Ammo testing at 25 yards

A

ahighe

Marksman
Jun 29, 2016
148
1
The question appeared in the Non-Technical Forum. However, I think the content is more appropriate here.

Ammo testing at 25 yards
I think testing at 25 yards can be an effective tool for evaluating ammo, equipment modifications, etc. For me, it is essential. I don't have access to an indoor range, and wind at my local facility typically dominates, obscuring differences between alternatives. I might get one or two opportunities per year to shoot in light wind, and even then only for an hour or two.

The major advantage of shooting at shorter distances is that the influence of the wind is disproportionately less. It decreases by roughly a factor of four at half the distance (it's actually more complicated than this) whereas the inherent ammo dispersion decreases by roughly a factor of two. In technical terms, the signal-to-noise ratio improves by roughly a factor of two.

That being said, I can understand why some people have misgivings. One must overcome at least three potential obstacles to obtain meaningful results:
1. Be able to measure smaller impact variations when targets are placed closer.
2. Based on results at shorter distances, predict what will happen at longer ones.
3. Must shoot enough rounds to draw valid conclusions.

The last potential obstacle is not unique to testing at shorter distances. Anecdotal (i.e. insufficient numbers of) observations using any measurement system most likely will lead to errant conclusions. Though once in a while, anecdotal observations may turn out to be correct.

1. Differences are small.
Typically I'll put one round into each of 120 bulls printed on an 8-1/2" x 11" sheet of card stock. Without rushing, that usually takes about 45 minutes. Then I'll scan the target, and analyze using TDS OnTarget software.

For example, below is the TDS analysis of the superimposed impacts for 100 rounds (two boxes) of a particular lot.

T1_1116_04257.PNG

Figure 1 - TDS analysis summary for 100 rounds of lot 1116-04257.

If I had taken the time to ensure that the scope was zeroed, all would have scored "100". LOL.

Unless a lot is horrific, just about all good target ammo will appear similar. That isn't particularly useful, although the plot qualitatively indicates that the distribution is rather symmetrical. Trying to measure groups is equally futile.

On the other hand, even when using a basic inexpensive home printer/scanner/FAX machine, precise locations of each impact can be determined. For example, the vertical location of each round is shown below. Open, and filled-red, circles correspond to the first and second box, respectively.

VertPosition.jpg

Figure 2 - vertical impact locations of 100 rounds of lot 1116-04257.

The small slopes are statistically insignificant.

It's a fair question whether the impact variations are real. How much uncertainty do the scanner and/or the software introduce? After scanning and analyzing the target, I rotated it 180 degrees, re-scanned, and re-analyzed. A plot of the second versus the first set of locations is shown below.

CapMeasSys.jpg

Figure 3 - second versus first vertical impact locations of 100 rounds of lot 1116-04257.

This is an excellent example of a capable measurement system. The error it introduces (SD = 0.0038 in) is much smaller than the range of measured values.

Conclusion: It is possible to accurately measure impact locations at 25 yards. This permits us, among other things, to calculate standard deviations and mean radii of distributions. As shown in Figure 1, SDx = 0.044 in, SDy = 0.046 in, and MR = 0.056 in.

Next: What to expect at 50 yards?

Thanks for reading.
Albert




copyright 2019 Albert Highe
 
Last edited:
How does testing at 25 yards translate to score at 50 yards?

I assert that we don't need to know the precise conversion factor. If enough rounds are measured, a lot that performs better than another at 25 yards will produce a higher agg at 50 yards. As a rule of thumb, SD and MR will scale roughly as a factor of two. Certainly actual performance requires confirmation at 50 yards.

That being said, based on the analysis in Figure 1, I expected the ammo would agg around 2100 in good conditions. So far, in light wind, I've shot two targets placed at 50 yards. Their scores were 2000 and 2200 (agg = 2100!). These results are encouraging. But knowing the variation one can expect among any collection of 100 rounds, I consider the agreement to be a happy coincidence. I'll discuss in more detail later when I address the third obstacle.

The most direct way to determine the scaling factor requires measuring impact locations simultaneously at 25 and 50 yards. I know of no one with this capability. However, one of the reasons I recently sent a barreled action to the Lapua Test Center is that they simultaneously measure locations at 50 and 100 meters. It isn't precisely what we want, but the distance ratio is the same factor of two. At the very least, it might provide some insight what to expect.

The Lapua Center evaluated 16 lots of Center X and 12 lots of Midas +. Similar to the Eley Test Center, initial screening consists of measuring group size of 10 rounds. However, if early fliers guaranteed the group would be large, testing was terminated early - a couple of times after shooting only four rounds.

After the initial screening, groups of 10 were repeated for those lots that appeared promising (about half). Representative examples of the results are shown below.

Midas50m100mExamplesFT.jpg

Figure 4 - Representative reported results for Lapua ammo. Group sizes measured simultaneously at 50 m (small circles) and 100 m (large circles). A, B, C, and D show data for different lots.

Figure 4A shows one of the lots that was terminated after four rounds. It appears the pattern seen at 50 m has pretty much been enlarged by a factor of two at 100 m. The ratio of group size is 1.94.

Similarly, the patterns in 4B and D expanded. Qualitatively, relative positions of individual rounds were preserved. However, the ratios of group size are 1.60 and 2.17.

On the other hand, the behavior in Figure 4C appears different. Nine of the rounds remained relatively close to one another, but the tenth moved disproportionately away from the rest. The size ratio between 100 and 50 m is nearly 2.5.

Figure 5 displays the results for two 10-round groups for each of two different lots. In 5E, at 50 m, the group sizes are different - expected variation. Although both groups spread out at 100 m, the first grows by 1.79X while the second is nearly 3X! In contrast, other lots, like the one shown in Figure 5F, show similar size increases for both groups.

Midas50m100mExamplesFB.jpg

Figure 5 - Representative reported results for Lapua ammo. Group sizes measured simultaneously at 50 m (small circles) and 100 m (large circles). Two 10-round groups are shown for each of two different lots in E and F.

The bar chart in Figure 6 displays the size ratios for 47 groups.

GroupSizeRat.jpg

Figure 6 - Ratio of group size at 100 m and 50 m for 47 lots.

The average ratio is 2.20 with a SD of 0.29. I can think of a number of reasons why the ratios differ so much. Variation among groups containing few members likely is a major cause.

Continued in the next post....







copyright 2019 Albert Highe
 
Last edited:
Lapua Test Center results continued...

Since each ARA target is the sum of scores for 25 bulls, we should evaluate ammo using methods that consider the locations of all impacts, e.g. Standard Deviation and Mean Radius.

Fortunately, Lapua provides some of the desired information. For example, for those 16 lots that have two groups of 10, their data from 50 m is combined and analyzed. Figure 7 shows the analysis of three of them. Among other things, each shows the calculated standard deviations along horizontal (Sx) and vertical (Sy) directions, and a bar graph that tallies the distances from the center for all 20 rounds.

HistExamples.jpg

Figure 7 - Detailed analysis for 20 rounds of each of three different lots.

Unfortunately, the same analysis isn't applied to the distributions measured at 100 m. However, using TDS software, the locations can be measured, and corresponding values can be generated, from the plots Lapua provides. Then is it straightforward to calculate the ratios. (Figures 8 and 9).

SDx50_100mBar.jpg

Figure 8 - Ratio of horizontal standard deviations at 100 m and 50 m for 16 lots.

The average ratio is 2.36 with a SD of 0.27. The ratio appears to be a little larger than the one for group size. But based on the uncertainty in each type of measurement, the two ratios are indistinguishable.

SDy50_100mBar.jpg

Figure 9 - Ratio of vertical standard deviations at 100 m and 50 m for 16 lots.

The average ratio is 2.98 with a SD of 0.41. The average ratio and the variation are larger. The increased vertical spread at 100 m compared to horizontal spread may be significant. Additional testing would be necessary to be confident that the difference between the two orientations is real.

Is anyone willing to share their results from testing at Lapua? I'd be happy to analyze them.

Conclusions: Between 50 and 100 meters, group size and standard deviation appear to increase at a rate somewhat greater than the increase in distance. Vertical spread could be greater than horizontal.
Between 25 to 50 yards, standard deviation may be proportional to distance.


Thanks for reading.
Albert




copyright 2019 Albert Highe
 
Last edited:
Albert,
I have some test lot data from Lapua. Would you like me to email it to you or post it on your thread?

I tried to send you a private message but error message said your inbox was full.
 
I'm definitely a proponent of 25 yd testing and I did a lot of it before I built my tunnel. With enough data at both 25 and 50, I can't honestly remember when 25 yd results didn't closely match 50 yd results.
Fact is, I may be the guilty party that suggested Albert give it a try. LOL

Without enough data, or by not following Albert's "three potential obstacles to obtain meaningful results", I can also understand along with Albert why some have misgivings and get different results than what I see.

FWIW, I'd still be testing at 25 yds if I hadn't constructed my tunnel.

Non-linear dispersion with increased distance:

I did some digging thru my data and found 57 10-shot groups provided to me from fellow shooters over several years that have kindly sent me the results of their testing through the Meyton electronic targeting system at the Lapua tunnel with the target sensors placed at both 50 meters and 100 meters.
I probably have well over 100 groups currently, but didn't want to spend the time searching for the rest.

With the data I gathered, the increased dispersion at those two distances is a ratio of 2.14.
Simply put, 2.14 * 50 meter grp size = 100 meter grp size.

Just as Albert demonstrated, statistically sound analyses demand that statistical uncertainties for the 2.14 ratio be included, and the proper way of expressing the ratio above at the 1 sigma level with a Standard Deviation of 0.37 is: 2.14 +/- 0.37.

1 sigma obviously means +/- 1SD, and that means the ratio for increased dispersion with distance will be between 1.77 and 2.51 68% of the time.
2 sigma means +/- 2SD, and that means the ratio for increased dispersion with distance will be between 1.40 and 2.88 95% of the time.

Don't get bogged down by the math or any unfamiliar terms, because what I'm saying is simply that if someone were to conduct their own testing with an electronic targeting system, they would see that ratio range anywhere from 1.40 to 2.88 for individual grps in a long series of grps but the average would be 2.14.
With more data, averages and standard deviation will most likely change some, but 57 10-shot grps from multiple rifles and a total of 570 rds is a pretty decent sample size.

In a recent discussion with Albert, he was somewhat surprised/confused why the intuitively correct ratio wasn't near 2.00 for double the distance, and why the ratio varied so widely.
Well believe me, I experienced the same confusion when I discovered this phenomenon some 20 or 30 years ago!

After years of research and teaching myself to become the best amateur ballistician I'm capable of without working toward a PHD on the subject, I'm "nearly" certain I have the answer along with thousands of rds of data to back up my theory.

In a nutshell, I offer the following:

Precision is definitely non-linear as the distance to the target increases for RF or cast bullet projectiles. It "is" linear except for special or very rare circumstances with jacketed CF projectiles, but that goes beyond the scope of this discussion.

The cause of this non-linearity in RF/cast is due to it being impossible to either manufacture a soft lead projectile with no Cg (Center of gravity) Offset and/or having the Cg Offset change due to variations in obturation near the chamber, and as it travels down less than "whatever" a dimensionally perfect bore is for the purposes of having the projectile exit the muzzle with the least amount of imbalance.

CF, on the other hand, does maintain near perfect projectile form and balance because the projectiles are incased in rigid relatively hard bonded jackets with jacket thicknesses held to tolerances magnitudes greater than can be achieved with naked soft lead projectiles, and that causes much less susceptibility to the bore dimensions creating Cg Offsets.

Note: I'm talking of only Cg Offset/imbalanced projectiles and not the yawing coning motion of the projectile that's actually the norm for almost all projectiles exiting the barrel.
In general, and excepting those rare circumstances that may be considered exceptions to the rule, with the correct barrel twist both CF and RF stabilize after short distances and a few cycles. In the case of RF, the projectile stabilizes well before it reaches 25 yds.

As an interesting side note to the above:
I've always believed one of the attributes of a Hummer barrel is the "whatever" the perfect chamber/bore dimension are that result in a nearly perfect balanced projectile exiting the muzzle.

Yeah, I know that's everybody's Holy Grail and what everyone has always wondered about for decades, but unfortunately I've yet to come up with a way to measure Cg Offset because I don't know of a method to either statically or dynamically test soft lead projectiles for Cg Offset as is possible with CF projectiles. Even if I could, the Cg Offset would change after the projectile traverses the length of the bore anyway.

It's just one more failure out of probably hundreds over 40 years of testing, where I've been unable to answer some of the things I'm most curious about. LOL

Landy
 
Would it be possible, does there exist equipment, that would allow one to measure, optically a slugs exit relative to bore alignment as a method to determine if a slug exits with or without any instability ?
Is it possible that cg issues would show there? Perhaps a system to measure bore alignment and then the slug early in flight relative to that alignment data ?
 
Waddcutters

Has anyone of the intrepid experimenters out there ever tried a match grade wadcutter? I know that it would be commercially infeasible and all the arguments about feeding and ballistic inferiority etc. but it could at least partially eliminate cg imbalance.

Sorry for the meanderings but it's a slow day at work.

Dennis
 
tim simbari said:
Would it be possible, does there exist equipment, that would allow one to measure, optically a slugs exit relative to bore alignment as a method to determine if a slug exits with or without any instability ?
Is it possible that cg issues would show there? Perhaps a system to measure bore alignment and then the slug early in flight relative to that alignment data ?
Click to expand...
Tim, I'm nearly certain a military or government Ballistic Research Lab such as the one in Aberdeen could easily provide us with the answers to this question as well as many others were curious about, but they have no incentive to do so.

Since the advent of modern ammunition, almost all research into lead based projectiles and their behavior in the transonic range of velocities (Mach 0.8 to Mach 1.2) ceased. Just about the only exception I'm aware of was when Robert McCoy and the Aberdeen Research Lab tested the Federal UM1 at the bequest of both some governmental entities and the US Olympic committee. Even in that case, they barely scratched the surface as compared to the research they've done with modern jacketed projectiles.

Another point since you mentioned instability, I believe we already know when the Cg Offset is troublesome based on the location of shots at 50M vs their location at 100M. As Albert mentioned, some of the groups or at least some of the shots end up at 100M in locations where they shouldn't be and sometimes the shape of the group bears no resemblance at the two distances. At other times, with some groups, they're nearly identical. I believe this is the primary reason for the Standard Deviation being so high when analyzing the data from Lapua.

With enough money and a lot of help from others more qualified than me, I "think" I could provide many answers. Care to donate maybe say $100,000 so I could at least begin? It's also likely I'll need many more donations near that size. LOL

Landy
 
Dennis D. said:
Has anyone of the intrepid experimenters out there ever tried a match grade wadcutter? I know that it would be commercially infeasible and all the arguments about feeding and ballistic inferiority etc. but it could at least partially eliminate cg imbalance.

Sorry for the meanderings but it's a slow day at work.

Dennis
Click to expand...
Hi Dennis,

I don't see a wadcutter as being advantageous except for one possible scenario, and the loss of ballistic efficiency probably negates the benefits.
That one scenario being where the BHN (Brinell Hardness Number) is such that the ogive slumps during acceleration. This was revealed in McCoy's testing of the Federal UM1 at Aberdeen where they were able to utilize some custom ammunition with both soft and hard lead.

You would also have to hope the slumping of the ogive would be consistent over its entire area, and that's another unknown.

Landy
 
Hi Landy,
Wonderful post - full of good info.
But I have to correct one thing you said
"I may be the guilty party that suggested Albert give it a try."

You are not "guilty". I'm thankful you urged me to try testing at 25 yards.
Thanks.
Albert

HuskerP7M8 said:
I'm definitely a proponent of 25 yd testing and I did a lot of it before I built my tunnel. With enough data at both 25 and 50, I can't honestly remember when 25 yd results didn't closely match 50 yd results.
Fact is, I may be the guilty party that suggested Albert give it a try. LOL

Without enough data, or by not following Albert's "three potential obstacles to obtain meaningful results", I can also understand along with Albert why some have misgivings and get different results than what I see.


Landy
Click to expand...
 
HuskerP7M8 said:
Tim, I'm nearly certain a military or government Ballistic Research Lab such as the one in Aberdeen could easily provide us with the answers to this question as well as many others were curious about, but they have no incentive to do so.

Since the advent of modern ammunition, almost all research into lead based projectiles and their behavior in the transonic range of velocities (Mach 0.8 to Mach 1.2) ceased. Just about the only exception I'm aware of was when Robert McCoy and the Aberdeen Research Lab tested the Federal UM1 at the bequest of both some governmental entities and the US Olympic committee. Even in that case, they barely scratched the surface as compared to the research they've done with modern jacketed projectiles.

Another point since you mentioned instability, I believe we already know when the Cg Offset is troublesome based on the location of shots at 50M vs their location at 100M. As Albert mentioned, some of the groups or at least some of the shots end up at 100M in locations where they shouldn't be and sometimes the shape of the group bears no resemblance at the two distances. At other times, with some groups, they're nearly identical. I believe this is the primary reason for the Standard Deviation being so high when analyzing the data from Lapua.

With enough money and a lot of help from others more qualified than me, I "think" I could provide many answers. Care to donate maybe say $100,000 so I could at least begin? It's also likely I'll need many more donations near that size. LOL

Landy
Click to expand...

Ahhh.....no
I figured there was the possibility tied to some rather expensive equipment usually associated with Govt.
Gotta figure out some way to get a DARPA grant for testing ?
I actually had a couple clients working on secure DARPA projects, AI and miniaturization , worth a look . :confused::confused:
I have long believed that the VAST majority of barrels, even very good barrels, have slugs exiting with , at best, very slight instability and the very best will have a slug really eliminate any type of ?excited? state well before they hit the muzzle. Never prove it but, oh well.
 
Last edited:
vertical

Is it likely that measuring the velocity at the muzzle, at 50 and at 100, calculating the BC of each shot and comparing it to the position on the target, it might explain the extra vertical?
 
gjs500 said:
Is it likely that measuring the velocity at the muzzle, at 50 and at 100, calculating the BC of each shot and comparing it to the position on the target, it might explain the extra vertical?
Click to expand...

Most definitely,

That would partially explain some of the ratios above 2.00 at double the distance, but not all of it. And, it should be much less significant for 25 and 50 vs 50 and 100.

Landy
 
My curiosity got the best of me and I decided to conduct a "Proof of Concept" test by shooting thru targets placed at 25 and 50 yds in my tunnel.

BTW, curiosity is a dangerous personality trait considering it was only in the mid-20's when I did this and I damn near froze my ass off! It also didn't help that it seemed like it took forever to get the 2 targets lined up and I swear it must have taken at least a dozen trips downrange to the 25 yd target before I got it centered reasonably well.

Although I've been led to believe it's possible to do this type of test with supersonic CF bullets from the research I've done over the years, I was highly doubtful it would be possible with RF ammo because of its low velocity and blunt nose.

In order to combat or at least minimize the paper deflecting the bullet as it passed thru the 25 yd target, I used 20# paper in lieu of the 110# card stock I normally use. I also printed the target at 50% of actual size so I'd have 25 POA's for precise aiming points.

Before I positioned the 25 yd target, I had set up my rifle on my one-piece rest pointing at the #13 bull at 50 yds. When positioning the target at 25 I was attempting to get the X-hairs to point at the same #13 bull, but gave up before I could get it perfect.
I made 3 more trips to the 50 yd target after each of the first 3 sighter/foulers because it was too damn cold to set-up my spotting scope to insure I was close enough, and then ran the target.

Here's pics of the Rifle and 2 targets after the 25 rds and 3SF's had been shot:






After looking at the numbers, my "Proof of Concept" test appears to be somewhat successful in proving there's paper deflection from the 25 yd target because based on the data, Ratios and Standard Deviations were much higher than Albert and I saw for the Lapua data we both analyzed.

Albert's numbers with his Lapua data: 2.20 Ratio-0.29 SD
My numbers with my Lapua data: 2.14 Ratio-0.37 SD
My Proof of Concept test for paper deflection with 5 shot Grp ES's: 2.51 Ratio-0.42 SD
My Proof of Concept test for paper deflection with Mean Radius: 2.75 Ratio-0.71 SD
I can also say my potential score at 50 yds on the ARA target is very poor compared to my average score.

While many may think this is proof positive for paper deflection, I'm reluctant to say the same because this is only a small data set and I didn't calculate the confidence levels. There's also the possibility that mid 20 deg temps aren't conducive for serious testing.

The 25-shot distributions on an Excel scattershot chart:

 
Continued

These are the 25 shots separated into 5-shot grps so they'd be similar to the Lapua images. They're color coded red, white, blue, dark green, light green for shot#'s 1,2,3,4,5.

As can be seen, the pattern of shots remain similar at both distances, but there are quite a few shots landing at 50 yds that you couldn't predict based on their location at 25 yds.
We're most likely dealing with a combination of factors causing the discrepancies. Paper deflection, Cg Offset, ballistic trajectories with increased distance, or possibly something I'm not aware of.



I thoroughly enjoyed this testing session on 10/27/19, even with the temps being far too cold, and I decided to investigate the results I got in greater depth.

I got hold of Jim Stensvad in McCook and told him if he was bored, I could use his help. This last Sunday (11/3/19) was predicted to have temps in the mid-50's and we spent most of that afternoon testing.

Part II to from last Sunday following sometime in the next few days.

Landy
 
You must log in or register to reply here.
Top Bottom
Back