Realistic synthetic motion is required in applications  ranging from commercial entertainment to surgical training.  However, generating realistic motion for complex objects is a  difficult task because of the large amount of data that must be  specified and because humans are very good at detecting unnatural or  implausible motions. I have explored one possible solution to this  problem: using physically based methods to automatically generate  motion for animated objects through the numerical simulation of  their physical counterparts. In particular, I have developed a  series of techniques for modeling the behavior of passive systems  such as water, cloth, and breaking objects, as well as techniques  for coupling multiple, heterogeneous systems together. In this  talk, I will emphasize recent research on modeling fracture  propagation in a dynamically restructured finite element mesh in  order to animate objects that can crack or tear. Because my goal is  realistic motion, I will also discuss evaluation techniques such as  user testing and side by side comparison with high-speed video  footage.