Accuracy: bullet jump vs pressure and velocity

[cdoc42]

I'm working up a load for my 6.5mm Creedmore (Kimber Model 84) and I'm not getting the accuracy I expected. My dilemma rests with the need to seat the bullets such that they fit in the magazine, the available space being no longer than 2.800 inches OAL.

A Berger 130gr VLD at 2.800" OAL puts the bullet 0.098" from the rifling. Using IMR 4350, H4350, Varget, BL-C(2), Hybrid-V 100 and Superformance I cannot get 3 shots less than 1.5 to 3.0 inches center to center, even though 2 shots with Varget were 0.175" (shot #3 was 2.5" away) and Superformance 0.236" (Shot #3 was 1.5" away).

The same pattern exists for a Hornady 140gr SST. At 2.800" OAL it is 0.08" from the rifling.

I have been using loads, 1.0 gr less than maximum, listed in the 50th Edition of the Lyman manual. Another thread suggested H4350 as the standard for this caliber with a load as high as 41.5 to 42.8gr with 140gr bullets, while the Lyman book lists 39.5gr of IMR 4350 as max (velocity 2585fps; 61,000psi) with a 24" barrel. My barrel is 22 inches.

After 40 years of reloading, I'm stuck. Maybe I'm just getting too old to figure it out. Most often my most accurate loads in the 9 different rifle calibers that I have are 0.015 to 0.02" from the rifling.

But this got me to wondering if bullet jump or pressure and velocity are the determinants to accuracy. That would mean I need to leave the seating at 2.800" OAL and simply increase/decrease velocity and pressure. For example, in the Lyman book, the most accurate load with 140gr Hornady A-Max is IMR 4350 at maximum 39.5gr, which gives a velocity of 2585fps and 61,000 psi. But just above it is W760 40.2gr giving a velocity of 2583fps but a pressure of 60,500psi. So much for velocity being the determinant. Another load that gives the same psi of 61,000 and similar velocity (2548fps)should be as accurate as IMR4350 but apparently not. So much for pressure being the determinant. So back to bullet jump I go, I guess. Should I just settle for making this rifle a single shot?

 

[F. Guffey]

I am the fan of the running start; I want my bullets to have that jump. When it comes to bullets I have bullets, some have round noses others have pointed noises. If I have a problem with a short box and a rifle that likes long overall maximum length ammo I use bullets with round noses. This allows me to get the nose of the bullet to the end of the box and reduce the running start to the rifling.

And then there is that thing about stuffing the bullet into the lands with no jump, I believe it is cute but there is noting that excites me about wondering if the bullet is going to get started before the pressure gets serious. I want my bullet past the rifling before the bullet has a chance to slow down.

And then there is that thing about ‘how the rifle got that way’; I do not know, the ugliest jump/running start free bore I have ever seen was over .250”. The man that built that magnificent rifle ask me; “How did it happen?”

We loaded 100 rounds for the rifle and found a narrow window where the rifle was most accurate, any more powder or any less powder and accuracy went south.

F. Guffey

 

[Jim Watson]

I don't guess you could try a few rounds loaded closer to the lands in single shot mode. That would tell you if it were the seating or some other factor.

 

[jmr40]

I own a Kimber in 308. They CAN be accurate, but they can also be picky about which loads they shoot accurately. They can also be hard to master and shoot well.

Not knocking Kimber rifles at all, I like mine, but they were never intended for match grade accuracy. For the guy who really wants or needs a 5 lb rifle for carrying into rugged steep terrain they are a bargain for the money. And once a load is found that they like they tend to be accurate enough. If the trigger puller has the skills to shoot a 5 lb rifle. Not everyone does.

I've always found my best loads by "borrowing" loads suggested by others rather than what is listed in the books. I don't load for 6.5 Creedmoor, at least not yet. But you're taking the right approach. You might want to ask around on several forums for load suggestions. I'd not duplicate a load I got off the internet unless I confirmed it in a manual, but if you run across a lot of people claiming that X grains of Y powder with Z bullet is working well for them, odds are it will work for you with some minor experimenting with the exact charge weight.

I've never worried a great deal about the exact distance to the lands. MOST bullets shoot better fairly close, but with some jump. I always load to fit the magazine. If they chamber without hitting the lands that is where I start and most of the time where I end. I may experiment with seating deeper, but if they won't fit the magazine I don't want a single shot.

 

[Unclenick]

, But this got me to wondering if bullet jump or pressure and velocity are the determinants to accuracy. That would mean I need to leave the seating at 2.800" OAL and simply increase/decrease velocity and pressure. For example, in the Lyman book, the most accurate load with 140gr Hornady A-Max is IMR 4350 at maximum 39.5gr, which gives a velocity of 2585fps and 61,000 psi. But just above it is W760 40.2gr giving a velocity of 2583fps but a pressure of 60,500psi. So much for velocity being the determinant. Another load that gives the same psi of 61,000 and similar velocity (2548fps)should be as accurate as IMR4350 but apparently not. So much for pressure being the determinant. So back to bullet jump I go, I guess. Should I just settle for making this rifle a single shot?

Bullet jump and barrel time are two factors involved in determining accuracy. When you fire a gun, the initial recoil moments and the pressure deformation of the steel causes the muzzle to bend slightly, and the bending point is usually about 2/3 of the way down the tube or a little more, and that becomes the node around which the muzzle vibrates. You want the bullet to exit when the muzzle is at or just before the sinusoidal peaks or valley in this bending where the rate of change in the muzzles position is at its slowest. This is so the normal shot-to-shot variation in barrel time isn't enough that the muzzle's position is appreciably different at the moments at which different bullets exit. This keeps their angles of departure the same.

Different powder burn rates produce different barrel times even when the muzzle velocity is the same. The earlier the pressure peak occurs in a particular powders characteristic pressure cure, the shorter the barrel time because the acceleration introduced by the peak is sooner, starting the bullet down the rest of the barrel at a higher velocity, causing it to take less time to get to the muzzle.

Given a particular powder, barrel time is affected by both bullet jump and powder quantity. When you tune a charge size, the higher the peak pressure, the shorter the barrel time. You are tuning the peak and muzzle pressures simultaneously, but in different proportions when you alter powder charge, with the peak pressure going up and down more with a given charge adjustment than the muzzle pressure does. When you adjust seating depth, the peak pressure varies due to difference in gas bypass around the bullet, but the muzzle pressure doesn't change very significantly, so seating depth lets you tune the peak almost independently of the muzzle pressure, thereby effectively changing the pressure curve shape together with barrel time. Both prove useful in dialing accuracy in.

There is an unclear additional interaction there that may be explained by other dynamics that are changed with seating depth, but that have never been clearly identified to my satisfaction. Perhaps the change in curve shape covers it, but I'm not convinced. The symptom is failure to tune by powder charge alone. Since barrel time is changed by both tuning methods, in theory, adjusting one or the other, within the limits of available adjustment range, doesn't always get you accuracy. Changing the charge alone should give you any barrel time you want, but you can tune that regardless of where the bullet is seated. And yet that seldom gets you to the ultimate accuracy load. You almost always have to tune seating depth, too, as this tech article by Berger explains. The Berger article also explains the best loads seem to be possible to obtain over a seating depth range, so folks fiddling with trying to find the best depth to the nearest thousandth of an inch are probably just fooling themselves into believing random variations in group size are real. Most people don't realize how much groups can vary randomly, as your example of two close shots with one way out demonstrates. The way out shot better demonstrated the gun's grouping and the two close together were a random coincidence.

Below is a mathematically modeled 9-shot group, followed by a target that is the first three of the nine, then one that is the second three, then one that is the last three. The RED ×'s are the mean hole location (group centers), so you can see how the average changes with sample size. If you had fired the three 3-shot targets as separate groups, wouldn't you be tempted to believe that the two shots close to each other in the upper right target and the three in the lower left target revealed what the gun and ammo can do, and the thrown out shot in the upper right target and the group in the lower right were somehow your fault? Sure you would. You'd be wrong, though. The firing order and grouping were from a random number generator. This is just how randomness works. If you flip a coin a hundred times, don't be surprised to find five or six heads or five or six tails in a row somewhere in the 100. If you want to know why this is so, ask God. But while you are waiting for the insight to be granted, if you can stop fighting it and accept that this is just the way it is, you begin to appreciate that getting three or even five shots in one group doesn't really give you good enough confidence what the gun may be relied upon to do. If several small groups in a row are all small, that's different, but then you are really comparing the larger total number of shots fired, and not just the three shots in one group, so you have a lot more data. Ten shot groups are much better. Denton Bramwell, who makes his living in statistical manufacturing process analysis and controls thinks something closer to 15-shot groups tells him about as much as he needs to know. In the old days, before Fisher's Anova was published (1920's), 30-shot groups were used because that's about the smallest number you can graph into compartments and start to see a symmetrical bell curve reveal itself.

Finally, note that if your gun needs bedding, lugs lapped, recrowning or receiver blueprinting, or if your chamber is off-axis with the bore, you may never find a tight shooting load. Load tuning always presumes the gun is in good shape, and you can do worse than to buy some Federal match ammunition for it and see how that shoots to get some idea whether the gun needs work or not. Federal has a knack for coming up with loads that shoot at least pretty well in about any chamber and barrel length. Indeed, Dan Newberry's whole OCW concept was inspired by the fact such loads can exist.

 

[jdscholer]

Do you actually need the performance that a VLD type bullet provides, or even sub MOA accuracy?

If your purpose for this rifle is hunting with a light rifle, and acceptable field accuracy with a magazine fed hunting bullet, I'd try a few loads with various "mo stubier" bullets as F Guffy suggested.

If you simply want to enjoy the best accuracy that your little rifle might provide, just figure that you've got a single shot. jd

 

[Mike / Tx]

Since all of my rifles are sporting type hunting they all feed from the magazine. I use this as the max length no matter the caliber or bullet.

I choose a powder in the med-slow to slow end and work up to a velocity that fits my needs. Usually this is somewhere in the 2800-3100fps area depending on caliber. If I see nothing resembling pressure like shiny spots on bases or primer issues I will then start to seat the bullets a touch deeper in .005" increments. Usually things will start to either spread out OE come together within two moves.

As for jump, my friends 300 RUM is shooting the Berger 185 with .125" of jump to the lands and gets stellar accuracy with them. I only worry about seating too deep with shorter light bullets, and with them, there is usually jump to start with.

 

[Metal god]

Not to high jack but I have hard time excepting computer generated stats .

The firing order and grouping were from a random number generator. This is just how randomness works.

I'd agree true randomness is just that . How ever as shooters with reloaded ammo and specific firearms all lumped together . Are our groups really random ? I say no therefore showing a computers random generated anything is not helpful . Did the computer take into account , rifle hold , cant , was the charges weighed or thrown , wind , bullet type ( match or junk ) , trigger flinch , new or old brass , bullet hold . The list is really endless of how our groups are not random at all . Are shots go exactly where we put them based on are abilities and gathered components .

This is not a knock on anything ( really ) I'm just seeing more and more of these theoretical graphs popping up more and more . I find my self having a hard time discerning truly informative and helpful ones over ones that are just a bunch of pretty colors on paper .

 

[cdoc42]

Unclenick, your post stirred some additional thoughts, but these may need some "tuning" to be sure I properly understand your discussion.

Changing a charge of a given powder or changing to a different powder entirely changes the barrel harmonics. As such, the velocity changes and may be thought to be the reason for accomplishing an accuracy goal. Hence the reason that high velocity is not always associated with better accuracy. If this is true, I should be able to limit my powder inventory and simply adjust the charge until I reach the maximum safe charge but have not yet achieved desired accuracy, then look for a different powder that will allow me to experiment with higher velocity. That would have prevented my use of the 4 powders I started with in my Creedmore accuracy search.

The Berger approach speaks just to seating depth change to find the "sweet spot" but ignores the potential need to address barrel harmonics through powder selection.

Your discussion clearly points to the need to do both.

That stuck a cord with respect to Browning's "BOSS" approach, which I migrated to when the concept was first marketed. I ended up with a .22-250, .270. 7mm RM and .338 Win Mag, all with the "BOSS." The success I had finding the "sweet spot" for a given load lead me to believe dampening of the barrel harmonics was the key to success, preventing the need to pursue the same end result by using multiple powders and bullets. That approach was not consistently successful, however, but I now recognize the next step should have been seating depth adjustment.

As an extension of your discussion, I wondered if any negative effect of a heated barrel on reduced accuracy is simply explained by changes in barrel harmonics. Allowing the barrel to cool just dampens the harmonics much like the "BOSS" system -? How long does it take for barrel vibrations to subside enough to be insignificant?

We have all experienced the "first shot" phenomenon, wherein it doesn't go where the following shots do. While we explain that being due to the need to "dirty" the clean barrel,or too much lube left over from cleaning, I wonder if it is more likely due to initiation of barrel harmonics.

All that being said, your discussion, as well as the thoughts presented by others in this thread, will change my approach to pursuing accuracy. I'll start by looking for a powder that provides the highest velocity at the lowest pressure (for safety), but instead of starting 10% under max and working up principally for safety reasons, I'll recognize that somewhere on that ladder is a barrel harmonic sweet spot. If I'm lucky enough to find it with that powder, the next step would be to follow the Berger recommendation for seating depth.

 

[Unclenick]

Cdoc42,

The Berger article isn't 100% clear. They used to recommend all loads be developed with their secant ogive VLD's touching the lands, as they explain. From that, you can infer that powder selections will have been tried that way first, and taking the best load you get that way, you then try the different seating depths to see if it tightens some. When you find a depth where it tightens, you then retune the charge weight.

Once you find a best distance off the lands for a bullet in your chamber, it tends to be consistent with other loads. You can verify that by nudging it in or out, but its jump becomes your new starting point for seating depths developing other loads.

The BOSS doesn't dampen vibration, per se, but rather changes the timing of initial muzzle flip. There is a good animation of an FEA model of a rifle with a barrel tuner firing and all the bends exaggerated 500× at Varmint Al's site. You see it force the muzzle to reach peak position in its first quarter cycle of what will become it ringing vibration after the bullet leaves. You can think of the firing event as plucking the string, and you don't really care how long the vibrations take to dampen out after that because the bullet is gone, and as long as the vibrations are over before you fire your next round.

Read Al's conclusions about what causes the initial deflection of the muzzle. "Vibration" is planted in our brains now, but it isn't an accurate description of what is happening. He does, however, have great animated FEA's of all the different vibration modes a barrel can go through, probably done while they were still figuring the mechanism out.


Metal god,

Yes, you are seeing more computer modeling and statistical evaluation. Bryan Litz uses it a good bit in his books to find what different assumptions do to hit probability at different ranges by overlaying the random grouping on other influences. If you research the subject, you'll find things like bullet quality, neck tension variation, and charge variations all contribute their own bell curve distributions to what his happening overall. So do small differences from one primer to the next, and so on. The human factors are not so neatly random, though.

What my own random target generator is doing is acting like a machine rest gun used to evaluate ammunition accuracy. I was playing with a windage input arguments at the time, which shifted things a bit left, but otherwise, if you take a look at photos of actual ammunition accuracy groups of 100 or 200 rounds, barring instabilities or other non-standard problems, its output is indistinguishable in appearance from those, as are Litz's (I got the idea of making Excel generate those targets from his books). The fact human factors are not included is OK, because once you know what equipment can be expected to do, its up to the shooter to develop his skills to get as close to that mechanical result as possible.

I should point out the plot I made has to be viewed with a grain of saltpeter. As I stated, it's random and round, meaning the standard deviations are equal in all directions. It's also important to mention, as I did posting that same image in another thread a short time ago, but which I apologize for failing to mention here, is that the set of targets is cherry picked. That is, I "fired" a number those 9-shot groups with the F9 key, looking for one with this kind of extreme range of three-in-a-row group sizes. So it's not a typical result, and isn't random in that sense (even though the target POI's were the output of a random number generator set up to generate a normal distribution, or bell curve distribution). What I wanted to show is the kind of result that can occur sometimes, and how that can deceive the load developer into thinking he sees something he does not. The common example is a fellow who spends half the day at the range firing 3-shot groups, then finally gets one like the second group in that illustration and declares "that proves what the gun can do; the others are due to my errors". Nope. He just kept shooting until random luck provided something he wanted to see.

That brings up another important lesson about people blaming themselves for all the errors. I've recently been reading F. W. Mann's century old classic, The Bullet's Flight from Powder to Target. He wrote that the Creedmore clubhouse had a wall with 60 different shooter explanations for POI errors, each one different. Shooter's are funny in that they seem to relish finding a new fault in their own shooting. I suppose it gives us something to work on improving.

Here's a more typical target one, except its the first three sets of 3-shot groups out of a 1000 shot group. Worthy of note here is how large the 1000 shot group is compared to the size of the three shot groups. Then you ask, how tightly does my gun really shoot? Well, the more shots you put in a group, the more chances you give for the less likely and more extreme points of impact to occur, so sample size always increases group size, on average. It's another way of saying a small sample size will tend to get you holes where the probability of find one is highest. But also look at the shots furthest from the center of the 1000, and realize that they occured in a string of 3 as well, so some unlucky stiff will get a couple of holes much nearer the center and one way the heck out in the boonies.

The problem of counting on 3-shot samples came up when board member Pond, James Pond was trying to shoot an OCW round robin. The plot below graphs his actual data. He could not distinguish from the locations of the centers of the individual groups of three (purple diamonds) where his sweet spot was. However, by going to a running average of three sets of three in a row (9 shots; three of each load increment-yellow dots) the resulting 9-shot group centers showed what charge weight everything clustered around right away. You can see it would also have been obvious if the fourth group of three from the bottom hadn't had is center move way off the to left, but it did. Just one of those low probability outlier holes deciding to show up right at that particular time is all it takes to skew a thing like that, and it does happen from time to time, and that's all it takes to make you think you couldn't possibly be near a sweet spot if you get a group that large. That assessment as in error in this instance.