OCW test gone awry. Velocity all over the place!! Chrono?

[Bart B.]

Dan Newberry's OCW works, but it takes at least 300 yard test range or more to do well using it. That's when you'll see faster bullets strike the same place as slower ones because they're all leaving on the muzzle axis up swing. Depending on the barreled action's whip characteristics, a grain or two different in charge weight can put the bullet exiting at a more favorable place.

His OCW method is based on Chris Long's theory about a shock pulse going from the chamber to the muzzle then back and forth. The bullet's supposed to exit when the pulse is back from the muzzle so the muzzle's not bulged out or shrunk in from that pulse. I think that theory's flawed.

Problem is, there are two pulses originating at the chamber; one starts forward and the other backwards. They both go in opposite directions until the reach the end of the barreled action then reverse. They make about 4 round trips before the bullet exits and they're about 15 inches apart when they start.

The pressure pulse that shocks the barrel lasts for about .001 second. I don't know how long that pressure pulse is Chris Long envisions. But rifle barrels with a front sight band tight on the muzzle squeezing the bore several millionths inch don't hurt accuracy. Neither does Garand barrels at their muzzle with their bore and groove diameters opening up .001" or more from cleaning rod wear. The same lot of good commercial match ammo shoots well under 1/2 MOA at short ranges in different barrel lengths and different bore/groove diameters. Those shock waves will often be at the muzzle when bullets exit. Accuracy is not dependent on a given barrel length for a given lot of ammo.

Therefore, I think those shock waves have about zero effect on accuracy when they're at the muzzle when the bullet exits.

[Pond, James Pond]

Well, something I don't understand is how a method such as the OCW technique is better or worse than an other.

Ultimately they all just raise charges by set increments, don't they?

How is it that, say the OCW, would be flawed at 100m but just loading round 1gr or 0.5gr powder charge increases is somehow more representative at 100m?

For me shooting at 100m is very handy. My scope is 12x max, meaning a target at 300m is far less distinct (IE being able to discern which part of the target to hold on) and therefore far harder to see hits.

[Bart B.]

Yes, the methods are similar in that they use different charge weights. But if the difference between them is too small to be seen at the testing range, you won't see their effect. And several 5-shot groups shot with the same sight settings aimed at the same place will not have their centers at the same place nor will they all be the same size.

A 1 grain powder charge difference changes muzzle velocity about 60 fps and the bullet drop at 100m about 3mm; at 300m the difference is about 25mm. Your group sizes at 100m are masking the difference. And their sizes are too close to each other for the different charge weights to show any real difference.

At short ranges using that black dot on the paper as your aiming point, put that dot in the upper right hand corner of your scope's reticle crossing. You'll have much better repeatability of aim doing that. Or paste white squares on the paper that appear twice as wide as the reticle wire width then center it on that square.

[243winxb]

Pressure signs http://www.shootersforum.com/handloa...ure-signs.html

Increase in powder charge achieves unexpected velocity. (Average velocity will tend to increase by the same number of feet per second per grain of additional powder over the normal operating pressure range. If your next charge increment fails to produce the expected additional velocity or produces too much additional velocity, pressure may be high. Poor grouping usually accompanies this symptom. It is also caused at reasonable pressures by uneven bolt lug contact (lugs need lapping), in which case still further charge increments go back to producing orderly velocity increases and grouping improves. Suspect this last situation if the charge at which the velocity anomaly occurs is in the middle of a published load range. Otherwise, back the charge off 5% from where the issue started.)

[Unclenick]

A couple of things:

The 100 m groups change with precision, but as Bart said, it is hard to see a difference that is due to drop, alone. That is why Creighton Audette always used 200 yards, and why Randolf Constantine recommended taking that out to 300. The downside is that you need a big, bright objective on a high power rifle scope or a good spotting scope to keep track of which hole was made by which shot at that range. We who've shot Highpower almost all seem to have a nice bright Kowa spotting scope somewhere among our gear. If you don't have that, the longer ranges are harder to shoot an Audette Ladder with.

The OCW is a little different. You don't put all the shots on the same target as you do an Audette ladder, so keeping track of the individual shot holes can be skipped unless you are looking for heat walking or fouling shifting POI or some other trend that requires noting each hole's individual location. It's true that a 300 m target is a little harder to see, but that's what Newberry uses most of the time, so I think he's found an advantage in the extra drop easier to see, too. However, you can still find non-drop dispersions at 100 m. It just takes more effort.

The pressure wave concept is an interesting theory, but the OCW working or not working doesn't depend on it. It only depends on the groups you get. The OBT theory just tries to suggest where you might find such sweet spots. Dr. Lloyd Brownell actually noted the existence of the pressure waves on his strain gauge readings back in the 60's. He believed they were initiated by the impact of the bullet jump to the throat, which slows it down and lets pressure build before it is fully engraved and accelerating again. These waves are fast, but not large, and you can see them in the pressure measurement I made, below. You only see them clearly after the rest of the pressure has dropped, and there the negative half of the ring is missing here because the Pressure trace doesn't register below zero signals. The pressure wave excursions are mixed with the pressure curve starting from where it is around 30,000 psi and up in this plot. They are what cause the curve to be lumpy and rippled rather than smooth, like a QuickLOAD plot. Before 30,000 psi, the rising curves are smooth, but 30.000 was where the bullets were slamming into the throat. For most cartridges the ripples start at more like 12,000 psi, but the M2 ball used here has age-related high bullet pull, so the start pressure is high, too.



If Brownell's theory of creation of the pressure wave is correct, a bullet seated out to touch the lands should not create a pressure wave. I have not made the comparison in that same rifle to see. It's something to put on my bucket list. It would suggest Mid Tompkins, who always soft seats, doesn't have to concern himself with pressure wave effect.

Whenever you have the problem of telling one group average from another, you can apply what is called Student's t-test. It will tell you whether or not two averages are different to a specified degree of confidence, or if they fail to show their difference is more than just random to that same degree of confidence. Conversely, you can use the t-distribution to tell you how much confidence you may have that two averages are actually different and not just randomly different. This gets you around eyeballing the statistical significance of the difference by putting an actual number to it. You may or may not be interested in this, but Excel's data analysis package includes a t-test generator you can use by activating the ranges of data you mean to examine and telling it what confidence level you want to test.

The 100 m groups do just fine at identifying spread due to something other than drop, such as barrel vibration. However, the 3-5 shot bulls by themselves have too few samples to make the difference clear. This is partly described to a statistic called the standard error, which tells you how much to expect the group centers to move around from one group to the next. You find it by dividing the standard deviation of each shot radius from the center by the square root of the number of shots in the sample. Mathematically, it is the group-to-group standard deviation of the standard deviation of individual shots. Why it matters is shown in a graph of bullet vertical position from one of James's previous groups.

The graph tells you the same thing about barrel vibration an Audette ladder does, but does it at 100 m and does it for both horizontal and vertical at the same time (the OCW advantage over the Audette ladder). Note that the purple diamonds are the centers of his 3-shot groups, and if you looked just at their locations and at the similar sizes of the groups (not shown here), you'd have no clear clue as to where a ladder sweet spot might be. But when you look at the yellow circles, it is very clear which two are closest together, both horizontally and vertically, as the charge weight increases, and are therefore where the Audette vertical sweet spot is.

Those yellow circles are what are called running averages. Each one represents 9 shots. The first combines the first 3 3-shot groups 1, 2, and 3. The next combines 2, 3, and 4. And the next combines 3, 4, and 5, etc. Each time you create a point, you dump the first 3-shot group and add the next one to make the next 9-shot data point.

By combining the group centers that way, you get a larger, 9-shot sample for each data point, but most important here is you get a standard error about 5 times smaller. This is because you not only have the overall standard error smaller by a factor of 1.7, because of using 9-shot groups instead of 3-shot groups, but because two of the 3-shot groups making up each 9-shot group were also part of the previous group, you are only changing a third of the holes that influence the position of the center at one time. That reduces the average change between 9-shot groups by another factor of 3. The result is the mean position jitter is much reduced, leaving the average change in position much more evident. In this case, it's any charge from 40.1 to 41.0 grains that is within the sweet spot range, and the horizontal shift from 40.4 to 40.7 grains is almost certainly just random. If James sets his measure to average 40.55 for this load combination, he had a range of ±0.45 grains he can err without seeing his groups grow appreciably.



James, I'll recommend a tool for you to help gather data for these kinds of analysis. It is a $12 piece of software called OnTarget. There is a free version you can get (bottom of the page) to try it out, and you have to pay for the full feature version. In it, you import an image of your target. You want to have a rule included. You then calibrate the software with a little "Set Reference" tool that you click at each end of a known distance in the photo (the ruler graduations), then enter the real world difference in the little pop-up window that appears after the second click. You then tell it your bullet diameter and the range of the target and another tool lets you set your point of aim and place bullet-size circles over each hole. It then calculates the groups stats automatically. In the "tools" menu you can then export individual hole data to a csv file that Excel can read. You can then do all the stats you want on the hole positions, including the kind I did, above, in Excel.

If you don't like going the Excel root, the same company offers a more sophisticated program that you print targets from. When you scan those targets, it recognizes them and finds the hole centers and does all manner of common stats for you without you doing anything else. It also has the ability to virtually combine multiple groups into one. But this program costs a good bit more ($75).

Finally, on your chrongraph, it does sound like it was off. Most chronographs use consumer grade electronics that are out of spec below the freezing point (32 °F or 0 °C), so you can't use them when it's colder than that. Batteries also have more internal resistance in cold weather. If it was cool when you shot the groups, I would consider buying the Energizer lithium batteries for the instrument. They have much, much better power delivery at low temperatures than alkaline or carbon-zinc batteries do.

I would keep an optical chronograph at least 15 feet from the muzzle for rifle shooting. The screens can be tricked into false triggering by ejecta from the gun. If it were a heavy magnum I would go for 18 feet. An easy way to set one up is with a laser bore sighter. You take the bolt from the gun and put it on a piece of paper that says "Remove Bore Sighter before putting me back". You set the gun up on bags so the sight stays on your target. Turn on the bore sighter, pace off the chronograph distance, then set the instrument up, holding your palm in the centers of the front and back screens, alternate, to find the laser and to adjust the chronograph mount so the screens stay centered that way. When you go back to the bench, a quick look through the scope will tell you where the rifle needs to point to shoot through the centers of the screens.

All the usual lighting condition warnings apply. Also, no light colored ground under the unit to cause bullet glints on a bright day (use a dark tarp if need be).

[Pond, James Pond]

Tell me, Unclenick... Have you ever thought of writing a book? I think some of the best chapters were in that post already!!

I confess that stats just need to look at me and I get confused! I'll need to really sit down and digest your description of how to do that 9-shot plot. Not that it was not clearly explained, just that I need to get my head around it.

I've been referred to OnTarget before and it certainly sounds interesting. I'd like to get it, but it does not accommodate my old version of Mac OS.

I'm actually thinking of getting a cheap laptop PC with Windows 7 or something and just use it to run all the programs my veteran Mac will not: Mapsource, Quickload (if I get it) and something like OnTarget. Until that time, I guess I will just have to crunch numbers as I can!

Regarding the Chrono, it was about 3m from the muzzle, so possibly a bit close. It was a warm, sunny day, though and the diffusers were in place.

[JeepHammer]

I was right with you all up until you decided to shoot the computer,
Not the rifle/target...

All the therotical application of vibration data, wave pulse band width ect. means nothing since that's not what puts lead on target.
APPLICATION.

I saw desk pogues, math geeks & computer nerds by the hundreds try and figure out what was 'Supposed' to shoot, hundreds of prototypes built,
And in the end, EVERY weapon/system ranged the target, then used actual test fire ballistic data to put lead on target.
The projectile doesn't know its supposed to act some different way, it simply follows the laws of phsyics as they exist at the time of firing.

Vibrations in steel move at LEAST 8 times the speed of sound,
The composition and grain of the steel will change harmonics up to 20%.
That was the basis behind spending BILLIONS on ceramic, composite and single crystalline structure barrels.
A set rate of harmonics, since every batch of barrels will have a composition difference, and even the machining process will makes barrels from the same batch have slight differences...

And you can't very well test fire each and every barrel for several thousand rounds and still call it 'New' to collect data for each barrel in invintory,
It's not practical, and its certainly not efficient.

--------

Accuracy shooters found out long ago, its just range time...
1/10 grain changes make thing much longer than they have to be.
Bump up or down a grain or 5 each way from a median starting load,
When it starts to group, then drop to a single grain and run them up/down by a single grain to see where it shoots best.

When you get it shooting, then start in with electronic data collection...
That will save you a ton of time and aggravating/conflicting results.

The velocity, and everything else doesn't matter if you can't put lead on the 10 ring. Accurate is more important than a couple hundred FPS, or even 1,000 FPS,
Or what some program says you *Should* be shooting...

Chronographs will read stupid with just a few degrees change in sun angle.
They will OFTEN read 30 or so rounds from the same batch at nearly identical velocities, and a half hour later they will read the same ammo all over the place through the same rifle.
It's just faulty data collection, not the ammo or firearm...

When I shoot through a chronograph I shoot 10 shot strings for a reason,
If they are ALL jumping around, its probably the ammo, so I shoot a *Standard* batch I know are consistent to check the function of the unit.
If it shows them jumping up and down, its the chronograph,
If it shows the 'Standard' batch to be consistent,
Then I know its the ammo I'm testing.
It's just not a good day for the chronograph some days...

[Bart B.]

The same rifle and ammo will easily produce a 100 fps average muzzle velocity spread if shot with 12 different people. It's amazing how different we all hold a rifle against us. And all those velocities will be lower than if the barreled action was fixed in place like SAAMI universal receiver mounted test barrels.

[JeepHammer]

Some of the Marine shooters got VASTLY different results in 'Range' conditions than the data collection guys got when firing the same barrels/actions/ammo from fixed test fixtures...
This drove the data guys CRAZY with a Capitol 'C'...

The field guys left the ammo set in the sun often times,
This gave them higher muzzle velocities.
On cool or cold days, no sun, the guys kept their mags/ammo in their shirts/coats.

The field guys left the gillie wrap on the barrels, or the left a damp towel on the barrel...
You don't know in the field if your final firing position will be in direct sun or shaded, so you shade the barrel at all times for consistency.
If its really hot, not just sun on a reasonable temp day, a damp towel keeps the barrel from expanding.

All field craft, and a computer won't know that stuff...
Neither did the data crunchers!
Results were all over the place and drove them nuts!

[Bart B.]

Shooting .308's and 30 caliber magnums at 1000 yards, I learned that rounds heating up in hot barrels could be easily accounted for as follows.

For every 30 seconds a .308 Win round stayed in a hot barrel chamber, I'd come down 1/4 MOA.

For every 20 seconds a 30 caliber magnum round stayed in a hot barrel chamber, I'd come down 1/4 MOA.

Longest I've waited with each between shots was about 3 to 4 minutes. After that hot round shot, I came back up most of the original correction then fired a shot, then came back up to the normal elevation zero.

There's been a few folks shooting long range at the Nationals when a check fire was announced and everyone unloaded their rifles until a boat cleared the impact area at Camp Perry. At commence firing, they'd load a round in a much cooler chamber, shoot it, call it good, only to have the spotter show up in the 9 ring at 6 o'clock. That lost point has cost a few to lose the match they would have otherwise won.

All of which is why I let rounds set in the chamber for 15 seconds before shooting when testing for accuracy.

[JeepHammer]

I was doing a thread on accurizing an AR upper where I was going to show much the same thing.
Shooting the rifle in a stationary jig, clamped down,
The first shot cold bore, then shoot 9 rounds behind it for a 10 shot group,
Then show how that string takes off in a particular direction...

Then square up the upper, and do it again, showing how the string doesn't take off nearly as violently headed for points unknown.

The barrel will shift one way or another, just a little,
But you won't get that string headed off to someone else's target!

With ARs, its usually the front face of the upper receiver not being true with the bore centerline,
Heat expansion pushes harder on the high spots which shifts the bore,
While low spots sometimes don't even expand enough to contact the barrel or exert any real pressure on the barrel,
So you get a dramatic angle shift as the barrel/upper heat up.

If you square/bore align everything, the shift is DRAMATICALLY reduced,
And you start to see exactly what you were talking about,
Bore expanding, rounds heating in the chamber and strings moving up/down instead of sideways or some strange angle.

Bolt rifles don't usually display this since the barrel screws directly into what is usually a lathe cut mount face, but with the barrel nut arrangement on ARs clamping the barrel against an angled face, its a real issue.

[WWWJD]

For what it's worth, I've never found an accuracy node with powder charges over 44gr using 4064, Varget, RE15 and 160-190gr bullets. All of those tests kinda felt the same way; waste of powder and bullets.

For whatever reason, I could almost bet that regardless of bullet weight and powder (within reason), my most accurate load will be with 43.0 to 43.6gr of any of these powders with this particular rifle. Shooting OCW, I really have to split hairs to pick the best of the bunch. Either side of that range, groups get wild and never settle back into another node.

To Bart's point (here or in another post), I've nailed my reloading method down so consistency may be outweighing some other factors at this point.

[Bart B.]

Bolt rifles don't usually display this (walking shots as the barrel heats up?) since the barrel screws directly into what is usually a lathe cut mount face

If that's true, how come commercial factory rifle's receiver faces are rarely square with the receiver's barrel tenon thread axis and have been notorious for decades for walking shot holes on target as the barrel heats up?

And the fact that once they're squared up properly, no more shot stringing happens?

[JeepHammer]

That's the $64,000 question...

This is a personal opinion,
But I think it might be in the manufacturing process procedures as things went from qualified gun smiths doing the machining to hired line workers doing a particular process then handing the pre cut blank to the next machining station with the next 'operator' doing their bit, and so on...

First off, virtually all 'Production' bolt rifles with round receivers start with a piece of raw tubing chopped off to length, then the barrel face is ground flat with a guide down the hole in the tubing.

When I watched this process, the first thing through my head was the tubing didn't even get blown out with an air hose!
There HAS to be chips from the CNC saw/coolant in there!

There is no way for that face to be aligned when its ground with lubricant/coolant & slivers/chips in there,
They MUST square that up on a lathe at some point...

Nope!
Goes into the milling center AS IS, the milling center indexes and runs a sizing reamer through the bore to standardize size,

So again I think, now they will face off the barrel end square with the bore,
But guess again!
Never touched a lathe the entire manufacturing process!

A big threading head in a separate process chews, rather than cuts the barrel threads, while indexing/clamping the OUTSIDE of the tubing!
They didn't even use the inside, reamed to size bore to index off of!

Even so, you still aren't going to get a 0.040" pencil eraser to pencil lead raised spot on those receivers like you will with the 'Drip' clinging to the front of an AR upper when its anodized,
Nothing quite like all the barrel nut force, and recoil force, and heat expansion force being concentrated on a pencil lead size 'Drip' on the front of an AR receiver.

In the days before CNC you already knew the saw didn't cut the piece off square,
You already knew the tubing wasn't DOM, so the inside was not anywhere near to true,

So the first thing you did was chuck it up,
Bore the inside for true and even, then face the tubing off at 90 degrees.
Problem solved,
And while it was chucked up, might as well LATHE cut some threads true with the receiver instead of forcing a tap into that tubing with questionable alignment and results...

The FIRST THING a gun smith does in a custom shop is pull the barrel,
Stick the receiver in the lathe,
Check the bore of the receiver, the front face, and the threads to see if everything is square & true!
They don't even bother with a testing fixture in most places since its so common to have bore, threads, face at three different angles...

And we all do it when tuning up production receivers, its just that common...

When I build an AR upper, the FIRST thing I do is use a upper receiver bore jig with a 90 degree square face,
And I lap the front face to check for square since you really can't get accurate readings trying to compress that anodizing.
A quick had lap tells me in one minute EXACTLY what that front receiver is doing, and I have yet to see one that is anodized and true/square.
100% failure rate with anodizing,
And a good 70% failure rate with the high dollar units that don't have the stupid thick/uneven anodizing...

Either way you have to do it yourself or hire a custom shop to square up what should already be true/square in the first place...

You would NEVER find one of the old Weatherbys with an angled receiver bore, out of square face or tap cut threads headed off in the wrong direction!
It's strictly a modern mass production thing that a real master gun smith simply wouldn't allow to happen, no matter how many pennies were saved in production...

I guess milling and turning is too much to ask for,
So everything gets milled and that's it...

[Unclenick]

I can confirm much of what JeepHammer said about manufacturing. I've done a little work for one of the major gun makers in the past (NDA, so I can't say who) and seen the factory. Receivers were cut from stock on a CNC machining center and the faces were just milled and the receiver threads were a bore and tap operation for speed. No lathe facing or single-point threading. Nor did they chamber barrels in a lathe principle machine, but rather the barrels were muzzle-down in an indexing collet chuck on a Burgmaster. When a barrel was in position for reaming, coolant was pumped from the muzzle up through it. The spinning roughing reamer came down and pre-reamed it, and then the Burgmaster tool head turret retreated and indexed to a finishing reamer and repeated its plunge. The problem with this approach is a deflecting reamer, say, from getting dull, can cut an oval chamber, where the same deflection on a lathe gets you an oversize chamber breech, but one that's still round.

They also straightened barrels as needed. I suggested straightening blanks before contouring so the metal thickness around the bore would always be axisymmetric, but I don't think that suggestion took.

When the pressure wave theory first came out I thought it was a great idea as it seemed to fit observations of multiple sweet spots. However, as JeepHammer says, the speed of sound in steel varies with the alloy and grain orientation. I measured it on several barrels and got anywhere from what would be about Mach 14 to 17 for STP air on the long axis. It's enough variation that the theory's proponents who claim to predict sweet spot barrel time correctly within 2% are going find that span (4%) will cover node overlap for the different possible sound speeds most of the time. So its success is a sort of self-fulfilling with that tolerance. Once you have an established accuracy load, though, working backward from that it does often seem to get you close to other sweet spots, particularly in thin, whippy barrels. It's analogous to the way establishing actual bullet drop at some reasonably long range, lets you find come-ups for other ranges. But you still need to shoot a baseline calibration.

Anyway, the bottom line is there is no getting around testing a gun to see what it likes to shoot best. Dan Newberry is correct that a good sweet spot load is one that tolerates a lot of slop in exact powder charge. That's not just because of powder dispensing control, but because the pressure changes that occur with variation of powder charge are mimicked by things like change in temperature. So if you want a load that has the most immunity to changes in conditions, you want one with a wide charge weight sweet spot window.


JeepHammer,

I wasn't quite sure what you meant about "shooting the computer". I thought perhaps you were referring to that program I linked to. It doesn't shoot or try to predict anything. It just finds group sizes and individual hole locations by working from a scan or a photo of the target. It's convenient in that it does the measuring for you and can spit the measurements out in a form Excel can read directly. That lets you apply statistics to act like a sort of noise filter to make it clearer what the trends in POI are. But that doesn't tell you how to shoot. It's just computerized 20/20 hindsight about how you were shooting.

[JeepHammer]

Unclenick,
You went past what I have done personally.
I've seen every component made, I've seen every component tested in about every way the 'Egg Heads' could come up with,

The projectiles, barrels, and ballistics are still the same as they always were...
No big break troughs, just very slight incremental advances in STATISTICAL ANALYSIS.

After all the billions spent during 'Star Wars' research, and all the trillions spent building and testing new weapons systems...
The only real advance is data collection and analysis.

None of the 'Super Duper' ballistic weapons research projects, with big changes in materials or design, ever lived long enough to see standard production.
The old steel barrel, non-ferrous projectile contact point is still king,
And all advances in 'Sighting' are based on actual trajectory observations,
Added accuracy has yet to be designed into the projectile or barrel through some space age design or use of materials.

They still shoot the thing, in about all conditions,
Track the round, log the data, and refer to those data tables instead of coming up with,
"This design/load will be dead accurate" and that statement being so.

I saw guys 'Shoot Computers', design something that was 'Perfect' on the computer models,
Just to have it shoot like crap in the real world...
I saw it for over 10 years, so I'm pretty sure it's still happening.

We are still shooting the same basic design firearms,
The same basic design ammo,
And the same basic distances since the Civil War,
And the accuracy has only slightly improved.
Considering the technology advances in everything else...
I'd say we are up against one of those physics curve apexes and all gains from this point forward are going to take a computer to identify...