Can case volume be inferred by its weight ?

[Brad Y]

Gday Dave and welcome back.

So reading between the lines would you agree that if you purchased a large batch of cases and could weight sort them into very close sub batches you would be more than happy to use them and not go to any finer batching methods?

[Brad Y]

There’s a few things I would take from this test if I were going to do my own. But leaving the primers in there wouldn’t be one. Those swish 21st century primer plugs would be my choice. Even though his test with the single case had 10 repeats, the trend was looking impressive for how he could reproduce the weights. Food for thought if anyone wants to do a repeat at it.

[ajvanwyk]

Dave,

Thank you for sharing! Peter made his paper available in the thread on SD/ES reduction (I believe) and was certainly an interesting read. My conclusion, which may or may not be supported in fact, is that the pressure variation and resulting velocity differences get lost in the noise of all the other variables.

Your graph may indeed also support one of my observations when sorting by Volume, as I initially started with cases sorted to within 0.01 gram, I noticed very little difference in volume and concluded that the required time /effort to perform this task was not worth it in the majority of situations. The spread was 0.4gn across 150 cases.

I am interested in an earlier comment, where your analysis saw larger volume variations resulting in a similar velocity variation to my calculations. I've placed the QL screenshots below, as a comparison only and would be grateful for your insights.

Screenshot_1.jpg

Screenshot_2.jpg

Screenshot_3.jpg

Albert

[DaveMc]

Brad - Our conclusion at the time - weight sorting is the be all and end all of volume sorting. Once we found an accurate volume measurement to test with (metho and primer pocket plug) it showed clearly how accurate it was and how inaccurate some of the other methods were. It is also soooo much easier to implement. It is the ONLY method I personally would use to volume sort the brass - but with a few of my personal caveats and only for long range work
1) It must be from a single batch of brass to avoid any discrepancies in case head dimensions or brass density. DO NOT MIX BATCHES. (some different batches would be ok to mix if they have uniform brass density and case head dimensions but determining this is a difficult task.)
2) It can be done on new or used brass but ideally the brass is trimmed first, (neck turned, primer pockets and flasholes uniformed etc if that is part of your regime)
3) Beware of uncleaned brass (it was surprising how volume/weight and variation in internal powder/carbon deposits in a case)

I am also very happy with the bb or #2 lead primer pocket swage/plug, the 21st century ones look great too. But anyone wishing to do a volume test with water, alcohol or spirits etc - please hone your test on one case first. in fact I would recommend you get 10 cases and weight them with filling each one 3 (or more) times with fluid of choice. Weigh the initial filling and compare to the subsequent ones as the case internals are wet out. I will do that today for a few cases and post the results. BUT you need to do this to validate your own methods.

In all these methods there are always several sources of variation and several sources of measurement errors.
1) Measurement errors - e.g. the scales will have some measurement errors no matter how accurate. The liquid filling will have some measurement errors due to bubbles (micro or macro), meniscus or leveling errors (no matter which technique) etc.
2) Variation - there will be variation in brass uniformity (head dimensions, brass density, trim length, primer pocket etc etc. There will be variations in liquid density over time and temperature etc etc.

All these variations and errors add to give us variance in the data - well they don't add actually - they work together like this. You sum the variance of all the variables (s.d. squared) and then square root for the standard deviation of the end result.
ie s.d.(population)= sqrt(s.d.(variable1)sqd + s.d.(variable 2)sqd + etc etc).

In the above test we measured the one case 20 times and had a maximum error of 12mg (0.19 grains) initially but reduced to +/- 6 mg (0.95grains) with improvements in technique. So the combined variation (measurement and other) due to scales, metho density, technique etc made up a significant proportion of the variation in that first graph. In other words the relationship between true case volume and case weight in that brass is a lot better than the relationship between brass and metho weight.

Now from our testing looks like brass weighing (if done correctly) is actually more accurate than most of the other methods I have seen. - It is hard to improve on and if you want to try and find a better method then please verify your results - test test test - verify verify verify.

[DaveMc]

Hi Albert,
Ok first can we please clarify- is this a 284 case - the volume looks rather high? and a 30 inch barrel?

[DaveMc]

First comment relates to % change - if using a SAUM case compared to a 284 there will be less % change in a SAUM to our results. or 308 cases for same volume increase but this is marginal.

I ran the same test in Gordons software last night and his gave me a 10m/s (33 fps) change with every grain in a 284 case with case volume around 66.5 grains H20. His software has seized on my computer now and can't shut it down so will leave to someone else to try. lets say he shows an increase of 3fps with every 0.1grain of case volume (approximately 1 grain case weight).

We did several tests at the time. Unfortunately due to chronograph reading errors the correlation was only so- so but all tests still showed a consistent relationship (slope) of around 2fps per grain case weight with Rsqrd values around 0.3-0.45. This is what needs repeating with a labradar. I verified at the time on Quickload and also showed around 2fps per grain 2209. - sometimes remember that we will also start only burning a percentage of the powder before exit so there are limiting factors on muzzle velocity.

Could you please rerun with a 284 case and 180 vld bergers and 30 inch barrel - as that is what was used at the time for our testing. Also do a bigger variation (+/-1 grain in volume) - say 65, 66 and 67 grains H20 volume and 51 grains 2209 so we move away from rounding errors. Thanks.

[ajvanwyk]

Dave,

Happy to do that.... BTW, GRT does freeze from time to time and I guess is part of the teething problems of a beta, still great though !?

You are correct in the fact that my calculations are for SAUM, 183gn Sierra with 2209 and a 31.6-inch barrel. I agree that this would provide variations and actually noticed that by also changing the powder in this combination....

Based on the supplied criteria, I get the following for a .284... (Note: the COAL in the system is very short, 2.800". I therefore, increased it to 3.000 but was only a guess as I don't have real-world experience with a .284)

284.jpg

284 1.jpg

284 2.jpg

284 3.jpg

[Gyro]

Are the velocities going to be checked from the volume/weight checks and if so I suggest they really need to be done from a machine rest ? Why not just get agreement on the volume checking method then leave it there ? Regards the correlation between the case weight and its volume then seems unless we are to check each and every case we need to "assume" that's not an issue ( especially in the same batch ) and I presume many are doing that already ?

[pjifl]

It is great to see the discussion getting serious again.

Just to jump ahead, it is very easy to use a ballistic program such as

https://bergerbullets.com/ballistics-calculator/

to predict that a 20 f/s increase in Velocity from 2820 f/s should place a 180 Hybrid 7mm about 1/2 minute higher on a 1000y target.

That is outside the X ring. And remember that one needs to stay near the middle 2/3 of the X ring to have any tolerance to wind reading errors.

OK - we may have partial compensation, and the effect may be swamped by atmospherics on a poor day or shooter/rifle errors but over 30 or 40 shots the effect surely has to be considered. The game is now about who can claw more Xs in their score.

Just because you may not be seeing the effect of 20 f/s change on the target, does not mean that someone else is not benefiting from better Velocity control.

For now, best simply concentrate on the wt/vol problem but it is not a futile quest. If nothing else, understanding these relationships allows us to predict what is a sensible tolerance.

[DaveMc]

Thanks Albert - that is almost perfectly in line with Peters and my earlier real world observations, case weighing and metho measurements as well as chronograph results (magnetospeed) despite significant variation the slope was always very similar. A labradar and well controlled velocity observations against metho and case weight should indeed show this (but needs to be done to put it to rest once and for all. - Happy to do it on a machine rest if it reduces one of the variables although in our work a top level shooter, heavy rifle and decent rest can achieve very low standard deviations anyway. What this needs is a large sample size to be defining and conclusive.

But all evidence so far points to: 2fps per 0.1 grains of case volume or approximately 1 grain of case weight in a 284 case with 2209 powder.

Case weighing certainly easier than any other method and less room for error and will give you the results you need for long range work.

I will point out though that many batches of Lapua brass I have measured for my 308 had under 2 grains spread - for me and my 308 shooting that was not worth the effort to sort.

I only do this for 900-1000 yard + shooting.

[Gyro]

I think it's important to remember that the 284 variants and the 7Saum ( and long bullets ? ) are NOT famous for low ES/SD numbers anyway, which is a detail that perhaps screws with the tests we are discussing here ?

[DaveMc]

I could argue differently Gyro.....

The sd's that were being achieved by some of our top level SAUMS (and 284's) were truly amazing - massively improved by Labradar testing.....It is one of the reasons they have become the go to caliber for 1000 yards - not just wind bucking, accuracy and low recoil/manageability but superb elevation. 308 s a great caliber too - I had the luxury of seeing data from 30+ of our top shooters in both fields. sds of 2-5 were being achieved by a lot of shooters.

[pjifl]

When I was shooting at my best, and with barrels in their prime, I was achieving a Vsd of 3 to 4 with a SAUM. That was usually taken with a 12 shot string. 2209 powder.

Overkill for the shorter ranges.

Peter Smith.

[UL1700]

I think it's important to remember that the 284 variants and the 7Saum ( and long bullets ? ) are NOT famous for low ES/SD numbers anyway, which is a detail that perhaps screws with the tests we are discussing here ?

I don't have experience with the 284 but the 7SAUM one should be achieving single figures with ease, I guess it depends on you definition of low? Our Dasher achieved a 12 shot ES of 3 last year where as the SAUM is still up around 15 but it only has 100 rounds through it so still plenty of work to do.

[DaveMc]

But Gyro - yes we would need to do this with good gear and good loading techniques to get the variation as low as possible and achieve a strong correlation coefficient.

Alternatively we could do as you suggest and just be happy that there is a strong correlation between weight and volume (within brass batches) based on good accurate alcohol or spirit weighing - then rely on the quickload data and internal ballistics to tell us how much that will effect your combination - then make a decision as to whether it is worth it for you.....that is also a simpler path forward.