[ODE] Difference in scale of body vs collision geometry

[Jon Watte]

The gravity scales by mass, so you don't need to scale that.

The impact on density shouldn't matter, because ODE does not 
actually use density anywhere inside stepping or constraint 
resolution.

Thus, if you scale weight by 1/N, then you have to also scale 
force and torque applied by 1/N, and that should do it. Note 
that gravity is a measure of acceleration, not force, and 
thus shouldn't be scaled!

Cheers,

			/ h+


-----Original Message-----
From: vincent ngai [mailto:vincent.ngai at gmail.com]
Sent: Monday, December 06, 2004 6:27 PM
To: Jon Watte
Subject: Re: [ODE] Difference in scale of body vs collision geometry


Could you give me more details? Scale the force applied and lower gravity?
What about the impact of having a geometry scale that is vastly
different from the mass?

Sorry I'm a newbie at this.

Thanks


On Mon, 6 Dec 2004 15:46:11 -0800, Jon Watte <hplus-ode at mindcontrol.org> wrote:
> 
> > Am I right to say that I would have to scale the linear measurement units from
> > 1m to 1.0 to 1m to 0.0001 units?
> 
> Wouldn't it be sufficient to scale torque and force?
> 
> Cheers,
> 
>                        / h+
> 
> 


-- 
Regards,
Vincent