[ODE] friction anisotropy?

Graham Fyffe gfyffe at gmail.com
Mon Apr 4 19:50:31 MST 2005


Hey all.  It seems to me that the way ODE handles friction necessarily
causes an anisotropic friction with respect to the direction of
sliding on the surface.  What I mean is, if I take a simple block and
slide it along a ground plane, the friction in a diagonal sliding
direction will be greater than the friction in an orthogonal
direction, because ODE's friction is a pyramid instead of a cone.  Is
my take on this right?  Or am I on acid?

The other point I want to ask about with ODE and friction is this:  I
slide my block and also have it spinning, like if you were to slide a
book across the floor and give it a little spin when you release it. 
In real life, the book will stop sliding and stop spinning at exactly
the same time as it comes to rest.  In ODE, the book (or block) stops
spinning WAY before it stops sliding.  What's the deal with that?  Is
it related to using a friction pyramid?  Is it related to having only
a small number of contact points instead of integrating over the
entire contact surface?  That last thought seems wrong, since putting
thumb tacks on the bottom of the book shouldn't change this bevavior
:P  Maybe there's a bug in my code...

Thanks ahead for any help on this one!

- Graham


More information about the ODE mailing list