Check out this Graph

[Guest]

While tuning my Stormrider I moved the second barrel band to positions measured in mm from the end of the air tube to the front of the band. I was careful to keep the measurement accurate and then I graphed the Group size, the group width, the group height and the mean radius against the position of the band. I got this very interesting graph. In the end I set the band to 63mm and shot three 5 shot groups which were under 15mm in diameter. The graph predicts a little worse than that but nails the placement of that barrel band.

 

[JimNM]

OldSpook - that is superb data. I like theory - but data is King!

 

[nervoustrig]

Good stuff! The overall group size and mean radius were smallest at 130mm so why not use that position? Or am I reading the data wrong?

 

[Guest]

Good stuff! The overall group size and mean radius were smallest at 130mm so why not use that position? Or am I reading the data wrong?

No, I think you are reading it right. Let me draw a couple of lines on that graphic. I might try 115 and 135 (I think I can get it that far back).

So if we look at the markup. It looks like 63mm and and 115mm are the two points on the graph which are going to yeild the "roundest" group. Notice that the blue line tracks spot for spot either a height maximum or a width maximum. There is another node forming off graph probably around 145 or 150 but I can't get the barrel band into that area without cutting the stock. That ain't happening. It does look like 115 mm should do as well as or better than 63mm. I might give that a try but it is nice from a strength perspective to have that band as far out as it will go. I did end up confirming the graphs by shooting additional groups with the band set to 63 mm. I documented that at the bottom of this thread https://www.airgunnation.com/topic/on-tuning-a-stormrider/ The groups were slightly better than predicted averaging about 14mm or 0.55 inches. So a little smaller than a dime. Not earth shaking for a PCP but good enough to take squirrels at 50 yards or so. The gun is shooting the Griffin 13 gr LDC at 840 fps for 20.4 FPE. 

 

[biohazardman]

Really shows harmonics at werq good job.

 

[nervoustrig]

Okay thanks for explaining the strategy. It makes sense.

I’m curious is either of these favorable locations happen to correspond to the 4/7 or 3/7 factors. I was looking at a picture of the Stormrider and comparing to my Plinkster (CR600W) and soon decided it was futile to estimate with so many potential little differences.

Will you measure the distance from the center of the original barrel band to the muzzle (include the muzzle brake)?

Then how far from the original barrel band would each of these favorable locations be for the second band? 

 

[Guest]

breech=0

first band at 101mm

second band at 318mm

end of barrel at 434mm (pellet leaves rifling)

muzzle at 597mm

first band to end of barrel 333mm

second band to end of barrel 116mm

first band to second band 217mm

333/7*4 = 190.3mm

333/7*3 = 142.7mm

I'm not sure how to make that math work but I suppose there is a point here we can make. It is that the assembly on this rifle is a complex mechanism when it comes to how vibrations will resonate within it. We have two cylinders which are affixed at three points, the breech, the first barrel band and the second barrel band. Resonant frequencies will exist between the breech and the first barrel band, between the first and second barrel bands, and between the second band and the end of the barrel. The position of the barrel relative to the rifle is fixed at those three points. It is the movement of the barrel at the point where the projectile exits the rifling which we are trying to minimize. One could probably calculate the wavelength of the complex vibrations but the graph measures them well enough.

So looking at the graph again.

We can just measure the wavelength of the horizontal and vertical components. I get 25 for the vertical component and 23 for the horizontal component. Now then those might both be 24mm easily enough BUT there is that spot where the convergence between the two does not seem to occur at about 102. It should be there and it isnt but then shows at 115. Those first two convergences are just about exactly 13 mm apart then we miss one and then we get one again at 26 mm (115mm). I bet that convergence is the best one to use and I will test that and get back to you if the wind drops off this afternoon.

 

[JimNM]

What will be the result of two barrel bands at the two best nodes? Is there room to try that? 

 

[nervoustrig]

Each successive barrel band will increase the resonant frequency which means that anti-nodes (areas of high amplitude) are occurring more frequently. By itself, that would be a bad thing because it means there is a greater random probability that the muzzle will be in a slightly different place as the pellet emerges into the world, owing to slightly different dwell times (in this context, I’m referring to how long the pellet spends in the barrel…not the valve dwell time that is more commonly discussed). However it will also be at a reduced amplitude so there is typically a net benefit, just as you might expect by starting with an extraordinarily stiff barrel in the first place.

With that said, the law of diminishing returns begins working its way in pretty soon. For most barrels, even ones that we may be tempted to call soda straws (e.g. the long 7/16” OD barrels Crosman uses on the Maximus or Discovery), you quickly reach the point where adding another band has no discernable benefit in the presence of the other factors that cause random dispersion/scattering of pellets on target (wind, pellet defects, basic accuracy of the barrel, etc.). 

 

[nervoustrig]

Thanks for the pic and detailed measurements of the barrel bands and other features. I think the single biggest factor that makes it hard to predict the optimal location is the presence of the muzzle brake. It detracts from the otherwise homogenous nature of the bare barrel. For example, theory predicts the bare barrel would benefit most from the second barrel band being at approximately 291mm as opposed to the 318mm where you have it. I’m inclined to think the effective length of the unsupported portion (from the perspective the vibration mode translated to the muzzle) lies somewhere between the end of the barrel and the end of the muzzle brake, and that is congruent with your actual location. 

Or it may be a classic case of reasoning the data to fit the conclusion FWIW, I’m not totally pulling this reasoning out of thin air. For example, if we treated the overall length (meaning including the muzzle brake) as the relevant measurement, that would predict the optimal location of the second band at 384mm. That’s much further forward than your actual location, and not as close as the previous prediction that assumed only the bare barrel. Differences of 66mm and 27mm, respectively. However my guess is the presence of the muzzle brake only increases the effective length of the barrel by a small amount because its vibration characteristics are most influenced by the long and relatively thin section of barrel between the original barrel band and the muzzle brake.

At the very least, it’s reassuring that the location is not closer to something that could be considered a halfway point. That would be a head scratcher for sure.

On a more general note, an interesting aspect of harmonic tuning is that it’s not just about minimizing the amplitude of the vibration at the muzzle, but also recognizing that it’s beneficial to arrive at a state of tune that places the pellet at the muzzle just as the muzzle is at a peak (or a trough) of its vibration. That’s because the muzzle spends a greater amount of time at these locations, meaning it’s more likely to be at a consistent place as pellets arrive to the muzzle at slightly different times. If instead pellets are arriving at the muzzle nearer a zero crossing, the muzzle is in the process of swinging with great velocity toward a peak or trough. Of course, lacking a way to observe this behavior directly, there’s a bit of trial and error associated with it. That’s why it can be intensely frustrating when accuracy inexplicably goes to pot, but it is simultaneously encouraging to realize that you’re perhaps never more than one tweak away from changing a “pretty good” to an “outstanding”.

 

[Guest]

On a more general note, an interesting aspect of harmonic tuning is that it’s not just about minimizing the amplitude of the vibration at the muzzle, but also recognizing that it’s beneficial to arrive at a state of tune that places the pellet at the muzzle just as the muzzle is at a peak (or a trough) of its vibration. That’s because the muzzle spends a greater amount of time at these locations, meaning it’s more likely to be at a consistent place as pellets arrive to the muzzle at slightly different times. If instead pellets are arriving at the muzzle nearer a zero crossing, the muzzle is in the process of swinging with great velocity toward a peak or trough. Of course, lacking a way to observe this behavior directly, there’s a bit of trial and error associated with it. That’s why it can be intensely frustrating when accuracy inexplicably goes to pot, but it is simultaneously encouraging to realize that you’re perhaps never more than one tweak away from changing a “pretty good” to an “outstanding”.

Good stuff. I need to think on this a bit. I see no wind outside today but it is sprinkling. Supposed to let up later. I may get in some groups with that band at 115. If I do I will post below.

Also the 4/7 node does exist. It is present on the graph at the 90mm position. I was measuring the location on the graph from the end of the air tube backwards. That makes all of this harder to visualize. As it turns out the end of the air tube is the same point where the pellet exits the rifling. So the air tube is as long as the barrel.

On the presence of the moderator: It no doubt has some effect on the fundamental harmonic of the barrel (or that section after the second barrel band).

The waves we seem to be measuring here are not at the fundamental harmonic though. I will get back to this.

 

[ackuric]

I would like to see the original data set ran 5 to 10 times, noting the max/min/avg of the tests prior to concluding that there were no other offending variables effecting the results of said tests, meaning if they're as conclusive as presumed, you should get identical results every single test you run with different barrel band positions. Other than that, I certainly look forward to further tests and hope you continue down the path! GL.



-Matt

 

[Guest]

I would like to see the original data set ran 5 to 10 times, noting the max/min/avg of the tests prior to concluding that there were no other offending variables effecting the results of said tests, meaning if they're as conclusive as presumed, you should get identical results every single test you run with different barrel band positions. Other than that, I certainly look forward to further tests and hope you continue down the path! GL.



-Matt

I'd like to see that also but it won't happen with these 13 griffin slugs. I agree it would be nice for someone to take this ball and run with it. I might do it myself using something less expensive as my projectile. If I do, I'll make sure all the data are available as they are here. Given the time, I would like to do that.

As to "identical" results, no, you never get identical results, you get (at best) statistically similar results. We have a large enough data set, in this case 9 shots at each of 10 different band positions. Here is why I believe that, each of the band positions has nine samples. We are calculating the center of the group for those nine samples and calculating the mean radius of the group and the horizontal dispersion and vertical dispersion. These numbers are relative to the center of the group, not the point of aim. A total of 90 data points to establish the curves we plotted on the graph. There is no doubt that there is human error in those groups; however, when three intersections on the graph are regularly spaced to within a millimeter that's a fairly strong indicator that the maxima and minima are where they appear on the graph. 

Also the graph is predictive. It says I should get better groups at certain locations and I DO get better groups at (at least) some of those locations. I shot groups again at 63mm and 115mm today. A careful look at the graph suggests that I should get slightly better groups at 115mm than I get at 63mm. So lets look at those targets. I shot two targets. The target which has the 4 five shot groups shot with barrel band at 115mm also has some random shots on it from other pellets. I circled the pertinent groups. The third group on that target has a shot outside the circle marked with a "?". That shot was the first shot in that group and should be inside the circle. The other target has 4 five shot groups on it shot with the barrel band at 63mm. I ran the targets though "On Target TDS" and got the following data:

63mm: Mean Radius 6.3mm; Average CTC 17.0mm

115mm: Mean Radius 5.7mm; Average CTC 17.4mm

I then tossed out the worst shot in each group and ran the targets through TDS again and got:

63mm: Mean Radius 4.8mm; Average CTC 11.6mm

115mm: Mean Radius 3.5mm; Average CTC 8.7mm

I would have liked a stronger confirm than that but the truth is the predicted difference is just about what you would expect from the graph. The truth is the graph says a band placement of 63mm will give a group size of about 18mm, and it does. It also says a band placement of 115mm should give a group size of about 17mm and it does.

Here are the targets.

Anyone wanting calibrated images let me know. I will take all the pictures again with a ruler in the pictures so that you can check my numbers.

 

[Guest]

Akuric, I thought you would want to look at this.

I ran a series of 10 shots per point using the Crossman Premier 10.5s staight from the tin. I got a very different graph. You had said you would like to see that same test run multiple times. I could not afford to do that with the Griffins, well I suppose I could have but ... would rather not. I shot the comparison graph so as to see if we got a similar graph with a different pellet. Very different graph on the right side. There is a similar trend displayed. In both graphs mean radius improves as me move the band closer to 120mm from the end of the rifle. There are very different oscillations going on the left from the right. I'm going to run a series with the H&N FTT which is 8.6 grains. I'll add that to this graph when weather permits ... won't be today. 

 

[Guest]

Shot this a couple of days ago. The group at 115mm was well under a dime. There was a right to left wind gusting to about 10 or 12 mph. So that would explain the width of the group at 115mm.

I got my best group at 115mm with the CP 10.5s as well. So that is interesting. I realize it is not statistically certain but at this point I am still developing a theory. Maybe tomorrow I can shoot some other pellets to get a feel for how picky that sweet spot at 115mm is actually. If other pellets prove consistently accurate then I will do more testing.

Here is that target, the additional data was taken from the targets already documented. I did not graph the group at 133mm.

 

[scratchcoobster]

Great idea, and good work. I am going to try this to get the correct placement of my barrel bands. Impressed with what the actual data shows.

 

[Guest]

Great idea, and good work. I am going to try this to get the correct placement of my barrel bands. Impressed with what the actual data shows.

Well actually Ackuric is right. There isn't enough here to be positive of what we are seeing. It seems like a good way to "pattern" a rifle systematically if you want some structure. It did work for me in this case. We have had quite a bit of wind lately. Today I shot anyway.

I took seven diffrent pellets out and shot a ten shot group with each of them. Then, because of the wind I tossed out the worst shot in each group and plotted the results. They are not earth shaking but the CP 10.5gr came in again right about where they should be. I will try to refine the band position now to see if I can get another improvement.

Looks like there are pellets the rifle prefers with the band at that position and regulator set at 120 bar.

Likes the CP 10.5s, Polymags, H&N FTTs and the JSB Exacts. Does not like the JSB Heavies or the Baracudas.

Targets:

 

[Guest]

Well clearly I've been chasing rabbits. After I got the barrel band settled in, I decided to sleeve the barrel in carbon fiber. So I bought a half meter of the stuff 16mm OD and 12mm ID. Then I figured since I was doing that I should clean up the barrel bands. So I took a grinder to those, and files, and wet sand. I got some West System epoxy and fitted it all together then glued it up. I can still remove the barrel from the rifle as a unit. The barrel bands are now integral and can't easily be removed.

So about that. I found that a band placement of 115mm to 120mm gave me the most consistent results with the largest number of pellets. Then I went back in and cranked the regulator back to about 117 BAR from 120 BAR. I just wanted to pick up a couple of shots on the end of the string. It now gets 20 shots on a fill and then falls off the regulator pretty hard. There are 4 projectiles the rifle seems to prefer at this point. They are:

Griffin 13 gr Slug @810 FPS for ~19 FPE

CP Heavy 10.5 gr @890 FPS for ~18.5 FPE

H&N FTT 8.64 gr @945 FPS for ~17.1 FPE **BEST GROUPS**

JSB Exact 8.44 gr @960 FPS for ~17.3 FPE

So good enough selection. The number two pellet was the JSB Exact and they shoot well enough. Here are composite groups taken from 5 five shot groups I did today on two different targets.

Dog picture is gratis

More on the rifle here:

https://www.airgunnation.com/topic/on-tuning-a-stormrider/