Re[2]: [ODE] Feedback from contact joints.. ?

[Amund Børsand]

Nate W <coding@natew.com> skreiv:

> On Thu, 19 Jun 2003, Amund [ISO-8859-1] Børsand wrote:
> 
> > Hi, I'm trying to implement differentials in my car, and then I need
> > to know how much resistance the wheels get from the ground. First I
> > thought I could do this by getting the feedback from the wheel hinge
> > (I'm currently using hinge2's with Fmax to rotate the wheels), but I
> > need to know the resistance _before_ the wheel has turned.  So I
> > thought what I needed was to know how much force the ground applies to
> > the wheel.

> Seems to me that the information you need will be present in the hinge
> feedback.  

Yes, sounds reasonable. Don't remember why I ditched that thought...

> > Or maybe this doesn't solve my problem anyway?

> For a differential, aren't you more interested in how much the wheel has
> rotated, rather than how much force is on the wheel?  E.g. if one wheel
> has turned .2 radians and the other has turned .3 radians, you know your
> differential input shaft turned .25 radians, and the spider gear (if thats
> the term?) allowed +/- .05 radians difference between the two.

Yes, but the problem is that this works the other way around. Or
actually, both at the same time! I need to know how much resistance the
wheels have before I turn them! I think... or maybe I'm beginning in the
wrong end here.

The point of differentials is that they split the torque equally
(contrary to popular belief), so that if I put 300 NM on the input shaft,
each wheel should recieve 150 NM (times the gearing ratio in the diff.,
but that's another story). I think it's all wrong using the Amotor in
the first place, now that I think of it... Guess I'll have to experiment
a little with applying my own torque to the wheels. But thanks for the
reply, at least it got me one step further!