[ODE] Controlling object speed externally
Jon Watte (ODE)
hplus-ode at mindcontrol.org
Wed Oct 3 11:13:38 MST 2007
Additionally, make sure you add sufficient rotational inertia dampening,
else you'll get bouncing/jittering.
The problem with the original description is that Crestez wants to use
ODE for collision response, but not for movement. Unfortunately,
collision response is the same thing as movement. Thus, it might be
better to add forces and torques to the bodies, each step, to make them
go where he wants them to go (within limits of simulated engine power).
Bram Stolk wrote:
> What are your ERP and CFM values?
> Use a non-zero CFM and a ERP smaller than 1.
> On 10/3/07, *Crestez Dan Leonard* < cdleonard at gmail.com
> <mailto:cdleonard at gmail.com>> wrote:
> I'm using ODE by completely controlling object position/speed/etc
> my code and only letting ODE control them after the first collision. I
> want to use only ODE for collision detection and response, not motion.
> Is this acceptable or should I try to use motors or explicit forces
> instead? It seems to almost work.
> I have a problem: When objects crash they start "wobbling" on the
> ground. I disabled all bounciness but objects seems to oscillate a lot
> on the ground. For instance a large box will first "pivot" on the
> and then go back and pivot on the back edge and raise to almost the
> same height. They take way too much time to stabilise. I saw this FAQ
> I tried to implement "rolling friction" with something like
> AddTorque(-0.1 * GetAngularVel()) but it didn't work. Shouldn't the
> inertia tensor be involved here somehow? Or perhaps it should be
> implemented for each contact point? I have no idea.
> Perhaps it actually happens because I control the objects by hand
> before collision?
> ODE mailing list
> ODE at ode.org <mailto:ODE at ode.org>
> Zapp: Captain's log, stardate...er..
> Kif: Ohhh. April 13th.
> Zapp: April 13th. Point 2.
> ODE mailing list
> ODE at ode.org
More information about the ODE