Abstract In this study, topology optimization of a thin-walled square tube subjected to an axial crushing load was carried out to enhance its crashworthiness performance, utilizing the modified Bidirectional Evolutionary Structural Optimization (BESO) procedure. In this way, by systematic searching, topology optimization finds the best material layout for the structure to satisfy the objective function. In each iteration, a new pattern of material-void through the tube walls can trigger various folding mechanisms. Thus, the energy absorption history trend might not be smooth and the ultimate pattern proposed by the algorithm is less likely to be the optimal one. With an emphasis on the BESO hypothesis regarding direct removal and addition of elements, the modified BESO algorithm excavated the optimal design in two steps. First, material-void patterns were produced, considering specific energy absorption as the element efficiency criterion. Then, the pattern with the least material usage which satisfies an energy absorption constraint was selected as the optimal design. Such an optimum layout also boosts the tube performance respecting peak crushing force, since the embedded voids through the structure serve as a crush initiator. The optimization code was developed in MATLAB, which was automatically integrated with ABAQUS software performing the nonlinear crushing analysis. To validate the finite element model, an experimental investigation was carried out on a simple square tube. Having performed the topology optimization and some fine-tuning process, the topologically optimized tube was constructed and experimentally tested. Results showed 16.21% weight reduction enriched with 36.57% enhancement in the peak crushing force by sacrificing less than 3.5% energy absorption, which represents substantial improvements from a crashworthiness viewpoint. The obtained interesting results showed the high potential of the BESO algorithm for the design of thin-walled structures under large deformation and out-of-plane buckling.
Read full abstract