"Since radiation is disabled for parts within bays..." - umm, should it?

[RidingTheFlow]

... and I wondered why my payloads in closed bays started exploding after long burn.

You do realize that even when bay is closed, parts within must radiate *towards bay walls*? This way there would be no way for parts to be much hotter than bay walls at outer space temperature (unlike it seems to be now - appearing like when you close the bay it also becomes stuffed with magical mineral wool).

Update: it seems to be worst during physics warp - I can confirm that warp of 4x causes payload temperature skyrocket during burn if bay is closed (and if warp is disabled, temperature actually starts dropping). Something weird is going on here.

Edited June 23, 2015 by RidingTheFlow

[Darnok]

Warp x4 is causing lots of weird things like additional aero forces or part connections are stronger during warp, but if you would try same flight path without warp your craft would be destroyed just by hitting air.

[Red Iron Crown]

Afaict this is a simplification, necessary because parts don't radiate heat to each other. If radiation in bays was enabled the parts would radiate as if they were out in space even when the bay was closed. It is certainly true that parts inside could be much hotter than the bay walls IRL, vacuum is an extremely effective insulator so the only mechanisms for transfer of heat to the walls are conduction and radiation. Eventually they would reach an equilibrium, but it would be a slow process.

As for the physical timewarp, I haven't really tested anything with it yet, can't comment.

[RidingTheFlow]

It is certainly true that parts inside could be much hotter than the bay walls IRL, vacuum is an extremely effective insulator so the only mechanisms for transfer of heat to the walls are conduction and radiation.

Ok, so I have a cargo bay with payload inside:

- fuel tank

- docking ports

Please explain, where heat comes from making temperature of inside parts raise and eventually exploding? Unless there is heat producer *inside*, parts should never get *hotter*, they must only cool, even if slow. But they get hotter for me, esp. if I timewarp during burn - looks like they somehow receive more heat from engines than manage to dissipate, and then boom. If I open cargo bay, then their temperature starts dropping, so it definitely tied to open/close state somehow.

[Red Iron Crown]

Heat from the engine is being conducted to the parts in the bay, and continues to do so after the burn is complete until equilibrium is reached. Once the bay is open they are able to radiate their heat away.

[RidingTheFlow]

Last I've checked, heat goes from hotter part to cooler part, not other way around.

How come parts inside bay become hotter than the bay part they attached to? There is no heat pump inside attachment node.

Edited June 23, 2015 by RidingTheFlow

[Red Iron Crown]

Because the bay is ignored during conduction (sort of). Compare the temperature of the tank to the part in front of the bay.

This can be mitigated by having the last part before the bay be one with a high thermal mass (large fuel tanks work well for this).

[RidingTheFlow]

I do have large fuel tank (adapter) before the bay, and its still has lower temp than payload. I can't see any heat source close to payload which can cause it to overheat.

Edit: I've did more testing and it appears to be caused by 2 factors:

- bay must be closed

- physics warp must be active

Edited June 23, 2015 by RidingTheFlow

[steve_v]

Last I've checked, heat goes from hotter part to cooler part, not other way around.

Ah, yeah...

Parts accumulate heat from everywhere else, yet cannot radiate at all if in a bay? Whut? How can a part in the cargo bay get hotter than the bay itself, with no internal sources of heat?

Parts in a cargo bay should both conduct and radiate heat to the cargo bay walls, and thence into space. A bay is either an insulator or it's not, at the moment cargo bays operate like magical heat pumps. Thermodynamics fail.

Edited June 23, 2015 by steve_v

[RidingTheFlow]

Maybe its timewarp bug. At least I think at the moment - haven't managed to repro it without warp, maybe its just too slow. However, if bay is open, it seem not to overheat even in a 4x warp.

Edited June 23, 2015 by RidingTheFlow

[steve_v]

To me, the simplest kludge would be to simply disable all heat conduction into parts in the (closed) cargo bay... if the doors are a perfect insulator, make the walls likewise. The point of a cargo bay is, after all, to protect the cargo... not roast it.

I have already had more than one payload explode because I forgot to open the bay doors, and the probe core sucked all the heat from the main engines. This is IMO a bug, and one that need fixing ASAP.

Admittedly, I tend to physics warp during the coast to apo on ascent, so it may well be related to warp. More testing needed.

Edited June 23, 2015 by steve_v

[RidingTheFlow]

Did more testing, definitely look like a bug.

I've used HotSpot mod to visualize the heat flow & temps. If timewarp is x1, it all works fine:

However, as soon as I enable just x2 physics warp, I notice that *conduction flux* starts getting crazy on cargo bays (even on actual *outside* parts) - it oscillates like crazy between extreme values. Looks like some physics solver experiencing extreme instability. This visible as rapid flashing of parts via HotSpot overlay:

And looks like that heat flux going crazy eventually causes bay temp & parts inside (parts inside even more) to raise - since parts inside heat more, this cause them explode first.

Since this looks like a definite bug, can we move this to bug reports forum? Thanks

[Red Iron Crown]

It does indeed look like a timewarp-related bug. Rather than move the whole thread, I would suggest starting a new thread for it in support (and the bugtracker, if you're inclined) with your data and logs.