Abstract:
This paper describes the optimization of shell buckling in the presence of geometric imperfections, which are defined using Karhunen-Loève expansions. The methodology, implemented in LS-DYNA, is applied to the collapse behavior of thin-walled cylindrical shells under axial compression. Two objective quantities are used in the design: minimizing the average peak normal force and the average internal energy. The study demonstrates that LS-OPT can be used with a Monte Carlo simulation to generate an experimental design and perform the optimization. The results show that while minimizing the average internal energy, the optimal design has a substantially increased hole size and increased shell thickness, resulting in a heavier design but with more robust behavior within the design constraints.
