[ODE] Re: Joint forces

[nlin@nlin.net]

paul.geisler at web.de wrote:

> It would be nice to have funcions which can report forces which traverses
> between two connected bodies over a specific joint. I got this idea,
> cause i want to simulate gr oups of objects which can be checked for
> "force overload" inside their joints, so parts can "brake apart" on
> their demand. So you could test the joint forces, and if they
>  exceed a defined "break limit", the joint gets deleted, and the body
>  could break apart.

Question to physics gurus: ODE uses the maximal coordinate (as opposed to
a generalized or reduced coordinate) method, where constraint forces are
computed to ensure that each constraint (joint) is met, thereby effectively
cancelling illegal accelerations with constraint forces. Is the "force which
traverses between two connected bodies over a specific joint" exactly
equal to this computed constraint force? It seems like this would be
a wonderously simple and elegant physical interpretation of the computed
constraint force - but I am not sure if this is an oversimplification.
Is it?

-Norman