1) The LCP formulation
2) The Impuls Based Method like described by Mirtich
3) And the Propagation method (penalty forces/impulses) like implemented in Bullet by Erwin and described in D. Baraffs paper.
My first question is how do method 2) and 3) work with articulated structures where I describe the joints like in classcial constrained dynamics (J*v = 0).
My second question is what your experiences are. When does the LCP works well and when the impulse based - maybe somebody likes to share his experience. In our current game we have contact / joint ratios of about 5000 / 20. So I wonder if one of these methods should be prefered in the presence of a great number of contacts?
Regards,
-DirkStatistics: Posted by Dirk Gregorius — Fri Jul 29, 2005 3:07 pm
]]>