2D Ball Collision (velocity vector calculation)

This is a discussion on 2D Ball Collision (velocity vector calculation) within the Game Programming forums, part of the General Programming Boards category; Hey, I'm working on a project for one of my classes - while the part I'm stuck on isn't centric ...

  1. #1
    Registered User
    Join Date
    Aug 2010
    Location
    Rochester, NY
    Posts
    196

    2D Ball Collision (velocity vector calculation)

    Hey,

    I'm working on a project for one of my classes - while the part I'm stuck on isn't centric to the project, it is indeed a part, which is why I'm telling you about it initially.

    In said project (it's in C), I have to simulate a collision of 2 balls in 2D space. Now assume all the collision detection and such has already been done (it is), I'm having trouble calculating the new velocity vectors (I'm pretty crappy at math, haha).

    I have 2 versions where I attempt to calculate the exidence vectors.
    Code:
    
    // Calculate the ball deltaX and deltaY
    float dX = balls[i].X - balls[j].X;
    float dY = balls[i].Y - balls[j].Y;
    
    
    // Normalize each delta
    float normalDX = dX / sqrt(pow(dX, 2) + pow(dY, 2));
    float normalDY = dY / sqrt(pow(dX, 2) + pow(dY, 2));
    
    
    // Calculate dot product and get each magnitude
    float magnitude_I = (balls[i].velocityX * normalDX) + (balls[i].velocityY * normalDY);
    float magnitude_J = (balls[j].velocityX * normalDX) + (balls[j].velocityY * normalDY);
    
    
    // Apply initial magnitudes to opposite balls
    // (Collision is perfectly elastic)
    float I_final_mag = magnitude_J;
    float J_final_mag = magnitude_I;
    
    // Apply velocities to X/Y components of each ball
    balls[i].velocityX += (I_final_mag - magnitude_I) * dX;
    balls[i].velocityY += (I_final_mag - magnitude_I) * dY;
    
    
    balls[j].velocityX += (J_final_mag - magnitude_J) * dX;
    balls[j].velocityY += (J_final_mag - magnitude_J) * dY;
    Code:
    float dX = balls[i].X - balls[j].X;
    float dY = balls[i].Y - balls[j].Y;
    
    
    // Normalize each delta
    float normalDX = dX / sqrt(pow(dX, 2) + pow(dY, 2));
    float normalDY = dY / sqrt(pow(dX, 2) + pow(dY, 2));
    
    
    // Calculate dot product and get each magnitude
    float magnitude_I = (balls[i].velocityX * normalDX) + (balls[i].velocityY * normalDY);
    float magnitude_J = (balls[j].velocityX * normalDX) + (balls[j].velocityY * normalDY);
    
    
    // Get the X and Y components of resultant magnitudes for the change in velocity
    float newDY_I = sin(magnitude_I);
    float newDX_I = cos(magnitude_I);
    float newDY_J = sin(magnitude_J);
    float newDX_J = cos(magnitude_J);
    
    
    // Apply the changes in velocity to the balls
    balls[i].velocityX += newDX_I;
    balls[i].velocityY += newDY_I;
    balls[j].velocityX += newDX_J;
    balls[j].velocityY += newDY_J;
    Where:
    - I/J are just counters, representing different balls
    - velocityX/velocityY are the X/Y components of the velocity
    - X/Y are the locations

    All the examples I've found of calculating the new vector are in OO or Java which are blackboxing the vector math.

    Both of these solutions yield crappy results, I'm not sure where the issue is, I'm not sure if it's something I'm doing wrong in the implementation, or something wrong with my thought process, but something is definitely wrong.

    My thought process for this is:
    1. Get vector between 2 ball centers
    2. Normalize vector
    3. Take dot product of that and the velocity vector
    4. Apply resultant to X/Y components of velocity
    - WHere 3/4 must happen for both balls.

    Maybe somebody sees something that's obviously wrong and can point me in the right direction or something? Not looking for answers here, just help.

    TIA.
    Last edited by Syndacate; 02-13-2012 at 06:15 PM. Reason: Fixed color

  2. #2
    Programming Wraith GReaper's Avatar
    Join Date
    Apr 2009
    Location
    Greece
    Posts
    1,625
    ( Just noticed this thread, sorry )

    Let's see, the math for sphere elastic collision( without rotational forces ) are:
    Let Norm1 be a unit reflection of Sphere1's velocity vector onto the collision plane
    Let Norm2 be a unit reflection of Sphere2's velocity vector onto the collision plane

    Let Vel1 be the velocity vector of Sphere1
    Let Vel2 be the velocity vector of Sphere2

    Let Mass1 be the mass of Sphere1
    Let Mass2 be the mass of Sphere2

    Let P1, P2 be the momentums(scalars) of Sphere1, Sphere2 respectively ( Momentum = Mass * Velocity )
    P1 = Mass1 * |Vel1|
    P2 = Mass2 * |Vel2|

    Let A1 be the acceleration of Sphere1 on the time of collision
    Let A2 be the acceleration of Sphere2 on the time f collision
    Let Time be the amount of time both sphere's retain contact
    ( Force = Mass * acceleration && Force = Momentum / Time <=> acceleration = Momentum / (Mass * Time) )
    A1 = P2 / (Mass1 * Time)
    A2 = P1 / (Mass2 * Time)

    Then newVel1 = Vel1 + A1 * Time = Vel1 + (P2 / Mass1) * Norm1
    And newVel2 = Vel2 + A2 * Time = Vel2 + (P1 / Mass2) * Norm2
    Hopefully I didn't confuse you.
    Also hopefully, I didn't do anything wrong... ^_^
    Devoted my life to programming...

Popular pages Recent additions subscribe to a feed

Similar Threads

  1. 3D velocity issues
    By gesangbaer in forum C++ Programming
    Replies: 5
    Last Post: 04-02-2010, 08:58 AM
  2. Collision detection between bezier and ball.
    By CornedBee in forum Game Programming
    Replies: 1
    Last Post: 04-27-2006, 01:49 PM
  3. Velocity of an object
    By hdragon in forum Game Programming
    Replies: 1
    Last Post: 03-16-2006, 01:35 PM
  4. orientation from velocity
    By ichijoji in forum Game Programming
    Replies: 2
    Last Post: 08-28-2005, 08:22 PM
  5. Velocity vector troubles
    By VirtualAce in forum Game Programming
    Replies: 8
    Last Post: 01-18-2005, 11:40 AM

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21