[ODE] External vs. internal joint forces

Arndt Twickel ode at twickel.ws
Wed Jun 13 17:03:58 MST 2007


>On 13_Jun_07__20:31, Michael Goldish wrote:
>> I would like to know if it is possible to get information about
>> internal  vs.  external forces/torques in joints. By internal forces
>> I mean forces  applied by a motor of the same joint, by external
>> forces I mean all other forces.
> My solution to this problem is to use AMotor joints in addition to
> hinge  joints (or whatever you use).  [...]
> Use dJointSetFeedback on the AMotor joints and hinge joints separately,
> and hopefully you'll receive the correct information.

This sounds like an interesting option. Have you (or anyone else)
already tested this setup? Does dJointGetFeedback on the AMotor joint
only return the forces/torques originating from AMotor itself? Are
external forces only acting through the other joint, e.g. hinge
joint? Even if the setup works as desired (which would indeed be very
cool), two general problems remain:

1. It would only work with rotational joints, not with sliders (no
    AMotor equivalent exists for sliders).
2. The additional constraints add to the computation time, especially
    if many of these joints are used.

> If I may, I would like to ask a related question: Once you have the
> internal forces, how do you compute how much work has  been done?

Together with the angular or positional displacement of the joint you
should be able to use the standard formulas:

    work = torque * angle  or  work = force  * distance

> If the answer is too long or complex [...]

But maybe I don't see the real problem here??


More information about the ODE mailing list