Multibody simulation

Simulation of the dynamics of physical systems is an important aspect of the engineering discipline for approximating the dynamics of real life. The simulation of complex multibody systems to an acceptable degree of accuracy involves the mathematical modeling and computer implementation of systems such as mechanisms and vehicles comprised of multiple parts. In this dissertation, new algorithms are developed for multibody simulation using a rather general mathematical model. Both open-tree and closed-loop topologies are implemented. Constraints, specifically, joint constraints, are investigated. A new algorithm is developed that projects the original configuration space into the unconstrained orthogonal subspace, thereby reducing the dimension of the system equations without resorting to complicated transformations. The reduced set of equations not only increases the simulation speed, but also improves the numerical accuracy of the simulation results by reducing the number of calculations performed. Constraint forces can easily be obtained if required for analyzing the multibody system. Algorithms by themselves are not immediately useful to users. A program was developed to implement the algorithms. The program, which was written in C/C++, incorporated the use of Microsoft Windows Application Programming Interfaces (Windows API), Microsoft Foundation Classes (MFC), and OpenGL graphics language. The system states are integrated by applying standard numerical techniques for integrating a set of first-order differential equations. Accelerations and constraint forces are obtained using direct and/or iterative techniques for solving a set of simultaneous equations.; With today's powerful computers, a graphical interface becomes feasible to serve as the communicator between the program and the user. The software therefore includes a graphical user interface. Concurrent graphical animations of the motion of the system simulated are created. These are important to the user because they provide access to the multibody dynamics core of the program. This complete software effectively integrates all the algorithms and allows them to be put to good use. Applications of the program to several example problems ranging from single body systems to a complex fourteen-body missile launcher system and the results are presented to illustrate the effectiveness. Results are compared with analytical solutions for simple systems and results obtained from other multibody program.