Part clipping causes reentry/heat damage/heat indicators

[hieywiey]

I'm not sure where this goes, but whatever.

Part clipping in 1.0 makes the little heat bar thing appear and can make parts glow, is this intended or is it a bug?

[Temeter]

I also noticed some of my planes having hot wheels after reloading. Seems like there is some minor bug with heat.

[diegzumillo]

I'm not sure. Sounds like intended behavior. Heat doesn't come only from atmosphere, it also spreads from part to part (heat diffusion). So if you have part A that is overheating, and part B is in close contact with it (or inside) it should overheat as well.

[hieywiey]

I'm not sure. Sounds like intended behavior. Heat doesn't come only from atmosphere, it also spreads from part to part (heat diffusion). So if you have part A that is overheating, and part B is in close contact with it (or inside) it should overheat as well.

No, it's not that, it always generates heat (on the launchpad, in orbit, everywhere) regardless of the reentry heat setting and completely ignores the Ignore Max Temperature cheat. I think it sees them as having some kind of wierd friction due to part clipping defying (our) laws of physics. This is one of those times when gameplay is more important than realism, imo.

[diegzumillo]

Ah, sorry, I misunderstood. That does sound like a bug.

[B-STRK]

I've observed the same thing. I've been using girders either clipped into a SPP fuselage to provide a mount for a wheel, or as a structural frame with parts clipped inside it (which I used for example in a small aircraft, and the vtol engines are mounted to the girder). In both cases they turn out to be the hottest part of the craft--either absorbing heat from something connected to it, or generating heat during reentry (and with a heat shield and some other parts between it and the reentry).

I believe I've read somewhere that girders could act as radiators of sorts to pump out LV-N heat, so for a moment I thought at least with the aircraft the absorbing-heat thing was sort of the expected behavior. The reentry thing was bugging me, though--nothing I could do could stop the girder from blowing up (even with a SpaceY 5m heat shield all but covering the cross-section of the craft, and either stock or Procedural fairings enclosing the whole craft). Point is, the first or one of the first things to blow up is a girder that is clipped. (Well, then again, even "exposed" wheels/tracks also heat up like crazy, even if in a fairing and the cross-section's wholly covered by a heat shield, but those wheels are also clipped to the girder somewhat...).

Maybe clipping's making temp calculations for parts go crazy, and for those parts that have certain heat properties, they go haywire?

[monophonic]

Do you see any indication that the clipped parts might be shaking, even ever so minutely? E.g. flickering pixel or few around the part when the craft is otherwise completely at standstill?

Sometimes clipped parts start shaking for no apparent reason. (It is actually ghost forces appearing within the physics engine, which probably has not been originally built with clipping in mind.) It has happened even to the extent where the entire ship falls apart from the forces. Even with minimally visible shaking the parts could travel at considerable speed per physics frame, and get heated by atmospheric drag because of that. Floating point inaccuracies can even magnify the effect.

This is still conjecture however, I haven't done any debugging to see if something even remotely like this is actually what happens.

[B-STRK]

Do you see any indication that the clipped parts might be shaking, even ever so minutely? E.g. flickering pixel or few around the part when the craft is otherwise completely at standstill?

Sometimes clipped parts start shaking for no apparent reason. (It is actually ghost forces appearing within the physics engine, which probably has not been originally built with clipping in mind.) It has happened even to the extent where the entire ship falls apart from the forces. Even with minimally visible shaking the parts could travel at considerable speed per physics frame, and get heated by atmospheric drag because of that. Floating point inaccuracies can even magnify the effect.

This is still conjecture however, I haven't done any debugging to see if something even remotely like this is actually what happens.

Couldn't tell from visual inspection, especially during reentry. I've also got Joint Reinforcement installed to minimize any flexing of joints, stack or clipped, which I guess would make visual observation even harder, if there's any flexing in the first place.

What's disconcerting from the reentry angle is that this is happening even if the said part in question, not to mention the whole craft, is completely hidden behind a heat shield and fairings. Theoretically, shouldn't this mean that everything inside the fairing should be shielded from aerodynamic stress so as long as the fairing and shield hold? I can post a few pics of the craft(s) in question, if it helps.

[B-STRK]

Okay, so I did a couple of run throughs with my own craft (it's modded, BTW, which I know would normally not fit in this subforum--mods, I understand if you guys relocate this if it turns out to be the case!--but since this thread is already here, and it might be because of stock behavior--or maybe not?) in order to show what's going on. Like hieywiey, the issue involves parts that were clipped. The album below covers the high points (full description of what's happening over at the main Imgur site):

Javascript is disabled. View full album

A couple more things to add: just leaving the craft + fairing (so those girders are shielded inside) on the runway in sunrise, the temps still climb. And, on another craft, where I have clipped a couple of Near Future static solar panel girders (PX panels) onto a short stock girder, plus a Mechjeb box and a surface scanner, plus mirror-mounted a couple of USI ducted fans (I was making the aircraft for that carrier), it was the girder and the solar panels that were pegging the temp gauges (the heat, I assume, came from both the fans and the sun).

I'm not versed in how the dynamics are coded, so I can't get too technical, but I could break down the possible hypotheses:

1. hieywiey is right, and there's a bug in how the temps in absorption and radiation are calculated when parts clip. Don't know how that works, though...
   
2. It's not quite a bug, but it's a painful consequence of clipping. This would be monophonic's theory a couple posts back: If a clipped part is shaking within the part clipped, the two should be generating friction. When I spawn the rover on the runway, though, I couldn't observe any of the flickering or shaking that monophonic asked me to look out for. (Plus, like I said above, I had KJR to stick things together hard. Unless, of course, that could be the issue?)
   
3. This is all intended behavior, and I am being punished for depending too much on girders to serve as my wheel bogies. As the heat experiments in the Add On Development board have seen, girders are one of those parts that could be used as radiators for an LV-N on full (not as good as wings though...). Which means they are designed to absorb and throw off heat as fast as they could. It's just that in reentry, there's too much heat to go around.
   
4. Maybe it's because I'm using HyperEdit? (Huh?)
   

#3 could be the right answer (that it's not a bug or a clipping issue), but it raises some more questions: then, why didn't the other girder, up in the rover's nose, peg the temp charts? (Maybe because of all the parts between it and the heat shield? But, like I observed in one of the pictures above, I had a prior experiment where more girders were mounted on a tank between the heat shield and the rover, and all this still inside a fairing, to serve as a heat soak, which didn't work.) Sure, maybe a heat shield won't stop all the heat from getting through to the contents, which is kinda counterintuitive (plus, as observed in a heat shield rebalance mod, a dead stock heat shield actually insulates completely the part it's attached to, which is also counterintuitive... hence the mod's existence), but then what about all the other parts in between the shield and the girders? Shouldn't the fairing also be able to absorb some heat, without transmitting it to what it protects inside?

So for now (and I'm thinking of waiting for 1.0.3, anyway, just to be sure), the only solution I got is to go with the active braking module in the last few pictures--which also confirms that a lot of the heat that killed the girders is coming during the reentry phase, heat that's not supposed to be transmitted to the craft ideally. It's not efficient, though, having to design a vertical drop deorbit module, and an active braking module, plus the skycrane, and send them all towards Duna, Eve, or Laythe (a lot of dV to spend... which I do have in sandbox, but still, the principle of the thing...).

I hope someone can figure out if this is not supposed to be happening, or this is actually the intended behavior. The advantages of fairings and heat shields is supposed to be that this isn't supposed to happen, after all... (Unless, of course, girders were not meant to be wheel bogies)

Thanks for any analysis!