[ODE] Re: ODE Digest, Vol 12, Issue 16

[tt]

Hi Marco

I think your way of determining the forces doesn't work for our 
application since we need the forces on a wheeled mechanism.
At the moment we're using ODE in a quasi-static way, i.e. we're moving a 
certain distance and then we're waiting long enough for the mechanism to 
rest in order to get out the effects of bouncing/oscillating. Once the 
robot is at rest we're applying and increasing a force to a wheel in the 
direction of gravity and as soon as the wheel is lifted, we get a good 
approximation of the weight on this wheel. Of course we have to do this 
in a loop for all wheels and it's not real-time, but we get the results 
we need.

c'ya,
Thomas

>
>------------------------------
>
>Message: 2
>Date: Sat, 14 Aug 2004 11:07:38 +0200
>From: "Marco Vienup" <Marco at vienup.de>
>Subject: [ODE] Real contact forces (=weight distribution)
>To: <ode at q12.org>
>Message-ID: <LPEAKBJFPPMBIBHOADFLEEEOCHAA.Marco at Vienup.de>
>Content-Type: text/plain;	charset="iso-8859-1"
>
>Hi Thomas,
>I dont´t know if I was the person, but I had the same problem.
>I needed the correct forces on the feet of a walking robot.
>So I placed a thin body with a very small mass at the bottom of each leg.
>I connect the leg and the thin body with a fixed joint.
>With the dJointGetFeedback function I can determine the forces which affect
>to the leg.
>How do you determine the forces at the moment, and why they are wrong ?
>I didn´t understand it :)
>
>Greetz
>Marco
>
>---
>Marco Vienup
>Mail: Marco at Vienup.de
>
>  
>