LAUSR

Abstract A robust design optimization algorithm is proposed for variable angle tow composite structures in the presence of uncertainties in the constituent material properties and applied loads. The proposed algorithm uses a stochastic perturbation method to propagate these uncertainties through to the simulated structural response, measured in terms of buckling load. The expected value plus a selected number of standard deviations of the response in the form of a bi-criteria problem. To describe the curvilinear fibres, two types of fibre path function, namely linear- and nonlinear-variation formulae, are adopted to illustrate the proposed methodology. A comparison between the resulting robust designs and deterministic designs is made, and changes to the final designs of fibre tow paths arising from the inclusion of uncertainty are discussed. It is shown that the robust designs out-perform the deterministic designs under real-world situations that include uncertainties.