MULTIPLE DESIGNS APPROACH FOR CONTINUUM TOPOLOGY OPTIMIZATION

Topology optimization can provide creative conceptual designs for structure of industry product in the preliminary design stage. However, the traditional topology optimization approaches focus on searching for one optimal solution which may be invalid due to the refinements of models or the additional design requirements. This paper presents the multiple designs approach (MDA) to get two or more diverse topology designs simultaneously in conceptual design, which can reduce the risk of lacking full knowledge of the designs by providing multiple designs. This paper gives general optimization model formulations for MDA in which weighting function is used as the objective function to evaluate the performances of multiple solutions and diversity measure is used as constraint to make difference between configurations. A kind of diversity measure is presented in the paper and its physical significance and features are also discussed at the same time. This paper solves two compliance minimization problems based on variable density method as examples and gives detailed optimization model formulations and sensitivity analysis. The parameters of objective function and constraints in MDA and latent performances of different solutions are also discussed in the examples. The results show that MDA could propose multiple diverse designs for detailed design stage.