I had a partial case head separation using a 1.8oz buffer

[DeNinny]

Nope. If the case fails prior to pmax, you still have considerable amounts of powder that has not been converted, and now you just provided a very tiny aperture for it to continue its burn through.

Yup. Because that pressure is just going to keep tearing the aperture open more and more which will keep the pressure down. That torn aperture is even weaker than what is was before it tore, so any pressure that builds up will simply tear it more and dissipate in the process.

The powder burn wouldn’t magically stop once the case ruptured in the imaginary event you’re describing.

I never said that powder burn magically stops. See above explanation of reality. The burning powder will continue to create enough pressure to continually tear the ever weakening brass.

It would continue to be converted in that tight space formed by the case in-battery, the bolt face, barrel extension, which is why you see in-battery failures turn AR-15s into grenade-like devices.

Nope. At least not always. In this failure it continued to uncork the bullet and tear the brass at the feed cone score line. Which is why this failure only blew out the magazine and didn't grenade.

Remember that the projectile would still be static or in a retarded start-pressure position in this model, obstructing the bore now that a new point of relief has been found.

This is the crux of where I think we are in most disagreement. I don't see this happening because again the gas will have had to go down the barrel, up the gas port, and return back to the gas key for the bolt to even begin rotating off the lugs which would be the start of this failure model. It doesn't do this until the bolt carrier starts moving, basically. And by that time the pressure of the gas is way too low to cause a failure like OP's or any failure that physically tears a wall in the casing. At best the extractor is just going to rip the metal where it is physically in contact with the brass. But up into the head sidewall and take out that flap? No way. That model is not possible if that same area of brass already survived a higher pressure earlier. And it did by the definition of this failure model.

Also remember the brass is designed to resist pressure up near pmax and preferably above it. So any pressure well below pmax will have a very, very, very low likelihood of rupturing the brass especially if the location isn't even in contact with the extractor.

I'm sorry but this is why the pics don't line up to the extraction failure model. Not how I see them at least.

I'll say again,

This is why bolts, barrel extensions, carriers, and upper receivers turn into fragments with an in-battery catastrophic failure.

The tightness of the space doesn’t allow clean, well-guided venting of the continued burning propellant.

I'll say again, you are assuming a specific failure does the same or similar type of damage. And again, there are other viable failure models within the laws of science where the damage is just at the point of failure with only limited "damage" like the OP experienced. There are going to be varying degrees of a "blowout" basically. Because there are a lot of variables involved.

In the OP’s demonstrated failure, there is no soot residue emanating from the failure point, which means that the propellant was already converted.

This fact does not refute the blowout failure model. I'd venture to guess that the blowout in battery failure model is going to occur very close to Pmax and so of course I see that all the propellant is nearly converted if not fully. This point is in full support of the theory. And pictures.

If you have ever seen in-battery failures, there is a ton of soot around the chamber and remaining parts fragments.

Yes, because these particular in battery failures had the conditions for that. I've seen them too!

His brass is squeaky clean, with a very coherent failure node aligned perfectly with the extractor.

The brass can be squeaky clean with a blowout as stated earlier. And the failure node is aligned perfectly at the feed cone score line at the extractor side, simply because at the extractor side there is less support to resist the brass as compared to the lug sections. Again, the brass can expand the most at the extractor.

The pictures of the failure indicate all this exactly.

The fact he was using a 1.8oz buffer should put all of this to rest.

That is early unlocking all day long, a literal recipe for it.

The fact that this failure aligns with exactly what Bill Alexander designed around and tried to prevent with his own grendel barrels should put all of this to rest.