Implementations of parallel grid assembly for moving body problems

Abstract

The development of a parallelized aerodynamic simulation process involving moving bodies is presented. The implementation of this process is demonstrated using the Beggar code. Coarse grain parallelism of the solution process is obtained using domain decomposition and functional decomposition techniques on a shared memory, MIMD computer achitecture. Explicit message passing is implemented using PVM. Separation trajectories for a three store ripple release are compared with previous calculations on single processors and execution times are presented along with an analysis of parallel efficiency. Overlapping of the computation of the flow solver and the grid assembly process improves the parallel efficiency of solving moving body problems. © 1998, American Institute of Aeronautics and Astronautics, Inc.