SV: [ODE] Iterative solution

Billy Zelsnack billy_zelsnack at yahoo.com
Tue Mar 18 12:46:01 2003


>Russ Smith said:
>as far as constraint order goes, i'm not convinced it makes any
>difference (but i've not verified it one way or the other). the
>general non-blocking SOR iteration to solve A*x=b is:

I find that a randomization of the contacts before iterating works well.
Randomizing during each iteration works a little better, but most of the
benifit is gained from randomizing just once. Sorting by relative approach
velocity works well too.

-billy