Buckling-constrained topology optimization using feature-driven optimization method

Structural stability has attracted increasing attention in topology optimization because of the buckling effect under compression load. In this work, the feature-driven optimization method is developed for structural topology optimization involving buckling constraints. First, the finite cell method is extended to linear buckling analysis. A stress relaxation strategy is proposed to effectively remove the pseudo-buckling modes from low-density regions in combination with the adaptive quadtree/octree-based integral scheme for the geometric stiffness matrix. Then, the feature-based topology variation model is constructed to drive topology optimization. The boundary integral scheme developed in our previous work is adapted to the sensitivity analysis of the buckling load factor, which effectively avoids the time-consuming domain integral. Finally, the influences of solid and void feature definition, initial layout, number of design features, and minimum feature size are systematically studied. The advantages of the present method are demonstrated with the help of numerical examples.