Still here

Hornady 2266 into water. 18" of penetration and dented the back of the last milk jug.

55gr, boat-tail, soft-point, w/cannelure, core is lead and just barely soft enough to scratch with a thumbnail. The apparent thickness of the shank is due to the bullet being shaved on the bias.

Bullet fired from a 16" barrel. Loaded in front of an appropriate charge of an appropriate powder. Guestimated velocity at impact of 2900fps.

The next item on the agenda is to stair-step down the load and find out where my friend's AR stops feeding reliably. We will be looking for the min-load where 50% of the rounds and 10% of the rounds don't feed.

The plan is to use H-4895 in this test because that powder is "friendly" down to 65% case capacity, or about 16gr of powder. I expect the gun to cough up a hairball well before we get down to 16 grains of powder.

One of the variables is the shooter's mass that is coupled to the body of the firearm. The worst case, and the easiest to ensure consistency, is to hold the firearm as loosely as possible. That is, limp-wrist it.

The purpose is two-fold. One is to back-engineer what the fine folks at DPMS design their ARs to. That might be useful if I ever decide to build my own rifle.

The other is to have a baseline before my friend starts junkifying his rifle. Suppose the rifle, as designed, will cycle reliably down to 20 grains of powder and my friend installs a new buttstock/buffer assembly. Suppose that after that modification it pukes at 24 grains. Since a very "typical" load is 25  grains of powder it may cycle just fine when everything is nominal but my friend just gave away 80% of his margin; margin that can be eaten up by running a little bit dry or a little bit of dirt.

Any comments from folks who have real experience with ARs will be much appreciated.