Robust topology optimization under multiple independent uncertainties of loading positions

Abstract

The topology optimization problem of a continuum structure is further investigated under the independent position uncertainties of multiple external loads, which are now described with an interval vector of uncertain-but-bounded variables. In this study, the structural compliance is formulated with the quadratic Taylor series expansion of multiple loading positions. As a result, the objective gradient information to the topological variables can be evaluated efficiently upon an explicit quadratic expression as the loads deviate from their ideal application points. Based on the minimum (largest absolute) value of design sensitivities, which corresponds to the most sensitive compliance to the load position variations, a two-level optimization algorithm within the non-probabilistic approach is developed upon a gradient-based optimization method. The proposed framework is then performed to achieve the robust optimal configurations of four benchmark examples, and the final designs are compared comprehensively with the traditional topology optimizations under the loading point fixation. It will be observed that the present methodology can provide a remarkably different structural layout with the auxiliary components in the design domain to counteract the load position uncertainties. The numerical results also show that the present robust topology optimization can effectively prevent the structural performance from a noticeable deterioration than the deterministic optimization in the presence of load position disturbances.

Topology Optimisation in Structural Steel Design for Additive Manufacturing

“…Among recent advances, one can highlight the efforts to account for uncertainties in the loading intensity and position with the method by Wang and Gao [460], and the compliancefunction-shape-oriented approach by Csébfalvi [461] and precursory work [462], the generalized material interpolation scheme by Chan et al [463], as well as the RBTO approach by Nishino and Kato [464].…”