1. Attention: We have put together a thread with tips and a tutorial video to help with using the new software. Please take a moment to check out the thread here: Trapshooters.com Tutorial & Help Video.
    Dismiss Notice

Pic of air flow and force on a moving pellet

Discussion in 'Uncategorized Threads' started by Dr A C Jones, Aug 29, 2008.

Thread Status:
Not open for further replies.
  1. Dr A C Jones

    Dr A C Jones Member

    Joined:
    Jan 29, 1998
    Messages:
    367
    This is a little something I have been working with on and off over the last year or so. The CFD (Computational Fluid Dynamics) tool is a standard commercial tool used for all sorts of things like planes, cars, pipe flow, air conditioners, anything with a moving gas or liquid. Modeling a (yup, just the one) pellet is a little tricky because the code is not tuned for this. The results are very sensitive to how the air flow breaks away from the pellet. An airplane wing is easy by comparison because the flow is smooth. As the picture shows, in the case of the pellet there is a lot of turbulence behind the pellet and complex flow features over the surface - even more so if the speed increases. Getting these right is crucial for accurate predictions of drag.

    This is a model with the relative speed of the pellet to the air of Mach 0.9 which is about 312m/s. Note that the air is moving and the pellet is stationary, that's just the way the code works. Note also that even though the pellet is subsonic there are flow features that are up to 400m/s. Mostly pellets are "transonic" which means they are somewhere between subsonic and full-on supersonic, this is what this picture shows. This is why it is no big deal about pellets passing the sound barrier. The air flow at Mach 0.95 is not much difference to that at Mach 1.05.

    This first picture shows the air speed as it passes over the pellet. Differences in air speed are closely related to pressure so note the short distance of these effects. Once pellets are a few mm apart they are effectively aerodynamically independent. (The pellet diameter in this case is 2.4mm).


    [​IMG]


    This next one shows some more details of the air flow. The arrows are coloured to show the velocity (in m/s this time) and the surface of the pellet is coloured to show the surface shear stress, i.e. force on the surface of the pellet from the moving air.


    [​IMG]


    That's science!!

    Andrew.
     
  2. JBrooks

    JBrooks TS Member

    Joined:
    Nov 6, 2006
    Messages:
    3,707
    Musket Ball.

    Pellets have friends. Think of them as kind of a Nascar field.
     
  3. rifle guy

    rifle guy TS Member

    Joined:
    Jan 25, 2007
    Messages:
    103
    Could you add a dimple or a flat spot to that and see their effects on the flight? very neat stuff
     
  4. Juno

    Juno TS Member

    Joined:
    Jan 13, 2008
    Messages:
    635
    Pretty colors!
     
  5. joe kuhn

    joe kuhn Furry Lives Matter TS Supporters

    Joined:
    Jan 29, 1998
    Messages:
    7,112
    Location:
    Naperville, IL
    Doc - invent some pellet shape that flows through the air better so we don't have to push it so hard to start with. Less recoil, more effective force downrange. Is that what you're after? Why, Doc, why? Just asking, not criticising.

    Joe
     
  6. oz

    oz Active Member

    Joined:
    Jan 29, 1998
    Messages:
    1,817
    excellent thread. however pellets are not perfectly round. I think of them as a knuckleball. oz
     
  7. omahasportingsupply

    omahasportingsupply TS Member

    Joined:
    Jan 7, 2007
    Messages:
    1,138
    Good information. In your research have you found the evidence to prove or disprove the theory that mixing shot sizes gives you an inconsistent pattern? If you could disprove that theory, many of the posters on TS.com could mix their 7 1/2s and 8s in their loads and would quit asking which size to use. LOL Just kidding as it will never happen. IMHO If you post it they will come. Omaha
     
  8. recurvyarcher

    recurvyarcher Well-Known Member

    Joined:
    Apr 26, 2006
    Messages:
    6,450
    Dr. Jones, computational fluid dynamics is a discipline that has been around for quite some time, not a commercial program...LOL!!! I'm sure that you are referring to a software program, though.

    Since you have this tool at your fingertips, what would be extremely interesting is if you could model two types of shot pellet; one that is coated with graphite -vs- one that is uncoated. This would provide some useful information for all of us. Do you have the time and inclination to do that and post the results? It would be much appreciated.
     
  9. phirel

    phirel TS Member

    Joined:
    Jan 29, 1998
    Messages:
    9,556
    Great illustrations and I want to make things much more complicated. We are primarily interested in what is going on in a cloud of shot. Trying to visualize many shot interacting is beyond my abilities.

    Pat Ireland
     
  10. Dr A C Jones

    Dr A C Jones Member

    Joined:
    Jan 29, 1998
    Messages:
    367
    First, let me point out that getting the models to the current state has taken about 6-12 months on and off. Although it looks like "just one measly sphere" the model takes about three days to run on a workstation.

    In some respects the model is quite limited. All it does is work out the average flow, turbulence levels, temperatures etc. In reality, the turbulence is constantly changing with eddies constantly being shed from the back of the sphere. Modeling this is 1000s of times more numerically intensive. It might even be impossible.

    To see why this is important, take a quick look at the link above about the aerodynamics of paintballs. Just read the headings in blue and you will appreciate just how complicated this apparently simple subject can be.

    However, in its current state the model gives drag figures that are consistent with Shotshell Ballistics for Windows (about 20%) which was my first objective.

    In response to the questions:

    skeet2day - 7.5 vs 8s? It should be obvious that altering the size of the sphere by a small amount will not make a profound difference. All that will happen is the overall effect of deceleration will alter slightly. This is a combination of frontal area, pellet mass, and how the drag coefficient changes with pellet size (basically how the air flows over and breaks away from the pellet). It is easy albeit time consuming to rerun the model with a smaller pellet. But we know from experience of downrange velocities that the difference between 8s and 7.5s is small.

    JBrooks, omaha, Pat - broadly effects of pellets interacting: Look at the top picture. The pellet is 2.4mm across (0.095-inches). The extent of the disruption of the air flow is ~3-pellets widths out the side, 2 to the front and 5 to the rear. Once pellets are further apart than this they are effectively independent. There isn't any interaction.

    recurvy - pellet coating: Probably no effect. The CFD code assumes a stationary boundary layer of gas on the object surface. At a molecular level this is meant to represent the molecules that stick to the surface. I know what you are thinking now, "but what if these molecules could slip better rather than stick". It will not make any difference. The coefficient of friction of air is much less that Teflon anyway.

    *** I just had a thought though: Since the air is getting pretty compressed does its effective viscosity increase? I don't know. Maybe someone knows if supersonic planes / missiles get a coating of something.

    *** Whatever, the CFD code cannot model this effect.


    Rifleguy, oz - broadly misshaped pellets: I have a model of a pellet with a flat on it. Let me just say that things "get interesting". The problem with "interesting" is whether or not it is correct. Normally when you use CFD code you have some real tests to correlate with. I haven't got this luxury so when I get an odd result that still delivers plausible results I need to be very cautious before making my Eureka shout and running down the street in my birthday suit. I am also very aware of how little the CFD code is modeling. See the paintball link above and this link:

    http://www.iihr.uiowa.edu/projects/low_reynolds/

    with this pic at the bottom:


    [​IMG]


    flicher - effect of velocity: Probably none. Again if you look at the first pics, what can happen if the velocity changes a bit? The drag goes up or down but not much changes. The pics above are for Mach 0.9. I have done Mach 1.5 as well. The pressure, temperature, velocity profiles get greater and more defined, the overall deceleration increases but there's no catastrophic change. Of course there is a possibility of some effect not modeled coming into play (see the paint ball link again). My practical advice would be to drive pellets as fast as possible so you can use smaller pellets all subject to recoil not putting you off. Unfortunately your trap rules stop you going down this route.

    Joe - a better pellet shape? That'll be a bullet then. Incidentally, a bullet complete with rifling marks is easier to model than a pellet.

    Andrew.
     
  11. recurvyarcher

    recurvyarcher Well-Known Member

    Joined:
    Apr 26, 2006
    Messages:
    6,450
    "I need to be very cautious before making my Eureka shout and running down the street in my birthday suit"......LOL!!!! That was funny.

    Thanks for the very interesting post. It's quite fascinating.
     
  12. Hauxfan

    Hauxfan Well-Known Member

    Joined:
    Jan 29, 1998
    Messages:
    5,078
    Thanks. That was really interesting and very nice of you to post it here for everyone to see.

    Hauxfan!
     
  13. hmb

    hmb Well-Known Member

    Joined:
    Jan 29, 1998
    Messages:
    9,417
    Andrew,

    When the pellets pass through the sound barrier I believe they encounter turbulence which disrupts the pattern of the pellets. Am I right or wrong about this concept? HMB
     
  14. Dr A C Jones

    Dr A C Jones Member

    Joined:
    Jan 29, 1998
    Messages:
    367
    HMB, Depends what you mean by turbulence. There is turbulence all the time. The wake behind the pellet is turbulent, the air being pushed around is turbulent. The pellet is having a pretty buffeting time whatever speed it is traveling at. I'm not sure what effects supersonic has on a tiny round sphere. When I run the model at Mach 1.5 I get higher and more defined pressure and velocity contours. This is a trait of so called supersonic effects. What I don't get is the huge pressure waves you see on aircraft. It could be because the model is wrong (CFD models tend to be diffusive. This means they tend to smear sharp effects) or simply down to scale. A plane is of course millions of times larger than a pellet. Maybe if I ran the pellet at Mach 1000000 I would get the shock wave. Even a bullet is ~150 times+ heavier than a pellet so when traveling at Mach 1.5 it is having to move a lot more air.

    Bear in mind that a pellet is not at all like a bullet. A bullet flies at a slight angle (due to spin and drag) and has rifling interacting with the air. As the shock wave moves up/down/around the bullet it will have different effects and may nudge the bullet one way or another that affects its trajectory thereafter. With a single bullet your are wanting the absolute best repeatability with every projectile.

    A pellet and a pattern are quite different. The pellet is pretty much symmetric. As the pressure waves move around the pellet they do so equally so there is no effect to push the pellet off course. Even if there was an effect, all it does is add to the randomness of the pattern. Random is random so a little bit more mixing up doesn't make any difference. The only downside is that the additional mixing will (on average) increase the pattern spread. How much? Probably tiny but that's a guess part supported by the fact that patterns are not affected by other things very much.

    Andrew.
     
  15. joe kuhn

    joe kuhn Furry Lives Matter TS Supporters

    Joined:
    Jan 29, 1998
    Messages:
    7,112
    Location:
    Naperville, IL
    Doc - Hope you saw the video of the tank shooting ball bearings. The dynamics was pretty interesting with some pellets moving away from the center, some falling back and away, and other scenarios. Overall the shot cloud changes shape pretty dramatically. It would be neet to draw the circumference of the cloud and see it change as everything moves down range. And then you've got a 3D version to think about.

    Good to hear from you.

    Joe
     
  16. JerryP

    JerryP Active Member

    Joined:
    Jan 29, 1998
    Messages:
    1,753
  17. Tron

    Tron Supporting Vendor Supporting Vendor

    Joined:
    Jan 29, 1998
    Messages:
    6,653
    I sure hope that Neil doesn't get wind of this thread. He'll be on you like a numerator on a denominator.

    Tron
     
  18. Dr A C Jones

    Dr A C Jones Member

    Joined:
    Jan 29, 1998
    Messages:
    367
    Joe K, It's a little known fact that before the best tank commanders take delivery of their barrels they send them off to a well known barrel guru for "tweaking". If you look closely at the tank video you'll see the results . . . .

    Andrew.
     
  19. HSLDS

    HSLDS Well-Known Member

    Joined:
    Nov 2, 2006
    Messages:
    6,335
    Location:
    S-E PA
    Dear Dr. Jones,

    Wow - I appreciate the work...

    I know this is asking a lot, but have you thought of extrapolating this for a shot stream, or even for a wad filled with shot?

    I do appreciate the work you've put into in this so far, but the next logical question would ask how this impacts upon a group of pellets traveling together in a shot stream (both generally for the stream and on the interaction of the individual pellets), and how this impacts upon patterns, wad choice, etc.

    I guess you could even apply it to choke selection/design as well.

    Food for thought, and thanks again for sharing this.

    David D
     
  20. silverarrow

    silverarrow TS Member

    Joined:
    Jul 1, 2008
    Messages:
    6
    Doc

    What program did you use for the CFD work. I am a senior engineering student and this stuff is awesome to me. In my fluids class we didn't get to do a lot with CFD and wish we had done more. Great simulations I have worked some with loading simulations with gears and such I know how long coming up with these numbers take. So nice work the results are fascinating.

    -Clem
     
Thread Status:
Not open for further replies.
Search tags for this page
images of moving air flow
,
pellet tron forum
,

picture of air pellet sizer

,
shotgun pellet speed