Precision AR-15 Ammunition

As a former military armorer, I was in a position to conduct many studies on the AR-15. Meticulous record keeping helped. I have passed along many of the results of my research in these pages and they helped cut average group sizes of the Army Reserve Shooting Team’s National Match AR-15s fully in half. Among the successes, I also tried things that flat didn’t work. There are folks that swear by some of these ideas and this article will likely ruffle some feathers.

Good ammunition is critical to precision shooting and we’ll examine that here. Let’s start with sorting, something handloaders love doing. Often, meticulous sorting is wasted time. Any sorting procedure needs to improve group size or velocity consistency and has to be accomplished in a "reasonable" amount of time with affordable tools. Sorting brass by weight is popular. Theory says heavier brass must have thicker walls and hold less powder, raising pressure and velocity.

Many of the big military teams weight sort their brass, assuming the volume theory is valid. Unfortunately, this is done with no evidence to validate it. Some folks weight sort in 0.5 grain increments but it is just as easy to sort in 0.1 grain increments and that is what I did. I have used about 30 little plastic bins laid out to sort cases by the thousands. The finished sort fit a near-perfect bell curve with most brass falling uniformly in the middle of the range.

As a test, I selected five cases incrementally from the low weight side and another five from the heavy side. Those ten cases were from the same lot but at the extreme ends in weight ranges. Then I took another ten of the most uniform mid-range weight and loaded all twenty rounds exactly the same, weighing powder charges to within 0.02 grains. Machine rest testing and chronograph revealed zero difference.

I repeated the experiment using different guns so that both stainless and chrome moly barrels with cut and button rifling. Still no difference between the uniform weight brass compared to cases purposely selected from the heavy and light weight extremes. I did this with both Lake City and Winchester Cartridge Company military brass with crimped-in primers. Primer retention in even the heavy brass was fine. I also tried some Lapua brass, having found the weight spread from their cases to be very close.

For years, the Army Marksmanship Unit had used the same brass (new Lake City with crimped-in primers) as most other military teams, but then switched to the more expensive Lapua. I suspected they knew something that I didn’t. Maybe it was just the specific lot I received, but the average case weight from Lapua was heavier and thicker in the neck area. I knew that more neck thickness with tighter neck tension might raise pressure.

I also noticed that the variation in case weights in .223 caliber was much higher than I had experienced in other calibers I had tried like .308 or 6.5-284. I refused to weight sort the cases because I figured if you are going to spend the money on imported brass it better pay back in having to perform less work. However, accuracy and chronograph testing showed no improvement over the brass we had been using. Worse, there were primer retention and pressure issues.

Lapua comes unprimed and lacks the primer crimp found in military brass, forcing us to reduce powder charges back by a half grain to retain primers. Spending way more money and dropping about 50 FPS in muzzle velocity with no performance enhancement isn’t smart. Compass Lake Engineering has a post on their website cautioning about pressure increases with Lapua brass so I am not the only one to document this.

Guns can "see" gross changes in brass weight by pressuring up more but it appears that minor changes in case weight (0.1-0.5 grains) don’t show up in velocity or accuracy. All of my test uppers were sub-minute guns that would have shown differences if they existed. In fairness to those of you who shoot other disciplines, my testing was done with AR-15 Service Rifles with short 20" barrels and fat gas gun chambers, two characteristics known to produce fat groups and velocity spreads.

Using different guns from other disciplines may change your results. Thinking that my short 20" Service Rifle barrels might not be burning gunpowder uniformly enough, I repeated the tests with Match Rifle uppers having 26" barrels. No change in results. About a year later, I was building a Palma AR-15 for a customer with a 32" Bartlein gain twist 5R barrel and a tight bolt rifle chamber. This was the very first time that weight sorting of brass appeared to pay off.

Brass of a uniform weight outperformed brass with the maximum weight difference by 7 FPS. That’s not much and may easily be within the testing margin of error. A 7 FPS velocity change at 1,000 yards will only move the bullet two inches vertically. I hope to repeat this test in another long-barreled AR-15 in the future to see if these results hold true.

5-284 (left) and .308 (right) proved more consistent than .223 casings in the center. Above center: Gross changes in cartridge casing volumes can be seen by the guns. Significantly heavier and thicker Lapua brass coupled with a lack of primer crimp resulted in blown primers. Powder charges had to be reduced. Primers will blow out when pressures rise too high such as when using brass with low volume or from "cooking" ammo on the range.

Blown primers were the frequent result of hard seating and cooking ammo by leaving it in a hot chamber for too long before shooting it. I did a literature search to see what others had found on the brass sorting topic. I knew that researchers had conducted scientific studies to see if variations in cartridge case weights could be correlated to case volume.

Race car auto mechanics have spent a fair amount of similar effort to uniform the heads of engines and the heads on my ’66 Charger were uniformed when I had the engine built for it. The approach the cartridge case folks used was to weigh a case with a spent primer in it to act as a plug, then fill with water and weighed a second time to determine the true capacity. None of the studies I’ve read to date have established a direct relationship between case weight and volume.

Most studies concluded that there was, at best, only a casual relationship. The whole theory of an alleged relationship between case weight and volume and resultant velocity relies heavily on a strong tie between weight and volume and the studies I’ve reviewed blew holes in that. I am always anxious to learn but lacking any counter evidence my own tests have shown no improvement in velocity consistency or accuracy from brass weight sorting. My advice is to not waste your time.

Perhaps one of the reasons that the weight sorting of brass fails to predict velocity changes is the percentage of the case weight in the head. About 40% of the total weight of a .223 Remington case is in the head, which doesn’t hold gunpowder so variations in the weight of the case head are not reflected in the case volume. The water technique would certainly have allowed a person to identify cases with uniform volumes.

For a military shooting team trying to load close to 40,000 long range rounds a year, this can’t be done in a time efficient manner. Perhaps the water method would be attractive to Bench Rest competitors using the same 10 or 20 cases over and over. Tied with the notion that one can predict case volume by weighing the cases is gauging the thickness of the brass. Brass varies a bunch in its thickness within the same lot as well as from one side of any given case neck.

The brass weighing theory is sometimes extended to say that cases with thick neck walls will extend their dimensions down to the case head but I have never seen any actual evidence presented. I performed a small test to see if the claims are accurate. Using a neck thickness gauge to locate the thickest side of the neck of a fired .223 Remington case, I marked that side of the case with a Sharpie.

Then, I clamped the case in a vise and sawed through it, deburred the stump of the case, and used a set of calipers to locate the thick side. The thick side of the stump was right in line with the thick side of the neck. A sample size of only one isn’t a good test but this agrees with the urban legend I chose to cut my time and material losses and quit with what data I had.

If the theory that thick neck cases would have less volume, would pressure more, and achieve higher velocities was true, then sorting cases by average thickness would be worthwhile. If that theory has been studied, much less proven, I have not read about it. So, I gauged the thickness of some new primed military brass at the neck and selected ten that averaged on the skinny side, ten more that averaged thick, and another ten selected at random.

The velocity spread of the random batch was no worse than those that were hand selected. If the thickness theory were true the ten cases that gauged thin should have developed a slower velocity and they didn’t. The cases that had thick necks should have developed faster average velocity and that didn’t happen either. And the ten randoms should have had a higher standard deviation but that also failed to happen. I don’t recall how many times I repeated the test with different uppers but nothing ever became of it.

None of the results fit the theories and I concluded that not only is it a waste of time to separate brass by average thickness it is an even greater waste of time than weight sorting because it is slower to perform.

Bullet on the right seated crooked with run out because the casing neck was thicker on the left side. Above center: Two neck thickness gauges. Author prefers the one on the right. When I tried Lapua .223 brass, I found it to be heavier and measured thicker than Lake City, and it pressured more.

It appears that large weight and thickness changes can be "seen" by short-barreled gas guns with fat chambers but subtle changes in either (such as one would have using a single lot of brass) aren’t seen by the AR-15 in Service Rifle configuration. There was one other thing that I noticed from the study.

The brass that gauged thick resisted bullet seating to a greater degree and when I measured my loaded cartridge lengths those rounds averaged a few thousandths of an inch longer than ammo loaded in brass selected at random. The brass that was selected with thin necks resisted bullet seating less and came out shorter by just a few thousandths of an inch than brass selected at random and not sorted. There is light at the end of the tunnel regarding measuring neck thickness, however.

I said before that brass thickness at the case neck varies a bunch from one side of the neck to the opposite side. I’m talking 0.005" here and often more. That degrades performance in two ways and they will be additive. First, no matter how good your seating die is, when a bullet is seated in such a case neck the thick side is going to resist seating to a greater degree than the thinner side.

That is guaranteed to produce a finished round where the bullet is not aligned on the center axis of the case, that is, increased run out. The bullet will not enter the back of the barrel straight on and performance will suffer. Second, when around is fired from a case with a large difference in neck thickness the thick side will try to hold onto the bullet, causing bullet tilt upon moving forward. Such a bullet will hit the back of the barrel crooked and not shoot well.

Deal with this by gauging neck thickness of new cases. Really good brass will have uniform thickness but you’ll find very few of those. Almost as good is brass that varies no more than 0.001" from the thin side to the thick side. Put all the cases that gauge 0.000-0.001" in one pile for your most important longer distance matches and all the stuff measuring to 0.002" in a second pile for your second most important matches.

If you are desperate, make a third pile for the up to 0.003" variance, though your best gains will come from the first two piles. The select brass will have less run out, shoot better groups, and have more uniform length. If you have determined a bullet seating sweet spot, this helps hit the harmonic sweet spot more consistently. Outside neck turning accomplishes the same thing as gauging but creates uniformity rather than identifying it, though turning may be a little slower.

I ran a study on the effectiveness of turning necks with a slow hand tool and found a 21% reduction in group size. However, the motorized neck turning tool I used made groups larger! I believe I was getting some chatter because the tool was designed to trim case length and modified by the manufacturer to perform neck turning. In fairness, I may not have mine adjusted properly. I have fooled with it over the years and never gotten it to perform right.

The necks look good but their performance just isn’t up to the standard of the hand turner. Guns are specific in their response to neck turned brass. Some like to have their necks turned and respond with greater accuracy while others aren’t so keen on the processed brass. This may have to do with the specific dimensions of chamber necks. I advise you to follow the neck turning tool instructions and then compare neck turned and not-turned brass.

Like sorting by uniformity of neck thickness I also recommend sorting by head stamp and number of times fired. The comparison of pressure characteristics between Lapua and Lake City brass reveals why. Sort civilian brass from military stuff. Keep each brand to itself. Most of the WCC brass I have used was a little heavier and thicker than Lake City and didn’t hold primers as well. Both have crimped-in primers but the WCC doesn’t seem to hold them as well, possibly because the WCC is a little softer.

I have had to reduce powder charges about 0.3 grains in order to prevent blown primers. Differing brass has differing neck malleability, influencing neck tension. There are various theories on light and heavy tension for accuracy.

I am of the light neck tension camp. Neck tension can have some influence on the burning of the powder. One thing that just about everyone believes is that uniform neck tension is best. The military anneals case necks while some civilian companies don’t. So it is reasonable to have differences in neck tension. As brass is fired repeatedly its hardness will also change and that will impact neck tension. Sort brass by both head stamp and number of times fired for best results.

Meplat trimming was the flavor of the week some years back. The little hollow point tips of match bullets of some brands have uneven surfaces and some long range competitors thought such unevenness was causing the ballistic coefficient of each bullets to be unequal, resulting in vertical displacement. Hand operated devices were made up to shave a tiny bit of the bullet tip and make them even. Some guys got in a hurry and tried to hook these meplat uniformers up to electric hand drills with horrible results.

One guy submitted some he had for testing and it looked like a beaver with dull teeth had gnawed on the end of his bullets. The rounds were actually so bad that they shot worse groups compared to the same ammo lot not boogered on. I don’t hear squat about meplat uniforming these days and it seems to have fallen out of favor. Not long after meplat uniforming came bullet pointing.

Where meplat trimming was supposed to uniform ballistic coefficient, pointing was more directed to improving aerodynamic properties by changing the profile of the tip. The tool looked like a conventional die and fit in a press. The bullet was run up into the thing and reshaped via swaging. One of our Team shooters bought one of the kits and pointed some 80 grain Sierra and 90 grain Bergers, then loaded normally for a range and machine rest test with both tests revealing accuracy compromised slightly.

Promised ballistic coefficient gains should have shown up vertically on long range targets but couldn’t be found. I thought I was seeing more signs of pressure and got range reports of blown primers when non-pointed bullets weren’t. I suspected the swaging process might be causing swelling down in the bullet’s bearing surface, so I compared them. The pointed bullets averaged 0.0003" fatter. Not much, but another team member measured the same results.

Much like delivering a dry fly to a trout, I suspect that both meplat trimming and bullet pointing require a fair amount of finesse. If you are getting good results from either and have objectively tested your work don’t let me talk you out of it. If you haven’t tried these applications I don’t recommend it. Good match bullets have little weight variance (barely 0.1 grains) and my typical ten round batch machine rest testing found zero difference in bullets sorted by weight.

However, measuring bullet base to ogive dimensions with a bullet comparator and caliber-specific bullet insert can find quite a spread. The sorting crowd posits sorted bullets may have more similar total ogive profiles and should produce loaded rounds of more uniform lengths. For those seating to a harmonic sweet spot this should improve performance.

Rounds of more uniform length should position the base of the bullet more uniformly relative to the powder charge, resulting in more uniform burning of powder and more uniform velocities. I tested these theories and found if my overall cartridge length was any more uniform it was, at best, very subtle. It appears that the greater determinate of cartridge length is the thickness of brass at the necks. Cases with thick necks resist bullet seating to a greater degree and tend to produce longer ammo.

If there is any improvement to be had in overall length from bullet sorting I believe that the variations in case neck thickness is masking this. Gunpowder can also influence seating depth where one is compressing the charge. Even for those of us who meticulously trickle powder charges, when they are dumped into the case there is a randomness in how they initially pack. Stick powder can show a fair amount of variance.

When the bullet base is coming down into contact with the top of the powder charge those rounds where the powder lies higher in the case will resist seating more and come out a little long. My chronograph readings with sorted bullets was no better than with bullets selected at random. Nor were my machine rest groups any tighter with sorted bullets. Even if there is any validity regarding positioning of the bullet base in the case I don’t believe short-barreled, fat-chambered gas guns can reveal any difference.

From what I have seen on the chronograph, the bullet seating depth has to be moved many hundredths of an inch to make any noticeable change in velocity. Does this mean that we throw the bullet length sorting out along with

brass weight sorting? Not necessarily, but if bullet base-to-ogive measurements can be used to accurately predict variations in ballistic coefficient this would only show up at extreme long ranges. I’m not aware of a study to prove or disprove the base-to-ogive measuring theory and inquiries to Sierra and Berger confirmed this. Both companies admitted there have to be some variations in the actual bullets in any given box but neither could say how much their products vary.

Sierra felt there might be something to be gained by sorting bullets by base-toogive measurements when shooting at 1,000 yards but Berger didn’t address a recommendation one way or the other. It appears there is room for further research into this topic. Derrick Martin’s The Complete Guide to ARto do this with one chronograph set at the muzzle and a second set 500 yards downrange. Not all that esoteric but definitely something requiring the right test equipment, meticulous record keeping, and a fair amount of time.

Hard seating is a perfect example of something that does not work. What I am about to describe is just plain wrong. I am a great believer in observing shooters in other disciplines and applying lessons learned but this doesn’t always work. In certain shooting disciplines riflemen with bolt actions sometimes soft seat their ammo, loading bullets a little long with modest neck tension and chamber each round singly while carefully closing the bolt.

The front of the bullet stops in the throat while the case moves up around the back of the bullet. Since the bullet is driven into the back of the throat, pressures rise and loads are set accordingly. Some shooters have tried the equivalent in a gas gun. With our 600 yard ammo loaded long so each shooter could seat to the depth ideal for their chamber, a few tried to emulate the bolt gun procedure.

The shortcut version of soft seating was to lay the round on the top of a magazine and trip the bolt catch, allowing the bolt to fly forward. This hard seating jammed the long-loaded bullet into the back of the throat. If a cease fire was called, extracting ammo dislodged the bullet and spilled powder. The bullets often had to be beaten out with a cleaning rod inserted from the muzzle. When hard seated rounds were fired it caused more problems. Most military shooting teams load their long range ammo hot.

We used brand new L. C. brass with crimped-in primers fired only once. Guys tried to use that hot ammo in their hard seat effort and blew primers, lodging them under the trigger or up in the barrel extension, rendering the rifle inoperable. I had to get in there with picks and tweezers, sometimes removing the trigger. Testing showed groups got fatter and that the incidence of fliers was worse.

I set up a mini study using across section of Team uppers, machine rest testing ammo seated to the sweet spot and hard seated ammo finding that hard seated rounds always shot worse. On to factory ammunition. Even if we found that weight sorting brass, bullets, or base-to-ogive worked, it doesn’t help when using factory ammo. I thought that maybe sorting loaded rounds by weight might help. I already knew that weight sorting bullets was a waste of time and that the bullets only varied by about 0.1 grain anyway.

I knew that the greater variances in the loaded rounds would come from weight differences of the brass and the factory dispensed powder charges. So I took a known lot of factory match ammunition loaded with 77 grain bullets and weight sorted them in 0.1 grain increments.

Selecting ten round samples from the top of the bell curve and comparing to randomly-selected ammo from the same lot revealed no accuracy difference but the weight-sorted rounds did improve maximum extreme velocity readings by about nine FPS on average. While most long range shooters would kill to improve the consistency of their 1,000 yard ammo velocity, the USAR only used 77 grain magazinelength ammo out to 300 yards where this difference has no benefit.

What about manipulating overall cartridge length in factory ammo? Knowledgeable handloaders have known for decades that each gun will perform best with ammo is seated to the harmonic sweet spot. This normally occurs when the bullet is seated out far enough to be in close proximity to the throat of the chamber. I usually find the best sweet spot to occur between 0.000-0.010" of jump in the AR-15. Unfortunately, seating bullets out further is not an option with factory ammo because it would not function.

The standard for seating ammo length to feed well in AR magazines calls for an overall cartridge length of 2.25". A Wylde chamber, when new, jumps such ammo about 0.090". Even if we could pull the bullet forward to eliminate that awful jump and get it

Magazine-length ammo can have bullets bumped just a few thousandths of an inch deeper to hit the harmonic sweet spot of the firearm, resulting in tighter groups. closer to optimum range it won’t fit the magazine. However, we can seat magazine length ammo deeper. Magazinelength ammunition is can be nudged a few thousandths below 2.250" and still function. Every gun has its own specific bullet seating sweet spot and the process for identifying these sweet spots is the same.

With 77 grain magazine-length ammo, take ten rounds at random and shoot a group. Seat the next ten rounds 0.002" shorter and test. If it gets better, seat 0.002" deeper still. Continue until groups start opening up. Normally, you’ll find that sweet spot before seating 0.012" deeper than factory length. Rifles will feed as reliably with ammo set to 2.238" as it will with the original factory specification of 2.250" but finding the precise bullet seating sweet spot can shrink your groups noticeably.

Note, this technique is very lot specific. Retest if changing lots or brands. Also, make any changes in seating depth before straightening rounds for concentricity. If you straighten first and then reseat, you might induce runout. Managing concentricity, experimenting with bullet seating sweet spot, and rigorously applying shooting fundamentals are your best solutions when employing factory ammo.

Of course, finding the sweet spot has to be done on a range and such testing may be taking away from practice time that could be spent honing your skills. Now, cooking ammunition. One of the most dramatic examples of cartridge heating I’ve seen occurred when the Army Reserve Team competed in Australia. The Ausies had a much more caviler attitude about their on-range ammo storage and just dumped off a pallet of ammo on the range with no shade, no range shack, or cover tarp.

On a particularly warm afternoon foreign teams were blowing primers. A National Guard shooter blew an extractor and a South African competitor had his rifle just plain blow up! Fortunately, like most of the rest of the firearms that I have seen blow up, the shooters escaped unscathed. The Army Reserve Team complained about the ammo and its storage and the Ausies just laughed, claiming our chambers were too tight despite being as-issue 5.56mm NATO spec.

Other countries also shooting M16 variants had the same problem but the ADF Steyr AUGs seemed immune. Turns out, the AUG have a 1:9 twist barrel while all the rifles having pressure problems were 1:7 and that was the difference. I was able to take a small quantity of the Ausie ammo home and contacted the Army Ammunition Plant at Lake City for testing. They determined the Ausie ammo exceeded safe U. S. chamber pressure specs even when not baked for hours by the sun.

Because of the different burning characteristics of their powder, however, that high pressure was very temporary and by the time the gas was metering through our gas ports the pressure there had dropped to below U. S. specifications for the M16. This accounts for why our guns experienced high chamber pressure problems and operated sluggishly. The Steyr AUGs area piston-operated bullpup and the ammo worked fine for them. So much for armchair theories about ammo interchangeability on the battlefield.

Sure, their ammo would chambers and fire in our guns but not with ideal reliability and safety. Because of this, the National Guard Team started to carry their own ammo to foreign matches thanks to the resources of the Air National Guard and their planes. The Reserve Team lacks this luxury and continue to shoot the host nation’s ammo. I built two M4 flat tops with 14.5" gain twist barrels for the Team.

One barrel had a gentle rate of gain that went from 1:11 twist just forward of the chamber and terminated at 1:9 at the muzzle and the other had a much more aggressive rate of gain starting at 1:20 and finishing at 1:10. Both were hammers and fired U. S. 62 grain ball ammo as accurately as any barrels I ever used. This positive experience with gain twist roughly two decades ago carried forward and all the ARs I build today have gain twist barrels on them.

Overheating cartridges in Australia would not be the Army Reserve’s only incident. If anything delays your normal cadence of fire during single fed slow fire events, extract the round from the chamber. Leaving that cartridge in a hot chamber while holding up will "cook" it and run the risk of changing the burning characteristics of the powder. As a general rule, don’t shove rounds into the chamber and close the bolt until you are ready to fire.

Shade your ammo when shooting and use a little restraint in your powder charges when reloading. Don’t be on the ragged edge of blowing primers to get peak performance from these guns. Sitting on the Assembly Line with a spotting scope I’d watch guys that I knew were "cooking" ammo in hot chambers and could predict their fliers before they shot them just by how long they left rounds in hot chambers before firing!

Ammo loaded with ball powder (like the Ausie stuff) has a greater reputation for being susceptible to cooking but ammo with stick powder will do it too if you are not careful. AG