AE-FF2 Airstream Protective Shell is not protected by the 2.5m Heat Shield

[antipro]

AE-FF2 Airstream Protective Shell is not protected by the 2.5m Heat Shield.
Probably cause it has a slightly wider diameter, I don't know.
Is it intended to be that way on purpose or is a bug?

the ss down here show 2 versions of a lander:
1) shield + service bay + shell = shell explodes.
2) shield + adapter = lander survives.




 

[Guest]

Can't see your pictures just blank. I think your guess is right, if the heat shield has less diameter than that which it's protecting, then it can't protect what it cannot cover.

[antipro]

evidently I have not explained myself well. I try to turn the question in another direction.
I attached 2 tanks(max temp 2000k) radially on the lander, so that they protrude for a quarter of their length outside the shield protection.
anyway they stick out more than the 2.5m shell.
result = lander survives, the tanks just warm up a little.

So why the 2.5m shell(max temp 2600k), which stick out few cm only, evaporates so soon that seems it has no protection at all?
does not matter if it has a service bay in front or it is inside the service bay, it always explodes.






 

Edited March 10, 2021 by antipro

[Guest]

Ahh are you talking about ablation rate? So a heat shield ‘suffers’ more in the heat than a gas tank which seems weird? Yes could be a bug, although by design an ablator is designed to ‘take a beating’ as one of the ways it works is to ‘carry’ the heat away through convection.

One thing to check is the setting (pause menu settings) for heating and reentry heating of parts. On some of the easier settings, part heating is reduced (obviously easier) but you have the option of turning it up (above 100%) if you enjoy a challenge.

[antipro]

2 hours ago, FruitGoose said:

Ahh are you talking about ablation rate? So a heat shield ‘suffers’ more in the heat than a gas tank which seems weird?

no and no.
I only showed how another part does not explode if is covered for half or 3/4, after you said:

22 hours ago, FruitGoose said:

it can't protect what it cannot cover.

pointing out the fact that the shell protrudes much less.
and since this exact same problem had arisen in the past with the known terrier engine bug,
I just wanted to know if it's the same type of bug.

[Guest]

Haha a useful input from me in this thread then - I'm just confusing things more so apologies. Hope you solve it 

[OHara]

3 hours ago, antipro said:

problem had arisen in the past with the known terrier engine bug,
I just wanted to know if it's the same type of bug.

No, not the same problem as version 1.8 had with the Terrier engine.  The dimensions of the (base of the) fairing are correct in PartDatabase.cfg

You can use alt-F12 debug options to show the temperatures in KSP thermal simulation, if you are so interested.  The debug output is not completely self-explanatory, but there are people on the forum who have figured it out.

KSP keeps track of the temperature of the exposed portion of the skin, and the conduction of heat from the skin to the insides is quite slow.  So it does not matter much what fraction of skin is exposed; the exposed portion of skin can fail with little help from the rest of the part.

It still seems strange that the exposed skin on those tanks survived better than the exposed skin on the fairing.  To figure out why, I think you would have to look at the debug numbers.

[antipro]

4 hours ago, OHara said:

It still seems strange that the exposed skin on those tanks survived better than the exposed skin on the fairing.  To figure out why, I think you would have to look at the debug numbers.

I've added a battery so now those tanks protrudes even more. and still they don't explode.



then I did the thermal debug thing but I still can't figure out why the shell explodes, I only see its "Exposed skin" value grows up very fast till it reaches its limit,
according to the KER "Critical Skin Temperature" value.

but maybe you can:
https://www.dropbox.com/s/bk17do9f74l5ix0/heat debug.mkv?dl=0
 

[OHara]

I think that what you see is just an unwanted consequence of how KSP divides the part into coarse chunks in order tracks the heat (link).  I see similar behaviour in version 1.3.1, except that  the service bay was bigger back then so it feels the heat.

The slightly-sticking-out part has exposed area about 1% the area of the heat shield, so it gets about 1% of the 'convection flux' which makes sense (70kW being about 1% of 7800kW in my image).   That heat warms just 1% the thermal mass of the skin (1% of 54kJ/K being 0.5kJ/K in my image) so that bit of skin heats up about 70/0.5 = 140 K/second.   The skin can dump heat to the internals, but not fast enough.

In effect, that 1% of the skin experiences very close to the heating you would see if all of that part's skin were facing the shock-compressed atmosphere.

You can cheat the system by wiggling around a little.  Then the fraction of skin exposed goes up briefly, tricking KSP into spreading the heat over a larger thermal mass.  I think that effect is why nobody has  noticed during the past five years in enough detail to report a bug.

The simulation of heat that came with KSP 1.0 is in many ways just detailed enough to be mysterious.

Edited March 11, 2021 by OHara