Blended Composite Optimization combining Stacking Sequence Tables and a Modified Shepard’s Method

Abstract

This article presents a computationally-efficient optimization tool for stacking sequence design of blended composite structures. In this tool, blended laminates are designed using stacking sequence tables (SST), coupled with a suitable genetic algorithm (GA). The SST approach guarantees complete-blending, ensuring manufacturability of the optimized design. The concept of successive structural approximations is implemented to improve computational efficiency. Optimizations are carried out on the approximations of responses rather than actual responses themselves, thus reducing the number of expensive design analyses. A recently-proposed modified Shepard's interpolation enriches the quality of the approximations used, by constructing multi-point approximations using the elite designs of the previous iterations. The generality and efficiency of the algorithm is further improved by directly approximating panel loads, thus enabling implementation of a wide range of stress-based design criteria. An analytical multi-panel blended composite problem is presented as an application. The results show that completely blended and feasible stacking sequence designs can be obtained, having its structural performance close to the theoretical continuous optimum itself.