Contact handling, bullet method or Bender method??

[Dirk Gregorius]

http://www.gphysics.com/archives/28

Here you also find the link to Claude's work you were missing lately. I think Erin implemented also a PCG a while ago and skipped it for some reasons, e.g. you cannot stop it as easily like PGS after a fixed number of iterations. Maybe you PM him and ask him directly.

I also think that the links mentioned earlier don't use CG in the context as we use PGS. David Wu used minimization techniques and the latest work of O'Brian is about FEM and destruction. The later might be similar to the Baraff cloth stuff.


Cheers,
-Dirk

[dneckels]

In the particular impulse method I was used (from J.Bender) there is a prediction step for the contact penetration (using a local RK4) that made it somewhat different from SI, as well.