This mod is not known to work with the latest version of Kerbal Space Program. Proceed with caution.
This is a behaviour-only mod. It adds no new parts, it just fixes KSP's stock heat shields so they fail when depleted and in extreme re-entry profiles. You now need to care a little about your re-entry profile, and using those "for novelty use only" 0-ablator heat shield parts will just get you killed.
(Credit to the Deadly Reentry crew; I reused their CC-BY-SA code to help make heat shields fail when depleted.)
ImprovedHeatShields also makes the inflatable heat shield a bit more delicate. It's amazing for what it should be amazing for - slow, grazing aerocaptures and aerobraking manouvres. But now if you try to use it for ballistic re-entry at high speed or of a heavy vehicle, you'll probably die. (This is optional, you can omit the ModuleManager config for it).
The max temperature of service bays is also lowered a bit, so they don't function as similarly magic heat shields. They're still very tough, just not "drop one into Eve at 4000m/s" tough.
Think of this as Deadly Reentry Lite. No Kerbal injury, no part damage, etc. Just heat shields that can fail, and saner temp tolerances on service bays.
KSP's heat shields are magic. Really. Try re-entering from Minmus on a steep trajectory ... with your heat shield tweaked to have zero ablator. You'll be fine. You can even use a 1.25m service bay, it's almost as good a heat shield.
I want heat shields where I have to care about the re-entry profile, at least a bit. Don't bring enough ablative material and your shield will be exhausted, with fatal consequences. Slam too hard into the atmosphere and the shield will overheat and fail - pyrolysis only gets you so far in terms of heat protection.
These shields aren't realistic. They're just slightly less absurdly unrealistic. Pyrolysis in KSP works like releasing some kind of super-compressed refrigerant that absorbs heat, or some kind of mega-endothermic chemical reaction. The thermal energy accumulates in the part until it reaches a threshold, then some ablator burns and poof, bye bye heat. The heat shield its self is a magic insulator, whether or not it has ablative material left. Instead what should happen is that the pyrolysis should greatly lower the thermal conductivity of the part ... until the ablator runs out.
To simulate that I borrowed Deadly Reentry's approach (and code) for resetting the heat shield's conductivity and lowering its max temp when it is exhausted. We presume it's weakened by being burned up. The max temp is also lowered overall, allowing the shield to fail when used in extreme re-entry profiles. They're really still magic fridges in physics terms, but the behaviour works out well enough.
Raw stats are from the beginning of time until now. Each follower and download entry represents one hour of data. Uneventful hours are omitted.