Parallel Computing on the Navier-Stokes Solver with the Multigrid Method

This paper is aimed to present the combination of the parallel computing and the multigrid method on the Navier-Stokes solver. The combination is based on the concept of the object-oriented programming (OOP), which consists of 4 independent modules: Grid Generation, Navier-Stokes Solver, Multigrid Method and Parallel Computing modules. The multigrid method is implemented by employing the full approximation storage (FAS) scheme for numerically solving the non-linear Navier-Stokes equations. The overall computation is performed by using the parallel computing in which a number of computers are concurrently computed for the same task but on different subdata. The two-dimensional laminar flow in a cavity at Re=1,000 is used as a test case. It is found that the computational time is decreased significantly when employing the combination of the multigrid method and the parallel computing.