[ODE] joint friction

Max Lungarella max.lungarella at aist.go.jp
Wed Nov 6 17:07:02 2002

hi russ,

yes, your model sounds perfectly reasonable to me. thank you very much. but what about ball and socket joints? is there a way to set something similar to an "fmax" parameter for them? was not able to figure out any function dealing with it.

thanks a lot!


> > i may be asking something that is documented somewhere.
> > but anyhow ... is there a way to enable friction in the joints?
> there is a good and simple way to do this. set the joint 'fmax' parameter
> to some (positive) value. set the desired velocity parameter to 0. now the
> joint will exert a force up to fmax to bring the joint velocity down to 0.
> this is actually a pretty good model of coulomb friction for a joint.
> changing fmax to be higher will give more friction - and there's no
> stability problem with choosing any value of fmax >= 0.
> russ.
> --
> Russ Smith
> http://www.q12.org/