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

Momentum and Collision in 3D

  1. Aug 14, 2010 #1
    Hey, I've done quite a bit of Mechanics work in the past... But I really don't even know where to start here.

    For some work I'm doing, I need to calculate the resulting velocity vectors when 2 objects collide. These 2 objects can be any shape, any mass and going at any velocity independently. (The simulation is in 3D, so x, y, z)

    I suppose I'd have to calculate the normal of the two objects, do something with their velocity vectors and masses... And I suppose it would also depend on the objects' bounciness/stiffness etc...

    If anyone could give me the names of some equations I can look into? Or any sort of help... Really just a place to start would be great.

    Thanks for any help,
  2. jcsd
  3. Aug 15, 2010 #2
    Commonly called a Billiard ball collision.
  4. Aug 15, 2010 #3
    I'm Googling that term now, apparently it's also called "Elastic Collision" which is a great help (I had no idea of what to even start searching before) - but all of the equations seem to be confined to a 1D plane of movement.

    Would it just be a matter of making the velocity vectors 3D? Here's to hoping so - it's just things rarely are so simple when transferring over to 3 dimensions in Maths :rolleyes:

    Ahh, actually - I just found this Wikipedia article
    Which explains about 2D and 3D at the bottom. It's not as simple, but neither is it beyond my knowledge - so I'm happy there :P

    Thanks a lot for the help; I genuinely wouldn't have known where to start looking without that
    Last edited: Aug 15, 2010
  5. Aug 15, 2010 #4
    Perfectly OK to just go to 3 dimensions. Or as many as you like ;-)

    You just have to resolve the velocities and momenta into 3 orthogonal directions.

    Of course, you rarely have to do that in physics problems because any collision between two particles is only in 1 plane, even when it's a glancing collision.

    You might want to consider inelastic collision - in the real world things tend to splat more than they bounce.

    If it's not a serious project ( a game for example) you could equally well 'fake it'. Just use random direction changes and keep the total momentum/ energy constant (or lose a bit - inelastic). No need to meticulously work out contact angles.
Share this great discussion with others via Reddit, Google+, Twitter, or Facebook