This project focuses on research in multibody dynamics, and addresses computational issues therein. It is directed toward improving computational procedures in three ways: 1) by developing more efficient and more accurate analysis procedures and algorithms; 2) by developing consistent approximation procedures; and 3) by developing improved methods for applying the technology. The research is motivated by a desire to address problems in robotics, biosystems, flexible structures, cable dynamics, and colliding system dynamics. Despite rapid advances in multibody dynamics, there remains a great deal to be done. Disagreements exist regarding the best method for obtaining the governing dynamical equations, how redundant and constrained multibody systems are to be studied, how to incorporate flexibility effects into the analyses, and how to use supercomputers and parallel processing in multibody analysis. Kane's equations of dynamics will be used in this project. They are amenable to efficient automatic generation of the dynamical equations, and useful in studying constrained and redundant systems. They also have advantages for systems with flexible members. This research proceeds on five fronts including the use of absolute angular velocity components as generalized speeds, automated modelling, impact analysis, subcomponent analysis of small mass portions of the system, the use of orthogonal complement and projection methods, development of Class VI computer procedures, and the use of animated computer graphics.
