A Study on Stacking Sequence Design of Composite Fan Blades Using Multi-level Optimization

Abstract

In this paper, the multi-level optimization method is proposed to design the optimal stacking sequence of composite fan blade more efficiently. The fan blade model used is a laminate with inconsistent stiffness because the thickness is not uniform throughout the model. This feature complicates the optimization problem and makes it difficult to apply optimal design algorithms that use only continuous or discrete variables. For this reason, we conducted the literature survey about combining the optimal design algorithm with continuous and discontinuous variables and developed optimization algorithm for the composite fan blade. Moreover, the design space of discrete variables is reduced through the optimal design of continuous variables. In the first-level, response surface method is used, with the design variables being the volume fraction of plies of the predefined angles of [0°/±45°/90°]. The optimization problem in the first-level aims at minimizing the tip displacement of the fan blade and constraints on other structural responses are imposed. At the second-level, the permutation genetic algorithm is used with the fiber orientation of each ply as design variables. In this level, we perform the optimal design of the stacking sequence using the results determined by prior level, while the composite design rule and compliance constraints are applied.