Updated - - -

Even if it isn't just a display bug as Rathlon suggests below, it probably isn't a good idea to hold your breath while you wait for it to be fixed. I gave up on my re-entry heat mod (the original re-entry heat mod that predated deadly re-entry) when I couldn't get my heat effects (which were based on simplified compressible flow, convective heat transfer and radiative heat transfer models and which were surprisingly consistent with real world re-entry heating data that I could find, given the simplicity of my physics model) to correlate with the visual effects that were added in v0.19 of the game. When I first introduced the mod several months prior to v0.19's release, Moach even went over the code with his brother (HarvesteR) and Moach passed along some suggestions from Harv to me following their conversation. Clearly they know that there are alternatives to the way they coded it, but KSP's developers have said over and over again that they are making a game first; it isn't intended to be a physics simulation.

Fair enough. If possible, if its not just a display bug (which i dont think it is for the reasons ive given above), i might even consider writing a mod to fix it... though i have little experience with that.

- - - Updated - - -

Even if it isn't just a display bug as Rathlon suggests below, it probably isn't a good idea to hold your breath while you wait for it to be fixed. I gave up on my re-entry heat mod (the original re-entry heat mod that predated deadly re-entry) when I couldn't get my heat effects (which were based on simplified compressible flow, convective heat transfer and radiative heat transfer models and which were surprisingly consistent with real world re-entry heating data that I could find, given the simplicity of my physics model) to correlate with the visual effects that were added in v0.19 of the game. When I first introduced the mod several months prior to v0.19's release, Moach even went over the code with his brother (HarvesteR) and Moach passed along some suggestions from Harv to me following their conversation. Clearly they know that there are alternatives to the way they coded it, but KSP's developers have said over and over again that they are making a game first; it isn't intended to be a physics simulation.

Since you've written your own mods for re-entry, do you think it is still behaving physically accurate (not 100%, but at least order of magnitude)?

Clearly, 20,000K for 1500 m/s at sea level is a bit high....

Edited April 30, 2015 by arkie87

[PakledHostage]

Since you've written your own mods for re-entry, do you think it is still behaving physically accurate (not 100%, but at least order of magnitude)?

I don't know. I haven't played KSP since [Gasp!!!] v0.19 was released. Now I'm just an old curmudgeon who hangs around on the Science Labs. And while I haven't seen such numbers myself, 20000 K does seem way too high (like close to an order of magnitude too high) for 1500 m/s at sea level.

[arkie87]

I don't know. I haven't played KSP since [Gasp!!!] v0.19 was released. Now I'm just an old curmudgeon who hangs around on the Science Labs. And while I haven't seen such numbers myself, 20000 K does seem way too high (like close to an order of magnitude too high) for 1500 m/s at sea level.

Exactly; it's what makes me think the game is using the wrong formula for convective heating i.e. adding a separate heat source term when velocity gets too high, rather than just make air temperature increase, and part temperature approach air temperature via convection.

[stupid_chris]

You do realize that you are interpreting the values wrong?

Temp is the current temperature of the whole part. Temp ext is not the atmospheric temperature. It's the stagnation themperature, which is the temperature air would have if all it's kinetic energy would be transformed into thermal energy. Now the latter is rather hard to actually calculate, so the "hack" used to calculate it is to say that 1m/s=1K, which is actually rather close. So yes, you can expect Temp ext to be pretty close, if not the same as the velocity.

tl;dr: Make sure you know what you're even looking at before blowing stuff out of proportions.

Edited May 1, 2015 by stupid_chris

[arkie87]

You do realize that you are interpreting the values wrong?

Temp is the current temperature of the whole part. Temp ext is not the atmospheric temperature. It's the stagnation themperature, which is the temperature air would have if all it's kinetic energy would be transformed into thermal energy. Now the latter is rather hard to actually calculate, so the "hack" used to calculate it is to say that 1m/s=1K, which is actually rather close. So yes, you can expect Temp ext to be pretty close, if not the same as the velocity.

tl;dr: Make sure you know what you're even looking at before blowing stuff out of proportions.

Yes, that is exactly what i was saying (and everyone else; no one is confused about that)... have you read the whole thread?

No one ever questioned why atmospheric temperature (i.e. external stagnation temperature) was so high....

We are confused why there is heating when readout for Temp_ext (stagnation temperature) is BELOW part temperature...

NathanKell proposed a reason here: http://forum.kerbalspaceprogram.com/threads/117511-Serious-Bug-in-Convective-Aerodynamic-Heating?p=1886854&viewfull=1#post1886854

IMHO, there is still a bug since Temp_ext is not used above Mach 2 or so, so it is misleading, to say the least...

Futhermore, I'm not convinced that the form of hypersonic convective flux is correct.

convective flux = density*velocity^3*convection_constant

What is the physical basis for this form (any sources?)?

Regardless, there also seems to be a numerical problem: at hypersonic velocities, flux is independent of part temperature so part can be heated to infinity, even above the stagnation temperature of the compressed air. this is clearly not physical

The only hand wavy explanation i can give is that since radiation heat transfer coefficient grows with T^3, (flux grows with T^4), radiative heat transfer coefficient will be dominant anyway... though i'm not really convinced that is the case

(for example: h_radiation(T=2000K)~=5.67e-8*2000^3=450 W/m2-K.... that is not large at all...)

Edited May 1, 2015 by arkie87

[arkie87]

I found a good reference for the convective flux relationship:

convective_flux = density*velocity^3*convection_constant

See Page 22

Interestingly, the derivation assumes that the wall temperature is much less than stagnation temperature. Thus, this formula correctly predicts heating when wall is cold and heating is just beginning, but as T_wall approaches stagnation temperature, convective flux will not approach zero, as it should...

Edited May 1, 2015 by arkie87

[Nibb31]

So there's a bug in KSP? Big deal. Report it, and get on with your life.

[arkie87]

So there's a bug in KSP? Big deal. Report it, and get on with your life.

Until just now, i wasnt sure if there was a bug, or if there was, what it was or how to fix it...

That's why i posted this, hoping someone could help me pin it down or explain that there isnt one...

If you dont have anything nice to say, please dont say anything at all

Edited May 1, 2015 by arkie87