Stacking sequence optimizations using GA with zoomed response surface on lamination parameters

Abstract

Stacking sequence optimizations for composite laminates are combinatorial optimization problems often handled by genetic algorithms (GA). Unfortunately, GAs often require large computational resources for individual evaluation. To overcome this problem, this paper proposes use of lamination parameters in an optimization process approach with GA. The optimization process requires precise approximation of design space using response surfaces to obtain the true optimal result. However, stacking sequence optimizations are discrete optimization problems and it is difficult to judge optimality of an obtained result on the lamination parameter space. For zooming response surfaces, we require an optimality check method. In this study, laminates adjacent to the obtained optimal result are introduced on the lamination parameter space and an optimality-check-test method is proposed. The test method is applied to a buckling load maximization problem of a laminated composite plate; its effectiveness is confirmed by example.

Keywords:

• GENETIC ALGORITHM
• RESPONSE SURFACE
• DESIGN OF EXPERIMENT
• COMPOSITES
• OPTIMIZATION