[ODE] Damage!

Erwin de Vries erwin at vo.com
Mon Jul 22 12:04:47 2002


Or even better just provide a callback function for these things. Just some
thoughts..

Provide a function:
void ImpulseCallback(dBodyID Body, dVector Pos, dVector Normal, dReal
Impulse)

The normal and Impulse members could also be a combined in a Impulse vector.

It might also be a good idea to simply provide a once per step resultant
impulse callback instead of 1 per contact point. Or provide both. =P

Erwin

----- Original Message -----
From: "Anselm Hook" <anselm@hook.org>
To: "Erwin de Vries" <erwin@vo.com>
Cc: <ode@q12.org>
Sent: Monday, July 22, 2002 03:53
Subject: Re: [ODE] Damage!


>
> This is the right way; my last post was inaccurate.  The forces are
> computed explicitly but not persisted outside of the solver - you have to
> find where they're computed and then persist them for your use later.
>
> -a
>
> On Sun, 21 Jul 2002, Erwin de Vries wrote:
>
> > > i think i need to know what happened inside the contact joints during
last
> > > iteration...?
> >
> > Thats also my thought. You need feedback from the dynamics solver.
Imagine
> > you have a car driving through a city at 100km/h. When you hit a
building
> > you are dead. When you hit a 10kg box you are (pretty) ok. Now you could
> > probably fiddle around with the body's mass and inertia tensor to get
> > something pretty convincing but its all a hack. I'm also looking for a
good
> > way to solve this problem.
> >
> > Erwin
> >
> > _______________________________________________
> > ODE mailing list
> > ODE@q12.org
> > http://q12.org/mailman/listinfo/ode
> >
>
>