The Navier-Stokes equations are central to applied scientific research. The complete set of three-dimensional Navier-Stokes equations is very complex and thus requires a substantial amount of computer time as well as memory in order to obtain an accurate solution. The scalability in both processing power and memory space of distributed-memory parallel computers give promise of solving large scale three-dimensional scientific problems based on these equations. In this paper, we describe the implementation and performance of a distributed three-dimensional Navier-Stokes solver in Parasoft's Express. We have run the solver on both the IBM Victor Computer (a 256-node transputer based system) and a token ring networked IBM RS/6000-520 workstation. Our test results demonstrate that distributed multiprocessing allows researchers to solve large scale computational fluid dynamics problems and can improve their productivity with reducing turnaround time.
ASJC Scopus subject areas
- Chemical Engineering(all)
- Fuel Technology
- Energy Engineering and Power Technology