That is why I am adjusting the energy density a bit. The predicted velocity using the IMR4895 model was higher than actual chronographed average of 5 I know we are admonished not to fool with powder characteristc files, but going from 4080 to 4000 kJ/Kg is only a 2% change and it fixed the velocity error.
The other thing I tweaked was starting pressure. My rounds are not crimped in, and the 168 gn projectile I am using has no cannelure, so I thought their standard 3856 should go down to about 1000 psi.
I left the other characteristics alone except solid density. I was trying to match beam scale weights and bullet lengths to % of case fill, which I had measured carefully using H20 and weight.
I wanted to get some feedback from others about the change to see if that was kosher. As you know RG4895 does not have a powder characteristic file. Its source is Radway Green pull which could have come from a bulk powder lot, not canister grade.