Uncertainty Quantification Based Multi- Objective Optimization for Crashworthiness Design

This paper presents a methodology for uncertainty quantification based multi-objective optimization of automotive body components under impact scenario. Conflicting design requirements arise as one tries, for example, to minimize structural mass while maximizing energy absorption of an automotive rail section under structural and occupant safety related performance measure constraints. Uncertainty quantification is performed in broadly using two methods: reliability based approach and robustness based approach. A design is called reliable if it meets all performance constraints in presence of uncertainty and robust if it is insensitive to such uncertainty. Mainly aleatory or stochastic uncertainty is addressed in this paper. Deterministic, reliability-based and robustness based multi-objective optimization solutions are compared. A genetic algorithm based multi-objective optimization software GDOT, developed in-house, is used to come-up with an optimal pareto-front in all cases. The technique employed here treats multiple objective functions separately without combining them in any form. A decision-making criterion is subsequently invoked to select the “best” subset of solutions from the obtained non-dominated Pareto optimal solutions. The pareto optimal set obtained in case are compared and contrasted and observations made comparing reliability based approach vis-à-vis robustness based approach.