Comparative performance of frontal(direct) and pcg(iterative) solver based parallel computations of finite element analysis

摘要: PARAM-10000 is latest of the series High Performance Parallel Super Computers developed by India. It employs a distributed memory and message passing architectures built as cluster SMP workstations using 4-way from Sun Microsystems.The work discusses about Finite element Analysis Programs with Direct Frontal Solver Iterative PCG (Pre-conditioned Conjugate Gradient Method) for 2d 3d analysis. The Comparative these Computation FEA (Direct based) bench marked on PARAM-10000. Studies have been conducted medium sized two three dimensional linear finite analysis problems elastic structural To increase problem size(dof), automated meshes were generated 8-noded serendipity elements 20-noded brick Each has analysed 1 to 4 Processors Detailed Speed-up Efficiency studies reported conclusions are drawn machine.