[ODE] Weird Dynamics

[Martin C. Martin]

If you're doing this as a way to learn about ODE, this is a great
project.  But if you're trying to create that simulation and don't care
about learning ODE, you'd be better off writing it yourself.  Your
simulation is essentially 2D, collision detection between two sphere, or
sphere-plane, is trivial, and finding the center of mass and moment of
inertia of the group is also pretty easy.  So is translating the contact
forces to the center, and computing the velocity of the group from the
velocities of the two groups that are being merged.  In fact, this strikes
me as a good assignment for a "physics modeling" course.  And doing it by
hand will certainly give you a good understanding of the basic issues in
physical simulation; you'll be able to appreciate something like ODE a lot
more, and understand its weird behaviour a lot better.

BTW, you want your sphere-sphere ERP (i.e. your fixed joint ERP) to be
very small (say 0.1 or 0.2), because you want sphere-sphere collisions to
be inelastic, or nearly so.

- Martin