Page 3 of 11 FirstFirst 1234567 ... LastLast
Results 61 to 90 of 321

Thread: Paintball Spin Physics - Getting to the final Answer

  1. #61
    Join Date
    Mar 2002
    Location
    Sacramento, California
    Posts
    158
    Originally posted by AGD
    Glenn,

    If a barrel did it's job perfectly every time and 100% of the spread was due to external forces then it would be a waste of time to try and improve it.
    Tom, I have to agree with that statement but it is based on some pretty large and asumtive "IFs". It seems to assume that the barrel is the only factor besides external forces that can have an effect on the placement of the ball. One could also state that if changing only the barrel on a particular piece, produced a measurably improved shot pattern, it was not a waste of time to try to improve on it. It would also be a safe assumption that the way the ball was presented to the external forces, had a significant influence on the effects of those same external forces. Maybe such a barrel was just better suited to the forces behind the ball or maybe just better suited to the balls themselves. In either case, the result was to present the ball to the external forces in such a manner as to allow better consistency of placement of an inconsistent projectile.

    [i]
    The problem, as I see it, is that people spend 300 dollars for a barrel not knowing if it will make a 1% or 50% difference. [/B]
    I know exactly where you are coming from and I whole heartedly agree. New and/or different is not always better.
    Even more of a problem is that those same people will often find a way to justify the expence whether it provided an improvement or not. Style points have become far to important to some. [/B][/QUOTE]

    [i]
    In general I have to ask you, how much of an increase in accuracy have we really seen in 15 years? Given the fact the barrel prices have increased by 10x and are now honed and sized to perfection, what are we getting for the money? [/B]
    That is a tough question to answer with any sort of brevity. In the context of this thread, there has been very little improvment in the overall accuracy potential of a paintgun in relation to barrels. Although, I believe that the typical paintgun of today will show measurably better results than the typical 'gun of times long past. It is much harder these days to produce significant improvments to shot groups by changing the barrel because of the general improvements to barrel production as well as the evolution of valving. Essentially, the guns that I built 15 years ago are every bit as accurate (for any one shot) as what we are producing today but I get better shot groups these days due to the better consistency of the air supply being used.

    [i]
    While you may be willing to spend big dollars on a 1% improvement most will not or at least would like to know what they are getting.

    AGD [/B]
    To the contrary; I am neither willing to spend my money in that regard, nor am I willing to take my customers' dollars for something that may not be of any real benefit. I've never been a proponent of having a bag full of barrels for your paintgun. Especially since most of my equipment is produced with non-changable barrels. However, we do offer services to ensure that a 'gun and its barrel are tuned to compliment each other and provide the opportunity to put the ball where you need it. I still believe that the valving of the 'gun is at least equally important as the barrel when in the pursuit of optimum performance.
    Glenn Palmer aka Paladin
    Do it right or don't bother.

  2. #62
    Join Date
    Nov 2002
    Location
    North of Phoenix, AZ
    Posts
    482
    Originally posted by yeahthatsme
    if somebody would send me a rifled and straight bore barrel and a way to secure a marker so that it wont move when fired i would be perfectly willing to test the grouping differences between rifled and straight bore. i would use an automag with a max-flo air tank. so if i can get the stuff i will do the test.....
    You would need the barrel to be exactly the same except for the rifling though. Unless you can get armson to make you a non-rifled stealth barrel this might be hard for you. You may also be able to get a hold of two old J&J custom barrels.
    Originally posted by AGD
    What are some joys and struggles of your career?
    The joys are when you make it work well.
    The struggles are when they want it to be a different color

    AGD



    http://groups.yahoo.com/group/thefifthmarker/

  3. #63
    Join Date
    Oct 2000
    Location
    Chicago
    Posts
    5,947
    All,

    Now I just want to point out to everyone that Glenn and I are having a very pointed discussion about the subject at hand. We are both presenting facts, opinions and points of view.

    It is refreshing to have such a discussion without it breaking down into name calling and disregarding the others opinions. This is why I love Glenn, he is a no BS guy. Doesn't take it and as importantly doesn't give it.

    This type of discussion represents what Deep Blue is suposed to be all about.

    Thanks Glenn.

    Tom

  4. #64
    Join Date
    Oct 2000
    Location
    Chicago
    Posts
    5,947
    Ok back to the battle,

    Glenn, I am purposely leaving the valving air blast thing out of this becasue it's one of the things on the table to be examined at the end.

    You do make a point that todays groups may in fact be closer due to the better propellent, aka compressed air, than a decade ago. I also thought about why you may be getting a full powder scrape off the inside of the barrel. It could be because todays balls are much rounder than 10 years ago. I know they could be off by 20 thou in 92.

    I will absoulutely give you the valve tuning issue if it concerns tighter velocity spread out the barrel. If it affects down range flight you would have to explain that to me.

    AGD

  5. #65
    Join Date
    Dec 2001
    Location
    New Hampshire
    Posts
    943
    Tom,
    Since we are considering that paintball production has evolved to the point where they are much "more round".
    I'm curious to persue the "ball distortion" issue. Wouldn't we just want to do the powder test with a "very" large bore barrel to find out whether or not the paintball distorts to fit the barrel? Or does it bounce around? Or what ever....
    Thanotos

    http://www.factcheck.org

  6. #66
    Join Date
    Mar 2002
    Location
    Hundred Acre Wood
    Posts
    454
    Originally posted by than205
    Tom,
    Since we are considering that paintball production has evolved to the point where they are much "more round".
    I'm curious to persue the "ball distortion" issue. Wouldn't we just want to do the powder test with a "very" large bore barrel to find out whether or not the paintball distorts to fit the barrel? Or does it bounce around? Or what ever....
    It would be interesting to do more testing to see if the results Paladin got are consistent across the board. Based on the results described in AGD's tech tip on barrel sizing, I'm betting that the ball will only expand to a point and that in a very large bore barrel the ball would bounce back and forth.


    From the tech tip:
    If the barrel is too big, the ball ricochets back and forth down the tube. We used to say it looked like Zebra stripes in there. Hence big barrels do NOT create an "air bearing". Barrels that are too small scrape most of the powder off and this creates excessive FRICTION. Tighter barrels that were too long were found to slow the balls down due to this friction. In other words, when you cut these barrels down, velocity went up. Remember the 8-10" acceleration distance, these barrels were 14" long and unported.

  7. #67
    Join Date
    Nov 2001
    Location
    NorCal
    Posts
    4,772
    Tom,
    Did you ever photograph any paintball flights that were fired from a barrel with "paint" in it? I think we have all observed enough evidence to believe that "paint" in the barrel adversely effects accuracy, unless you are trying to curve it around a tree.

  8. #68
    Join Date
    Mar 2002
    Location
    Sacramento, California
    Posts
    158

    Thumbs up

    Originally posted by AGD
    All,

    Now I just want to point out to everyone that Glenn and I are having a very pointed discussion about the subject at hand. We are both presenting facts, opinions and points of view.

    It is refreshing to have such a discussion without it breaking down into name calling and disregarding the others opinions. This is why I love Glenn, he is a no BS guy. Doesn't take it and as importantly doesn't give it.

    This type of discussion represents what Deep Blue is suposed to be all about.

    Thanks Glenn.

    Tom
    Ditto, what Tom just said!! ^ Thank you too. Always a pleasure !!!

    Now, I have a couple more tests that I want to try to duplicate today before I get into the "goggles on" envrinment of Tom's next post.

    Quik Question: Anyone, that is following this thread, have any experience with or a connection to freefall parachuting, aka Skydiving ??
    If so, try following some round and/or not so round objects in feefall sometime. Makes for some very interesting observations regarding the effects of air flow over various objects and their shape orientation to the line of flight. Especially at the transition from acceleration to achieving "terminal velocity" speeds ranging from 150 to 300 FPS. I haven't done any serious jumping in many years and I sure wish I had some video of some of the stuff we did.

  9. #69
    Join Date
    Dec 2001
    Location
    Northern VA
    Posts
    316

    Test #114 Analysis

    Analysis of the test #114 image

    I am calling the first image of the ball outside the barrel "position 1". "Position 0" is inside the barrel.

    Front view spin:
    Position 1: 36.3 degrees
    Position 2: 24.6 degrees
    Position 3: 14.4 degrees
    Position 4: Washed out
    Average spin: 11.0 degrees per strobe flash, spin vector away from camera, "top-spin" from the shooter's point of view.

    Underside view spin:
    Position 1: -3.2 degrees
    Position 2: -22.1 degrees
    Position 3: -36.0 degrees
    Position 4: Washed out
    Average spin: 16.4 degrees per strobe flash, spin vector down, "left-spin" from the shooter's point of view.

    Axial spin is negligible.

    Based on the measured spin directions, the Magnus effect should cause the ball to drift downwards and towards the camera. Here are the position measurements from the test #114 image.


    Front view position of ball:
    Position 0: 29 mm above string (bottom of ball)
    Position 1: 30 mm above string (bottom of ball)
    Position 2: 31 mm above string (bottom of ball)
    Position 3: 32 mm above string (bottom of ball)
    Position 4: Washed out

    Underside view position of ball, "towards the camera" and "in front of the string" is "down" in this portion of the image because of the mirror used to image the underside:
    Position 0: Indeterminate. Ball does not appear to line up with barrel exit.
    Position 1: 9 mm in front of string (front surface of ball)
    Position 2: 9 mm in front of string (front surface of ball)
    Position 3: 7 mm in front of string (front surface of ball)
    Position 4: Washed out

    The ball actually rises slightly and moves away from the camera. The Magnus effect predicts exactly the opposite of what we observe in test #114. There is an anti-Magnus effect that rears its head for smooth spheres, but its effect is generally weaker than the forces predicted by the Magnus effect. Given that test #101 showed that the Magnus effect resulting from a spin of 15 degrees per strobe flash was resulted in too small a deviation to measure with the scanned images provided, and given that the spin in test #114 is comparable in magnitude to the spin in test #101, it is reasonable to conclude (without having to go through all the math) that the deviation imparted by the Magnus effect in test #114 is also too small to measure in the scanned image provided.

    Not only is the deviation imparted by the Magnus effect in test #114 too small to be measured, but the deviation is also predicted to be in the opposite direction of the measured deviation!

    The rise of the ball as measured in the front view image appears linear (to within measurement error), which seems to indicate that it is either a result of not quite perfect alignment of the barrel and string, or an initial upwards "kick" that occurred just as the ball exited the barrel.

    The motion of the ball away from the camera is not quite as linear, but is still reasonably so to within measurement error. The ball may have acquired a slight "kick" away from the camera as it exited the barrel.

    What could cause a ball to be kicked slightly in a direction as it exits the barrel? I suppose slight kick could be imparted to the ball as a result of the differential pressures encountered from a nonuniform air release. I took a look at the smoke*.tif images to see if I could detect any nonuniformity in the way the gas escaped, but I couldn't make any conclusions from those images.

    Based on tests 101 and 114, I think that the Magnus effect is not what causes the shot-to-shot inconsistency seen in modern markers; the spin resulting from a barrel not specifically designed to induce spin (i.e. Flatline, etc.) is not sufficient to deflect the ball from its normal course. The inconsistency may be the result of nonuniform release of air behind the ball as it exits the barrel. In normal gameplay, pockets of air turbulence may be a factor.

    BJJB

  10. #70
    if acuracy is decided by how the liquid in the ball spins or dosen't spin then to get the fill to spin would be to solve a great problem. Mabey making the paint thicker would help the fill to spin when the shell spins. Then the fill would spin with the shell and we could just start rifling the barrels like actual rifles.

  11. #71
    Join Date
    Oct 2000
    Location
    Chicago
    Posts
    5,947
    BJ,

    NICE JOB!! Good analysis. Ok so NOW we are seeing real data that points in a direction, it just seems that the direction is not what everyone expected. So is the way of science.

    So please lets assume for a second that this 114 data generaly represents what's going on with paintballs. (I happen to know it does)

    The next claim is that maybe the ball is kicked sideways right out of the barrel. If this were true you would expect to see it deviate in one direction and that direction would be significantly increased in the second camera trap. If you look at the picture called "3D interpretation" you will see the balls position plotted as relative movement from one pic to another. If you look at the first position in the second trap (flash #6) you will see that the ball trended LEFT from the #5 flash postion but it hits WAY RIGHT at the final target backdrop.

    So if we interpret the data correctly the ball "S" curved in flight. This is not consistent with a ball knocked off course at launch.

    We are getting better lets stick to it.

    AGD

  12. #72
    Join Date
    Jan 2002
    Location
    North of Seattle
    Posts
    176
    Originally posted by AGD


    So please lets assume for a second that this 114 data generaly represents what's going on with paintballs. (I happen to know it does)

    AGD
    So, the conclusion based on BJJB99s nifty numbers work is that spin does not have an effect on accuracy? And... this conclusion is supported by all the other unseen data?

    I don't entirely agree with that statement... but, I cannot disprove it. So I'll assume it's true until it can be proved otherwise.

    So... where does that leave us? Why don't balls fly straight? Obivously there is something happening during the balls flight to cause it to wander. They had the same problem a few hundred years ago with muskets... everyone knows how they solved that problem.

    Have the same accuracy tests been tried with nylon balls? Assuming the solid ball is homogeneous... that should remove any inconsistancies associated with the paintball. From my playing around with shooting nylon balls I found they hooked like crazy. I might have been shooting a bit hot which could contribute to the inaccuracy. Have you seen similar results with nylon balls?

    Why is a paintball shot from a barrel coated with paint from a broken ball so inaccurate? Is there spin? Does the paint coating the ball cause it to hook?

    What was the flash *duration* on the strobes used for your photographs?

    again... more questions than answers. Sorry.

  13. #73
    Join Date
    Dec 2001
    Location
    Northern VA
    Posts
    316
    Originally posted by Redkey

    So, the conclusion based on BJJB99s nifty numbers work is that spin does not have an effect on accuracy? And... this conclusion is supported by all the other unseen data?
    I think the conclusion is more along the lines of "the levels of spin seen in the test 101 and test 114 data are not sufficient in and of themselves to affect accuracy to the magnitude seen in those tests." This is not to say that spin has no effect on accuracy; the Flatline barrel turned on its side clearly shows that spin does affect accuracy. But then the level of spin induced by a Flatline barrel is (I assume) significantly greater than what tests 101 and 114 show.

    Originally posted by Redkey

    Why is a paintball shot from a barrel coated with paint from a broken ball so inaccurate? Is there spin? Does the paint coating the ball cause it to hook?
    I think it may be a combination of spin and a nonuniform coating of fill on the outside of the ball which causes the horrible accuracy one sees when firing a shot from a paint-filled barrel. The spin would likely be greater than that seen in tests 101 and 114 so we'd be closer to the Flatline barrel regime in terms of shot curvature. The nonuniform coating of paint on the ouside of the paintball might result in an effect similar to golf ball dimples, increasing the turbulent flow and thus magnifying the effects of spin even more. Of course, this is all guesswork on my part.

    Originally posted by Redkey

    What was the flash *duration* on the strobes used for your photographs?
    The ball from test 101 appears very round in flight, perhaps only a pixel or two "longer" in the direction of flight. Assuming this is from the flash duration, it amounts to about 1 mm of travel during the flash exposure. At 280 fps, we get 1 mm of travel in about ten microseconds. Thus I'm betting that the strobes are around 10 microseconds in duration (or possibly a bit faster if some of the ball elongation resulted from a failure to recover from possible deformations during firing).

    BJJB

  14. #74
    Join Date
    Dec 2001
    Location
    Northern VA
    Posts
    316
    Originally posted by AGD

    The next claim is that maybe the ball is kicked sideways right out of the barrel. If this were true you would expect to see it deviate in one direction and that direction would be significantly increased in the second camera trap. If you look at the picture called "3D interpretation" you will see the balls position plotted as relative movement from one pic to another. If you look at the first position in the second trap (flash #6) you will see that the ball trended LEFT from the #5 flash postion but it hits WAY RIGHT at the final target backdrop.

    So if we interpret the data correctly the ball "S" curved in flight. This is not consistent with a ball knocked off course at launch.
    What is the vertical and horizontal scale (preferably in inches per pixel) used in the file labeled 114axis.TIF? If my measurements from the in-flight shots of test 114 are reasonably correct, and if I'm applying them correctly to 114axis.TIF, then the "WAY RIGHT" you're talking about amounts to about 2.25 inches of deviation from the point of aim at the final target backdrop.

    A total horixontal in-flight deviation of less than 2.5 inches might be explained away by unstable air, either in the form of density gradients or residual air currents from people having moved through the test area prior to testing.

    BJJB

  15. #75
    Join Date
    Mar 2002
    Location
    Sacramento, California
    Posts
    158
    Originally posted by AGD
    BJ,

    So if we interpret the data correctly the ball "S" curved in flight. This is not consistent with a ball knocked off course at launch.

    We are getting better lets stick to it.

    AGD
    Since the ball seems to have a rotation to the right, it makes sense that it would have "S" curved in flight even if it had gotten a little "kick" to the left as it exited the barrel. Based on the orientation of the seem of the ball to the line of flight, as seen in test #1:01 any muzzle blast could easily move it to the left a bit and then allow the rotation to move it back to the right.

  16. #76
    Join Date
    Oct 2000
    Location
    Chicago
    Posts
    5,947
    BJ,

    YES!! You correctly state that spin is a continuum. At the level we see in 114 it has no substantial effect but with the Flatline or wet barrel the magnus effect overcomes whatever the mysterious force "X" is.

    So lets talk about spin for a minute. Spin is a FORCE or magnus effect or whatever you want to call it. This force has the characteristic that at high spin rates (flatline) it always pushes the ball off the centerline of the barrel. So I am going to make the broad statement here and say that spin has a force vector randomly perpendicular to the line of flight. I think this is reasonable but I am open to debate.

    If your ball spun every shot (again flatline rpm) but the barrel did not induce the spin in one direction like the flatline just in random orientation, you would end up with a target grouping that looks like a doughnut. This is because the spin wants to pull it off the centerline. Ok hold this thought.

    We now know some things about the mysterious force "X". We know it has a magnitude because it can overcome the force of low rpm spin. We also know its magnitude is lower than flatline rpm spin. Now remember the doughnut pattern from spin? We get that because the spin force is not completely random, its always pointing away from the centerline. When we look at standard shot groups only affected by X (clean barrel, low rpm spin) they are randomly distributed, this leads me to interpret this force X as random, meaning it might not affect the ball at all on one shot but throw it off the next. This I think we agree is completely consistent with what the paintball world sees. So if you agree with what I say here, we are looking for force X to have a magnatude that varies randomly from zero to something less than flatline rpm on every shot.

    Importantly we see it overcoming the slower spin WHILE THE BALL IS IN FLIGHT. This is the next point of debate so I will start. If the X force happened in the barrel or say a foot from it, then the spin no matter how small should affect the flight path in a direction consistent with the axis of rotation as it flies down range. To state it another way, if the ball in flight was not being affected by X then spin should be the major factor causing deviation. From another point of view, the force X has to be happening while the ball is in flight because in 114 it pushes the ball in two different directions while its going down range. So in order to argue against this you have to explain how something in the gun or barrel can affect the ball down range as we see in 114. The one thing the barrel can do is impart spin to the ball and that affects it down range but since we have that under control you have to come up with something else. Fire away.

    Redkey,
    Yes we are basing this on unseen data, if when we get to the end, someone wants to see more I will gladly post it. I am quite confident that the brains on this forum, once they comprehend all the factors, will be satisfied with the answers. Like you said they had this same problem with the musket balls its not new stuff

    Yes we did test nylon balls, the shot group is #5 in the 8 box shot group pic. The scale of the in flight positions is proportional to the diameter of the ball. So if the diameter of the ball is .680. It deviates to the left less than one ball width in the 114 flight path.

    I see that I forgot to post the tiff files for you to download hi res versions. I will try and do that asap.

    Glen,

    You have to explain the muzzle blast effect in context of the smoke pictures. It is not apparent to me that there is a pressure jet exiting the barrel right when the ball leaves.

    Damn this is good stuff....


    AGD
    Last edited by AGD; 12-04-2002 at 04:30 AM.

  17. #77
    Join Date
    Oct 2000
    Location
    Chicago
    Posts
    5,947
    I appologize, I thought I had posted the tifs when I first put up the pics but I must have forgotten. Please download the tif and look at the pics, they are much clearer than the lousy jpg's in the thread. Sorry about that, you cant get good data from lousy pics. Here is the link and its in the data thread.

    AGD


    spin data

  18. #78
    Might I propose someone take a look into the varying ballistic coefficients of paintballs as the mysterious "X" force.


    Considering what Mr. Kaye posted above on Force "X", let's look at paintballs. Paintballs have a terrible BC. Might we also say that the BC of every paintball fired is different from the next because of their inconsistant nature. So far, unless handloaded, we cant control the orientation of every round that enters and exits the gun. Might not the Force "X" be the minute diffences in the BC of each and every ball fired? And might not the fact that even Nylon paint has bad accuracy be a result of the poor BC of paintballs in general? If you have a really good Paint/Barrel match flyers still occur. Possibly due to the fact that the tip/width/tail of every projectile fired is minutely different from the last or the next? The Flatline would override this as it imparts the hard spin along a verticle line that overcomes the BC and changes trajectory? Fill on the ball would do so also as it induced a spin?

    A few sites on BC that might make more sense....

    http://www.chuckhawks.com/bc.htm

    From Mr. Hawks...
    "Ballistic Coefficient (BC) is basically a measure of how streamlined a bullet is; that is, how well it cuts through the air. Mathematically, it is the ratio of a bullet's sectional density to its coefficient of form. Ballistic Coefficient is essentially a measure of air drag. The higher the number the less drag, and the more efficiently the bullet cuts through the air. So for purposes of flying through the air efficiently, the bigger the BC number the better.

    BC is what determines trajectory and wind drift, other factors (velocity among them) being equal. BC changes with the shape of the bullet and the speed at which the bullet is traveling, while sectional density does not. Spitzer, which means pointed, is a more efficient shape than a round nose or a flat point. At the other end of the bullet, a boat tail (or tapered heel) reduces drag compared to a flat base. Both increase the BC of a bullet."



    http://www.aeroballisticsonline.com/ballistics/bc1.html


    http://www.uslink.net/~tom1/calcbc/calcbc.htm



    This may be completely off the wall and entirely wrong, but it's the best a feeble mind like mine can come up with. And, I do realize that paintballs and bullets may be an "Apples and Oranges" comparison, but BC is what describes the change in the shape of projectiles over the years in firearms. I think someone did mention that the firearms world solved this problem long ago, or something to that effect.


    Respecfully,

    Pstan
    Last edited by Pstan; 12-05-2002 at 05:32 PM.

  19. #79
    Join Date
    Oct 2000
    Location
    Chicago
    Posts
    5,947
    Pstan,

    Good comeback! Yes BC can affect things in flight. Problem is that all spheres have the same drag coefficient with is something like .7 or .8. The nylon balls being perfectly round to better than 1000th of and inch and missing the seam should fly straight and true but they don't. In fact the shot group for nylon balls is hard to differentiate from regular paint.

    This would seem to sugest that the seams, drag coefficient differences and small size variations are not a substantial portion of our X force.

    AGD

  20. #80
    Join Date
    Oct 2000
    Location
    North Plains, OR
    Posts
    4,956
    I could sit here and rattle off all of my Fluid Mechanics knowledge, but this explains things pretty well.

    http://www.princeton.edu/~asmits/Bicycle_web/blunt.html

    Pstan- The difference between the bullet and the paintball is that one is a streamlined body and the other is considered a bluff body, respectively.

    Tom- Not for all spheres. Take a look at the graph. It's a function of the Nr (Reynolds Number), but for all paintballs shooting through air you can consider it the same.

    I haven't been reading this thread, but when I find the time I'll look it over and see if I can add some input. It seems like an entertaining conversation thus far.

  21. #81
    Join Date
    Oct 2000
    Location
    Chicago
    Posts
    5,947
    Black,

    Yes of course you are right about the DC being different at different R. You all should look at the site Black pointed out. Very informative especially figure #3....


    AGD

    The trip wire thing is cool too, I will have to look at that more.

  22. #82
    Join Date
    Dec 2002
    Location
    Hinsdale IL
    Posts
    9
    Larry De La Brandais turned me towards this very interesting discussion thread, and I thank him for it. You may not. Although I haven't done much on the paintball physics subject for awhile, I still remember some of the stuff I put together, and certainly still have an interest in it. (As a reminder: http://home.attbi.com/~dyrgcmn/pball/pballphys1.html) After seriously wading through the many very interesting arguments and points raised on this thread, I thought I might as well join the confusion. The variety of expertise and depth of thought on dynamics issues is great to see. Warning, I am coming in very late in the discussion and I am trying to summarize a whole lot of stuff, so this will be long.

    On spin: The Magnus effect or force is small. Gravity and drag are the two big forces driving any ball dynamics. The problem is that even if the Magnus force is secondary it is sufficient to shift a trajectory enough so that you miss your intended target even when you are aiming with the utmost determination. The Magnus effect should not be considered along just any one particular axis. The effect is general and is based on the dynamics of air moving over a spinning object or ball in our case. Mathematically, it is defined by whatis known as the vector cross product between the spin and the motion of the fluid around the ball. In our case the ball rather than the fluid is moving, but that is simply a frame of reference difference. What it does mean is that any deflection from a true non spinning trajectory depends on the spin direction, the spin velocity (rpm), the ball velocity, the ball direction, and the relationship between the spin axis and the ball direction. The latter relationship dictates that the best we can do is talk about accuracy and spin in terms of a two dimensional splat pattern that will result around a central point unless we can absolutely control the spin axis direction. As I see it, the problem is not whether spin creates deflection, that is a given based on the scientific and practical literature. The real problem is determining how much spin will create how much deflection. (Note that I am not saying here that spin is the only cause of deflection.) On my physics page, I used some of the available literature to try and get some idea of what spin velocities are needed. Of course, none of the data available was for paintballs or for objects as small as a paintball. As someone who for most of his career was an experimentalist, I have a healthy skepticism for theory, especially when it has to stand alone as in my own calculations. I certainly have to agree with pstan's arguments along these lines. Actually seeing and being able to digest some of Tom Kaye's data is great, and I thank him for sharing it.

    Still on spin: Keep in mind that the relationship between spin velocity and ball directional speed is non linear. This is embodied in the lift coefficient curves. Take a look at the data on my site. The lift coefficient which is necessary to calculate the effect of spin, is not a constant. That means the force will not be a constant. Assuming that it is a constant will likely produce very misleading answers to understanding a true spin related trajectory, and definitely will not produce some of the weirder paintball trajectories (those that dip and rise). Several people have already discussed the lift coefficient data that I used in my calculations, where the lift coefficient is plotted as a function of V/U (rotational velocity/directional velocity). The importance of the V/U ratio seems to have been somewhat overlooked in this discussion. If you look at the lift coefficient curves I used, at low V/U there is not much affect (force) on the ball. The reason is that when the ball leaves the barrel, the linear ball velocity is very high relative to the spin velocity. But the inevitable drag force does a great job of rapidly slowing the paintball. Assuming the spin rate does not change (not necessarily true), as the ball proceeds downrange, the forward velocity drops and so the V/U ratio increases. According to the lift coefficient data, the ball would first experience a small negative force and then as it continues to slow a much larger increasingly positive "lift" force. The result is that we will see deflections from a non spinning ball trajectory that are going to be nonlinear, and definitely increase at a faster rate as the ball gets further and further and slower and slower downrange. When I was first doing my calculations, I had a lot of problems understanding this large deflection near the far end of the trajectory until I actually dumped the V/U data in my calculations as a function of distance and watched it rise.

    On non spinning effects that affect trajectories: Real paintballs almost always have seams. If you look into the literature of sports ball dynamics and even wind tunnel tests you find that seams or dimples play a very important role in ball trajectories, whether the ball is spinning or NOT. Any asymmetrical imperfection, whether it is a seam or a dimple, can create a differential pressure/airflow around the ball; this will affect its trajectory from an ideal absolutely smooth sphere. If I remember correctly, the magnitude of this effect can be on the same order as the Magnus force for a spinning ball. Although I mentioned this problem on my physics page, I stayed away from trying to deal with it because of its nightmarish complexity and the fear that I was already running the risk of losing readers at a high rate. But the upshot is that inaccuracy in a paintball can occur without spin. As mentioned by at least one other contributor here, there will also be a change in drag with respect to the seam, since the 2D face presented to the direction of motion will be larger. However, using nylon balls with smooth surfaces should minimize this "slider" type effect.

    Liquid and air bubble motion in the ball issues: I did deal with this issue on my site, but I would not say it is thorough. This issue is particularly important from several standpoints. The fill may help slow a spinning ball down due to internal frictional forces. That would actually reduce the Magnus effect. Second, if the fill is not uniform from ball to ball that will affect the paintball accuracy even if the pressure driving the ball is spot on for every shot.

    Now for my take on some of the generous data offered by AGD, and discussion by others:

    I am having a problem understanding the 114 data particularly the 114side image. Tom indicates that the distance was 60 ft for the data shown, but based on the distances shown and the distances between strobe shots mentioned in another photo, the distances come out closer to 88 inches. What am I missing?

    Bjjb99's data analysis is great. Thanks for helping us all byspending the time doing it. Bjjb99's analyses of the 114 photos prompted me to look a lot closer at them, because I found the pictures confusing. I think (hope) I understand now. The ball consists of three stripes. Two run parallel to each other around the ball; a third is at right angles to these and runs through one hemisphere only. Looking very closely at the photos and grapping a Christmas ornament from the convenient Christmas tree as a prop, I think I now see what is going on in 114. The axis of spin is very approximately 30 degrees from the horizontal plane and 30 degrees from the vertical plane extending along the direction of motion. Looking at the front view and drawing the spin axis through the ball it would be extending back into the page from the upper right side of the ball at about 30 deg from the plane of the page, and would be coming out of the page extending to the bottom right of the ball. If this is true, then the bulk of the Magnus effect would indeed be manifested in a vertical direction from the true trajectory with only a lower deflection in the lateral or horizontal deflection. The problem is the ball will also be changing position due to gravity.

    I also looked at the positions of the balls for the set closest to the muzzle, but only in the view from the under side. I measured the distance from the "laser line" to the bottom of the ball. At the high magnifications I was using to measure the tif images, I could see each pixel. This created a problem because I was not always sure which pixel to measure to. For the three positions I found 4.50 mm, 4.21 mm and 4.15 mm respectively. But the problem is that each pixel was approximately 0.35 mm wide. Hence the middle ball position also was measured as 4.56 mm. Since it appears the error in measurement is just about the same as the differences I see, it is not reasonable to say there is any change. To be quite honest, I am not sure there should be any:

    Bjjb99 figured about 20 revs/s or about 1200 rpm. That sounds good to me. When I looked at the data a little less critically then him, I reasoned the spin was between 1000 and 2000 rpm. Now how about the magnitude of the deflection expected form the Magnus effect. All I have to rely on is my handy dandy trajectory calculator (but correct?). What I find when I use: 1200 rpm spin, 280 ft/s linear velocity, 0 gun angle, 90 deg spin axis (90 degrees to ball direction vertical to maximize side deflection), is a predicted deflection of 0.000301 feet or 0.0036 inches at 3.0 feet downrange. Now why choose the side deflection of the Magnus effect when I have already agreed with Bj that the major deflection due to spin will be in the vertical direction? Because the effect of gravity on the ball's position produces a 0.0020 drop in the vertical position of the ball at the same distance, roughly an order of magnitude greater than the Magnus effect. The reason for this is as I have already said the Magnus force is particularly ineffective when the downrange velocity is high (low V/U), especially this close to the muzzle. Just for comparison, the same calculation shows that when the ball hits the ground about 116 ft from the muzzle, the side deflection is approximately 6". This is the long winded explanation that basically comes to the same conclusion as Bjjb99. You can't see any change due to spin. But the reason is because of the regime the measurements are being made.

    One other thing I find curious about Bj's data is the change in the angle between successive points downrange. At first glance, it appears that the spin rate is decreasing rapidly. However, it is clear that in the later strobe shots the ball is still spinning. One possible answer to this dilemma may be parallax between the camera and ball, or it may be due to camera lens distortion or both. Another is that since the spin axis is at a strange angle relative to the camera, that there it is simply a problem of representing a 3D operation in a 2D image.

    Could the deflection data that is plotted in the 114axis image be predicted from the known information? Well first I am having trouble understanding the scales in the figures. Second, it could not be predicted by me. To describe completely the direction of a spin axis in 3 dimensions, two angles are necessary. My calculator only considers one angle. This effectively means that one spin axis angle is locked such that I can only determine the maximum possible deflection envelope. This cannot be easily changed without rewriting a fair portion of the programs.

    So where does all this discussion put us, or rather, just me:
    First, the effect of spin should not be looked for close to the muzzle of the gun. Get downrange to really see the effects remember CL versus V/U. However, you do need the spin velocity near the muzzle to find what the initial spin velocity is.

    The data from Tom says several important things to me, and generates several questions:

    Paintballs do spin. Did all shots show the balls spinning or were there a large fraction that did not spin at all? I guess I would expect that all balls would have some spin but with a low average spin velocity.under 5K.

    Some or all of the spin on the ball is preserved in the first two and a half feet from the muzzle and is not damped out. The ball holds a spin even further downrange (depending on where the last three strobe shots were taken.)

    For me, the shot pattern of the nylon spheres is one of the most interesting sets of data. Even the most perfect balls show a shot pattern, but we still don't know why. It could be spin effects, but that is not at all certain. It should not be due to seam or dimple effects since from the discussion the balls are nearly perfectly round. Several questions come up as I contemplate the data; please don't take them as a harangue. Are air currents possible? Were there any attempts to look for trends in shot pattern versus shot number? What was the distance of the target from the muzzle and was the target center perfectly collinear with the barrel axis? Also what is the scaling? Are the circles accurate shadow representations of a paintball diameter? The reason I ask the latter three questions is because this gives me a way to possibly cross check the accuracy of my trajectory calculator under definitely controlled conditions.

    Odds and ends

    Someone remarked they don't understand why the last point does not end up at ground elevation in my trajectory calculator, but is usually slightly higher. The reason is that the calculation is stopped when the next point produces a negative elevation. I considered it too much work to have the calculator find the exact zero point, especially when I have no idea how really close the calculations approximate reality

    Thurman indicated that the speed of the paintball from a flatline barrel should be no more than 1500 rpm. I find this low considering the speed of the ball. On my site, I calculated that assuming a ball "rolled" down a barrel at 280 fps, it would end up with a spin of over 94K rpm.

    Enough for now. Thanks.
    Gary Dyrkacz, aka [5x5]

  23. #83
    Somewhere along the line (in this thread) I thought I saw it mentioned that just about every ball fired through a paintgun experiences the magnus effect to some varying degree. I think it was further stated that the problem is that there isn't anyway to control which direction the effect causes the ball to travel and is essentially random. A description stated that due to this occurrence that the shot pattern of paintguns would look like a donut - or, if a straight line was run through the exact center of the bore to the impact point, the Magnus effect would cause the ball impacts to hit at varying distances from the desired point in a complete ring around that actual point of aim.

    So if this is the effect of uncontrolled Magnus, why not try to control it? The flatline does (in an unintended fashion - they aren't trying to "control" anything - just to induce the effect to utilize the result) but is not "adjustable" and is severe. The Z-Body for the 'mag was designed to produce a similar effect as the flatline - but differs in how it accomplishes this. The Zbody uses a "friction plate" that is adjustable from zero to maximum (I guess where it would crush a ball) and I think most people turn it up to a high degree attempting to reproduce the Tippmann result.

    What if a similar system was used to put just the slightest amount on magnus on every ball fired from a gun? Instead of going to far and putting a "major" amount on - why not just apply enough to be certain that effect was always consistent? That should theoretically reduce the group size which is another way of saying "increase accuracy."

    This might not be the holy grail - but I'm wondering if the line of thought is reasonable. I guess my thought is, "if paintball guns are inherently inaccurate due to variances (inconsistencies) in the shape/size/form of the projectile, and those variances in the projectile are unchangeable and must be accepted - why not try to control how the projectiles behave by forcing their behavior to be more consistent?" This is based on the huge assumption that the varying magnus I described above is caused by the imperfections in the balls themselves.

    -Calvin
    From a poster at PB Nation:

    ""Jim, back to your cave. Bob Long is on the batphone..."

    MY FEEDBACK

  24. #84
    Join Date
    Feb 2002
    Location
    Woodstock, GA
    Posts
    606

    from precision shooting:

    When fired by a human, all shot groups are a donut. The shooter
    with the smallest donut wins.

    Observationally, paint behaves the same, this is due to the
    natural mechanics of the body.

    Question: AGD/Tom.. or Glenn,

    "Does a paintball marker, when held in a fixed mechanism, exhibit
    this same tendency, or is there a pattern grouping that is somewhat
    uniform over a reasonably large sample set?"

    -rob


    TAG Factory

    Powered by:
    http://www.TheAngelGuy.com
    http://www.RegulatorsOutpost.com (Largest Paintball Field in Northern GA)

  25. #85
    Join Date
    Dec 2002
    Location
    Hinsdale IL
    Posts
    9
    I believe that the Magnus effect would only produce a well defined donut target pattern if the spin velocities were locked into a small non zero range. I also suspect at least one of the angles defining the spin axis would have to be constant as well. If we set up a coordinate system centered on the ball with x axis being the direction of ball motion, y being in and out of the plane of the page, and z being the vertical axis, then the angle that must be locked is that in the xz plane.

    With respect to improving the accuracy by placing a small controlled spin on the ball,I am not sure. You still need only a little friction, or a little asymmetric leakage of gas around the ball to alter its spin axis. If the alteration happened to be along the same axis as the controlled spin the result would be to either speed up the spin, or reduce it. This still leads to dispersion in the trajectory, and a broader shot pattern. Despite this, the controlled spin has momentum, and I think a ball with a controlled spin might therefore resist a change in its spin axis to some extent. On the other hand, many times I have read complaints from people that the accuracy of a flatline barrel is worse than a good straight barrel. At face value, this field observation would tend to negate the hypothesis. However, the preception of inaccuracy could be related to not holding the gun perfectly vertical as has been discussed here. Another possibility is that people can just notice a wayward spin much easier because of the extra distance.

  26. #86
    Join Date
    Jan 2002
    Location
    North of Seattle
    Posts
    176

    Hey Gary

    I stumbled across your website a couple months ago... was meaning to shoot you an email to say hi and comment on your page.. never got around to it though.

    so...

    Hi

    I like your web page. I had a good laugh at your paintball weight studies... I wonder how many people know the difference between mean and median, not to mention the 95% (three sigma?) confidence level. Any plans for future testing?

    Later
    Geo

  27. #87
    I dont know much about the magnus effect, but i do know that a knuckleball goes crazy in flight. Assuming that a nylon ball was not spinning at all, would it not act in the same way as a knuckleball? How much did the nylon ball spin while in flight compared to a regular paintball. If the spin is comparable then i may be way off but i would think that a nylon ball with no seam would not have as much spin as a paintball with dimples/flatspots/seams.

  28. #88
    Join Date
    Dec 2002
    Location
    Hinsdale IL
    Posts
    9
    Redkey,
    I needed the data on paintball dimensions and weights for my calculations. Once I started measuring and really saw how non round a paintball could be, I realized that a portion of our problems with accuracy must be tied to those odd shapes and seams. It was also one of the few times we had more than one type of paintball around. You obviously can image how boring it was sitting there with a micrometer trying to measure the diameters. Another reason for actually putting the study on the web was that I realized that not a lot of people have balances with 10 microgram accuracy available to them. No plans to do any more studies.

  29. #89
    Join Date
    Jan 2002
    Location
    North of Seattle
    Posts
    176

    Ball Weights

    Here are some measurements I made. Min of 20 balls were weighed for each paint type. Yes, I know, 0.1 mg resolution is overkill. I also have some seam and pole crush load data around here somewhere.


  30. #90
    Join Date
    Nov 2001
    Location
    NorCal
    Posts
    4,772
    [5x5]. Glad to see you are still interested in things like this. Interesting discussion, isn't it? Hopefully Tom will release more data and you can contribute more. Thanks for checking us out.

    Larry (Hitech) De La Briandais

Posting Permissions

  • You may not post new threads
  • You may not post replies
  • You may not post attachments
  • You may not edit your posts
  •