[ODE] Assembling Bodies

[Jaap Stolk]

On 4/18/06, Wesley Smith <wesley.hoke at gmail.com> wrote:
> It would definitely be more efficient to just make them one body.
> I'll potentially be connecting N number of bodies.  Right now, the
> most I've seen is 4.  The thing I want to preserve when I connect them
> is the dGeom shape.  So, if I have 4 bodies connected, they bounce
> around like 4 connected spheres and not one bigger sphere with the
> mass of 4 spheres.  In any case, thanks very much for your help.  I'm
> practicing my ODE on a simpler example right now so I can get
> everything straight in my head.

I'm coding something right now, to see if I can make things connect
and disconnect during the simulation. (since I can't think without
code) will post what I have tonight (working or not) the biggest
problems I can see are how to combine the rotational velocities of
both objects.

If you use the add_mass function to add the masses of the spheres,
then the resulting mass should behave like the mass of the connected
spheres, since that function also properly sums the moments of inertia
(mass distribution) of the bodies.
If you then connect multiple geoms to the new body, it should also
bounce and interact like the connected spheres.