# How should I model rotation caused by bounce friction? [duplicate]

In my Java2D simulation, I have a disc which, when it bounces off the boundary of a 2D region, begins to rotate, as if it had gotten some spin from the wall.

I'm having trouble relating the collision vectors and the angular velocity of the spin that should result.

How should I approach this?

• The first answer says 'cross of incident and normal'. Its unclear if its velocity vector or position vector. – Madeyedexter Feb 2 '14 at 17:57
• It has to be the velocity. Position vectors are always rooted at 0,0 and so their direction doesn't have much meaning. (Consequently, crossing them with something wouldn't do anything very interesting.) – Anko Feb 2 '14 at 20:05