Quantitative analysis of brass volume, weight, and velocity.


Well-Known Member
Jan 16, 2018
See this thread for what motivated me to undertake this.

Short story, had some extremely undersized virgin brass, and wanted to determine how this would affect processing and load development. There is plenty of hearsay and speculation on the internet, which leaves one with a "Chevy vs Ford" or "Coke vs Pepsi" outlook. To be fair, actually testing this rigorously is a tedious process, so I suspect that's why there's not much quantitative information out there.

The following is for one specific rifle (264 WM) and brass headstamp (WW). Take it or leave as you see fit.

For those that don't like wading through data, here's the summary of conclusions for this specific experiment with like head stamped brass...

1) Brass volume does correlate with brass weight.
2) Brass volume does not correlate with velocity.
3) Brass weight does not correlate with velocity.
4) Virgin brass produces slower velocity than once fired, though its volume is substantially less.

Experiment Equipment and Methods.

Rifle - 264WM. OEM Savage Accustock/Receiver, New 24" X-Caliber, 1:8 twist, 5R pre-fit. 26 rounds through it at start, 46 after experiment.
Brass - Win Western, neck turned, flash holes uniformed. Half were once fired, the other were unfired. Purchased long ago at auction, so at least 10 years old, probably more. All were run through an RCBS FL die, then over a Sinclair expander mandrel. The virgin brass were essentially untouched by sizing, except for the necks. The once fired were FL sized with an ~0.001 shoulder bump.
Load - 66.0 ± 0.02 gr H1000, Fed 215M primer, 0.03" off lands.
Scale - Sartorious, 0.0001 gram resolution (0.0015 grain). Actual reproducibility, 0.0005 gram (0.008 grain).
Chrony - LabRadar. All data verified based on the individual .csv tracking files. N = 7 once fired, and 9 new brass as tracking was lost or poor for 4 shots. The remaining shots had robust tracking for ≥ 130 yds.
Shooting session - 30º F, 1-3 mph wind, lying prone in snow with front rest and rear bag. Each shot alternated between fired and unfired brass, in 4 shot strings. Barrel was cooled with a blower between strings. Ammunition was all at ambient temperature.

Brass Volume Determination

1) SS pin tumbled in 2% Citranox and H2O for 1.5 hours, then dried for 12 hours on heat register
2) Once fired had primers from first firing still in. Spent primers were inserted in unfired brass.
3) Each brass was weighed empty, then filled with de-ionized water with a transfer pipette untill the meniscus protruded above case mouth. A micro-pipette was then used to draw off ~0.002 grams (0.03 gr) of water at a time until meniscus was flush with the case mouth. Empty weight was subtracted from full weight to arrive at H2O weight. Reproducibility was ± 0.05 grains.
4) Each brass was labeled, and all weights, volumes, and velocities were recorded for each individual brass.
5) N = 20 once fired, and 10 new brass

Results - Weight vs Volume

1) Brass weight does correlate with volume. The correlation is stronger with unfired brass, but both fired and unfired show a similar slope, which is on the order of the difference in density of brass vs H2O (~10 to 1).
Weight vs Volume.png

2) Brass volume does not correlate with velocity in the expected inverse manner, if you squint, you actually see the opposite trend. Statistically significant? Probably not.
Volume vs Velocity.png

3) Brass weight does not correlate with velocity in the expected manner, and if anything shows a reverse correlation as above.
Weight vs Velocity.png

Results - New vs Unfired

1) Unfired brass had substantially lower volume (~4%, or 2 grains of H2O).
2) Despite the lower volume, new brass produced lower velocity than the once fired (~0.5% or 15 fps).
3) ES and SD was within error at 33/11 (OF) and 39/12 (UF).
4) Accuracy and precision were within error for 4 and 5 shot groups, and conditions weren't ideal, as I was hurrying to get the experiment finished as it got dark. Unfired had 1.04 MOA 5 shot, and 0.58 MOA 4 shot, and fired had 0.98 MOA 4 shot and 0.9 MOA 5 shot.

Conclusions and discussion

1) I'm not gonna worry much about this extremely undersized new brass. I'm getting very good accuracy, reasonable velocities, and there's clearly very little difference, beyond a few fps, between the once fired and new brass down range.

2) Volume has little effect on down range parameters in like head stamped brass. There's no conclusive proof it measurably affects velocity. That said, there's ample evidence that different head stamps can have VERY different pressure/velocity for the same load, which is often reflected in weight and interior volume.

3) Brass weight does not correlate with velocity. I did a previous experiment, which produced slightly more compelling results, but not much.

4) QuickLoad (see table below) suggests the smaller volume brass should produce substantially higher pressure (4800 psi) and velocity (62 fps) for the same load. The opposite happened in this experiment.

147 ELDMVol (gr H2O)67.5 (P)67.5 (V)Max (gr)Max (V)
Virgin Brass
Once Fired

I can speculate that this is because the new brass expands to the same volume as the once fired (i.e. chamber size) in the first few micro seconds after ignition. This probably saps a little of the combustion energy, and results in a slightly lower velocity. Now, if the two different pieces of brass had the same exterior volume, which fit snugly to the chamber walls, but 2 grains difference in interior volume, then QuickLoad's modeling would probably hold up.

Hope you find this useful. It's saved me from undertaking a bunch of hassle and fire forming and various other voodoo, which probably would just eat up time and components for no perceptible gain.
Last edited:
Excellent post and interesting finding's, Hmmmm.....30 degrees and laying prone in the snow, you are dedicated and most likely younger than I :) It supports what I have always believed that brass weight and volume does correlate.
It was a tedious test, but I thought it might be useful to others. I might do a bit more next time I go out, as I'd like to beef up the statistical significance.

Here's some Quickload data for the 147 ELDM. QuickLoad suggests there should be 24 fps extreme spread between the high and low volume of the once fired, which is double my standard deviation on 7 shots, so a 95% chance I should have seen it if it was there.

147 ELDM
V (gr H2O)V (fps)P (psi)
ES Once FiredMin Vol
Max Vol

@Savage 12BVSS I see you actually read through everything to catch the snow and weather. ;) It actually wasn't too bad. Wind was calm, and I was dressed appropriately.
H2O weight isn't a value of volume, but for this exercise it works.
I measure the CC's of every case after it's weighed, as you found out, case weight means diddly squat to volume. I use a Burette measured in millilitres with a alcohol, isopropyl for measuring, no miniscus is formed with this liquid.
Also, belted mags are all the same, virgin brass is smaller in head to shoulder length by design, the belt is what it headspace off of.
Volume only needs to be measured after the first firing for this reason.

H2O weight isn't a value of volume, but for this exercise it works.

It is if the density of DI H2O is known...

I measure the CC's of every case after it's weighed, as you found out, case weight means diddly squat to volume...

Actually, if you read my OP, weight and volume were the only thing that did show a reasonable correlation...

(hint: It's the first numbered point, and first graph)

Words escape me as to express my amazement, gratitude, thanks and admiration for your efforts and the invaluable data you gathered and shared with us.

Very, very very well done indeed.

I am unfortunately at the financial stage in life where reloading tools like a Lab Radar, 21st Century concentricity gauge, a truly accurate powder scale (I have 2 RCBS 1010's to double check my loads and are about as good as I can afford) are tools I LOOOOONG and lust for but with sons soon to be off to and in collage for longer than I care to think such reloading luxuries will have to wait.

I printed out all you posted and will forever be in my reloading for bottle neck calibers binder.
I do read everything in posts and yours impressed, reinforced the way I do it and makes the work I do seem worthwhile. Keep up the good work you have one reloader that follows your posts.

Recent Posts