I think the method I recall from Sinclairs loading hand book or some such was something like: (bullet diameter) + 2*(neckwall thickness) - ('n' thou desired neck tension). A real world example might be better. For my .308 Win, I'm using the following numbers:
.308 + 2(0.011) = 0.330
0.330 - 0.002 for neck tension = 0.328
Some people subtract a bit more for neck tension, but add in some (1-2 thou) for brass spring-back.
I normally just go w/ the numbers like I did above, and then order that one, and the size larger and the size smaller, so I have 0.327, 0.328, 0.329 bushings. Figure it's one of the less expensive loading items I have, and it's useable in any other .30 cal I have that I get a bushing die for.
Last summer, I spent an hour on the phone with Patrick Ryan at Redding on this subject...
I had just gotten a .40XB-300WM, and picked a bushing that was 0.003" smaller than the loaded neck.
But the sized cases ware WAY under the bushing size (now that can't happen, can it??) How can a 0.338 bushing give a neck OD of 0.333?? So I figured that the bushing was mismarked... and took it out and measured it. 0.338". I put the bushing on the sized neck and it rattled
That's a very badd thing!
Pat first said he hadn't heard it this, and asked for the neck size of the fired cases... and after an hour with Pat, we came to following conclusion.
It turns out that the recomendation for 0.002" under is based on cases that have been shot in tight neck chambers... the type most accuracy nuts use. In this situation, the neck only expands maybe 0.001" over... or may still be UNDER, and still be tight on the bullet, after firing.
But if you use a chamber that is the typical 0.005" to 0.010" over the loaded round (like most field rifles)... the neck expands to fill the chamber and is now quite larger.
When put through the bushing, the radius of the bushing (R=0.018"), causes the brass to first flow at a 45-ish degree direction to the center line... and when it gets to the bushing bore size, it overshoots several thou... thicker necks cause more overshoot.
So you can size a 308 cal case with a .338 bushing and wind up with a 0.330 OD neck
I doubt that there is a way to buy only one bushing... so except that several bushings just come with the territory
I settled on turning the necks on my 300 Ultra a couple thou thinner to work with the bushings I already had, cheaper at the time. My total chambers neck clearance was .006" and now it's about .008", so brass will work a little more... a consideration. Tight head space will keep things centered up with the bore still, so that part I'm not too worried about.
I figured the same thing was causing this too, it just makes perfect sense the brass would not just stay tight to the bushing bore as it rounded and entered into it. Various chamber overall clearances most definately explain the differences amoung everyones bushing size selection, I thought it was my new brass that was softer keeping things at the same OD as the bushings ID in the beginning. The new brass was already sized though, the expander had only been run through it to iron out the dented in necks so the cases entered the bushing very straight. This is likely why they stayed the same as the bushing size when sizing the first time and why they did not after firing and sizing again. Another reason for a good match chamber.