Dismiss Notice
Join Physics Forums Today!
The friendliest, high quality science and math community on the planet! Everyone who loves science is here!

Physics of a karate punch

  1. Jun 9, 2007 #1

    I was wondering i anyone could help me. I am trying to measure the speed of a karate punch in order to calculate the force of the strike.

    I have read several papers on the subject but each of these use a high speed video camera and software such a videopoint to analyse the data.

    Does anyone know of another way to measure the acceleration? I only have a standard camera at 30 frames per second?


  2. jcsd
  3. Jun 9, 2007 #2
    Measure the total impulse, by striking a heavy bag and seeing how high it swings.
  4. Jun 9, 2007 #3


    User Avatar
    Gold Member

    Welcome to PF, Steve.
    If you or one of your friends knows a bit about electronics, it should be quite simple to build a chronograph. Use a couple of strips of aluminum foil a set distance apart as your triggers and punch through them. When you break the first one, the timer starts. The second one stops it.
    I don't know any electronics myself, but I'm sure that someone here can supply you with a circuit diagram.
  5. Jun 9, 2007 #4
    This is only assuming that it would be a perfectly elastic collision. This would hard to assume.

    Danger has a good idea. You could also see if the physics department at your school has some photogates. You set them up near each other, measure the distance, and it will give you the velocity of your fist.
  6. Jun 9, 2007 #5
    you couldn't calculate the force that punch anyway if even if you did know the velocity, how much weight are you putting behind? you could use a strain gauge though.
  7. Jun 9, 2007 #6
    Thank you for the replies so far.
    I am trying to recreate a study which has been done many times before - calculating the force required to break a piece of wood/the maximum number of boards which could be broken. I have the data for the boards and I also have the data for the mass. I wanted to recreate the study using my own data however as I do not have access to high speed digital video I cannot use the software based calculations I have seen before. Another method I could use would be to use a camera with a strobe and an open shutter. I was wondering if anyone had used any other methods - just to measure the velocity.


  8. Jun 9, 2007 #7
    Why would you want to assume that? I suggested a method to measure the total momentum transferred by the punch. It's a basic physics textbook problem to use that same apparatus for calculating the velocity, energy or momentum of a bullet before it embeds itself (completely inelastically) in such an apparatus; the only assumption is that the punch/collision is brief compared to the period of the swing.

    A camera, ticker tape or photointerrupter circuit might measure the quickness of the punch, but as far as the strength is concerned, it seems like the total impulse would be of more interest than the peak force (which may likely be obtained only when striking a concrete wall in vain). You could use impulse to calculate how the opponent would be thrown backward, whereas basic force data isn't likely to give accurate predictions of, say, what damage will be done.
  9. Jun 9, 2007 #8
    Why not just sit increasing weights on top of a board? (As long as it's only a few cm from the ground, I don't think a local gym would object to one of its customers doing this experiment).
  10. Jun 9, 2007 #9
    weights apply less pressure than a fist.
  11. Jun 10, 2007 #10
    I still do not understand how a ballistics pendulum is going to help this experiment. If we look at the equation in discussion: [tex]M_f*V_f + M_p*V_p = M_f*V_f ' + M_p*V_p '[/tex], we see that we still need to know the final momentum of yourself. This seems like it would be hard, if not impossible to calculate without use of a high-speed camera.

    I mentioned the collision being completely elastic because then we could measure the velocity of the pendulum without regards to the arm, because I was thinking that the arm's velocity would be at zero. I made a mistake, and this is not the case. Definitely an oversight by me.

    Also, if we use the impulse-momentum equations, [tex]F\Delta T = M\Delta V[/tex], I think we would become much closer in finding the strength of a punch that would break a board, which is why I thought the velocity of the fist to be important. I'm not sure, but maybe the [tex]\Delta T[/tex] part of the equation would be the time the fist takes to go the distance between the wood's at rest state and the wood's position right before snapping? I think the impulse of the punch is much more important than the force being applied.

    P.S. I'm only a high school physics student, so this is only guessing from what I've learned this year.
  12. Jun 10, 2007 #11
    I think ur making it too complex.
    hold the camera still and record the punch.
    (assuming the arm moves with a constant speed),
    on computer screen view ur two consecutive images and measure the distance between the arm's positions on those two
    then multiply that distance by 30 to get the displacement in one second
  13. Jun 10, 2007 #12
    Only if we're trying to calculated the "transferred momentum" by measuring all the other initial and final momenta, which misses the point of a pendulum; the correct equation (energy conservation) relates the maximum height of the pendulum to its initial velocity (transferred momentum). I'm not suggesting this will help find the peak force with any accuracy, only that transferred momentum may be a more appropriate measure of a punch.
  14. Jun 10, 2007 #13
    I see what you're saying, I think. But I'm still confused about how we can use the conservation of energy equation if the collision is not elastic (maybe it can be made elastic enough so that the energy lost is negligible?). In order for us to know the kinetic energy of the fist using the height of pendulum, mustn't we assume that no energy was lost? Using conservation of momentum would seem like a good alternative because all collisions conserve momentum (but not necessarily energy); however, because there is so many different things acting here (mostly just the fact that the human structure is so complex and imperfect, it's hard to isolate the punch itself from the rest of the body) it seems like that, also, is going to be hard.

    I do definitely agree, though, that calculating the impulse/transfer of momentum will be much more useful in this scenario than a force.
  15. Jun 10, 2007 #14
    The point of a punch is to affect its target. I'm suggesting measuring the effect on the target. I don't know why you would want to know the "kinetic energy of the fist", especially since I understand karate punches aim to briefly hold the bones of the arm straight (so that the entire body's forward momentum, generated by the more powerful leg muscles, is directed into the target.. but what would an aikidoka know?) ..A quicker jab, with more kinetic energy in the fist, might also be a less effective punch, so (like whiteness of knuckles, and unlike perhaps energy dissipated into target) it wouldn't necessarily be a good measure of performance.
  16. Jun 10, 2007 #15
    I thought you were the one that brought up kinetic energy:

    I was trying to build off your ideas.

    Your comparison between quick punches and punches with a follow-through relate directly to what I was talking about with impulses (FT = MV). I am still confused about how we can calculate this without any special equipment.
  17. Jun 10, 2007 #16
    Assuming that the force is constant, why not just take a weight meter and attach something on the end to punch and then place the meter horizontally on flat surface (with something to hold it back so it doesn't fly off when you punch it). I dunno though, the idea seems so simple that I'm sure there's something wrong with it.

    Not sure if I got the term correctly but what I mean by weight meter is one of those meters where you place an object on the end of a spring and it measures the weight (force). You could also just use a regular spring.

    As for getting the maximum reading, use the camera?

    Wouldn't work if you wanted to be as accurate as possible though.
  18. Jun 10, 2007 #17
    If (perhaps unlike the OP) you did want to calculate the total impulse (I), which is equal to the total transfer of momentum, the idea is to strike a (not particular special) hanging bag (for a duration that is short compared to the period it starts to swing with). The initial momentum of the bag is equal to the total transfer of momentum (since it was previously at rest). Energy is conserved as the bag swings up (raising its centre of gravity some distance h, which is fairly easy to measure as the bag slows) so you can deduce the bag's initial velocity ([itex]\frac 1 2 v^2 = g h}[/itex]) and hence (by later weighing the bag to find its mass m) determine the total impulse imparted by the punch (I[itex]=m\sqrt{2gh}[/itex]).
  19. Jun 11, 2007 #18
    Oh, d'oh! After you've said it about 5 different ways now, I finally understand what you're saying. Before when you mentioned finding the potential energy of the bag, I thought conservation of energy for collisions. But now I see that we are just using conservation of momentum, and only the momentum transfered matters. This is most definitely the most accurate (by simple means) way to do this.

    Now that I read through the other posts, I don't know why I had such a hard time grasping that. :P
  20. Jun 11, 2007 #19
Share this great discussion with others via Reddit, Google+, Twitter, or Facebook