Lessons Learned on the Final Stage of the Journey to Subsonic Sub-MOA Accuracy with the 300BLK

[sundevil11]

You sir, are a gentleman and a SCHOLAR.

Thank you, 20X11. I appreciate that a lot!

Regards,

- tw

[sundevil11]

Please explain what your interpretation is of max COL. It can mean a lot of things.


The SAAMI spec for the 300BLK cartridge shows the maximum cartridge overall length as 2.26". This also happens to be the maximum cartridge overall length that will cycle rounds out of a magazine, per SAAMI spec. I'm aware some magazines and receivers have a problem with cycling at 2.26", but my rifle doesn't. It will handle a max COL of 2.26". At least it will now that I'm using a micrometer sizing die and a good caliper to size to within a tolerance of .001".

However, my study was based on rounds sized using the cartridge base to ogive (CBTO) length method. I don't want to take the time to explain that method but I will say a few things about it. CBTO length and COL length do not map one to one with each other. That's the whole point in using the CBTO method - it gives more consistent seating depth because that distance is measured from rifling to ogive. The length of cartridge tips vary from that. Therefore, if you size to a consistent CBTO, the freespace distance will be the same, or closer to the same. This means, of course, that there is no direct way to convert COL into CBTO. The only way that this can be done is empirically. The process used is to gather up a bunch of already loaded rounds for that bullet type, sort them by COL, and then take the average for all the different CBTO lengths for each COL sort interval. That produces a line with a localized variance on it. See Figure 6 in my original post. By generating a straight trendline from the variance line, you now have an effective COL to CBTO conversion formula. Just remember that there is going to be some COL spread for any CBTO length you choose. It’s unavoidable.

Anyway, here are the equations I used for that conversion (See Figure 6):

CBTO = (.97 * COL) - .528
COL = (CBTO + .528) / .97

Just remember, the equations from my data are only good for 220 gr. SMKs. They are based on a single 100 round sample set. The equations are only truly representative for that sample. For that one fixed sample set, maximum COL corresponded to a CBTO of 1.664”.

Using a CBTO equivalent to maximum COL might get you into trouble. The trouble isn’t necessarily in exceeding COL. The shooters who seat their bullets right on the rifling, or just a few .001” intervals away from it, exceed maximum COL all the time. The biggest problem is for people shooting semiautomatics who want to cycle their ammo out of the magazine. That may or may work with a CBTO corresponding to maximum COL. For me, it worked.

I mentioned before that a (16) count interval of evenly spaced .001” test points starting from maximum COL is sufficient for finding your best CBTO length. That interval is way generous. There will be approximately three relative minimums in that test interval. The best, will be the one closest to maximum COL. (See Figures 8, 11 and 12). The relative minimums will get worse the farther away you get from maximum COL. That’s inherent in the mathematics of Pi.

This is going to raise the obvious question of what about those ‘freespace zero’ guys who seat their bullets right on the rifling. I did mathematical modeling to consider that also, but for obvious liability reasons I’m not prepared to submit graphs or charts that might encourage that. The mathematical model I generated suggests that there is no advantage to be had by exceeding maximum COL unless and until you’re within only a few thousandths of an inch from the rifling. The reason for that is that the next and last really good relative minimum on the harmonic curve doesn’t occur until just a few thousandth’s of an inch before the rifling. The relative minimum just short of maximum COL is the next best.

Shooting with bullets seated on the rifling is going to wear a barrel out really quickly. I’m not prepared to state what sort of relative advantage that might produce. I don’t want to encourage that kind of reloading, but I’ll tell you this: The relative advantage isn't much.

[plant.one]

Sub-MOA should be good enough.

by your logic so should driving at 35 mph in the 55 mph zone. just because you can.


some of us prefer to drive 60 though. Others of us have places to be that necessitate speeds faster than 35.

much like driving 35 in the fast lane, some of us have better things to do with our time than writing up a bunch of useless crap and trying to dazzle people with bullshit.


thank you, drive through.

[bangbangping]

However, my study was based on rounds sized using the cartridge base to ogive (CBTO) length method. I don't want to take the time to explain that method but I will say a few things about it. CBTO length and COL length do not map one to one with each other. That's the whole point in using the CBTO method - it gives more consistent seating depth because that distance is measured from rifling to ogive. The length of cartridge tips vary from that. Therefore, if you size to a consistent CBTO, the freespace distance will be the same, or closer to the same. This means, of course, that there is no direct way to convert COL into CBTO. The only way that this can be done is empirically. This means you have to gather up a bunch of already loaded rounds for that bullet type, sort them by COL, and then measure all the different CBTO lengths for each COL sort interval. That produces a squiggly line. See Figure 6 in my original post. By generating a straight trendline for that wiggly line, you now have an effective COL to CBTO conversion formula, but just remember that there is going to be some COL spread for any CBTO length you choose. It’s unavoidable.

Anyway, here are the equations I used for that conversion (See Figure 6):

CBTO = (.97 * COL) - .528
COL = (CBTO + .528) / .97.

Man, you use a lot of words. The part of the bullet ahead of the ogive varies more than the BTO. Yup. CBTO is the critical measurement; COAL not so much as long as your round feeds. My question about what you were considering MAX OAL is based on this conclusion of yours:

The best, will be the one closest to maximum COL. (See Figures 8, 11 and 12). The relative minimums will get worse the farther away you get from maximum COL. That’s inherent in the mathematics of Pi.

I have actually taken the time to read through your entire post several times, and I just can't see how you can make that blanket statement. And aside from your last two or three groups, it looks like you have a broad area with consistently smaller groups in the middle of your graph. (I'll address the last few groups later)

I have no dog is this fight, and care nothing about your copyright, so I hope you take my criticism of your work as what it is: open and honest analysis from a guy who has been reloading for over four decades, knows a bit about mathematics (though probably not as much as you), and quite a bit about the scientific method, which, frankly, seems to be lacking in your work. At a high level, this is what I see:

1. As for the scientific method, there are a number of flaws:

• * Your testing covered six months, half a year of seasonal changes. Could that affect the results?

• * Day to day variations in conditions (yours as well as at the range)

• * Random headstamps and trim lengths. This could result in clusters of good/bad groups.

• * Given the point above, I assume other loading anomalies. For instance, what was the shoulder bump? Was it consistent?

• * Changes to the brass itself over the life of the test. You don't mention how much brass you had and how many times each piece was reloaded.

• * Barrel wear over the life of the test. You should do round-robin testing to eliminate this factor.

• * Changes in your shooting ability over the life of the test. Again, you should do round-robin testing to eliminate this factor.

• * Extrapolating from a single barrel/bullet to all barrels/bullets. Can't be done, regardless of your math, because you can't derive a universal equation from a single test. If you want to claim a universal truth, at least try several barrels, multiple bullets, and shoot sub and super. From a machine rest. And tens of thousands of shots, with all other variables eliminated or accounted for.

• * Human bias. You assumed what would happen before you tested, and then you tested your assumptions yourself.

That last one is YUUUGE and cannot be ignored. There are reams of scientific papers regarding the dangers of unknowingly introducing bias into a scientific test. Consider, for example, the use of double blind tests in drug trials. Given your assumptions going in, and knowing nothing else, I would expect your last groups to be the absolute best (or worst, depending on how you handle pressure) of the test. One way to eliminate bias would be to mark your bullets based on load and mix them up, then look for the colored holes in the target. Another may be to have a second person load your rifle so you don't know what you're shooting.

Comments on the test / conclusions itself:
1. In figure 7, Xsd is essentially a straight line with a very slight apparent downward slope. The Ysd is a cluster, and I don't believe any conclusion can be drawn. Essentially it centers around 1.25 with some pretty large variations with no clear pattern. Having shot and chronoed more than a few subs, my guess is that velocity spread had a big influence on vertical dispersion. (Your assumptions about trajectory are off by a factor of about two) Given your use of mixed brass and little consideration for the charge, I wouldn't be surprised to see ES up to 100 in some five shot groups. BTDT. But we have no idea about that, because you didn't clock your shots. (BTW, I'd like to see that graph extended over the entire range of the test).

2. Figure 8 has some large gaps in the data. Were these not not shot or were they discarded? Why?

3. Fitting a smooth curve to your data leaves a bit to be desired, especially around the left side. You explain that away with "overcharging", whatever that is. Given that you were shooting subs with IMR 4227, I don't accept this explanation. And even if it were the case, that should remove this portion of the data from consideration. You can't have bad data, make up an excuse, and then still use the data. Unless you're a climate "scientist", but I digress.

All the above aside, here's the fatal flaw:
You assume that where you place the starting point of the bullet directly affects where it exits in relation to barrel harmonics, and completely ignore every other possible factor. Countless things change both the timing and nature of the barrel harmonics, as well well as when the bullet exits in relation to them. If you start with CBTO and find a sweet spot, it all goes out the window as soon as you change any other factor. With all your math skills, I honestly don't understand how you can shrug off bullet velocity (for only one factor) in this dance toward accuracy.

Lots of hard work, to be sure. But I have to agree with the others that your conclusion is flawed.

[plant.one]

Lots of hard work, to be sure. But I have to agree with the others that your conclusion is flawed.

lots of output, but no useful result & flawed methods....


which begs to question, which govt research facility does he work at

[dellet]

Please explain what your interpretation is of max COL. It can mean a lot of things.


The SAAMI spec for the 300BLK cartridge shows the maximum cartridge overall length as 2.26". This also happens to be the maximum cartridge overall length that will cycle rounds out of a magazine, per SAAMI spec. I'm aware some magazines and receivers have a problem with cycling at 2.26", but my rifle doesn't. It will handle a max COL of 2.26". At least it will now that I'm using a micrometer sizing die and a good caliper to size to within a tolerance of .001".

However, my study was based on rounds sized using the cartridge base to ogive (CBTO) length method. I don't want to take the time to explain that method but I will say a few things about it. CBTO length and COL length do not map one to one with each other. That's the whole point in using the CBTO method - it gives more consistent seating depth because that distance is measured from rifling to ogive. The length of cartridge tips vary from that. Therefore, if you size to a consistent CBTO, the freespace distance will be the same, or closer to the same. This means, of course, that there is no direct way to convert COL into CBTO. The only way that this can be done is empirically. This means you have to gather up a bunch of already loaded rounds for that bullet type, sort them by COL, and then measure all the different CBTO lengths for each COL sort interval. That produces a squiggly line. See Figure 6 in my original post. By generating a straight trendline for that wiggly line, you now have an effective COL to CBTO conversion formula, but just remember that there is going to be some COL spread for any CBTO length you choose. It’s unavoidable.

Anyway, here are the equations I used for that conversion (See Figure 6):

CBTO = (.97 * COL) - .528
COL = (CBTO + .528) / .97

Just remember, the data from my data is only good for 220 gr. SMKs, and not even that. It’s only good for the fixed sample size I used to generate this conversion. For that one fixed sample, maximum COL corresponded to a CBTO of 1.664”.

Using a CBTO equivalent to maximum COL might get you into trouble. The trouble isn’t necessarily in exceeding COL. The shooters who seat their bullets right on the rifling, or just a few .001” intervals away from it, exceed maximum COL all the time. The biggest problem is for people shooting semiautomatics who want to cycle their ammo out of the magazine. That may or may work with a CBTO corresponding to maximum COL. For me, it worked.

I mentioned before that a (16) count interval of evenly spaced .001” test points starting from maximum COL is sufficient for finding your best CBTO length. That interval is way generous. There will be approximately three relative minimums in that test interval. The best, will be the one closest to maximum COL. (See Figures 8, 11 and 12). The relative minimums will get worse the farther away you get from maximum COL. That’s inherent in the mathematics of Pi.

This is going to raise the obvious question of what about those ‘freespace zero’ guys who seat their bullets right on the rifling. I did mathematical modeling to consider that also, but for obvious liability reasons I’m not prepared to submit graphs or charts that might encourage that. The mathematical model I generated suggests that there is no advantage to be had by exceeding maximum COL unless and until you’re within only a few thousandths of an inch from the rifling. The reason for that is that the next and last really good relative minimum on the harmonic curve doesn’t occur until just a few thousandth’s of an inch before the rifling. The relative minimum just short of maximum COL is the next best.

Shooting with bullets seated on the rifling is going to wear a barrel out really quickly. I’m not prepared to state what sort of relative advantage that might produce either because I don’t want to encourage that kind of reloading, but I’ll tell you this: it’s not much. Only competition shooters shooting at extreme distances should care. Sub-MOA should be good enough.

Holey wasted words and time

Let me get this straight. You spent all this time, money and energy proving that you could get MOA accuracy, using the bullet that the chamber was designed around, loaded at the length it was designed for and only had to vary your seating depth less than .016" after changing brass, powder and primer from the original specs.

I originally thought that you might be refining the works of Dan Newberry and Chris Long (I see you no longer reference them in your works, so assume you are trying to claim all your ideas/research as your own). Sadly, that is not the case.

Advances in the industry need repeat-ability. Sadly your results will not likely be repeated if you simply change powder, bullet or primer lots. If you move the target more than 25 yards or Move to another range shooting the same distance. MOA means more to most people than 100 yards.

Congratulations, you proved that the guy who worked out the cartridge and chamber dimensions had a good idea of what he was doing and that allowed you to build a mathematical model that should help you put together your next cartridge using a different bullet.

We all look forward to your next report

[bangbangping]

Another point regarding subs. Many of us learned a lot over the course of several months in the 200 yard sub challenge. At least a few of us had the idea (at least initially) that loading long would be the key to good groups....ummm...patterns. Dellet and his wife humbled us more than once by loading with the bullet sitting on top of the powder without regard to length. Regardless of what the barrel is doing, minimizing ES is essential to distance shooting with subs. You can try to time your bullet's exit 'til the cows come home, but if one bullet is travelling 1050 and the next at 980, you're not going to hit squat at distance. And it's tough to have a low ES with lots of space in the case (and without attention to every other detail). It would be interesting to see you test your "best" long load against some from, say, the 1.576" range at 200-300 yards. Then chrono the the loads and compare the results.

[sundevil11]

1. As for the scientific method, there are a number of flaws:

• * Your testing covered six months, half a year of seasonal changes. Could that affect the results?

• * Day to day variations in conditions (yours as well as at the range)

• * Random headstamps and trim lengths. This could result in clusters of good/bad groups.

• * Given the point above, I assume other loading anomalies. For instance, what was the shoulder bump? Was it consistent?

• * Changes to the brass itself over the life of the test. You don't mention how much brass you had and how many times each piece was reloaded.

• * Barrel wear over the life of the test. You should do round-robin testing to eliminate this factor.

• * Changes in your shooting ability over the life of the test. Again, you should do round-robin testing to eliminate this factor.

• * Extrapolating from a single barrel/bullet to all barrels/bullets. Can't be done, regardless of your math, because you can't derive a universal equation from a single test. If you want to claim a universal truth, at least try several barrels, multiple bullets, and shoot sub and super. From a machine rest. And tens of thousands of shots, with all other variables eliminated or accounted for.

• * Human bias. You assumed what would happen before you tested, and then you tested your assumptions yourself.

Human bias? I pointed out I started at the low end of the harmonic curve to test dispersion patterns on purpose. I admitted to that bias up front. If I had wanted to fudge my data I would have had to have known what sort of final result to expect in advance. I eliminated any possibility of that problem by starting at the other end. This is about all you can do in any scientific experiment when you are collecting and compiling your own data. Colored bullet holes are a fanciful idea, but even if it were possible, why. Each pattern was fired at a separate target patch.

The real point is that mathematicians and people who work in military ordnance have known about this barrel harmonics curve for nearly (300) years. Every USAF fighter pilot, including my father, is briefed on it in gunnery training. For him that was (50) years ago. I knew about this curve growing up, not just a year ago when I tried getting decent load information to refine my tests to find a sub-MOA subsonic load. Instead, all I got was heckle and jeckle criticism from guys like you. I got the twenty question treatment. That suggested to me that you guys really didn’t know, so I set about over the last year to collect the test data necessary to demonstrate one more time what the barrel harmonic curve looks like.

You argue my test results are flawed and then you cite a laundry list of items that you claim should effect my conclusions instead of the nature of my test data. However, if it were possible to refine and improve upon those laundry list items, it would only make my dispersion patterns better, not worse. This is actually called arguing both sides of the fence by the way. In legal circles it’s called collateral estoppel and is grounds for having your whole argument thrown out.

Your claim I should have control over the weather before my test data can be considered valid is laughable. This also means you don’t know the difference between the importance of a first order variable and something that can barely be considered a fourth order variable, and I’m supposed to take you seriously.

My cartridge lengths were the best I had and I admitted to the fact that there was some variability up front, but look at my results. I achieved sub-MOA data without perfectly uniform cartridge trim lengths. Cartridge trim length is a third order variable, by the way. Had my trim lengths been better, my patterns would have been tighter, which only supports my conclusion on barrel order harmonics – a first order variable.

Finally, I held back on my discussion of the harmonic curve between maximum COL and zero freespace, but you accepted my treatment of the subject at face value. If you had been able to do the math yourself, you would have been able to recognize it and you would have pointed it out. It doesn’t effect the conclusion, but it does effect the relative value of the result. This means you are trying to argue things dogmatically that mathematically you don’t understand.

You don’t know the difference between a first order variable and a third order variable. I’m through here.

And finally: "Driving at 35mph?" Mixed metaphor. I want my subsonics to stay subsonic.

[BoomerVF14]

[bangbangping]

Wow, and I tried to be nice. Clearly you have a problem with constructive criticism. As I mentioned, my math skills may not be quite as good as yours, but at least I understand that your data is shit and you can't seem to grasp why. You don't even understand or acknowledge my arguments, but just see red if someone questions your god-like math abilities. And I'm not arguing against barrel harmonics, you douche. I know they matter, and have known that for longer than you've been alive. I'm arguing against your crappy test methods, your bias, and your willingness to grasp at straws to support your theory that CBTO is everything, and that you can determine the best CBTO without regard to any other factor. You can't. It's not mathematically possible. If you disagree, please explain how you know the timing of the bullet exit in relation to barrel movement when you don't have a f*cking clue what your velocity (and velocity spread) is and also how that may affect vertical dispersion in a subsonic round. Or how you've tested your premise with varying loads and they all agree. Simple. Then go post a note card in the 200 yard sub challenge thread. We're waiting.

Human bias? I pointed out I started at the low end of the harmonic curve to test dispersion patterns on purpose. I admitted to that bias up front, which is about all you can do in any scientific experiment.

Seriously? I've never read a scientific study that started with "I'm biased. But I admit it, so it's all good." I actually gave you several methods eliminating it. That's what you should try to do.

I’m through with you.

Thank you. Carry on in your clearly superior manner.