Irene Tartaruga, Jonathan E. Cooper, Mark H. Lowenberg, Yves Lemmens
DOI Number: N/A
Conference number: IFASD-2017-193
Optimization strategies are widely used in the design process across all engineering fields. The consideration of uncertainties increases the difficulties in determining an optimum nominal design, and this is particularly true for systems whose numerical model is computationally expensive to simulate. Robust design optimization (RDO) and reliability-based design optimization (RBDO) are the two techniques commonly adopted to deal with the optimization of performance of systems under uncertainty. In the presence of a very computationally expensive numerical model, nonlinear behaviour and multiobjective problems, the traditional RDO and RBDO are not always suitable because of two main problems: the prohibitive computational cost and the neglect of higher-order moments common for the RDO and RBDO techniques. Thus it is advantageous to develop a method that can assure reliability and robustness in the designs of complex engineering structures. In this paper, novel optimization techniques based upon an evolutionary approach are presented and demonstrated considering the design of landing gear systems to eliminate ‘shimmy’, an undesirable limit cycle oscillation.