[ODE] Combining contact surface parameters

[Graham Fyffe]

You might want to use sqrt(A,B) for bounce as well, since it's the
coefficient of restitution.  Usually those are given as "a on a" and
"b on b", much like mu.  This kind of parameter goes well with the
sqrt idea.  bounce_vel, I suppose average would be fine since it's
kind of a hack anyway :)  For slip, it's vaguely friction related, so
I think you're good.  As for softCFM and softERP.. hmm.. I guess you
want the effective spring rate to be averaged, so you could convert
back from this:

	double erp = simStep * spring / (simStep * spring + damp);
	double cfm = 1.0 / (simStep * spring + damp);

then average out the spring and damps, then convert back...  : )

- Graham


On Tue, 1 Mar 2005 03:50:53 -0600, Alex Rubinsteyn <rubnstyn at uiuc.edu> wrote:
> Thanks Graham! That's a great idea.
> 
> I'm going to try the following system:
> 
> When colliding surfaces A and B
> 
> mu, mu2: sqrt(A*B)
> slip1, slip2: sqrt(A*B)
> bounce, bounce_vel: average(A,B)
> motion1, motion2: average(A,B)
> softCFM, softERP: max(A,B)
> 
> I'm really not sure about slip (FDS) and motion, since I've never used them
> before.
> 
> -Alex
> _______________________________________________
> ODE mailing list
> ODE at q12.org
> http://q12.org/mailman/listinfo/ode
>