Experiments with a Chronograph.

[Peter M. Eick]

I decided to try and answer a few basic questions today with my Oehler 35p chronograph. Each experiment will be documented separately, but in general, all shots were from either a Glock 20, with a KKM comped 6†barrel or a stock witness 10mm. All shot stings over the chrono unless noted were 20 rnds. The chrono was at 10 feet from the muzzle. Sky conditions were overcast, 8 to 12 mph wind from the rear, 30.06 BP and humidity of 99% (it is Houston for cripes sake). Finally, all numbers are presented in the standard Oehler formation. High velocity (+), low velocity (-), extreme spread (e), mean (m), standard deviation (s).



First question. Does sorting brass really matter a lick?

The experiment: 50 shot strings with calculations after the fact since the 35P does not do math past 20 rnds. The input data, Mixed brass of 15 Federal, 20 Winchester and 15 Remington cases or 50 Starline once fired, Remington 180 grn JHP’s cci300 primers 10.5 grns of AA7, COL of 1.250, crimp to .4227, loaded on my Pro2000. All charges individually weighted.

The data :
Glock with Mixed 1206+, 1116-, 90e, 1165m, 24s
Glock with Starline 1188+, 1117-, 71e, 1156m, 18s.

Conclusion: It probably does not matter a lick if I sort the brass if I am using Federal, Winchester, Remington or Starline cases. It is probably a waste of time statistically.




2nd Question: Does hand weighing each charge and trickling it to exactly the weight matter over just letting the uniflow dump it?


The experiment: 20 shot strings with the witness, 10.2 grns of blue dot, Remington 180 grn JHP, cci300 primers, starline brass, col of 1.250, crimp of .4227.

The data:
Hand weighted 1105+, 1053-, 52e, 1083m, 13s
Pro2000 dumped 1097+, 1040-, 57e, 1073m, 16s.

The conclusion: It does not really matter with blue dot. I guess I should run the same experiment on all my powders just for fun.



3rd Question: How much difference in velocity between the two guns?

The experiment, 2 different 20 shot strings. 9.0 grns longshot, berry 180 grn TMJ, cci300 primers, col of 1.254, and crimp of .4218.

The data
9.0 grns longshot, berry 180 grn TMJ, starline brass, cci300 primers, col of 1.254, and crimp of .4218.
Glock 20 1285+, 1182-, 103e, 1240m, 25s.
Witness 1203+, 1161-, 42s, 1189m, 10s

11.0 grns of aa5, 135 grn berry tmj, starline brass, cci300 primers, col of 1.250, and crimp of .4221
glock 20 1422+, 1317-, 105e, 1346m, 26s
Witness 1336+, 1267-, 69e, 1308m, 20s.

Conclusion: The glock is a faster gun, but the witness is more consistent. I also noticed that while shooting it, the witness just seemed more accurate. Since I was just dumping rounds into the berm while chrono-ing, I could hit rocks with the witness, but the glock they just went a bit wild.




Final question, without consideration to the rest of the reload or guns, what is the most consistent powder?

The data, all from 20 shot strings:
800x 22s, 25s, 21s, 27s, 29s
aa7, 24s, 35s, 28s, 25s, 24s, 18s, 24s
aa5, 26s, 20s, 20s, 17s, 25s
blue dot, 29s, 21s, 16s, 13s, 27s
Longshot, 10s, 25s, 24s,
Unique, 21s, 25s,
Universal Clays, 45s, 38s
Power Pistol, 14s, 15s

(Yes I actually shot 908 rounds of 10mm today. My thumb is sore from reloading the mags!).

The conclusion: Well it’s more complex then a simple conclusion, but it looks like power pistol is the best, then probably blue dot and 800x. Most of these rounds were run in the Pro2000 as is. This means it is complete composite of how much the gun likes the load all the way back to how consistent I ran the press. I was surprised how poorly Universal clays did and AA7. AA7 is a very fine dumping powder so maybe it just is not working well with these loads. Interesting because it seems to work well and is reasonably accurate. I was also surprised at how much better unique is then universal clays.

I guess my path forward is that I will conduct more rigorous test using starline brass, cci primers, and rem 180 grn JHP’s. I will only vary the powder for every load and see what it does over the chrono.

Any thoughts or comments?

 

[larryw]

My own tests match your own results.

Mixed brass did not lead to a noticeable difference on chrony or rested groups with pistol cartridges, 45ACP in particular. None the less, I keep my StarLine pistol brass separated for special use.

Once a powder is found that drops consistently from my 550B, measured vs. dropped charges do not make a noticeable difference over chrony or with rested groups. In addition to 45ACP with 231, WST, BE, TiteGroup, N320 and N340, I've found this to be true for 357 and 44 Mag using BlueDot, 296 and H110; 223 varmint/target loads using H322, 748 and N133; 308 NM loads using 4895; and 30-06 hunting loads using 4350. Key point being variation is small relative to total charge, and within safe margin if near max.

In many cases, I think more benefit is gained by watching seating depth and bullet/ogive variation.

 

[steveno]

I would say that as long as you aren't using a max load using different brands of brass with the same load mixing brass might not be a problem. if you are at max there is no way I would mix brass. 9 mm brass also seems to be inconsistant in thickness bewteen brands and that relates to bullet tension. if you get a case with a little less tension and the bullet somehow gets seated different in handling the 9 mm is not very forgiving. I have shot a lot of different factory loads in 9 mm but it came down to reloading everything got thrown away except Federal brass because that gave the most bullet tension.

 

[P95Carry]

Quite some test Peter - no doubt rather time consuming (as well as thumb trauma inducing!). Thanks for posting.

Very interesting indeed ... and I also reckon that mixed brass is a minor issue with most pistol rounds. I do tend tho with ''hand cannon'' stuff where loads get quite hot, to try and stick to the one case make - usually that is not a problem.

Rifle - another matter entirely IMO .. where the consistency does I feel show ... tho darned if I can support that with data like you have done!

 

[Kamicosmos]

Nice Report! Very enlightening. I have just started seperating all my pistol brass out into individual headstamps, but maybe I won't worry about it.

I would be curious about trying a similar set of tests with rifle loads though.

 

[griz]

That's pretty interesting. I might try a mixed brass test in 38 special just out of curiosity.
Thanks for the report.

 

[taliv]

nice work! very interesting, and thanks for posting


i will take issue with the mixed brass testing. please let me know if i misunderstood what you've done, but it looks like you just dumped all the brass together as "mixed".

as an example, my chrono work shows, for instance, that speer nickel 45 acp cases consistently add 20 fps vs winchester brass 45 acp cases.

if you've lumped everything together and some of those mfgs have more volume, and some have less volume, then all you've done is shown that starline cases are just a tiny bit roomier than the average of all other brands. right? that seems like useless data to me. i can't imagine what conclusion you would draw from that.



GENERALLY speaking, i think the main reason reloaders would care about matched headstamps in HANDGUN cartridges is to keep track of how many times the rounds have been loaded, as many of us watch for defects to start forming in a "batch" and then discard the entire batch of cases. i don't think the difference in case volume in commercial brass is generally a problem unless you were pressing your luck on +P stuff.


edit:

The input data, Mixed brass of 15 Federal, 20 Winchester and 15 Remington cases or 50 Starline once fired

if that wasn't clear above, just break out your "mixed" data so we can see the fed, win and rem numbers separately. your "negligible benefit" conclusion is probably still correct. i'm just saying you can't get to that conclusion by mixing all other brands into one number.

 

[Dave R]

Peter, thanks for applying the scientific method for our enjoyment and enlightenement. One of the best posts I've read in a while.

 

[Grump]

Accuracy???

Seems like one critically important variable was not measured--ACCURACY.

IIRC, the NRA tested mixed-brass lots with .44 Mag. reloads back in the early '60s sometime and found that a powder/primer/bullet combination that did close to 1-inch groups with one brand of brass opened up to 3-inch groups when they went to mixed brass. Don't remember if they did velocities, 'cause that was before the invention of photoelectric "skyscreens."

Quiz--Anyone less than 35 years old here know what we used for chronographs before skyscreens??

 

[griz]

I know the answer, but alas I passed 35 in the last mellinium.

 

[taliv]

i'm only 34, but i've heard interesting stories of my ancestors hanging pieces of metal up and measuring how far the impact swung them backward, and then using an abacus and a slide rule to figure out how much velocity it would have taken to move the metal x inches

of course, since they had to walk uphill (in the snow) to and from the target, the physics was a lot more complex than it is these days

 

[Peter M. Eick]

Tally what I did on the mixed brass was just take the number of cartraiges listed of each brand and shoot them as the fifty. So I had 15 rounds with federal stamp, 20 rounds of winchester and 15 rnds of remington brass. This was my "mixed" lot of 50. It was not really mixed, I just selected a reasonable sampling. I then shot 50 rnds of starline only. Those are the numbers.

Thanks for the comments guys. I will try more experiments with just varying primers on a single load and then a standard load and just vary the powders. That should be a fun one.

Finally on the accuracy. Do to the public range I shoot at, it is very hard to set up and shoot for accuracy and chrono at the same time. The logistics of handling the cease fires kills this. So, I either shoot for chrono or for accuracy. Yesterday I was basically chrono-ing and just dumping rounds into the berm.

 

[griz]

of course, since they had to walk uphill (in the snow) to and from the target, the physics was a lot more complex than it is these days

You got that right! I remember (kinda) shooting one day. I forget if it was 40 inches of snow when I was 25, or 25 inches of snow when I was 40. But the uphill both ways part really screwed up the trajectory.

 

[Ben Shepherd]

I usually "push the envelope" with most of my pistol loads. A couple things I've come to conclusions on:

1. I don't use remington pistol brass, period. It's given me trouble(lots) in 38 spec., 357 mag, 41 mag, 44 mag and 454 casull. After the last incident I went home and threw it all in the recycle bin at work.

2. If going for heavy loads I use federal or starline brass, although winchester will work ok.

3. If you're at least 10% below the book max and just loading plinking/practice ammo, then Mr. Eick and I agree.

All this is IMHO, and as always YMMV.

 

[BluesBear]

I agree with Tally.

Mixing several brands of brass and testing them as a single lot against a different brand really doesn't give an accurate conclusion.

You are just testing the average of one lot against the other.

Test again using an equal number of several different brands. Each brand should be from the same lot#.

Then, if the results are all similar you can conclude that mixing them together would be of no concern.

 

[alex-v.]

There is one way to look at the data for the single lot of mixed headstamp cases. If an extreme spread of 90 fps and 71 fps is acceptable then the results are OK.

The real test would be to create a couple of lots of mixed headstamp cases and test them against each other. And, if the experimenting is still fun, to go and test single headstamp single lot # cases against the other samples.

 

[The Bushmaster]

To keep track of batches and times loaded I will still religously match head stamps and where ever possible will try to get once used brass from the shooter to insure that batch of 50 are brothers. Besides. Neatness counts.

 

[BluesBear]

The real test would be to create a couple of lots of mixed headstamp cases and test them against each other.

Well it would be a test but it would not fit my criteria for a "real test". There are just too many variables involved in testing different lots of mixed cases.

You'll get much better resukts by sorting the cases. Loading them all identically and then testing the lots separately and then comparing the results.

It's quite possible that you will find one particular brand of cases that as a group will give you a larger deviation that the mixed lot does. But you'll never know unless you sort them and try it.

 

[alex-v.]

There are enough handgun shooters who reload mixed lots (mixed headstamps and mixed # of times fired). Often these seem to be the ones who enjoy action pistol type shooting and shoot at closer distances where minute of bullseye is not critical.

The test would be to load up lots where the headstamp is all the same and the number of times fired is known. These would be run over the chrono and the results compared against the same load but in mixed lots.

What would we see in the collected data? I think it would be that the mixed lots had a greater Exteme Spread in velocities and that the Average Velocities would vary more in the mixed lots. But, I really doubt that the Ave Vel would be that much off from the known lot tests.