Implicit model for rigid bodies.

ngbinh
Posts: 117
Joined: Fri Aug 12, 2005 3:47 pm
Location: Newyork, USA

Implicit model for rigid bodies.

Post by ngbinh »

This is the "best student paper" in Robotics Science and Systems conference in 2007.

This is can be say the most accurate formulation for rigid bodies system with Coulomb friction.

Some quick reads:

1. It couples collision detection and dynamics in one formulation. Currently we have collision detection -> dynamics, thus we only used the information of old time step and the constraint may not be satisfied at the next time step.

2. The formulation is highly non-linear CP and can only be solved by Wisconsin's PATH

3. It will work with any kind of super quadric shapes.

I will also post a demo soon.

Link and abstract:
http://www.roboticsproceedings.org/rss03/p23.html