How do ammo manufacturers weigh powder charges?

[mugsie]

Ammo manufacturers produce tens of thousands of rounds each day. So, how do they load them? I can't believe they weigh each charge, and that would mean 100's of scales, constant calibration etc. they must use a ball powder and throw it via volume rather than weight.

Does anyone have any knowledge as to how they do it?

 

[arizona98tj]

Powder elves....about 4,780 per shift last I heard.

 

[rockhopper46038]

Some of the automated loading machine manufacturers measure by weight (i.e. checkweighing the empty case then the full case and comparing the two) while others use volumetric plates. Camdex does the former, Ammoload the latter. Those are the two automated ammunition loading companies that I'm most familiar with.

 

[btg3]

Yes, volumetric. Powder charge may be weight after, but generally not prior to case fill. Of course, samples are periodically taken to confirm weight via scales. Statistical process control, rather than 100% weighing. 100% fill verification can be made with fill probes.

Also, not necessarily ball powder. Weight consistency for a specific powder depends on flow and loose powder density -- which can be affected by humidity, vibration, particle breakdown, additives, etc.

 

[C5rider]

Interesting information.

And, as I'm pining over whether my scales are EXACTLY perfect with the right amount of powder for each charge from my single-stage press, I also wonder how accurate these measuring systems could actually be?!!!

Are they THAT good, THAT quickly?

 

[mugsie]

That's what I was thinking C5rider - are they that accurate that quickly? I discount the powder elves because a while back the teachers union tried to unionize those guys - since most of them couldn't read, the contract was never ratafied. Guess that leaves volumetric. Scales I think have too much baggage associated with them, due to the constant calibration, possibly something hanging up on them, air movement etc. Has to be volume and it has to be a powder that packs relatively consistantly too otherwise the volume height would be different on each throw.

 

[btg3]

... I also wonder how accurate these measuring systems could actually be?!!!

I was process engineer for a system that was accurate to +/-0.0005 grams. It produced ~60 powder charges per minute and there were over 100 such pieces of equipment in operation.

 

[C5rider]

was accurate to +/-0.0005 grams

Man! To get that close, I'd have to shut off the air conditioner!

Thanks for the insights.

 

[beatledog7]

When I build handgun ammo, I calibrate the scale, adjust the powder throw, test it for consistency, then load. I recheck every tenth round or so to make sure the powder weight is still right. I verify that there's powder in every case before seating bullets.

Ammo manufacturers do what amounts to a very similar process.

 

[rodregier]

For most applications (with good production equipment) volumetric is sufficiently precise,
but I'll bow to BTG3's superior first-hand knowledge of actual deployed gear.

 

[Bart B.]

It's been known by accuracy afficianados that metering extruded powder to a 2/10ths or 3/10ths grain spread typically produces more accurate ammo than metering ball powder to a 1/10th grain spread.

.0005 grams equals .007715 grain. There are laboratory scales that do that. But no machine I know of meters powder to .0010 gram (0.01543 grain) spread unless it's got a micro slicer to cut powder particles into smaller ones.

 

[mugsie]

It's been known by accuracy afficianados that metering extruded powder to a 2/10ths or 3/10ths grain spread typically produces more accurate ammo than metering ball powder to a 1/10th grain spread.

Would you mind explaing this to me please Bart?

I also tend to agree with you since .0005 grams equals .0077162 grains, I don't see how this could be done economically on a large scale. I have managed clean rooms and labs requiring a high precision of measurements, and as the precision went up, so did the time required to do so. If machines threw 60 rounds per minute, and he had 60 machines, then each machine could take up to 59 seconds to settle down. I understand that, but they could no longer be ganged, they would need to be isolated so as not to interfere with each other. Even the wind currents from someone moving by them would affect the reading. Temperature, humidity all affect the reading. Doing it this way, while it could be done, would end up being tremendously expensive. I am sure it's done with volume and not weight.

Maybe the more expensive (as in taking longer to produce) ammo would weigh each throw, but certainly not commercial bulk ammo. Just wish I knew how they actually do it.

 

[taliv]

seems like slicing up powder like that would change the burn rate, but maybe not much

i weigh powder on an acculab scale that is precise to .02 grains and i typically load +/- .02 grains from my target so for example i target 42.52g of H4350 in my 6CM load and take anything that reads 42.50 - 42.54g. Over several thousand rounds, i've found usually one kernel of powder will get me into our out of that range. so i'd estimate the average kernel is about .02g (for 4350 which is extruded)

i suppose some of the tiny ball powder i use for blasting 223rem loads could meter much finer than that, but nobody i know uses them for accuracy, so there'd be no reason to even attempt to measure that close. going by velocity variance i've measured in factory ammo, i'd guess if they get within a half a grain of their target, most factories would be happy. And if they're outside that, they probably still just go with it and expect some guy on the internet to report another "flyer"

 

[Bart B.]

Mugsie,

The only difference I've seen put in print as to why ball powders are not popular in accuracy venues is those who used Oehler's System 43 measuring chamber pressure. Ball powder tends to have a big pressure spike at the beginning of the curve; extruded powders have a more gradual pressure built up. The belief is ball powder therefore slams the bullet harder into the rifling and more deforms it than when extruded powders are used. This was a few decades ago when I read the article.

Federal Gold Medal .308 Win. match ammo will shoot in well built rifles darned near 1/2 MOA at 600 yards with IMR4064 metered to a 3/10ths grain spread. No ball powder I know of has done that.

Benchresters have used H322 in match winning and record setting 22 and 24 caliber free-recoiling rifles. Nobody I know of using ball powders has equalled stick powders performance in that accuracy game.

Same for shoulder fired NRA match rifles used in 200 through 1000 yard matches. Best example was when Lake City army ammo plant switched from Remington operating it to Winchester in the 1980's. Winchester put ball powder in M118 7.62 NATO match ammo and it was not all that great in any well built rifle.

As one of a few former US Palma Team members developing loads for Sierra's new 155-gr. Palma bullet in early 1991, we tried several ball powders just to see what happened. While they metered most uniform, had the most uniform muzzle velocity and peak pressures, their accuracy was a long way from what IMR4895 and other stick powders produced with greater spreads in these ballistic areas. IMR4895 metered to a 3/10ths grain spread in brand new Winchester cases over Fed. 210M primers and up to 4/1000ths bullet runout shot well under 2/3 MOA at 600 yards in a couple dozen rifles from around the world. A test of 20 random picked rounds shot from a machine rested rifle all went into 2.7 inches at 600; a picture of that group was in a fall 1991 issue of Handloader Magazine.

I think there's a greater spread in primer fire and brimstone output than what a 2/10ths grain spread of extruded powder makes gas to push out bullets.

 

[GaryL]

I was process engineer for a system that was accurate to +/-0.0005 grams. It produced ~60 powder charges per minute and there were over 100 such pieces of equipment in operation.

I wonder if we've been in the same plant. Probably not, but I have installed equipment into a similar operation, a bit upstream. Anyway, in my experience standard grade factory ammo typically doesn't hold quite that tight, but the factory I'm familiar with has had their high end stuff subjected to rigorous test standards, and in one they passed a million round test without a single ammo related failure. I wasn't told who the customer was, only the basic results of the test. I can say that the equipment I was installing was there to ensure that standard was maintained.

BTW, all the stuff I saw was measured by volume. Statistical testing was used to verify every batch before shipment.

 

[Joe's]

About 3 years ago I read on a major ammo makers web site they guaranteed +or- .4 grs. accuracy. Doesn't sound that accurate to me.

Be Well,
Joe's

 

[Bart B.]

Sierra Bullets meters all powder charges used in cases to shoot their bullets for accuracy tests. A 3/10ths grain spread of extruded powder's good enough to put their best match bullets into sub 1/4 MOA groups in their indoor 200 yard range.

 

[NeuseRvrRat]

Bart just dropped some knowledge on us

 

[ArchAngelCD]

Probably something like this: http://www.youtube.com/watch?v=5LsiscYhujA

There is a Black Hills video using the same machine and the Tech said they drop the powder in 3 stages for consistency and they use a powder check die too.

 

[Bart B.]

When I toured Sierra's plant in California back in the late 1960's, their ballistic tech used a Lyman 55 powder measure dumping charges with a single stroke. He showed me his last set of test targets with 10-shot groups fired from a rail gun in their 100 yard range. All of those groups of Sierra 30 caliber 190 HPMK's were a bit over 1/10th inch; in the "ones."

When Sierra had a really accurate batch of 30 caliber match bullets shooting in the ones coming out of the production line's final stage (ogive pointing die machine at 90 per minute) as he grabbed a handfull then shot them, they would go into a special barrel. As soon as accuracy went back up to about 1/4 inch, the top few inches of bullets in that "special" barrel would be removed then the rest packaged 1000 per plain brown box with the type and lot number put on it. They were sold to a couple of folks who took them to high power rifle matches and sold them a bit below retail. They still had the sizing lanolin on them and one or two in the box would have folds in their jackets; they were not polished shiny clean and bright nor visually inspected like those sold in green boxes were. But they shot 30 to 40 percent more accurate than the green box versions. Sierra stopped this nice thing when their plant moved to Missouri. Virtually all of the high power rifle match winners and record setters used these ugly, greasy, bullets up through the late 1980's.

 

[Bush Pilot]

Ammo manufacturers produce tens of thousands of rounds each day. So, how do they load them? I can't believe they weigh each charge, and that would mean 100's of scales, constant calibration etc. they must use a ball powder and throw it via volume rather than weight.

Does anyone have any knowledge as to how they do it?

I'm guessing they don't use Lee dippers.

 

[btg3]

I wonder if we've been in the same plant. Probably not, but I have installed equipment into a similar operation, a bit upstream. Anyway, in my experience standard grade factory ammo typically doesn't hold quite that tight...

It was a low density powdered metal process for electronic components. Some of the powder was pricey at ~$200/lb. We used a 2-stage drop with vibration to fill a set volume. Via applied stats we learned that the process was quite capable, but we were over-adjusting the process based on error from the scales. We were able to determine that within-sample variation was below significance, thus the solution was to measure 10 samples at once, which gave the scales another decade of breathing room. The process immediately was more stable.

In other words, the accuracy of volumetric-fill exceeded the accuracy of the lab scales -- roughly by a factor of 10x for the equipment used.

 

[GaryL]

btg3 - ok, sounds interesting. So something like small volume sintered parts, like maybe capacitors. So if we crossed paths (could still be a possibility) - it wasn't at an ammo plant.

Unfortunately my experience at the ammo plant came a couple years before I started reloading, so I missed out on an opportunity to learn more about the process. The engineers I worked with were willing to answer some of my questions, but now I'd have a few different questions, since I was more into the shooting side of things then. So I occasionally post up things that I know are fairly generic and try not to give away any proprietary information I might have picked up along the way.