Evolutionary Identification and Optimization of Composite Structures

Abstract

Composite materials, especially composite laminates, play a significant role in the modern industry. Laminate is a material built by joining two materials and it usually consists of two phases: the matrix and the reinforcement. The laminate is typically build of many plies (laminas) having different ply angles. Laminates are popular due to two main reasons: i) the high weight-strength ratio (in comparison with the conventional materials); ii) the possibility to tailor the material properties to the designer requirements by manipulating several parameters like: components material, stacking sequence, fibres orientation or layer thickness [1]. If laminas are composed of the different materials the laminate is called a hybrid one.Two aims of the present paper are: i) to identify the material constants of the laminates; ii) to find the optimum stacking sequence of the laminates. The standard laminates as well as the hybrid ones are considered [2]. Different optimization criteria connected with the modal analysis and free vibrations are taken into account.To solve global optimization tasks and discrete optimization tasks as well as to avoid difficulties with the objective function gradient computation, the evolutionary algorithm (EA) is employed as the optimization procedure. To reduce the computation time, the distributed version of the evolutionary algorithm is used [3].The finite element method (FEM) professional software package with the laminate modeller is used to solve a direct eigenfrequency problem for the laminate plates. The numerical examples presenting the efficiency of the proposed attitude are attached.As it can be seen from the numerical examples, the proposed identification and optimization method gives positive results. Consequently, this method can be applied to different laminated structures in order to identify the material constants or to find the desired laminate properties for a given criteria.