Observed by @Starwaster, confirmed by own experimentation.

When a Mk1-2 Pod equipped with an SDHI Heat Shield and Paradock reenters Kerbin's atmosphere, it begins swinging wildly instead of maintaining an aft-first orientation as the stock Heat Shield does.

We suspect that recent atmosphere changes in KSP 1.2.x (which stock parts already account for) may be the cause.

@Starwaster initially suggested adding the CoPOffset and CoLOffset parameters from the stock 2.5m heat shield, but this had no significant effect.

However, adding the ModuleLiftingSurface definition does seem to help keep the pod reenter aft-first, although some tweaking will be needed since the pod and paradock still experience close-to-overheat warnings.

That's strange; when I changed the CoPOffset to 0, 1.1, 0, that did keep it oriented properly for me. I think I left CoLOffset alone. And I'm really not even liking that it's in there. Basically it moves the CoL to the bottom of the shield when it should be somewhere inside of it. But that's how stock does it and I'm really scratching my head as to why.

While mulling over this problem some things about stock aero have occurred to me that are really bothering me. If you look at a craft's overall center of pressure it's the sum of all CoP for all parts. But stack mounted parts occlude drag. Which does make a certain amount of sense except for the part where that means that it's cancelling out the CoP contribution of occluded parts.

I hope I'm misinterpreting something somewhere because that would mean that a command module type craft has its CoP entirely in the heatshield instead of the craft's centroid. It's no wonder we have to play games with the CoPOffset in the heatshields. That kind of tweak should be entirely unnecessary.

I'll try removing the CoLOffset and see what happens.

@Starwaster, I just tested the following:

CoPOffset only: pod orientation unstable during reentry

CoPOffset + CoLOffset: pod orientation unstable during reentry

CoPOffset + ModuleLiftingSurface: pod orientation stabilizes, but heatshield appears to tilt pod such that it is not fully pointing retrograde; pod and paradock experiencing overheat warnings

CoPOffset + CoLOffset + ModuleLiftingSurface: pod orientation stabilizes, but heatshield appears to tilt pod such that it is not fully pointing retrograde; pod and paradock experiencing overheat warnings

So it looks like ModuleLiftingSurface is critical to the new aero reentry system, but quite a bit of tweaking will be needed to bring the SDHI heat shield on par with the stock 2.5m one.

@Starwaster - I dun goofed.

If you look at the SDHI Heatshield CFG at any commit up to and including 30e9148, you'll find that there already exists some CoPOffset and CoLOffset values. In 30e9148, all I did was to add some new definitions at the top, not realizing the old ones were still there.

As of ecda9e1, I've made sure to remove the old values, and also adjusted the CoMOffset, since I noticed that the stock heatshield has its CoM based around the centroid of the shield mesh (whereas the centroid for SDHI's heatshield is up from the origin by 0.075m).

In any case, I ran some tests with the latest updated CFG, and it seems to be much more stable at pointing retrograde. Please test this out and let me know if it works okay.

looks good so far.

Bomoo also reports that it works on his end too, so I think we can wrap this one up.