Parallel strategies for reduced number of processors applied for CFD simulations on workstations

Abstract

This work is aimed to exploit the computational capabilities of nowadays workstations by building algorithms parallelized for a limited number of processors. For an engineer that is working in aerodynamic design, the possibility of having efficient and whenever available computational resources is a considerable advantage. By using adaptive algorithms to numerical solution conform to workstation's hardware configuration, acceptable processing time and enough accurate solutions configuration can be obtained. The algorithms were tested on different hardware platforms and encouraging results were obtained.