Nonlinear metamodeling and robust optimization in automotive design

We overview the methods for nonlinear metamodeling of a simulation database featuring continuous exploration of simulation results, tolerance prediction, sensitivity analysis, robust multiobjective optimization and rapid interpolation of bulky FEM data. Large scatter of simulation results, in crash-test simulations caused for example by buckling, is still a challenging issue for increasing predictability of simulation and accuracy of optimization results. For industrially relevant simulations with large scatter, novel stochastic methods are introduced and their efficiency is demonstrated for benchmark cases.