Alright, work will commence in earnest after work today. Here’s a list of things to overcome:
-Determine whether or not a physically accurate simulation will work best or to cheat reality in order to get the best looking results.
-Determine the best path for the sling mechanism. This comes in two parts… For a final run, a Cloth object will be best, however, for run-throughs and tests, a cloth object could be too computationally expensive to run. Therefore, tests will need to be run and experiments to determine an adequate proxy object.
-Sling wires: Maya Hair will likely work best.
-The Wall to be destroyed. Will need to develop a system to realistically destroy the wall on impact while maintaining a reasonable simulation time. This will need to include some sort of semi-dynamic crumbling system which uses cube primitives as proxy-collision objects.
-Keep sim time down! With an entirely dynamic driven or semi-dynamic driven system, this will get extremely computationally expensive. The best likely approach will be to run through the simulation of the trebuchet and the wall separately, baking them off and then positioning them accordingly.
-Setup shaders and render out the bad-boy.
I’ll be using Maya 6 Unlimited. Perhaps some Houdini though I don’t know about the interoperability of Houdini Apprentice with other programs.


