A grain parallelized structural MDO engine which includes shape and topology variables

Generality and flexibility of use in structural multidisciplinary design optimization problems is enhanced by consideration of multiple objectives and/or shape and topological parameters as design variables. Subprocesses such as interfacing between disciplines, model parameterization and model reduction, etc. are controlled via an appropriate workflow and data management system. In this paper an application management approach and its utilization for structural shape optimization is described. Furthermore the use of evolutionary algorithms and response surface approximations is discussed. Additionally grain parallelization of the solution process by carrying out analyses in parallel on high performance PCs is utilized.