Standard deviation and accuracy

[Dolomite_supafly]

Some of you may have seen a condensed version of this on another board.

SD's can contribute to accuracy but it is all about the barrel's harmonics. SD complements a gun's accuracy potential but does not make a gun accurate.

All barrels whip when fired. They all act like a bull whip except on a microscopic level. And this whip is what affects accuracy most, not the powder charge or even the bullet itself. If you have a bullet that exits at the top of the whip and the next bullet at the bottom of the whip the accuracy will be miserable even if you are using the highest quality bullet available. Now if you can get all the bullets to exit at the same part of the whip, regardless of velocity differences, they are going to shoot close to each other. It is very possible to have a bullet exit at say 2,800 fps impact at exactly the same spot as one that is at 2,900 fps or even faster. The reason is both bullets are exiting the barrel during the same part of the whip. And during the horizontal line testing I have seen it happen.

Barrel profile can also affect barrel whip. A thin barrel will have a larger whip than a heavy profile for obvious reasons. This is why heavy profile barrels tend to shoot better with a wider range of ammunition than thinner profiles.

That is why we accuracy test, to find a load(s) where the bullet consistently exits the barrel at the same time during this microscopic whip. And it is possible to find two or three loads that work really well even though they may be way off from each other velocity wise. That is because these different loads have their bullets exiting the barrel at the same point in the whip. And then the velocity difference, or how it gets to that velocity, is the only thing that is causing the accuracy not to be absolutely perfect.

Now if you are lucky, and I mean really lucky, you will find a load where the bullet exits at the same point during the whip and that load also has very small SD's. That would mean the node will lead to smaller groups then the small SD will only enhance those small groups. Now some of the best shooting loads I have had didn't have really good SD numbers. And likewise some of the worst loads I have ever tested had very small SD numbers. It is all about finding a load with a low SD that forces the bullet out of the barrel at the same exact time during the whip. But in most cases it is a compromise between the two.

I also believe certain loads, using specific powders and bullets, can also minimize the amount of whip a barrel exhibits. Some powders burn faster than others and this accelerates the bullets quicker than a slower powder. I think this acceleration affects the whip although I cannot prove it. The reason I believe this is FGMM is a consistent performer in a lot of different platforms. I also believe certain loads using specific powders and bullets can also minimize the amount of whip a barrel exhibits. Some powders burn faster than others and this accelerates the bullets quicker than a slower powder. I think this affects the whip. The reason I say this is FGMM is a consistent performer in a lot of different platforms.

There are a few things that affect SD. First, and the biggest IMHO, is the powder choice. Some powders for some reason have very small SD's while others have a very large SD. Second would be the bullets. Large SD's can also be caused by bullets with varying bearing surface lengths. Even within the same lot some bullets vary. And when a bullet has a long bearing surface length that increases pressures and that in turn increases velocities. But I do not believe bullets play a significant role compared to the powder selection. Primers can also cause problems with SD but I think it is miniscule compared to the powder.

Everyone has their own routine for accuracy testing. I do it this way to minimize my time at the range. I can shoot at my house so I can determine SD's without being at a range. With my testing all components are as close to being identical as humanly possible. Cases the same, bullets the same, primers the same and all are loaded in the same setting. The only variable when I test will be the powder charge and overall length of the bullet.

This is how I test loads for accuracy. I will load 3 rounds with the bullet I plan to shoot starting at the minimum load and at the maximum overall length. Then I load 3 more but I add .3 grains to the load and again at the maximum overall length. I do this,.3 of a grain at a time, until I start to see signs of pressure. I record the results then move on to the next powder and do the same. And once I find a load that has a SD I feel is small enough I start the load development.

And because the bullet exits at a certain part of the whip you need to find the position where the bullet exits at the same time. I normally load the 3 that had the smallest SD. Then I load 3 more with the same charge except I seat the bullet .01" deeper in the case. Then 3 more that are .01" deeper than the previous. Seating .01" deeper will change where the bullet exits during the whip of the barrel. I will generally start out seating .01" and stop at .10". That is unless I see pressure signs first.

Then I take these loads and shoot them at a horizontal line on the target. I am not worried about lining up with a vertical line at this point. I try 100 or maybe even 200 yards. You are not worried about the horizontal group size but vertical group size. That is if you have 3 rounds that are spread out over 2" vertically that means the bullets are exiting at different times during the whip. If you find that 3 rounds that are on that horizontal line, even if they are spread out horizontally, that means these are exiting very close to the same point during the whip.

Then once you have a powder charge that has a low SD and a overall length that has very little vertical dispersion then it is the time to fine tune the load by trying changing the charge weight by .1 grain and shoot some groups.

What you want to find is a load with the lowest SD that also has the least amount of vertical dispersion.

And when I say seating depth I do not seat to the same OAL but the same OAL to the ogive of the bullet that way the bullet is the same distance away from the lands. For my .224" bullets I have use a seating that is .221" so I can get as close to the bullet's ogive as I can when seating. When done loading I can have the OAL differ by as much as .05" but they are the same distance from the ogive to rifling.

Everyone has their way of doing things but for me this minimizes the frustration.

Dolomite

[RED333]

Dadburn that was deep, have to read it again, and again.

Looks like a bit of work to do to get a load for my Browning.

[peejman]

When you say "standard deviation" are you referring to velocity, group sizes, or both?

[Dolomite_supafly]

Velocity.

Dolomite

[gregintenn]

Good information. I do not have a chronograph, so I haven't concerned myself with standard deviations to this point. Might be something to play with in the future. I have found guns responding more to a change in seating depth than a change in powder weight. I also measure the overall length, as I don't have equipment to measure the length to the bullet ogive, nor the barrel throat. My method requires more shooting and blind luck than yours. Thanks for sharing.

[jaysouth]

Dolomite,

I am not smart enought to comment on most of your post except to agree that the smaller the standard deviation of a sample, all other factors being equal, the more accurate that load will be.

when casting bullets, I weigh a random sample and compute the standard deviation of the sample. I then separate the bullets into groups by the amount of standard deviation. some folks use a one grain weight separation, but I have found that segregating the bullets into SD intervals insures uniformity and gives better accuracy in the rifles that most of us shoot, at the ranges that most of us shoot.

By way of example, a couple of months ago, I cast several hundred bullets in a 200 gr. mold. The samples weighted from 194 grs to 204 grs in the sample. The standard deviation of the sample was 2.48 grs(rounded to 2.5). I then segregated the bullets into groups of 194 to 196.5 grs, 196.6 grs to 199 grs. and so on. This insured better accuracy that shooting the lightest bullet, 194 in the same shot group as the 204.

Assuming that a little knowledge is dangerous, I do not pretend to be an actuary, but this method has imporved my accuracy in cast bullets.

[nvrpc1]

Lots of good information. I do agree with the fact all of my 308 shoot very well with 168 and 175 FGMM and for me this is all I am concerned about. I have tried to obtain the MV of stock 168 FGMM in hopes by using IMR 4064 I can put the bullet in the same place by using 40 grain of this which is what the disassembly powder charge was on a 168 FGMM but the powder looked as if was IMR 4064 SC ( short cut ). Since I do not know I'm going to try 40 grains of 4064 and see what it does. I do know that 41.5 grains was to fast. I need to get 2016 FPS with the suppressor on using a Garmin doppler radar unit. If anyone has a magic load please let me know. I'm shooting a Remington 700R5 24" bbl with a Banish 46

[peejman]

 

[nvrpc1]

It all started by doing a search on the internet and asking a question about SD and how come a friend's high SD was shooting tighter groups than his low SD. Good information is valuable no matter how old it is. Still trying to find that powder / bullet / brass and primer combo that will replicate off the shelf FGMM 168 and 175 FGMM since that is all I will ever shoot or have a need for.  I have tried 178 gr ELD-X think that may be the way I want to go but FGMM will be what's available if times get hard. I do have one other question I've never asked before. Is 168 / 175 FGMM the same as what the military uses in the battle fields. Meaning are the SD and MV the same and is the powder in FGMM and LC loads the same.

[srszczer]

About 6 months ago I worked up some loads to simulate FGMM performance in my Ruger Precision Rifle. In my rifle for 168gr SMKs I had better luck actually with IMR 3031. I think when I disassembled some rounds I found around 41gr of powder in them so I went a bit to either side. My best load was 41.3gr of IMR 3031 which was less than 0.75 MOA at 100 yards. I haven't had a chance to try and verify that performance or work up some loads for 175gr unfortunately.

[BigK]

On 8/10/2024 at 2:16 PM, peejman said:

 

Says a lot about the usefulness of this site right?