Okay, this is a very limited test, but I shot my .220 Swift (80 gr AMAX w/32 grains of Varget) using two different kinds of bullets: annealed using the manual flame vs using my new Annealing Made Perfect machine. I alternated rounds:



Flame
3183
3169
3199
3196

Avg MV 3186.8
St dev 11.9
ES 30

AMP
3170
3173
3169
3166

Avg MV 3169.5
St dev 2.5
ES 7

I know that is a very limited sample, but it impressed the hell out of me.

Who'd a Thunk? An interesting finding.
Now do a test with "work hardened brass" cases, torch annealed and the gizmo annealed cases.
The energy required to seal the chamber with the brass must vary a bit.

quote:

Originally posted by Rapidrob:
Who'd a Thunk? An interesting finding.
Now do a test with "work hardened brass" cases, torch annealed and the gizmo annealed cases.
The energy required to seal the chamber with the brass must vary a bit.

The brass I used was pretty old.

I find it hard to be to impressed with "very limited samples". They may or may not be indicative.

Well, think about the odds of the first four samples of AMP cases performing that well. Not likely due to chance.

I am logging every shot I take with my LabRadar since I bought the AMP, and I am a believer.

Very interesting, thanks for the info.

quote:

Originally posted by AnotherAZWriter:
Well, think about the odds of the first four samples of AMP cases performing that well. Not likely due to chance.

I am logging every shot I take with my LabRadar since I bought the AMP, and I am a believer.

Two weeks ago I was shooting a load that I have previously fired hundreds of rounds for. Now with the AMP I am seeing my SD half what they were. It is so easy I now do it after every firing.

quote:

Originally posted by McKay:

quote:

Originally posted by AnotherAZWriter:
Well, think about the odds of the first four samples of AMP cases performing that well. Not likely due to chance.

I am logging every shot I take with my LabRadar since I bought the AMP, and I am a believer.

Two weeks ago I was shooting a load that I have previously fired hundreds of rounds for. Now with the AMP I am seeing my SD half what they were. It is so easy I now do it after every firing.

I'll take your data with the larger sample size any day, Mac.

From a statistical perspective at least a sample size of 25 is preferable.

The science aside, at least for a moment, it makes sense that a more controlled process would yield a more consistent result.

quote:

Originally posted by Dulltool17:

quote:

Originally posted by McKay:

quote:

Originally posted by AnotherAZWriter:
Well, think about the odds of the first four samples of AMP cases performing that well. Not likely due to chance.

I am logging every shot I take with my LabRadar since I bought the AMP, and I am a believer.

Two weeks ago I was shooting a load that I have previously fired hundreds of rounds for. Now with the AMP I am seeing my SD half what they were. It is so easy I now do it after every firing.

I'll take your data with the larger sample size any day, Mac.

From a statistical perspective at least a sample size of 25 is preferable.

The science aside, at least for a moment, it makes sense that a more controlled process would yield a more consistent result.

I agree the data limited, but that doesn't prevent one from running an F test for unequal variances. MsAZW is a Six Sigma Master Black Belt, so I asked her. She said more data is preferred, but it doesn't prevent you from running a statistical test. The result:

F-Test Two-Sample for Variances

Variable 1 Variable 2
Mean 3186.75 3169.5
Variance 188.25 8.333333333
Observations 4 4
df 3 3
F 22.59
P(F<=f) one-tail 0.014627006
F Critical one-tail 9.276628153


There is a 1.4% change that the SD difference between AMP brass and flame brass is due to random chance.

But hey, I will keep shooting...and annealing every shot using the AMP machine.

I totally agree that more data is needed. The only think with the AMP is it is so predictable and easy. That now I will actually anneal my brass on a consitent bases. It’s no big deal now to run 10-20 cases. I would never dream of setting up the cases, painting them with the temp paint to try and run a small batch with torches.

Was there an increase in accuracy to go along with it??

quote:

Originally posted by ramrod340:
Was there an increase in accuracy to go along with it??

Yes exactly what I want to know, I'm only interested in reducing extreme spread of muzzle velocity "if" it reduces group size.
Otherwise what is the point...

I would need to still shoot some more. It was out of a hunting rifle that I have been perfectly happy with the 3/4” 5 shot groups it normally produces for a very light rifle.

Guys if it works for you makes you feel better and you have time and $$ go for it. I gave up trying to make a .75" rifle a .625" years ago.

Enjoy

quote:

Originally posted by ramrod340:
Guys if it works for you makes you feel better and you have time and $$ go for it. I gave trying to make a .75" rifle a .625 years ago.

Enjoy

Killjoy!

quote:

Originally posted by ramrod340:
Guys if it works for you makes you feel better and you have time and $$ go for it. I gave trying to make a .75" rifle a .625 years ago.

Enjoy

I concur.

quote:

Originally posted by ramrod340:
Was there an increase in accuracy to go along with it??

I honestly didn't measure those groups (shot them on a steel gong at 500 yards and alternated shots so that as my barrel heated it wouldn't affect the MV).

Some loads shoot great even with a high variance in MV...at close range, but once you get out there in terms of range, you do see a higher vertical dispersion (what I love about the LabRadar - what other chrony can you take out in the desert and document the velocity vs bullet splash?).

I will also say this: the last ten shots from my two Lapuas at 800 yards have varied less than 1/2 MOA vertically. I don't have my logbook as I am in MN right now (deer opens Saturday!) so I don't have the exact number but I noticed an immediate improvement when I got the AMP. At 1190 yards my vertical dispersion is higher, but I believe that is because I wasn't taking into account differences in temperature. I was looking at my logbook and noticed on warmer days (95 to 110), I always shot higher than cooler mornings (60-70). The total vertical spread at this range is about 1 MOA.

I do agree with Mac; it is so easy to anneal now I do so for every case. But I also agree for normal hunting it doesn't matter.

quote:

Originally posted by AnotherAZWriter:

quote:

Originally posted by Dulltool17:

quote:

Originally posted by McKay:

quote:

Originally posted by AnotherAZWriter:
Well, think about the odds of the first four samples of AMP cases performing that well. Not likely due to chance.

I am logging every shot I take with my LabRadar since I bought the AMP, and I am a believer.

Two weeks ago I was shooting a load that I have previously fired hundreds of rounds for. Now with the AMP I am seeing my SD half what they were. It is so easy I now do it after every firing.

I'll take your data with the larger sample size any day, Mac.

From a statistical perspective at least a sample size of 25 is preferable.

The science aside, at least for a moment, it makes sense that a more controlled process would yield a more consistent result.

I agree the data limited, but that doesn't prevent one from running an F test for unequal variances. MsAZW is a Six Sigma Master Black Belt, so I asked her. She said more data is preferred, but it doesn't prevent you from running a statistical test. The result:

F-Test Two-Sample for Variances

Variable 1 Variable 2
Mean 3186.75 3169.5
Variance 188.25 8.333333333
Observations 4 4
df 3 3
F 22.59
P(F<=f) one-tail 0.014627006
F Critical one-tail 9.276628153


There is a 1.4% change that the SD difference between AMP brass and flame brass is due to random chance.

But hey, I will keep shooting...and annealing every shot using the AMP machine.

Don't get me wrong, I perform analysis on small sample sizes regularly. We would all love to have larger samples. But you didn't need to consult Ms AZW to know that....

quote:

Originally posted by Dulltool17:

quote:

Originally posted by AnotherAZWriter:

quote:

Originally posted by Dulltool17:

quote:

Originally posted by McKay:

quote:

Originally posted by AnotherAZWriter:
Well, think about the odds of the first four samples of AMP cases performing that well. Not likely due to chance.

I am logging every shot I take with my LabRadar since I bought the AMP, and I am a believer.

Two weeks ago I was shooting a load that I have previously fired hundreds of rounds for. Now with the AMP I am seeing my SD half what they were. It is so easy I now do it after every firing.

I'll take your data with the larger sample size any day, Mac.

From a statistical perspective at least a sample size of 25 is preferable.

The science aside, at least for a moment, it makes sense that a more controlled process would yield a more consistent result.

I agree the data limited, but that doesn't prevent one from running an F test for unequal variances. MsAZW is a Six Sigma Master Black Belt, so I asked her. She said more data is preferred, but it doesn't prevent you from running a statistical test. The result:

F-Test Two-Sample for Variances

Variable 1 Variable 2
Mean 3186.75 3169.5
Variance 188.25 8.333333333
Observations 4 4
df 3 3
F 22.59
P(F<=f) one-tail 0.014627006
F Critical one-tail 9.276628153


There is a 1.4% change that the SD difference between AMP brass and flame brass is due to random chance.

But hey, I will keep shooting...and annealing every shot using the AMP machine.

Don't get me wrong, I perform analysis on small sample sizes regularly. We would all love to have larger samples. But you didn't need to consult Ms AZW to know that....

She laughed

Reducing ES is important for long range. When I shot 1000 yd BR, we considered 20 FPS extreme spread to cause 2.5” vertical dispersion at 1000 yds. For most shooting getting ES under 20 fps isn’t that important. However, will this annealer take a 40 ES load to a 20 ES load?

Gentlemen,

I am out of the country, and Inthink this might be a good test for us to run.

How about running 100 rounds in both a 223 Remington and a 308 Winchester?

I will take 50 cases in each and shoot them as is.

I will run them through the annealing machine and repeat.

Any suggestions would be appreciated.

Looking forward to your findings Saeed.

quote:

Originally posted by Saeed:
Gentlemen,

I am out of the country, and Inthink this might be a good test for us to run.

How about running 100 rounds in both a 223 Remington and a 308 Winchester?

I will take 50 cases in each and shoot them as is.

I will run them through the annealing machine and repeat.

Any suggestions would be appreciated.

You don't say what the cases will be, pre-fired or new? If all are pre-fired and assuming they have been fired the same amount of times, I would take the 50 cases (for each cartridge), size and load 25 and anneal, size and load the other 25. This would take out the possibility of any variability in the number of times the cases have been fired when doing your test firing as then all 50 cases would be on the same number of firings with 25 annealed and 25 unannealed. I think batches of 25 are plenty enough to show meaningful data.
Makes sense?

We just got a Tikka T3x TAC A1 in 308 Winchester.

The brass is once fired RWS MATCH factory ammo.

I will prepare 100 cases, 50 will be used as is, 50 will be annealed.

Our standard prep is full length size, trim, uniform primer pockets.

Not sure what scope to use, will let you know.

IMO you really see a benefit ater a csae has been work hardened - which is a lot
More than once fired.

quote:

Originally posted by AnotherAZWriter:
IMO you really see a benefit ater a csae has been work hardened - which is a lot
More than once fired.

Yes. Cases that have been fired 2-3 times or less will not likely show the benefit as will cases fired 6-7 times or more. When I was shooting 1K BR, I would have good accuracy through six firings. After that I would start with new cases because 11 years ago brass was cheap. A fellow competitor with very expensive brass (Lazzeroni) frequently annealed and had no issues with accuracy at 40 firings.

While the test should be with work hardened brass, I am thinking best results would be achieved by annealing at least every 2-3 firings? That way the cases wood never work harden?

To clearly demonstrate the effect of annealing, I would suggest taking a set of brass fired multiple times, then divide the set into two and prepare/load each half the same way except for annealing. Then take both half-sets to the range and fire them on the same day interleaved with each other. This way the only variable is annealed vs. un-annealed and any distinct trend in the results can be attributed to the variable.

It would be generous of you to run this test and share the results with us!

What do you guys think of this idea? Instead of 50 cases take 20 cases split into two groups of ten each. Anneal one set every time and load the other 10 over and over until you have shot each case ten times. Document the results each time
you fire the ten.

quote:

Originally posted by AnotherAZWriter:
What do you guys think of this idea? Instead of 50 cases take 20 cases split into two groups of ten each. Anneal one set every time and load the other 10 over and over until you have shot each case ten times. Document the results each time
you fire the ten.

Sounds like a great idea.

Do I shoot 2 5 shot groups each time or 1 10 shot groups??

What would be more indicative??

Actually we have got away from the OP's original trial which was to prove the effectiveness of the AMP annealing method over flame annealing.
All we are doing now is showing that annealing is effective in improving ballistic performance by restoring proper case neck tension but we already know that is a benefit of annealing. The AMP may show very consistent results in this comparative annealed/no-annealed test but we are not expanding the sample size and doing the same testing as the OP?