OzFclass

Bias alert – I have never believed that case weight batching is worth the effort. I think there are a lot more important things that can affect accuracy, see my previous rant here > viewtopic.php?f=5&t=13394

-----
I recently watched an interview with Jack Neary Interview (approx 17min45s in) where Jack claimed that weighing brass was bunk, not worth doing, makes no difference, it’s rubbish. Neary is a well-known US benchrest shooter, Hall of Fame member who should know a thing or two.

However I have never seen anyone actually produce any sort of analysis or evidence to back up their belief either for or against brass weighing. So, I decided to do exactly that – the analysis part at least.
I didn’t want to do endless trips back and forwards to the range, Labradar in hand with a variety of different weight cases in sufficient numbers to make any testing statistically meaningful. So this analysis is based on not actually doing any shooting because I would have had to do a large number of firings (20+ for each case weight) to generate enough data for the data to be significant.

------
Rumour has it that changes in brass weight affect internal case volume to the point that it will result in pressure changes significant enough to lead to velocity changes and therefore POI changes. This analysis ignores the potential for changes in harmonics affecting POI.

This analysis assumes that all of the weight variance in brass is accounted for by a change in internal case volume, that is all the case rims etc are perfectly identical. It also assumes that the outside of the cases measure identically which it is rumoured that fired cases are. This is the worst case scenario in terms of variability of case internal volume.

Cartridge brass is 70% copper and 30% zinc. It’s basically a fixed ratio irrespective of manufacturer. Cartridge brass has a density of 8.53 grams per cubic centimetre.

Water is ubiquitous. Pure water (or near enough to ) is the same everywhere. It has a density of 1 gram per cubic centimetre. That density only varies slightly with temperature and that variation is too small to come into play here.

This is all very convenient for figuring out changes in internal case volume with changes in weight.
Brass is 8.53 times heavier than water for the same volume and given that the volume of cases is measured in grains of H2O the conversion between the two is easy.
A case with a capacity of 100 gn H2O will have a capacity of 853gn brass. A change in case weight of 8.53gn brass will mean a change in H2O capacity of 1 grain.

-------
I took 30 new cases of Lapua .284 Win from a previously unopened box. I weighed the cases and recorded the weights. The minimum was 192.96gn, the max was 195.04gn for a variation in weight of 2.08gn (brass).

Using the trusty 8.53 constant I can now work out what the weight variance means in terms of gns of H2O instead of brass.
2.08gn divided by 8.53 = 0.244 gn H2O. I can now use this info to plug numbers into an interior ballistic calculator that uses grains of H2O capacity to predict pressure and velocity.

------
I tried Gordons Reloading Tool (GRT) and QuikLoad to make predictions based on the load data I was using in my 284. In both apps a whole lot of parameters can be changed to predict/estimate the effects of differences in load parameters. However GRT overestimates the velocity produced by this load due to the old (2016) powder model used. Quikload is spot on for velocity estimation. The calcs are therefore from Quikload. Having said that even though the velocity is off the mark in GRT, the changes in velocity are very close to Quikload.


Load data used:
.284 Win., 180gn Berger Hybrid, moly coated projectile, 51.5gn 2209. Nominal case capacity of 66.3 gn H2O. OAL 81.3mm. Measured MV 2790 fps.

Results:

The table below shows the effects of changes in brass weight from a starting point of 66.3gn H2O capacity (shaded column) . Each column to the left has an increased brass weight of 2.0 gn (smaller case volume) and to the right a decrease in weight 2.0gn



The figures use extreme deviations in brass weight illustrate the concept. A case weight variation of 8 grains yields a velocity change change of about 20 fps. A case weight variation of 2 grains (the original Lapua sample cases) yields around 4fps.

So barring the possibility that I have made a fundamental error in logic in my view weighing is not worth doing. If I've made such an error, I'm sure you'll let me know.
Pete

EDIT: I have replaced the table as I has it arse about

In my experience weight sorting isn't worth the effort because you never know where the extra weight is, on the other hand sorting by volume (with a liquid) if you have the time and patience may yeald some (very small) return.
Matt P

If you want some conclusive evidence test ADI cases against Lapua against Winchester and see what you come up with .You can then draw a graph of speed against volume and extrapolate for different volumes

Barry Davies wrote:If you want some conclusive evidence test ADI cases against Lapua against Winchester and see what you come up with .You can then draw a graph of speed against volume and extrapolate for different volumes

Barry
would be interesting but I only shoot 284, so no ADI or Win available.
Pete

I have done the tests . The result is basically a straight line graph connecting the 3 different
cases. Generally it all adds up to approximately 1 f/s / grain variation in case weight.
I'd not check by volume.
I guess this could equallybe applied to a 284 . The upshot is that a few grains variation in case weight is of no consequence.

I look at it this way. Once the projectile leaves the barrel, it is in the hands of the Gods, regardless of what we do before releasing the trigger.
Happy New Year all, and hope good scores are prevalent.
Cheers,
Trevor.

Another test you might try is to load the same case several times to the same specification, chronograph it and note the speeds.
Of course you have a different primer, different projectile, different barrel condition( unless you clean between shots)probably a slightly different neck tension and probably a variation from one extreme to the other in chronograph accuracy.
What does it all tell you?

Barry Davies wrote:Another test you might try is to load the same case several times to the same specification, chronograph it and note the speeds.
Of course you have a different primer, different projectile, different barrel condition( unless you clean between shots)probably a slightly different neck tension and probably a variation from one extreme to the other in chronograph accuracy.
What does it all tell you?

Barry that would tell me that variability is built in, it can't be eliminated. There are just too many factors at play.
For some shooters I have spoken to, batching by case weight is a holy grail of must do's. I am not a believer, but I had to try to find out.

The point of the exercise was to get an insight into how much of the variability in muzzle velocity could be attributed to case weight variation. Using the same case would defeat the purpose. However it may be a useful comparison point.

Say 10 firings of the same case v's 10 cases of varying weights (all other parameters being equal)

But as always a few problems pop up. 10 is about the practical limit of firings for a single case, but 10 cases randomly chosen would not make for a statistically significant sample size to compare against. That is, with 10 cases the data wouldn't mean anything because the population is large (all the cases you are going to fire) and the sample is too small to be representative.
I chose 30 in my sample of case weighings because statistical mythology tells me that a sample size of 30 is about where samples sizes begin to be significant. That is, it is likely that all 284 Lapua brass weight (from that lot) will fall within the data range of sample. But there will be outliers - some that don't fit into the sample.
The sample I used had a spread of about 2 gn brass weight. I examined the theoretical effect on velocity of case weights up to 8gn above and below the nominal case weight. This indicated a variation of 19 and 17 fps respectively.
Given that the chances of a piece of quality brass like Lapua being that far out in weight (8gn) are vanishingly small, I am very confident that weighing cases is unimportant. I can live with that.

In F Class, how many 5's do you want to shoot v's how many 6's do you want to shoot? None and all, right. You want 100%.
So what level of uncertainty are you going to accept in the culling of cases? I guess buying a entire production run and culling it down to a hundred or so identical cases might do it.

How deep down the rabbit hole do I want to go? - not far, I just want to buy good components and shoot them.


Pete

That's correct Peter, variability is built in and you have absolutely no way of determining just where that variability is.
You can spend much time trying to find out --- for what? A few ft/ sec which matters not for a properly tuned rifle..

Barry Davies wrote:If you want some conclusive evidence test ADI cases against Lapua against Winchester and see what you come up with .You can then draw a graph of speed against volume and extrapolate for different volumes

I have tested Winchester vs Norma 300WSM cases. The Winchester cases which were 20gn lighter required 1.0 gn more powder to keep the same velocity. Im not sure what the velocity difference with same loads was but someone here could work it out roughly.

RDavies wrote:I have tested Winchester vs Norma 300WSM cases. The Winchester cases which were 20gn lighter required 1.0 gn more powder to keep the same velocity. Im not sure what the velocity difference with same loads was but someone here could work it out roughly.

Rod
let me know the projectile, approx powder load/powder type and I'll have a crack at it. That's a big weight difference.
Pete

PeteFox wrote:

RDavies wrote:I have tested Winchester vs Norma 300WSM cases. The Winchester cases which were 20gn lighter required 1.0 gn more powder to keep the same velocity. Im not sure what the velocity difference with same loads was but someone here could work it out roughly.

Rod
let me know the projectile, approx powder load/powder type and I'll have a crack at it. That's a big weight difference.
Pete

220gn Sierra with 62.0gn 2209 with Norma or 63.0gn 2209 with Winchester brass for 2830 fps with each

Questions for you Pete:

What tolerances do you weigh your powder to?

In answer to your question, yes, I do, but these were habits set when we only had ADI components to use and the weight variation in cases, projectiles and projectile Ogive length varied significantly

Aus9914 wrote:Questions for you Pete:

What tolerances do you weigh your powder to?

In answer to your question, yes, I do, but these were habits set when we only had ADI components to use and the weight variation in cases, projectiles and projectile Ogive length varied significantly

Aus9914,
I think you misunderstand what I did to get the numbers. As I said in the post, I did this without shooting. It was an exercise to determine whether (theoretically) the weight variation within a single lot of cases would make a significant difference in velocity/pressure. It didn't.

Now if we take Rod's figures above of using different manufacturers brass (or different lots) it might well make a difference. I'll get those figures soon.

Given that the velocity predicted by QuikLoad is always spot on for me (within 5fps), the methodology behind it must be correct. Therefore if the velocity is correct then the factors that determine velocity (pressure and volume) must also be correct. Change one of these and the velocity will change but by how much was the issue.

-----

However to answer your question if my target weight is 51.5gn, I accept 51.48 or 51.50. That is; to within a kernel.
This is probably in the same OCD dept as weighing brass but I have a machine to do it automatically for me.

Pete

If you'd like to spend some time Pete, you could do the following :

Take say 30 fired cases, not resized but uniform trim length and clean (inside and out) and number them. For each case put a used empty primer cap in it, weigh it empty, then full of water and record both weights. You can then use these figures to graph how volume varies with brass weight. Importantly, see if there is a linear relationship between volume and brass weight. If there is then you can get by with just weighing the brass to determine the volume.

My suspicion is that different brands and batches of brass will give different results. For example some may have most of the weight variation in wall thickness (good for our purposes), and some may have more variation in head and extractor groove diameters (not so good).