[ODE] problem using 0.0 as joint limit

[Tyler Streeter]

Here is a weird problem I've been trying to debug for
several hours:

I have a hinge joint with limits: low = 0.0, high =
2.879.  The initial joint angle is 0.0, of course.  I
set the desired hinge motor velocity to 0.0 with a
non-zero max torque value.  Immediately after starting
the simulation the joint gets forced to an angle of
about -0.995.  It seems really strange that it would
get forced away from an angle of 0.0 with zero desired
velocity, let alone being forced below the low joint
limit.  I'm not handling collisions between joined
bodies, so that's not the problem.  Also, if I set the
max motor torque to 0.0, the problem doesn't occur. 
This seems to imply that the joint motor is causing
the problem, even though the desired velocity is
always 0.0.

I was able to get around this problem by setting the
low joint limit to -0.01; that totally fixes the
problem, but I'd still like to know if this is a bug
or something.  Is there a problem with using 0.0 as a
low limit for a motored joint?  Maybe somehow it's
conflicting with the initial joint angle also being
0.0?

I've noticed that this problem also occurs with a
ball-and-socket with amotor joint, btw.

Thanks to anyone who can help.

Tyler Streeter


	
		
__________________________________
Do you Yahoo!?
Yahoo! Photos: High-quality 4x6 digital prints for 25¢
http://photos.yahoo.com/ph/print_splash