Movable objects 'stick' to immovable objects

Mar 28, 2010 at 11:17 PM


spheres rolling on a plane with elongated boxes for them to collide against


At acute angles or slower speeds, the spheres 'stick' to the boxes and roll along them rather than bouncing off. This happens at every combination of elasticity & roughness I have tried.

Are there any other settings I should be looking at that might affect the 'deflection' of the spheres?

Mar 30, 2010 at 1:19 AM

Thanks for the report, I'll see if I can set up a repro and figure it out. It sounds more like a bug than a config issue.

Apr 12, 2010 at 9:05 PM

I've had mixed success investigating this, but I think I get the gist of what you're talking about. I'm wondering if the MinRestitution property of PhysicsManager plays a part in this behavior. Basically it defines a cutoff under which objects will not bounce off each other but instead be treated as inelastic. The purpose of the property is to reduce jitters when elastic bodies are resting on each other. Setting it to zero might change this behavior, but could have the downside of causing some "resting" objects to "micro-bounce" indefinitely. It could be that setting it to a smaller value would help, though. Will keep investigating.