| A general framework for rapid implementation and testing of numerical methods for multibody system kinematic and dynamic simulation, dynamic sensitivity analysis, and kinematic workspace analysis is presented. Efficient numerical algorithms that take advantage of the structure and functionality of the mechanical system, as well as sparsity in mathematical models, are developed for solution of the differential-algebraic equations of motion and sensitivity equations.; The Cartesian formulation with Euler parameters for orientation is shown to yield the structure and simplicity required for consistent and systematic generation of the equations of motion, including higher order derivatives of terms in the equations of motion required for a broad range of analyses. Efficient numerical methods that take advantage of system topology, functional intent of the mechanism, and structure and sparsity of matrices in the equations of motion are shown to yield efficient simulation and analysis algorithms.; Optimization analysis for multibody systems is conducted in two areas closely related to real-time multibody system simulation. Parameter identification from experimental data provides a method of choosing unknown parameters in mathematical models such that its dynamic response matches experimentally measured data. Variable fidelity model correlation addresses the problem of selecting control inputs to a high fidelity model, suitable for accurately predicting loads and stresses for design purposes, to minimize deviation in its dynamic response relative to that of a lower fidelity model of the same system, suitable for real-time simulation. |
