How much bullet weight variance for bulk 55g fmj?

[Shadow9mm]

Hey, was tinkering with some bullet weight sorting while I was waiting for my tumblers to finish. Just playing around with some bullet Hornady 55g fmj.

How much variance should I expect with bulk bullets? To me it seems rather heavy on the outliers. Just pulled and weighed out of the box until 1 column filled up.

 

[44 AMP]

My old Cartridges of the World lists M 193 bullet weight as 56 -2 gr.

I take that to mean that anything 56gr or up to 2 gr less was considered in spec.

 

[ballardw]

Of those shown this is the summary information I get:

Analysis Variable : wt
Mean Std Dev Median Skewness Kurtosis
54.9464789 0.1412505 55.0000000 -0.1297171 -0.4394235

A t-test for mean weight = 55 gives a t score of -3.19 meaning that the probability the true mean is different from 55 is 0.0021. So we would reject a null hypothesis that the true mean weight is different than 55 and a typical alpha=0.05 .

Of course this assumes that your, likely rounded to one tenth grain, measures are accurate.

 

[Shadow9mm]

Of those shown this is the summary information I get:

Analysis Variable : wt
Mean Std Dev Median Skewness Kurtosis
54.9464789 0.1412505 55.0000000 -0.1297171 -0.4394235

A t-test for mean weight = 55 gives a t score of -3.19 meaning that the probability the true mean is different from 55 is 0.0021. So we would reject a null hypothesis that the true mean weight is different than 55 and a typical alpha=0.05 .

Of course this assumes that your, likely rounded to one tenth grain, measures are accurate.

I tend to think of most measurements as relative or comparative in reloading. Seeing as, while we have good equipment, its generally never been calibrated since it left the factory and may not be able to be recalibrated. It is out there, but i cant afford it.

 

[tangolima]

It is within +/-0.5%. I think it is pretty good.

-TL

Sent from my SM-N960U using Tapatalk

 

[JohnKSa]

It is within +/-0.5%. I think it is pretty good.

Yup, I had already done the calculation and was going to post pretty much what you did.

 

[FrankenMauser]

When I saw the title, I was going to say 1% is pretty standard. Less is better.
But smaller bullets, especially FMJs, seem to have a wider tolerance.
Looks like you got a good (enough) batch.

 

[44 AMP]

Also keep in mind that bulk 55gr FMJ are not made to be, or sold as match bullets and tolerance range within MilSpec guidelines is good enough.

One of the things you are paying for when you buy "match" grade components is a higher than standard degree of uniformity.

Bulk bullets, are good enough for their intended use. They aren't made to be the best possible, that's not the point, or the price point.

 

[Unclenick]

A bigger sample, or ones to 1/20 of a grain, might tell it otherwise, but the fact the first and last columns are taller than the second and fifth columns doesn't fit a bell curve shape very satisfactorily. The fact there are no shorter columns further up or down may just be due to them culling weight errors outside a certain range. So, despite the tall fourth column, I'd want to see a larger sample before trying to decide if that's a normal distribution or if we are seeing the output of multiple sets of tooling.

 

[Shadow9mm]

Also keep in mind that bulk 55gr FMJ are not made to be, or sold as match bullets and tolerance range within MilSpec guidelines is good enough.

One of the things you are paying for when you buy "match" grade components is a higher than standard degree of uniformity.

Bulk bullets, are good enough for their intended use. They aren't made to be the best possible, that's not the point, or the price point.

I didn't really have any expectations at all, was just killing time and though I would see what happened. Got me wondering what was "normal" tolerance for bulk grade ammo though, to give me some frame of reference.

 

[Unclenick]

I pulled a lot of Lake City M2 Ball in the past, and despite using a very uniformly spherical propellant, the powder charge had a span of something like 2%, IIRC. The bullet weight span was also about 2%, but in QuickLOAD, that bullet weight difference in a 24" tube makes a difference of about 10 fps in velocity. A 2% difference in the WC852 powder, however, makes about 54 fps difference, So, for bulk ammo, I'd worry more about powder charge precision than bullet weight precision.

 

[Shadow9mm]

10fps, off of a 50fps extreme spread is a 20% improvement. Even off of a 100fps extreme spread thats a 10% improvement. Thats significant imho. Im about half way through sorting about 1000 bullets now.

 

[44 AMP]

Its only an improvement if it improves the group on the target. Otherwise, its just numbers...

 

[BringEnoughGun]

I have done the same thing, playing with a new digital scale. I came up with similar results. When I measured a few of the lines of bullets again, some weighed the same as the 1st time, some weighed differently.

I asked an engineer at the Sierra Bullet factory one time how much did each bullet actually vary in weight from a single box. How much was acceptable? He looked at me and matter-of-factly said "There is no variance, they will all weigh the same". So I just figured that my personal shooting would never be up to the spec of me ever being able to notice a difference.

And you cannot believe 100% what our scales tell us. It is a matter of degree's, even though we would like it to all be an exact science......

 

[ballardw]

First introduction to a digital scale was in a chemistry class. You could see the weight difference of your finger prints if you handled the materials by hand but it was designed to measure such fine differences and may have been set up for plus/minus 0.001 gm accuracy (been a long time).

If the engineer at Sierra believes that "There is no variance" then he either is 1) not measuring with a sensitive enough scale or 2) not much of an engineer.

 

[tangolima]

I mentor junior engineers at work, including kids right out of school. You will be surprised how lag of common sense they can be. Take away internet, most of them are half brain dead. The few who aren't, they are the real gold nuggets.

Try tolerance, instead of variance, next time. It may click better.

-TL


Sent from my SM-N960U using Tapatalk

 

[Shadow9mm]

I mentor junior engineers at work, including kids right out of school. You will be surprised how lag of common sense they can be. Take away internet, most of them are half brain dead. The few who aren't, they are the real gold nuggets.

Try tolerance, instead of variance, next time. It may click better.

-TL


Sent from my SM-N960U using Tapatalk

Last i checked, tolerances effect variances. For example i am looking at the variances in bullet weight caused by the tolerances allowed during manufacturing.

You could say i have the ability to create my own tolerances by weight sorting the bullets.

But in this case i was simply trying to understand typically variance in weight for bulk fmj as a reference against what i have.

I have sorted around 1300 now and im all done. I was able to get just over 600 in a 55.2 to 55.3 weight range, around 500 in the 55.4 to 55.7g weight range, and about 200 in the 54.6 to 55.1g weight range. I did it this way as i generally load 500 at a time.

 

[44 AMP]

If the engineer at Sierra believes that "There is no variance" then he either is 1) not measuring with a sensitive enough scale or 2) not much of an engineer.

There is another possibility. It may simply be that you misunderstood his answer.

It is entirely possible that they all weigh the same, if his standard is "in tolerance". Looking at it one way, if everything in the box is in tolerance, then there is no variance.

 

[tangolima]

Last i checked, tolerances effect variances. For example i am looking at the variances in bullet weight caused by the tolerances allowed during manufacturing.



You could say i have the ability to create my own tolerances by weight sorting the bullets.



But in this case i was simply trying to understand typically variance in weight for bulk fmj as a reference against what i have.



I have sorted around 1300 now and im all done. I was able to get just over 600 in a 55.2 to 55.3 weight range, around 500 in the 55.4 to 55.7g weight range, and about 200 in the 54.6 to 55.1g weight range. I did it this way as i generally load 500 at a time.

He, the engineer, might not have heard of the term variance. But for sure he had come across tolerance. That's what I meant.

-TL

Sent from my SM-N960U using Tapatalk

 

[44 AMP]

Anyone speaking English ought to have at least heard the words "vary, variation, variance" and understand their basic meanings.

However, "engineer speak" can be narrow and technically defined, or broadly used, depending on the discipline, and of course the individual engineer.

Especially the individual engineer.