Publisher Summary This chapter describes the many software and hardware systems that were developed at Pratt & Whitney (P&W) to harness the power of hundreds of networked workstations. Historically, expensive and dedicated supercomputers were used to perform realistically sized computational fluid dynamics (CFD) analyses fast enough to impact hardware designs. With the large scale introduction of workstations a new and less expensive computing resource became available to replace the supercomputer for the large scale execution of production, design oriented, CFD analyses. The chapter describes the software created to facilitate all aspects of the parallel execution of a three-dimensional Navier-Stokes solver. The evolution of this software and the hardware required to make the system robust, reliable, efficient and user-friendly in a large scale environment are also presented. The impact this parallelization has had on the design process, product quality and costs are also summarized. The chapter demonstrates that the use of networked workstations for CFD applications is a viable, less expensive alternative to supercomputers.
[1]
T. A. Zang,et al.
Spectral methods for fluid dynamics
,
1987
.
[2]
Message P Forum,et al.
MPI: A Message-Passing Interface Standard
,
1994
.
[3]
C. R. Lejambre,et al.
Advanced Transonic Fan Design Procedure Based on a Navier–Stokes Method
,
1994
.
[4]
Jack Dongarra,et al.
PVM: Parallel virtual machine: a users' guide and tutorial for networked parallel computing
,
1995
.
[5]
C. Rhie.
Pressure-based Navier-Stokes solver using the multigrid method
,
1986
.
[6]
C. R. Lejambre,et al.
1995 ASME Gas Turbine Award Paper: Development and Application of a Multistage Navier–Stokes Flow Solver: Part II—Application to a High-Pressure Compressor Design
,
1998
.