BE-86 Testing
- Thread starter pctechdude
- Start date
pctechdude
New member
lol they are in that territory for sure, yet all those are within published loads 
and under max pressure! I will say they are stout and not for faint of heart!
The 180's are really close to chronographed Winchester 175gr Silvertips, if they were shot out of a 4" barrel they would be on par with the factory fodder.
More so these days I use Autocomp or Silhouette though in my 40's and 45's.
7.8gr of silhouette behind a 230gr jhp at 1.206" is a good load, right around 925fps in a 5" barrel. 8.2gr with same OAL nets a 1006fps from the same gun, A Springfield Range Officer 1911. But that load is +p.
and under max pressure! I will say they are stout and not for faint of heart!The 180's are really close to chronographed Winchester 175gr Silvertips, if they were shot out of a 4" barrel they would be on par with the factory fodder.
More so these days I use Autocomp or Silhouette though in my 40's and 45's.
7.8gr of silhouette behind a 230gr jhp at 1.206" is a good load, right around 925fps in a 5" barrel. 8.2gr with same OAL nets a 1006fps from the same gun, A Springfield Range Officer 1911. But that load is +p.
Pctechdude,
I appreciate that you made your own warning about over-book loads at the beginning, but the board has a required warning to be inserted in such cases. It's in the sticky at the top of the forum titled: REQUIRED READ for those posting Extra Heavy Load Information. In this instance I will edit it in for you. Please check stickies for basic forum information where they apply.
A couple of points: You said Federal HST and Winchester Ranger were loaded with BE-86. Did they publish that information somewhere (a Federal contract maybe?) or did a powder company technician provide that information? Usually powder choice is kept proprietary, and you often can't tell based on appearance not only because of appearance similarities among a number of powders, but also because ammunition makers can order lots made with variations not available to the handloader. An example would be Federal's use of special IMR4064 (published in an ATK document) with flash suppressant for the Mk.316 mod.0 sniper ammunition they make for the government. So the fact an ammunition uses a type of powder does not mean it's identical to what we buy for handloading.
The second point is that, as the Alliant tech mentioned, commercial loaders buy bulk grade powder normally (though I am aware of a couple of apparent exceptions) and not the canister grade sold to handloaders. Canister grade is more expensive because it is tweaked to keep burn rates in a narrower lot-to-lot range than bulk powder is kept to (though it can be the same by coincidence at times). This means bulk lots have wider lot-to-lot burn rate variation, and the ammo manufacturers adjust their charges to compensate for that using pressure test guns rather than relying on a recipe. The fact most handloaders don't have pressure test guns is why canister grade burn rates need tighter burn rate tolerances. If they didn't do that, then load manual recipes would be invalidated. Because of the burn rate variation issue, commercial loads can wind up with charge levels that are higher or lower than the canister grade version would need to produce the same peak pressure. Thus, pulling bullets to weigh charges does not necessarily provide a weight suitable for use with the canister grade version of the same powder. It requires validation testing in a pressure gun to know one way or the other.
You may have the above concerns covered, but I mainly don't want other readers to get the idea they can copy commercial loads directly without special instrumentation or validation from a trusted source.
Actually, that doesn't work out quite. It tells you what the average 6-shot SD is. It doesn't take into account that the mean velocity for each set of 6 will be different from the mean velocity for all sets of 6 combined, as would be used to find the overall SD, nor does it consider the higher square root bias the standard SD calculation has in a small sample.
I used Excel to create thirty shots averaging 1000 fps with an SD of 10 fps normally distributed around that mean value of 1000 fps, and divided it into five groups of six. Over 40 trials (got bored and stopped), averaging the five 6-shot SD's gave me results averaging -0.360 smaller than taking the SD of all thirty shots together. Of the 40, the error ranged from -2.23 lower to 0.59 higher, with a standard deviation of 0.56.
You can attack reducing the error a couple of different ways. One is to take the SD of the SD's, square it, then add it to the square of the average SD, then take the square root of that sum. When I did this, it reduced average error to less than +0.1 and the standard deviation of the errors came down from 0.56 to 0.51, with the error spread -1.446 to +0.652.
The other method is to eliminating square root bias from the estimates. To do this, take the 6 shot average SD (or do it for each 6-shot SD before averaging) and divide it by 0.9515. (An Excel formula for the bias factor for different sample sizes is below.) When I used that adjusted average 40 times, I got an average error just over +0.1, with an error standard deviation of 0.49, and an error spread from -1.435 to +0.769.
To try this out in Excel for yourself, the formula to make a cell produce a random number normally distributed around a specified average is:
NORMINV(RAND(),μ,σ)
where μ is the mean and σ is the real standard deviation for your theoretical infinite shot population.
In this case, for an average velocity of 1000 fps and a standard deviation of 10 fps, I used:
=NORMINV(RAND(),1000,10)
in each of thirty cells in a column to get a simultaneous set of 30 velocity variations in a bell curve distribution around 1000 fps.
For the other method, to get rid of small sample bias resulting from the way sample standard deviation is normally calculated, divide the SD from your chronograph by the result of this Excel formula:
(2/(n-1))^0.5*GAMMA(n/2)/GAMMA((n-1)/2)
where n is the number of shots in a sample.
Divide all your six shot SD's by that formula before averaging, or divide their average by it afterward and you will have a more accurate estimate of population standard deviation (that's what sample standard deviations try to be; an estimate of population standard deviation). Due to the standard deviaition in 6-shot mean values, there will still be error that hops around some, randomly, but at least the result will be more accurate on average.
I appreciate that you made your own warning about over-book loads at the beginning, but the board has a required warning to be inserted in such cases. It's in the sticky at the top of the forum titled: REQUIRED READ for those posting Extra Heavy Load Information. In this instance I will edit it in for you. Please check stickies for basic forum information where they apply.
A couple of points: You said Federal HST and Winchester Ranger were loaded with BE-86. Did they publish that information somewhere (a Federal contract maybe?) or did a powder company technician provide that information? Usually powder choice is kept proprietary, and you often can't tell based on appearance not only because of appearance similarities among a number of powders, but also because ammunition makers can order lots made with variations not available to the handloader. An example would be Federal's use of special IMR4064 (published in an ATK document) with flash suppressant for the Mk.316 mod.0 sniper ammunition they make for the government. So the fact an ammunition uses a type of powder does not mean it's identical to what we buy for handloading.
The second point is that, as the Alliant tech mentioned, commercial loaders buy bulk grade powder normally (though I am aware of a couple of apparent exceptions) and not the canister grade sold to handloaders. Canister grade is more expensive because it is tweaked to keep burn rates in a narrower lot-to-lot range than bulk powder is kept to (though it can be the same by coincidence at times). This means bulk lots have wider lot-to-lot burn rate variation, and the ammo manufacturers adjust their charges to compensate for that using pressure test guns rather than relying on a recipe. The fact most handloaders don't have pressure test guns is why canister grade burn rates need tighter burn rate tolerances. If they didn't do that, then load manual recipes would be invalidated. Because of the burn rate variation issue, commercial loads can wind up with charge levels that are higher or lower than the canister grade version would need to produce the same peak pressure. Thus, pulling bullets to weigh charges does not necessarily provide a weight suitable for use with the canister grade version of the same powder. It requires validation testing in a pressure gun to know one way or the other.
You may have the above concerns covered, but I mainly don't want other readers to get the idea they can copy commercial loads directly without special instrumentation or validation from a trusted source.
Nick_C_S said:
Actually, that doesn't work out quite. It tells you what the average 6-shot SD is. It doesn't take into account that the mean velocity for each set of 6 will be different from the mean velocity for all sets of 6 combined, as would be used to find the overall SD, nor does it consider the higher square root bias the standard SD calculation has in a small sample.
I used Excel to create thirty shots averaging 1000 fps with an SD of 10 fps normally distributed around that mean value of 1000 fps, and divided it into five groups of six. Over 40 trials (got bored and stopped), averaging the five 6-shot SD's gave me results averaging -0.360 smaller than taking the SD of all thirty shots together. Of the 40, the error ranged from -2.23 lower to 0.59 higher, with a standard deviation of 0.56.
You can attack reducing the error a couple of different ways. One is to take the SD of the SD's, square it, then add it to the square of the average SD, then take the square root of that sum. When I did this, it reduced average error to less than +0.1 and the standard deviation of the errors came down from 0.56 to 0.51, with the error spread -1.446 to +0.652.
The other method is to eliminating square root bias from the estimates. To do this, take the 6 shot average SD (or do it for each 6-shot SD before averaging) and divide it by 0.9515. (An Excel formula for the bias factor for different sample sizes is below.) When I used that adjusted average 40 times, I got an average error just over +0.1, with an error standard deviation of 0.49, and an error spread from -1.435 to +0.769.
To try this out in Excel for yourself, the formula to make a cell produce a random number normally distributed around a specified average is:
NORMINV(RAND(),μ,σ)
where μ is the mean and σ is the real standard deviation for your theoretical infinite shot population.
In this case, for an average velocity of 1000 fps and a standard deviation of 10 fps, I used:
=NORMINV(RAND(),1000,10)
in each of thirty cells in a column to get a simultaneous set of 30 velocity variations in a bell curve distribution around 1000 fps.
For the other method, to get rid of small sample bias resulting from the way sample standard deviation is normally calculated, divide the SD from your chronograph by the result of this Excel formula:
(2/(n-1))^0.5*GAMMA(n/2)/GAMMA((n-1)/2)
where n is the number of shots in a sample.
Divide all your six shot SD's by that formula before averaging, or divide their average by it afterward and you will have a more accurate estimate of population standard deviation (that's what sample standard deviations try to be; an estimate of population standard deviation). Due to the standard deviaition in 6-shot mean values, there will still be error that hops around some, randomly, but at least the result will be more accurate on average.
pctechdude
New member
Uncle Nick, thank you for editing the post and I apologize, for missing that.
As for the question around the Ranger and HST, I have confirmed with Alliant rep and an ATK employee that the powder is in fact BE-86. Also Alliant will tell you, as they told me, BE-86 has been around since late 70's early 80's as a bulk powder available to the commercial industry, this powder is not "new" by any means, only new to handloaders.
Winchester has not confirmed this lol, however I have pulled several of the ranger t's and replaced the powder with the same exact charge weight of new BE-86 and tested side by side. The performance, accuracy, recoil etc are exactly the same.
One of the engineers I have spoken with from ATK, has stated that BE-86 is a very forgiving powder much like Power Pistol, and from my testing the charges of BE-86 have mimicked such, essentially Power Pistol with flash suppressant which alters burn rate slightly.
As for the question around the Ranger and HST, I have confirmed with Alliant rep and an ATK employee that the powder is in fact BE-86. Also Alliant will tell you, as they told me, BE-86 has been around since late 70's early 80's as a bulk powder available to the commercial industry, this powder is not "new" by any means, only new to handloaders.
Winchester has not confirmed this lol, however I have pulled several of the ranger t's and replaced the powder with the same exact charge weight of new BE-86 and tested side by side. The performance, accuracy, recoil etc are exactly the same.
One of the engineers I have spoken with from ATK, has stated that BE-86 is a very forgiving powder much like Power Pistol, and from my testing the charges of BE-86 have mimicked such, essentially Power Pistol with flash suppressant which alters burn rate slightly.
Okay, Unclenick's post with explanation of the math is EXACTLY the root cause of why the Chrony-brand chronograph is an obnoxious machine. Actually, it is just one of the reasons that the Chrony angers me.
For a compact electronic device that does the volume of math it must, routinely, it is simply obnoxious IMO, to limit users to a ten-shot string for EXACTLY the reason Unclenick detailed.
That math is incredibly easy for a tech programmer of such a device. It ends up being an irrational morass of horrendous math for the average chronograph user.
I don't believe I have ever (successfully) been able to share this angst, at least in a manner where anyone else has ever agreed.
There is not much point in running l-o-n-g shot strings for loads that return lousy SD results. HOWEVER! When you manage to craft a load that returns impressive SD results, the only possible next step I can imagine is to attempt some long string results to really put those loads in the spotlight to see how they hold up.
To me, it's the difference between a handgun that manages to lay down one impressive 10-shot group on paper versus the handgun that you are dead sure certain will lay down that group EVERY TIME.
Chrony products? They don't care, as if it simply doesn't matter.
For a compact electronic device that does the volume of math it must, routinely, it is simply obnoxious IMO, to limit users to a ten-shot string for EXACTLY the reason Unclenick detailed.
That math is incredibly easy for a tech programmer of such a device. It ends up being an irrational morass of horrendous math for the average chronograph user.
I don't believe I have ever (successfully) been able to share this angst, at least in a manner where anyone else has ever agreed.
There is not much point in running l-o-n-g shot strings for loads that return lousy SD results. HOWEVER! When you manage to craft a load that returns impressive SD results, the only possible next step I can imagine is to attempt some long string results to really put those loads in the spotlight to see how they hold up.
To me, it's the difference between a handgun that manages to lay down one impressive 10-shot group on paper versus the handgun that you are dead sure certain will lay down that group EVERY TIME.
Chrony products? They don't care, as if it simply doesn't matter.
If I see some BE-86, I'll be picking some up to to test. That said,
My chrono just shows velocity for up to 32 shots. I do all my calculations in a spreadsheet anyway and throw out the occasional flyer. Also minimum of 10 shots for any kind of SD validity. I normally shoot 15 shots and get ES and SD. If I really want to know I'll shoot 30 rounds. Personally, after seeing all my 'data' SD is always close to a third of ES, so I don't really see the point of calculating the SD... That's just a personal observation though.
pctechdude
New member
Some More Testing - Includes other powders to for comparison
165gr Speer Gold Dot
6.7gr BE-86 Warning .2gr over Alliant's max load!
1.120" COL Length
Federal Brass
Win Small Pistol Primers
1053
1088
1046
1062
1109
1085
1079
1075
1089
1102
It had a little swing in the velocity, the ES was a little more than I like, I shoot for at most a 20fps Es and was short of the 1123fps advertised from Alliant using my stock Glock 22 with a 4.49" Barrel Brass was tossed about 5-6 feet to my right.
165gr speer gold dot
7.1gr Autocomp
Winchester Brass
1.125" COL
Winchester Small Pistol Primer
1115
1119
1131
1123
1119
1118
1126
1123
1119
1119
This load produced nice numbers ES of 16, with 4 shots hitting 1119fps. This is a max load per hodgdon, and had a little more snap to it than the BE-86 loading.
180gr speer gold dot
6.2gr autocomp
1.125" COL
Federal Brass
Winchester Small Pistol Primer
1059
1043
1050
1035
1042
1049
1052
1043
1041
1052
This load shot good, recoil was more push than snap, compared to the 165's. Was .2gr under max charge per hodgdon. This one the brass only went about 2.5-3 feet to my right. For a defensive load this would be a good choice in 180gr weights, unless you want more velocity and harsher recoil that is.
So far, in the one cartridge I tested in 40 with BE-86 it didn't exhibit the same performance and tight windowed velocities as in 9mm. But it is a small sample, I believe there is some room to adjust and find it's sweet spot. It may even be that the primer was the cause of the ES numbers, or something else entirely.
165gr Speer Gold Dot
6.7gr BE-86 Warning .2gr over Alliant's max load!
1.120" COL Length
Federal Brass
Win Small Pistol Primers
1053
1088
1046
1062
1109
1085
1079
1075
1089
1102
It had a little swing in the velocity, the ES was a little more than I like, I shoot for at most a 20fps Es and was short of the 1123fps advertised from Alliant using my stock Glock 22 with a 4.49" Barrel Brass was tossed about 5-6 feet to my right.
165gr speer gold dot
7.1gr Autocomp
Winchester Brass
1.125" COL
Winchester Small Pistol Primer
1115
1119
1131
1123
1119
1118
1126
1123
1119
1119
This load produced nice numbers ES of 16, with 4 shots hitting 1119fps. This is a max load per hodgdon, and had a little more snap to it than the BE-86 loading.
180gr speer gold dot
6.2gr autocomp
1.125" COL
Federal Brass
Winchester Small Pistol Primer
1059
1043
1050
1035
1042
1049
1052
1043
1041
1052
This load shot good, recoil was more push than snap, compared to the 165's. Was .2gr under max charge per hodgdon. This one the brass only went about 2.5-3 feet to my right. For a defensive load this would be a good choice in 180gr weights, unless you want more velocity and harsher recoil that is.
So far, in the one cartridge I tested in 40 with BE-86 it didn't exhibit the same performance and tight windowed velocities as in 9mm. But it is a small sample, I believe there is some room to adjust and find it's sweet spot. It may even be that the primer was the cause of the ES numbers, or something else entirely.