Weight optimization of large vessel

The aim of optimization was a weight reduction of the vessel. ANSYS was used for the FEM calculation. The optimization was performed with optiSLang.

The design space was described with approximately 30.000 optimization variables (sheet metal thickness) and 30.000 constraints (deformation and stress restrictions). A self-regulating evolutionary strategy was used as a stochastic search strategy for optimization. After 200 design evaluations, a significant weight reduction was obtained.

After 3.000 design evaluations, the genetic algorithm converged to its best possible design improvement.