OzFclass

Because no one seems to want to stick their head up and give a definitive answer, I decided to design my own investigation into the question
" Does the weight of brass case have a direct relationship to its volume, and if so can the case weight be used to batch cases by volume"

This is an important question because case volume is an important factor in the safety, velocity and accuracy of hand loading. The problem is that measuring case volume is tedious and prone to inaccuracies.
I set out to eliminate some the tedium whilst looking for a relationship.

I didn't set out to actually measure the volume of the cases I used. The data given aren't measures of case volume, please don't tell me the data are wrong and Lapua cases are bigger than that, the volumes I measured are not the volumes of each entire case, the reasons will be come apparent.

The equipment used: 30 x Lapua .308 three times fired in the same rifle. Cases are neck turned, trimmed to 51.1mm, have been de-primed, primer pockets cleaned, and neck sized with a Wilson bushing die. An A&D FX500i scale set to weigh in grains (gn). Resolution is 0.02 gn. I used water heated to 25 deg C initially that measured 22 deg C at the conclusion (I heated the water because I would have my hands in the water for an extended period and its cold in Tas) The change in density of the water was 0.08% over the time and is ignored.

Method and comments
I selected 30 cases randomly from a cohort of 200 because 30 is the normally the minimum statistically significant sample size.
I wanted to eliminate the influence of the meniscus on the results so I devised a simple stand that enabled the case be weighed neck down on the scale. It was simply a 50c piece with a 144gn ADI projectile glued to it.
I placed each case individually into the tub of water, swished it horizontally 4-5 times in an attempt to eliminate bubbles and then fitted the stand into the case mouth. I removed the case from the water neck down and dried it using a super absorbent cloth. Most and sometimes all of the visible water beaded off the case and the cloth easily wicked any remaining water away.
At this stage the primer pocket was still full (if it wasn't full the water was tipped out and the process restarted) and this water was removed with a cotton bud. Meniscus basically eliminated.
The seal between the projectile and case was very good and no overt force was used to seat the projectile into the case neck. After a minute or so in some cases a bead of water would leak but it didn't influence the results because it was still weighed with the case. see photo.


Now I realise ( as stated above) that I didn't weigh the entire volume of water that would fit into each case, but the volume not measured due to space occupied by the projectile is the same in each case, because all necks were resized using the same equipment, any variation I am treating as minimal and ignoring and in any case is much less than possible variations in a meniscus.
The data was entered in a spread sheet and the net weights calculated, graph produced and R[sup]2[/sup] value calculated.

Results

The results are given below in the table and graph.
Note that if you look along the 48.0 gr line at the bottom of the graph, almost the entire range of case weights are accommodated. This means that cases weighing in the range of 172 ~ 175 gn the water capacity can be approximately the same
The Pearson correlation coefficient R[sup]2[/sup] value for the data is 0.




.

Conclusion
The R[sup]2[/sup] value is zero. This means that there is NO relationship between case weight and case volume, at least for this batch of .308 Lapua large primer brass.
So weighing brass is a waste of time in terms of inferring the case volume.
The data also show the irrespective of weight, that the maximum volume variation was 1.04gn and 0.74 gn with one outlier omitted. Is the close enough? If so, just use the brass as is.

It is possible that there is some tolerance stacking in some of the data but I am confident (unless I have made a huge error in logic) they are as accurate as can be. These factors could be, change in water density and variation in projectile seating depth.

Pete

Nice work but you would have to be confident about your seating depth. For the sake of simplicity, assuming the plug is a uniform .308 diameter then you get a variation in volume of 0.0048ml or 0.0742gn per .1mm variation in seating depth

Π*(.308*25.4/2)2*0.1 = 0.00481ml
0.00481*15.432 = 0.0742gn

This means that in the extreme, a seating depth variation of 1.35mm would cover the full spread of H20 data

1gn/0.0742 = 13.5
13.5/10 =1.35mm

James

UL1700 wrote:Nice work but you would have to be confident about your seating depth.
James

James,
It was always going to be the sticking point. I tried to give all samples exactly the same treatment.
Initially I tried to estimate the meniscus on the cases sitting upright, but decided that the problem was always going to be the estimation, after a while they all started to look the same. Using your 0.1mm figure, who could estimate a variation of 0.1mm in the height of a meniscus. It would amount to the same error that you have posed.
I decided that a measurement was going to be more accurate than an estimation. Because all the necks were the same, any extra insertion of the projectile would require extra force (which I’m confident I didn’t apply). Failure to insert the projectile far enough would mean leaks.
Because any extra force would wedge the projectile into the neck due to the taper, that extra force would mean that the projectile would be harder to remove (I didn’t notice that either).
I’m relaxed about that potential for error.
Pete

Nice work Pete. I haven't studied your method yet ( and need to ) but if u have discovered another wives tail in this game then I will probably be the last person on the planet who's surprised.

Hi all, been a while -

Peter Smith, Marty and I did a significant in depth study of case volume some years ago (for the 2013 world championships). - including several methods of measuring volume vs weight. Some included the popular methods at the time, water weight, metho weight, powder weight , micro steel balls, etc etc and plotted against case weights. We also did some chronograph work with magnetospeed and velocity against measured case volumes and case weights. Peter has published a little bit of this data and the evidence was very strong for a good relationship between case weight and volume. (within a batch of brass at least). It is also by far one of the easiest methods to undertake. I will let Peter wade in here at his own time if he wishes but we had coincidentally just been rehashing this exact discussion.

We were thinking of revisiting some of it now we have the ability to measure velocities with more accuracy (labradar) and maybe this conversation and boredom may spur us (or me) into action to finish the work and put this baby to bed once and for all.

In summary - we found a strong relationship between metho volume and case weight. We had difficulty with almost all other volume methods. Water worked but was not as accurate due to bubbles and wetting of the inside of the case uniformly - metho wet the case out more easily. Cases need to be fireformed and ultrasonically or ss tumbled to remove carbon buildup. Peter succeeded in painstakingly measuring powder volumes but I struggled with that method.

The method I used was
1) fireformed, trimmed and neck turned cases (not in that order)
2) Ultrasonically cleaned (or stainless steel tumbled) and dried.
3) A #2 lead shot (or BB) was seated in the primer hole as a seal using a primer seater.
4) The sealed case with shot seat was weighed.
5) then the case filled with metho and drained (to wet the walls and corners) and then filled again - watching the meniscus carefully with loop and small wicking ss wire was used to adjust miniscus carefully and watch the evaporation drop the meniscus. and case weighed again.
6) I cross checked my method by doing 1 case 20 times.

Our work showed it was a worthwhile task for the high end shooters to sort their brass by weight within a batch of brass. In fact there is significant evidence to say weight sorting is more accurate than most of the other methods we tried. (sa long as it is within a batch of brass so the case head dimensions are similar.

It is good you are trying new methods Pete and I would encourage it. Your conclusions may well be right - for your batch of brass and your methods- but we will present some other data that had significantly different results. This is not unusual but what results to draw from it?? It is hard to draw results from a "no correlation" result - It typically means the variations (in measuring, techniques or other uncontrolled variables) swamp the relationships. Strong correlations however are very good indicators of the patterns.

Great work Pete. The only difference I did was sit a penlight battery on top of the case when the surface tension bulged above the case in order to achieve repeatable volume measures. Been down the path of the ball bearing, it works but butter fingers me kept stufing up. Great to see DaveMac enter the conversation. Lab Radar has been a game changer for the statistician. The metho method Dave uses and his method are sound. Your conclusion about volume matters Pete. It fits nicely with examination of pressure vessels and flow on effects in application transferable to rifles.

These two scatter graphs of the metho method before and after cleaning tell a BIG story. I had similar poor relationships with water and uncleaned cases as well.

The first correlation coefficient (R) = 0.88 (Rsq=0.79) with cleaned cases shows a strong negative relationship - ie increasing case weight indicates decreasing case volume. It should be noted that this was on a batch of winchester 284 cases with 5 grains of weight variation. A good batch of Lapua 308 cases may have under 2g of weight variation to start with with and therefore harder to get a good correlation over a narrower window.

Secondly the poor correlation against uncleaned brass (Rsq= 0.21) shows how a small variable (carbon build up in case) can affect such experiments. It is vital to control the variables before accurate scientific results and conclusions can be made. We cannot get much out of the second graph - apart from to say either unclean cases affect the metho measurement, or the weight to volume relationship or a combination of both.

What we can say though is the first stronger relationship is very solid evidence for a strong relationship between brass weight and case volume in that batch of winchester brass.

I recently volume sorted 200 cases using metho as a medium. i weight checked them prior to this test and the weight spread was 4.5 grains. For both tests the cases were ultrasonic cleaned and trimmed. When volume checked the spread was 1.2 grains, I made two groups with a spread of .6 of a grain in each group, interestingly the two groups split the total 50%. What does all this mean???? I will leave that to the more scientific contributors to work out.
Keith H.

To what extent does these small variations in case volume have on muzzle velocity? Can the best of our shooters read the total range conditions for every shot in the competition to the extent, that a few f.p.s. difference in velocity affects their score. I personally do not think so. But I may be wrong. I have shot some small groups in 1000yd. B/R. with un-sorted Winchester cases which are probably the worst for case variations. I think that we are deluding ourselves with all of this looking for something magical to happen to our shooting results. Anyhow it gives those playing around, and us others reading the results, something to do in the spare time we now have on our hands.
Keep safe,
Cheers,
Trevor.

Old Trev-39 wrote:To what extent does these small variations in case volume have on muzzle velocity? ...

This was discussed a few days ago on another thread. Reducing muzzle velocity variation is the primary objective of this research, so its a good question. Albert asked Quickload for a figure and for his SAUM loads it was about 1 fps increase for each 0.1 gn decrease in case volume.

For anyone who weighs cases, perhaps a good test would be to weigh about 200, take the heaviest 10 and the lightest 10, load them all the same, and measure velocities with LabRadar. Uniforming and cleaning the cases prior to weighing would no doubt improve the precision of the experiment. If there was a significant velocity difference between the high and low weight groups, then it would give a clear indication of what sort of velocity variation improvement you could expect for YOUR situation.

When we did this testing I also checked on quickload too for verification. All concurred that we were looking at an effect of about 2 fps (1.5-3 depending on cartridge, barrel length, powder etc) per grain variation in case weight. This is a little more than Alberts Quickload data but less than the article referred to in the previous posts. Peter did load some of the lightest cases and alternated his shooting against some of the heaviest and had some impressive variations in velocity that fit the conclusions well. So with batches of brass that only vary by 2 grains it may not be worth sorting for most - and certainly not for shorter ranges. Sorting brass that varies by 5-7 grains for long range shooting?? Up to you but it is one of the easiest sorting jobs in the game and only needs to be done once in the case prepping stage (Alans experiment - above is spot on - just alternate a heavy with a light to nullify barrel heating issues and post the plots.)

Sorting by volume with powder, water or spirits is however much more difficult to do correctly. I feel most people that sort this way are not validating their results and may actually not be sorting at all if doing incorrectly. Just filling cases with water or powder and weighing the results is likely to give you larger variation than you are trying to sort by. Check your method - do the same case several times and see how much variation you get.

I've sent enough rounds down range and for me I noticed a difference in x ring count . But that only works like anything else , when all of the other factors are in line . Otherwise it's a waste of time .

And wow , Dave Mac is back ! Congrats on your Dad winning the Qld Queens , could not have happen to a nicer bloke !
Regards Graham.

I just cross checked some of the figures in the new "gordons reloading software for 284 winchester and 2209 and for a 1 grain variation in case capacity his variation was 10m/s (33fps per grain H20 capacity)

Hi Sellsy, yes - Peter Smith dragged me out of my slumber on this one.... mainly because we did so much on it and really can put it to bed if we write it up correctly.. The missing pieces would be some nice verification with the labradar and some sorted cases which was not available to us then...and Dad has been shooting well - still got it - lucky enough to score some decent hand me up Barnard Saums from his son

Great insight Dave with regards to heating issues. These small variations have an impact when coupled with simple mistakes such as changing the head space by incorrect sizing which compounds pressure issues in a rifle that is not tuned specifically for a range such as 1000 yards. QL can be used to explore the pressure vessel effects, and results are significant. Trev, does BR have shorter strings? Heat issues might be less of a factor. Is 1000 yard BR range specific in that the same setup is not used across the course? Any mild positive compensation will assist to take the flutters out of a small velocity variation. How well that load performs at all other ranges will vary depending on the barrel lift profile. Why would we leave things to chance? We need to have confidence our gear is right so we can just concentrate on conditions. That means any variation is more likely to be associated with the conditions and we can tuck that into the memory banks to use to our advantage in the future.