Improving Standard Deviation (on Coated Bullets)

[kmc]

I went minor at major match this summer and set off to ensure that didn't happen again with high probability. I'm loading coated .45acp bullets on Dillon 1050, using Dillon dies. The powder is Win231 with Winchester primers and range brass. I'll get into the gorp below but the bottomline is I'm seeing velocities with double digit standard deviation--somewhere in the 17-23 range--over 12 rounds that is driving me crazy. The high SD is driving me to make loads that are too hot. I really would like to get it down to single-digit SD. I used to tune loads with jacketed bullets to a 172PF average and didn't pay much attention to SD. Maybe I need to stop on the coated bullets...

I've attached my resulting data below. So far I've checked powder drops (seem ok), OAL (seem ok), crimp (seem ok). I recently tried a different (coated) bullet, which seems to improve the issue somewhat. I've recently swapped in a Lee U (undersize) die to see if that helped. If anything, the U die may have made it worse. I've noticed quite a variation in weights on coated bullets compared to jacketed (I only recently switched to coated) and would have compared jacketed results as well--except I shot all of my stash!

 

I originally decided that I'd set my load to ensure every single round with high certainty made Major. I recognize now that is not the correct goal, especially with high SD. In any case, I set the load such that my 3 sigma minimum from the mean was >= 165 PF. Of course, given my high standard deviation, 3 sigma was significant as you can see in the attached data.

Rather than trying ensure every shot is >= 165 PF, I need to consider the probability of the average of the fastest 3 shots out of a random choice of 6 rounds being >= 165 PF, which allows some shots to be less than 165 PF. And I’d like that probability to be near certain. This makes for harder statistical math--I’ll have to think about that a bit. This problem largely disappears if I can force my SD to be single digit.

 

Are my velocity standard deviations typical? Are my variations in coated bullets typical? What other ideas do people have on minimizing SD? How are others approaching this dilemma to make PF but not shoot a cannon? Maybe I need to find a topic under coated bullets...

45ACP Load Results.pdf

[Sarge]

you should be able to get single digit SD's or very close. What will help immensely is sorting by headstamp and lubing cases. Try those few easy things and see if it improves.

[kmc]

7 minutes ago, Sarge said:

you should be able to get single digit SD's or very close. What will help immensely is sorting by headstamp and lubing cases. Try those few easy things and see if it improves.

I'll try a small batch sorting by headstamp (though, omg, will that drive extra work!) and all cases are already lubed with Hornady One Shot.

[Sarge]

2 minutes ago, kmc said:

I'll try a small batch sorting by headstamp (though, omg, will that drive extra work!) and all cases are already lubed with Hornady One Shot.

Sorting can be a royal PITA but cases vary just enough to alter case capacity etc. Different cases will also alter oal so case capacity changes even more. Not by much granted, but it takes very little to alter SD. Other things I do to get closer on SD is discard the first round or two of a batch when I start loading as they are always shorter than the rest. Then once I start loading I just keep pulling the handle at an steady pace. Even during load development I do the same. I will load a batch of 50 to chrono. First few I put in practice bucket then pull the handle until about 40 times and use those to test. Pulling the handle, removing round, measuring round, pulling again, etc is a sure fire way to have oals all over the place and therefore SD will suffer.

[superdude]

Your standard deviations (SDs) are typical if those are the results you get.

 

Some powders have wide SDs, some powders have narrow SDs.

 

You could try a different powder. True Blue is known to produce narrow SDs, though I have not tried it in the 45. Other folks might have other recommendations.

[MemphisMechanic]

Change powders. The way a particular powder burns in a particular caliber is by far the biggest factor in standard deviation.

 

Sarge loves telling everyone to sort by headstamp. ... and it does result in match-grade ammo, to be sure.

But remember that thousands of us have figured out how to load accurate reliable ammo without doing so, so don’t despair.
 

You are aware than 231 is temperature sensitive too, right? On a cold day you’ll lose noticeable velocity.

Personally I’m shaking my head at your “some of my rounds might go sub-165” mindset. I keep it simple, and make sure my slowest rounds are 5 points over the power factor floor. With more consistent velocities from a better powder choice, that will hopefully mean all of your ammo stays within a 170-174 pf window. Every single bullet.
 

Personally, I load a 147 to 135pf for Production with an SD of 7 to 8 using Prima V. My *slow* rounds are 133 power factor. It’s very accurate, and nearly as soft as any of the risky 128pf loads that guys are running out there.

This powder is also reverse temperature sensitive, and I did all that load development on a 95 degree day. It’s actually faster here in the winter.

 

Edited November 24, 2019 by MemphisMechanic

[Jim Watson]

I hate to lead you into the 15 ounce game, but the lowest standard deviation I ever got was with 452AA. 

Which would lead me to try WST. 

 

Do you have any jacketed bullets left to compare to coated?

[kmc]

I did more testing this morning under natural light with the Lee U die still in place. I first tested a control group of my same Blue Bullets, first with random brass, then with random but documented range brass. Then I ran a test with Blue Bullets and all Win headstamp range brass. All of these brass were lubed. I then tried two different bullet types both unlubed and lubed, using random range brass but documenting what was used. The first bullet used was Ranier 230gr (plated) and the second bullet used was Precision 230gr (moly coated).

 

I've attached the latest results below but I haven't answered my question yet--in fact, I'm further away. Both control groups were better than yesterday, for some reason. Both control groups were, in fact, more consistent than a single headstamp (Win). The Ranier were roughly equally consistent, though worse when lubed (no difference than unlubed if I ignore a single slower bullet). The Precision were nearly as consistent and no real difference lubed or not.

 

So the base measurement improved today. Headstamp sorting yielded no benefit. Two other bullets showed similar results as my improved base. Maybe there is some additional contributor that I'm not considering but no clue where to turn next. Thoughts?

45ACP Load Results.pdf

[kmc]

2 hours ago, Sarge said:

Sorting can be a royal PITA but cases vary just enough to alter case capacity etc. Different cases will also alter oal so case capacity changes even more. Not by much granted, but it takes very little to alter SD. Other things I do to get closer on SD is discard the first round or two of a batch when I start loading as they are always shorter than the rest. Then once I start loading I just keep pulling the handle at an steady pace. Even during load development I do the same. I will load a batch of 50 to chrono. First few I put in practice bucket then pull the handle until about 40 times and use those to test. Pulling the handle, removing round, measuring round, pulling again, etc is a sure fire way to have oals all over the place and therefore SD will suffer.

I agree on all points. I also throw the first 2-3 in the practice pile. I've tried to be consistent in this testing but since I'm loading specific bullets and brass. I'm not using the normal feeders, which makes my pace inconsistent, despite my efforts.

[kmc]

2 hours ago, MemphisMechanic said:

Change powders. The way a particular powder burns in a particular caliber is by far the biggest factor in standard deviation.

 

I hadn't considered the powder since I've been using win231 on 40 & 45 forever. It's interesting because when I went minor this summer, I was actually shooting 40 but using Win231 (and Blue Bullets). And yes, I know its inversely temperature sensitive. I load it in my temp controlled shop the, step outside and shoot through the chrono. So I don't believe the temperature is a factor. I just checked my cabinet--I've got a bit of N320, Solo 1000 (no kidding), Bullseye, Power Pistol, 700X, and Titegroup. Not to mention Autocomp and a couple of rifle powders... I'll have to do some research to see which of these would be a good choice, though leaning Titegroup at first blush. All I'd need is to get 50 rounds made up.

[kmc]

1 hour ago, Jim Watson said:

I hate to lead you into the 15 ounce game, but the lowest standard deviation I ever got was with 452AA. 

Which would lead me to try WST. 

 

Do you have any jacketed bullets left to compare to coated?

See my latest numbers. Not jacketed but plated anyway. I'm afaid I shot up all my jacketed bullets.

[superdude]

The first thing to make clear is that not all standard deviation measurements will be the same even if you use the same ammo in separate strings. Repeat the test and you'll get different SDs. 

 

I've seen SDs be over 3 times larger/smaller when using the same ammo but recorded in different strings.  For this reason, you're chasing your tail in this comparison.

 

If you're not happy with your SDs (and they are normal), try a different powder. 

 

I dug around and found some old 45 data with 185 grain grain bullets, and True Blue produced single SDs (in 3 of 3 15-shot strings).  

 

 

[kmc]

Swapped to Titegroup—cuz I had some—and problem is solved. I’m now at 7.8 SD and PF of 172. Three straight runs of single digit SD. I have no idea if Titegroup is temperature sensitive or if there are other downsides. Looks like its time to find out...and I’ve got Win231 for sale

[Sarge]

Lowest SD I ever recorded was a 4 using N 320 in 9minor

[RaylanGivens]

10 hours ago, MemphisMechanic said:

You are aware than 231 is temperature sensitive too, right? On a cold day you’ll lose noticeable velocity.

 

This powder is also reverse temperature sensitive, and I did all that load development on a 95 degree day. It’s actually faster here in the winter.

 

How did you discover that 231 is temperature sensitive?  Is there a table listing of some sort available?

 

I'm curious about n320...  Is it temperature sensitive?

 

[kmc]

1 hour ago, RaylanGivens said:

 

How did you discover that 231 is temperature sensitive?  Is there a table listing of some sort available?

 

I'm curious about n320...  Is it temperature sensitive?

 

Here’s a link with some info ==> https://www.thehighroad.org/index.php?threads/question-on-temperature-sensitivity.817116

[MemphisMechanic]

3 hours ago, Sarge said:

Lowest SD I ever recorded was a 4 using N 320 in 9minor

I got a 1.2 (!) using Prima V under a coated 124 with CCI primers.

 

I had to do the math by hand because four rounds in a row were exactly 1,081 fps... and the chrono kept rejecting them as duplicates.

 

Oh. And I was using mixed, random, worn out brass.

[MemphisMechanic]

11 hours ago, RaylanGivens said:

 

How did you discover that 231 is temperature sensitive?  Is there a table listing of some sort available?

 

I'm curious about n320...  Is it temperature sensitive?

 

Google is a powerful tool.

That’s it.

 

N320 is pretty damn stable. I wouldn’t worry about ammo loaded with it fluctuating very widely.

[RaylanGivens]

18 hours ago, kmc said:

Here’s a link with some info ==> https://www.thehighroad.org/index.php?threads/question-on-temperature-sensitivity.817116

 

Good info...  Thanks!