Abstract:
A study on the optimization of shell buckling, incorporating Karhunen-Loève-based geometrical imperfections. It addresses the challenges of ignoring geometrical and material imperfections in deterministic designs by modeling them as random fields. The work uses a stochastic analysis, specifically a Monte Carlo analysis, within the optimization process to quantify the variation in non-linear buckling. The study utilizes LS-DYNA for non-linear dynamic simulations of a shell with cut-outs, where a random field of geometrical imperfections is superimposed. The optimization set-up considers thickness and hole area as design variables and includes constraints on peak force and internal energy. The optimization process is fully automated using LS-OPT in Metamodel mode to call LS-OPT in Monte Carlo Analysis mode. Three optimization cases are explored: deterministic, stochastic, and robust, with the objective of minimizing mass or the coefficient of variation of the peak force. The findings indicate that including stochastic and robustness effects is feasible, though expensive, and that robust optimization leads to a heavier design while improving the coefficient of variation (COV) of the peak force.
