MAJOR Heating Bug

[DaMichel]

The first thing that seemed to have helped me is to let Steam check and redownload broken game files. There were actually some broken files in my install. But still i get overheating from time to time.

Second, as a simple workaround: ALT-F12, go to physics menu, thermal tab, and turn off the checkboxes for conduction, radiation and so on. Also tune down all the sliders to zero. This should pretty much disable the system. Unfortunately the settings are not persistent :-/

Perhaps it is possible to shut it off via ini files?

[Ryds]

The first thing that seemed to have helped me is to let Steam check and redownload broken game files. There were actually some broken files in my install. But still i get overheating from time to time.

Second, as a simple workaround: ALT-F12, go to physics menu, thermal tab, and turn off the checkboxes for conduction, radiation and so on. Also tune down all the sliders to zero. This should pretty much disable the system. Unfortunately the settings are not persistent :-/

Perhaps it is possible to shut it off via ini files?

That does it, thanks for the tip DaMichel. It's a shame there'll be no reentry heating but I've got DRE installed anyway so G forces should still play a part.

I'm just going to use 1.0.4 is a test platform for ideas and hope they finally fix it for the new Unity version.

[Starwaster]

That does it, thanks for the tip DaMichel. It's a shame there'll be no reentry heating but I've got DRE installed anyway so G forces should still play a part.

I'm just going to use 1.0.4 is a test platform for ideas and hope they finally fix it for the new Unity version.

Try copying this code to a cfg file named minMass.cfg

@PART
[*]:HAS[#mass[<0.0001]]
{
  @mass = 0.0001
}

Put that in your GameData folder somewhere

[DaMichel]

This is a pretty low mass. I guess some mod parts have such low mass because i cannot recall seeing a part that has less mass than 0.001 which is for, i think, the small cubic strut. Therefore, can you explain why?

[danfarnsy]

Watching Squad's bugtracker, they still seem to be having trouble reproducing this consistently. Grr. I have stuff blowing up all the time. I also get parts flashing white without the thermal overlay, which until I started digging into this I thought was just a lighting/graphics issue instead of an optical heat deal (like the red of the engines when they've been running two minutes). Small parts adjacent to larger ones with higher thermal multipliers, especially when something is loaded or opened, seem to be the most likely triggers. Suddenly it dumps a bunch of heat into a part it didn't have thermal access to before. The worst is small parts inside a service bay or adjacent/near one. Things have exploded on the launch pad so often, before I even get a chance to press the spacebar, that I've given up on launching the craft. And I NEVER put anything in a service bay which I intend to interact with again, because if I open it, my slightly larger ship turns into two smaller ones.

DaMichel's suggestion of numerical instability makes a lot of sense, especially looking at how a part which is thermally isolated by a service bay is suddenly exposed to a sharp temperature differential by opening the service bay (usually the diff in temp with the service bay itself). The thermal system looks at the temperatures of two adjacent parts, what the mutual heat conductivity between them is, and appears to transfer a certain amount of heat over regardless of the heat capacity it's dumping into. And real physics doesn't care about your heat capacity either, but real physics gets to use significantly smaller time steps, which is why we don't explode like Assimilated Bobs from Marathon when we step from an air-conditioned building into the summer heat.

Big tank to little stabilizer: "Hey, little guy, I've got a large surface area and just took some solar irradiance, increasing my temperature by 2 K compared to yours. I'll give you 1 K's worth of my heat!"

BOOM!

I'll be tweaking small parts in the mean time in my own install to have this line:

thermalMassModifier = 5

Edit:

This points to a relatively simple fix, though I don't know how much comparative overhead this would be, but before shuffling a quantity of joules from one part to the adjacent one, use some scale of the ratio of heat capacities of adjacent parts for conduction, and pre-emptively use that scale to gimp the amount of heat transferred. It would still scale up with temperature difference, of course, but if the thermal system is going to use a monolithic *temperature* for a Rockomax Jumbo Tank, then the system ought to scale back when the capacity differential between adjacent parts is so large. Conductivity is already supposed to take care of part of this, but the fact that a battery or an OKTO2 has such a small thickness and is then placed next to a large part that *isn't* treated as small pieces of sequential thermal mass, combined with the infeasibility of dynamically changing the step-size per unit-time, leaves us with a lot of exploding things that ought not to explode.

Edited August 4, 2015 by danfarnsy

[includao]

For awhile it seems they are never going to be able to reproduce this bug.

[Major SNAFU]

I am glad I found this thread.

I haveonly had the game for a short while, and I was doing pretty well. Then I installed FAR and I think made it too difficult too quickly - but I was getting on top of that and finally had some craft in orbit - then things started exploding. so far it has always been the octagonal struts.

A probe will just be in orbit for days, no issues at all and I will then change the view from map to spacecraft and boom. I also had a probe survive reentry, and was 100 meters from splashdown, decending at <5 m/s and then Boom! one of the octagonals suddenly blew up. Fortunately I have quick saved just prior to re-entry and what I discovered after many, many runs through is that if I stay in normal time and/or don't exceed 3X time acceleration, then I never have a misshap. If I go to 4X time compression, then it is almost certain that one of the octagonal struts will blow at some point during the final (chutes out and open) part of the decent.

I am thinking to back off of FAR for a bit and see if I can regain some of the fun, and then re-introduce FAR once I have more experience with the game.

The reason I installed FAR so quickly is that I didn't want to learn bad habits that I would have to unlearn with FAR.

[steve_v]

The reason I installed FAR so quickly is that I didn't want to learn bad habits that I would have to unlearn with FAR.

The transition from stock to FAR isn't the re-learning of aerodynamics it once was, now that stock isn't totally ridiculous.

FAR does seem to exacerbate the heating bug, but there's word of a partial workaround for it in the latest release, and DRE (lower thermal conductivity for the sake of harder) actually helps with the instability.

As far as I can see so far FAR + DRE is now slightly less likely to explode than stock... Still really needs to be fixed on Squads end though.

[Fendleton]

I had this crop up with BDArmory.....dropped a load of bombs on KSC, and it took revenge on me by heating my pre-coolers up to maximum and blowing my wing off. My experience seems consistent with the other explanations in this thread.

[Virtual Flyboy]

I am glad I found this thread.

I haveonly had the game for a short while, and I was doing pretty well. Then I installed FAR and I think made it too difficult too quickly - but I was getting on top of that and finally had some craft in orbit - then things started exploding. so far it has always been the octagonal struts.

A probe will just be in orbit for days, no issues at all and I will then change the view from map to spacecraft and boom. I also had a probe survive reentry, and was 100 meters from splashdown, decending at <5 m/s and then Boom! one of the octagonals suddenly blew up. Fortunately I have quick saved just prior to re-entry and what I discovered after many, many runs through is that if I stay in normal time and/or don't exceed 3X time acceleration, then I never have a misshap. If I go to 4X time compression, then it is almost certain that one of the octagonal struts will blow at some point during the final (chutes out and open) part of the decent.

I am thinking to back off of FAR for a bit and see if I can regain some of the fun, and then re-introduce FAR once I have more experience with the game.

The reason I installed FAR so quickly is that I didn't want to learn bad habits that I would have to unlearn with FAR.

Yep, octagonal struts are definitely on the .... list of randomly exploding parts. Launched and landed a manned Munar lander only to have the struts in the service bay holding my rover explode as soon I opened them. I was thrilled.

Edited August 5, 2015 by Virtual Flyboy

[Major SNAFU]

So the only way to make sure it doesn't happen at all is to open the 'cheats' physics section and turn off all of the heating effects, correct?

[Psycho_zs]

This happens on Linux too.

Decoupling, fairing deployment, docking, opening service bays, or just starting on launchpad can trigger this phantom heat in some part. Sometimes it is blazingly fast, sometimes it cooks ship slowly. Usually batteries in service bay are blowing up when it is opened. Or small cubical girder when fairing is separated. Now I have a nuclear tug docked to a station (engine off) with four hydrogen tanks slowly cooking themselves.

[westamastaflash]

I've had success by undocking and redocking to a station to get rid of the heat. It's a huge PITA. My LLL Communotron 64's all started heating up at the same time. Undock, Redock, and the skin temperature started going back down. ....

[Xyster]

Having this same issue. Causing quite a bit of grief. I'm glad others have mentioned methods they have used to try and work around these issues. I've tried just about everything short of re-writing the game. I linked a thread from steam with quite a few people having this issue to this thread. Hopefully we can find a temporary solution to this bug that is making me want to put this game back in the closet.

http://steamcommunity.com/app/220200/discussions/0/598199244893181717/?tscn=1439603100

[Starwaster]

Having this same issue. Causing quite a bit of grief. I'm glad others have mentioned methods they have used to try and work around these issues. I've tried just about everything short of re-writing the game. I linked a thread from steam with quite a few people having this issue to this thread. Hopefully we can find a temporary solution to this bug that is making me want to put this game back in the closet.

http://steamcommunity.com/app/220200/discussions/0/598199244893181717/?tscn=1439603100

Did you lower conductivity as advised? If so, how low have you lowered it? And did you make the change permanent?

[Gryphon]

I've been toying with raising the thermalMassModifier of affected parts. Am I completely off-base? So far, it seems to work, but I haven't really tested it. All I know for sure is that before, I got explosions from the Z-200 battery in certain situations. Since I set its thermalMassModifier to 3 (via a little MM patch), it hasn't happened again.

So far.

[Psycho_zs]

I've made a simple sed one-liner to cool down things in a save file. If you are on Linux (or OSX, is there sed on OSX?), this should help recovering overheated save files.

cat persistent.sfs | sed -e 's/tempExt[[:space:]]*=[[:space:]]*[4-9][0-9][0-9].*$/tempExt = 399/' -e 's/temp[[:space:]]*=[[:space:]]*[4-9][0-9][0-9].*$/temp = 399/' -e 's/tempExt[[:space:]]*=[[:space:]]*[1-9][0-9][0-9][0-9].*$/tempExt = 399/' -e 's/temp[[:space:]]*=[[:space:]]*[1-9][0-9][0-9][0-9].*$/temp = 399/' > cooled.sfs

Takes 'persistent.sfs' and swaps everything above 400K with 399K, dumping to 'cooled.sfs'. Tried to make a full-blown parser, but that came out painfully slow, so it is a sed hack.

Gryphon, show your MM patch.

[Castun]

I've suddenly ran into this bug on my current career out of nowhere.

It certainly seems like some parts get bugged and generate enormous amounts of heat. Even putting a new vessel on the launchpad instantly shows red with the temperature overlay enabled. Radiators don't make any difference. Using KER or the thermometer shows ambient temperature at launchpad is 288'C which is....wrong. This is also with a new test sandbox game.

[Psycho_zs]

It should be 288K, not 288C.

[Gryphon]

Gryphon, show your MM patch.

Heh. It's an exceedingly simple patch, and I am not certain it is a solid fix, but here is what I am using so far:

@PART[batteryBankMini]
{
	thermalMassModifier = 3
}

@PART[adapterSmallMiniShort]
{
	thermalMassModifier = 3
}

@PART[US_1C10_Wedge_Quadcore]:NEEDS[UniversalStorage]:AFTER[UniversalStorage]
{
	thermalMassModifier = 3
}

@PART[US_1C15_Wedge_Hexacore]:NEEDS[UniversalStorage]:AFTER[UniversalStorage]
{
	thermalMassModifier = 4
}

You'll note it includes two Universal Storage parts.

There are a few key issues or concerns with this workaround.

1) My testing is merely reloading the game and running x4 Physics warp to see if it explodes at the same place, in the same ship configuration. I don't really know if it's a fix, but I haven't had any of these parts explode after I applied the patch.

2) setting thermalMassModifier to 3 was just a guess, and

3) I had the Hexacore explode even after I set it to 3. That's why I set that part to 4. Then it was happy.

4) I am using FAR, which reveals the bug more often because it has a more accurate determination of the size of parts.

5) I am only at Tier 4 technologies (@90 Science cost, stock tree) in my game. Anything beyond that, I probably haven't flown yet in 1.0.4. Also, I haven't tested this outside of the Kerbin SOI.

6) I try to avoid clipping parts, so I don't know what effect that would have.

Edited August 22, 2015 by Gryphon

[Castun]

It should be 288K, not 288C.

Oh, well that makes a lot more sense then. 288K would be 15'C...so are all the parts being red normal for the low end of the temperature spectrum then? The exploding parts seem that much more goofy then. Jettisoning fairings or opening service bays seems to trigger it the most often I've noticed.

[Gryphon]

AAAAAND you might want to check on Claw's latest "Stock Bug Fix Modules," where he says:

[*] OverheatFix - Fixes overheating bug that causes thermal feedback and part explosion.

[Weatherman159]

Had a solar panel causing my whole ship to overheat and explode in a matter of seconds when extended. Setting conductivity to 1 solved this. Thanks to whoever posted that.

[Axelord FTW]

AAAAAND you might want to check on Claw's latest "Stock Bug Fix Modules," where he says:

It helped but it didn't kill the problem entirely.