There is an excellent article in Handloader #288 on cartridge expansion as a method for determining pressures. It is written by John Barsness, and is full of excellent information.
Since we have discussed this method of load development here before I thought I might share some of the key points the article makes in using this method to work up a load where there is no load data. It would also be great for those who feel current published data may have been lawyered down (an opinion I personally don't subscribe to, but if you are going off book, this is a good method to help you keep your fingers).
First, calipers don't cut it. You need a micrometer. You are going to need to accurately measure down to .0001", and your calipers just won't cut it.
Second, you need a reference load that is loaded at or very near max pressure.
Third, you need a supply of new brass of the same manufacture as that used by the reference load requirement #2. The article specifically calls out a test with the same load with rifle brass in 264 Win Mag. Factory ammo was pulled from brand X and the powder and bullet were loaded in brand Y using brand X primers. Brand X brass expanded .0005", while the softer brand Y brass expanded .0020" with the same load. Lucky for us, most of the guys loading actual full power ammo use starline brass, of which we can usually get a good supply.
With these requirements met, shoot a 10-20 round batch of the reference load in the target gun across your chronograph. With each shot, record the velocity and the case head expansion. When you are done, average the the values and now you have baseline case head expansion numbers for that brand of brass at SAAMI spec loads.
Begin working up your handloads. You are after the same averages, without ever exceeding the max recorded case head expansion. Continue to chronograph loads as well, and monitor other possible pressure signs, like flattened or cratered primers, smiles, etc.
Using this method it is possible to work up new loads safely, and closely approximate lab test results. The author of the article cites several cases where he was asked to get a "scoop" for Handloader on a new cartridge by being the first with data. In each case he used a variation of this method and produced data that fell within a half grain or so of the published data once it was released.
Grab a copy and enjoy!
Safe Reloading.