Thread: Simple Bouncing Logic

  1. #1
    Registered /usr
    Join Date
    Aug 2001
    Location
    Newport, South Wales, UK
    Posts
    1,273

    Simple Bouncing Logic

    Hello,

    I wonder if someone could give me a quick hand with the logic for an simple falling object that bounces inelastically, that is, it doesn't bounce back to the same height as before and will reach a point of rest. It's for a general 2D game object framework thingy that I fancied having a go at doing.

    Anyway, here's what I've got so far:-
    Code:
    tt = me->pvData; // get our pointer to extra info for this type
    if (tt->dVelocity != 0)
    {
        // We're on the move
        if (fabs(tt->dVelocity) < 0.1)
        {
            // Below minimum clout, if there's something below stop
            if (DetectRect(me, me->x, me->y + 1, me->w, me->h))
                tt->dVelocity = 0;
    
        }
        else
        {
            // Keep going
            tt->dVelocity += GRAVITY * 0.02; // 1/60 second
            me->y += (tt->dVelocity - (int)tt->dVelocity >= 0.5) ? (int)tt->dVelocity + 1 : (int)tt->dVelocity;
        }
    
    }
    else
    {
        // We're stopped.  If there's nothing holding us up...
        if (!DetectRect(me, me->x, me->y + 1, me->w, me->h))
        {
            tt->dVelocity += GRAVITY * 0.02;
            me->y += (tt->dVelocity - (int)tt->dVelocity >= 0.5) ? (int)tt->dVelocity + 1 : (int)tt->dVelocity;
        }
    
    }
    So far the object bounces once and then stops. When it collides with another object the velocity is inverted, like so:-
    Code:
    tt->dVelocity = -tt->dVelocity * tt->dVelScale;
    tt->dVelScale += 0.1; // try to give a ping-pong ball effect (retains more energy the closer it is to rest)
    Using dVelScale like that is probably not the best way to achieve what I want, but I want to use a simple model, people can adjust this value to give different effects.

    Can anyone describe how I can best check if the object needs to start/stop moving?

  2. #2
    Cat without Hat CornedBee's Avatar
    Join Date
    Apr 2003
    Posts
    8,895
    My physics book gives the following formula for the coefficient of restitution, which measures how elastic the collision is:
    e = - ( v1+ - v2+ ) / ( v1- - v2- )
    Where vn+ are the velocities of the two objects after and vn- before the collision. e = 0 means totally inelastic, e = 1 means totally elastic.

    And I just realized this has little to do with your question. But it might help you to make the collision more realistic.
    All the buzzt!
    CornedBee

    "There is not now, nor has there ever been, nor will there ever be, any programming language in which it is the least bit difficult to write bad code."
    - Flon's Law

  3. #3
    Registered /usr
    Join Date
    Aug 2001
    Location
    Newport, South Wales, UK
    Posts
    1,273
    Nono, I think it helps, I was actually looking at Wikipedia a few minutes after posting this at the COR and thinking "Ohhhhhhh...".

    I think it's a good idea to implement that into my "simple" model. The main point is just whether anyone else has had to deal with the possibility of something bouncing on a possibly moving object. I'm not after calculating normals, that'll come a bit later, just the best way to toggle between falling and not falling.

    EDIT: Now it looks like:-
    Code:
    tt = me->pvData;
    if (tt->dVelocity > 0 && tt->dVelocity < 0.01) // Only check if we're falling (positive velocity)
    {
        if (DetectRect(me, me->x, me->y + 1, me->w, me->h))
            tt->dVelocity = 0;
        else
        {
            tt->dVelocity += GRAVITY * 0.02;
            me->y += (tt->dVelocity - (int)tt->dVelocity >= 0.5) ? (int)tt->dVelocity + 1 : (int)tt->dVelocity;
        }
    
    }
    else
    {
        tt->dVelocity += GRAVITY * 0.02;
        me->y += (tt->dVelocity - (int)tt->dVelocity >= 0.5) ? (int)tt->dVelocity + 1 : (int)tt->dVelocity;
    }
    and
    Code:
    // This body's mass  = 10, COR = 0.8.  Other body's mass = 2, velocity = 0.
    tt->dVelocity = (tt->dVelocity * (tt->dMass - (tt->dRestitution * 2))) / (tt->dMass + 2);
    tt->dVelocity = -tt->dVelocity; // Er... bounce?
    Last edited by SMurf; 10-26-2006 at 04:49 PM.

  4. #4
    Registered User VirtualAce's Avatar
    Join Date
    Aug 2001
    Posts
    9,607
    You will have to integrate that in order to simulate the bouncing on a per-frame basis.

  5. #5
    Registered /usr
    Join Date
    Aug 2001
    Location
    Newport, South Wales, UK
    Posts
    1,273
    Not sure what you mean, Bubba. The larger block of code is called with the game loop (60 Hz), so acceleration due to gravity is simulated OK. The smaller block is called whenever a collision between the falling object and something else is detected, this should be an instantaneous calculation surely?

    Watching the object's velocity before and after a collision, using the same object parameters as before (positive = down):-
    6.120000 -> -2.972571
    3.327429 -> -1.616180
    2.703820 -> -1.313284
    2.646716 -> -1.285548
    2.674452 -> -1.299020
    2.480980 -> -1.205048
    2.574952 -> -1.250691
    2.709309 -> -1.315950
    2.644050 -> -1.284253
    2.675747 -> -1.299649

    It seems to oscillate...

Popular pages Recent additions subscribe to a feed

Similar Threads

  1. Actors, cues, event based logic.
    By Shamino in forum Game Programming
    Replies: 2
    Last Post: 04-27-2006, 10:58 PM
  2. Replies: 5
    Last Post: 01-31-2006, 01:54 AM
  3. ...deceptively simple...
    By Sebastiani in forum A Brief History of Cprogramming.com
    Replies: 0
    Last Post: 07-29-2002, 12:51 PM
  4. Circular Logic
    By DavidP in forum A Brief History of Cprogramming.com
    Replies: 1
    Last Post: 10-15-2001, 08:10 PM