Bi-directional evolutionary topology optimization based on stress minimization under design-dependent surface loads

Abstract

This article proposes an approach for structural topology optimization based on stress, under design-dependent surface loads. The algorithm has been developed by adapting the stress minimization version of the bi-directional evolutionary structural optimization method, with the inclusion of surface loads. The P-norm of von Mises stress is used as an aggregate function and volume constraints are employed. The stresses have been obtained using the finite element method applied to the continuum bi-dimensional elastic structures. The sensitivity analysis has been carried out by the adjoint method, including design-dependent surface loads. Three examples have been investigated. The first is the piston head problem, a benchmark for design-dependent pressure loads. In this example, the algorithm convergence is demonstrated. The other two are benchmarks for stress-based problems: the simply supported beam and the L-bracket. The proposed method is shown to be effective in reducing the maximum stress, generating easily manufacturable topologies.