ACADEMIA

MSC.Software today announced MD Nastran achieved two new leading performance benchmarks for engine analysis issued by the German Association for Research on Automobile-Technique (FAT). "Engine models have many millions of degrees of freedom (DOF) making them very large and requiring as much as several days processing time, that MD Nastran reduces to hours," said Reza Sadeghi, vice president, product development, Enterprise Computing, MSC.Software Corporation. The following results from the FAT benchmark models demonstrate the fast processing and performance results achieved with MD Nastran. Engine Block Linear Simulation with MD Nastran Model size: 25 Million DOF Compute resources: Itanium2, 1.4 GHz, 8GB RAM Elapsed time for 2 static load cases: 3 hours, 20 minutes Elapsed time for 20 eigenmodes: 24 hours (1 CPU); 8 hours (4 CPU´s) Engine Block, Head and Gasket Nonlinear Simulation with MD Nastran Model size: 10.5 Million DOF Compute resources: Itanium2, 1.6 GHz and 2x2 CPU´s DMP (distributed memory parallel), 9GB RAM Elapsed time for 2 load cases (pretension and pressure): 50,200 seconds (13.9-hours) "Processing speed and performance are crucial for cutting costs and enabling innovation," said Glenn Wienkoop, president and chief operating officer, MSC.Software Corporation. "The automotive industry understands these crucial issues and runs standard benchmarks used by its members for comparison of competing products. In this study MD Nastran was evaluated against other solvers and continued to highlight superior capability and performance time enabling more "what-if" scenarios that can increase innovation, reduce errors and accelerate time to market."