>>>radial expansion, spread over the entire circumference of the case, is minor (only about 2% when we expand from .450" to .460" diameter, for example - and only on initial firing unless we FL size), while axial stretching, confined to a narrow band (about .1" wide at most) somewhere ahead of the transition between wall and web, is much more significant and accounts for virtually all the cumulative thinning resulting in eventual separation.<<<

Please show us the number that you base the 2% upon.
I guess I disagree with your conclusion. When the entire .010 difference in diameter is accommodated by localized stretching of a case the amount of stretching .010 is significant to the thickness of the brass - normally near .030 thick.
Let me use a couple of extreme example.
1. What is the usual result of firing a 7.62 Nato round in a .30-06.
You get a blown out .308 case with no neck. There is usually no separation reported even thos the .308 case had little real headspace control.

2. What happens when a .270 is fired in a 7mm magnum?
Usually the rifle is wrecked because the case could not handle the radial stretching.
These are extreme example of case missfits but they serve to show the critical nature of having the case head fit the chamber closely.

There is another example of case to chamber misfit that exists with the 6.5 Arisakaicon. The Type 38 rifle is know to be very strong. It would exhibit very little spring with high pressure loads. After the first fire forming load it would have a fit from head to should to control the headspace. Yet ammo fired in these rifles give head separations due to the thinning of the brass where a huge step is formed due to the mismatch of the case head and the chamber.


One of my reasons for focusing on radial thinning is that on a an experimantal basis it is possible to use brass to eliminate most of it. By using larger brass you are left with only axial stretching forces.