Real-Time Physics Simulation Forum

simple collision detection in 2d

2006-08-29T05:16:57+00:00

If you use Baumgarte stabilization, it is quite easy to allow for overlap in an LCP formulation. I do not know of a successful physics engine that has completely eliminated overlap.

Statistics: Posted by Erin Catto — Tue Aug 29, 2006 5:16 am

simple collision detection in 2d

2006-08-28T21:56:15+00:00

the nonoverlapping condition is necessary to use LCP.

Statistics: Posted by fork — Mon Aug 28, 2006 9:56 pm

simple collision detection in 2d

2006-08-28T16:38:11+00:00

If you intend to allow overlap (which I recommend), then the separating axis test can work well. This might help you:
http://www.gphysics.com/files/GDC2006_ErinCatto.zip

Statistics: Posted by Erin Catto — Mon Aug 28, 2006 4:38 pm

simple collision detection in 2d

2006-08-28T05:49:49+00:00

What advice can you people give me for simple collision detection in 2d? For example, say I use bisection to find a time that was very close to the collision. So now I have two polygons that are almost touching. How can I generate the contact set from this? I am looking for the simplest thing possible. Speed is not an issue.

Statistics: Posted by fork — Mon Aug 28, 2006 5:49 am