If condition to check for contact of two balls

Hi,

I have a problem where I drop 2 balls and when they come in contact they should rebound. I have no problem in dropping or rebounding the balls. The main place where I have a problem is giving the condition of the contact. I am using an *if-else loop to check for contact * and I track the centre points of the balls at each time step to check for contact. In my code, the condition for contact never gets satisified even though I can see the balls merging. Any help would be appreciated. Here is the part of the code where I have stored the co-ordinates of the centers of the two balls in separate x and y variables and calculated the distance between the centers..

   %x1,y1 is the co-ordinate of center of one fiber and x,y that of the other
  x1=get(h2(1),'xdata');
  y1=get(h2(1),'ydata');
  x=get(h2(2),'xdata');
  y=get(h2(2),'ydata');
  d=((x1-x)^2 + (y1-y)^2)^0.5;
     if d<=(2*r)
-------------
-------------
-------------

[Image Analyst]

Are you using the debugger to step through your code? If you did, this kind of thing would be caught quickly. Also, please see the FAQ: http://matlab.wikia.com/wiki/FAQ#Why_is_0.3_-_0.2_-_0.1_.28or_similar.29_not_equal_to_zero.3F. And how are you incrementing x1,y1 and x and y? Are you sure they ever get close enough for an interaction (bounce or intersection)?

[Muthu]

Yes, I used the debugger and I found that it never goes inside the if-loop which means the condition never gets satisfied. As for x1,y1,x and y- the first ball will be dropped into a box and it settles after a point. Then I drop the 2nd ball and the co-ordinates of its centre point(which is saved in a plot) will be tracked at each time step for contact using velocities and displacements.

[Image Analyst]

If you're using the debugger then you will know all the coordinates and you can figure out why it's never satisfied. Take off the semicolons to make it easier to see the values.

[Jan]

Btw.: SQRT is faster than ^0.5, but for comparing the distance, it is much cheaper to omit the square root, but compare with the squared (double) radius.

[Muthu]

I tried seeing the values and setting the if condition accordingly but still it doesnt execute the if loop.

[Image Analyst]

Tell us the values of x1, y1, x, and y for which you think it should enter the "if" block but it doesn't. And tell us d and r in that case.

Answer 1

[Image Analyst]

I think your tolerances are too small, basically they're zero. Maybe widen it:

if d<=(2*r) * 1.05   % If it's within 5%, or whatever...

[Muthu]

Thanks, this was the problem. I totally overlooked it.

Answer 2

[Thorsten]

You forgot the "x = " in the third row. Should read

x = get(h2(2), 'xdata');

[Muthu]

Sorry my bad. It was in the code. I made a mistake while copying.

[Thorsten]

Ok. Then maybe r is changing inside your loop?

[Image Analyst]

So go up and correct it. Next time you can use control-c and control-v to avoid such mistakes. So your problem remains then.

[Muthu]

My r is not changing as it is the radius of the ball and its a constant. Both balls are of the same radius.