Cost effective design procedure for laminated composite structure based on GA

We present a design procedure for symmetrically laminated composite structure with GA (Genetic Algorithm) and finite element analysis. Design variables to enhance the rigidity of the laminated composite structure are the number of laminae, the fiber orientation and the stacking sequence. Here, the fiber orientation is selected among some prepared angles from a manufacturing viewpoint. In a conventional design procedure based on GA, an objective function is calculated by costly finite element analysis. Consequently, it is quite time consuming and unpractical. In this paper, to minimize the displacement at a point in the structure, we define an approximated rigidity function of the laminated composites by means of weighing coefficients and transformed elastic moduli in the prepared angles. The weighing coefficients are initially determined using the distribution of principal stress in the structure under an applied load. Then, we can guess that reinforcements should be oriented according to the weighing coefficients with which the rigidity function takes maximum. To find the global optimum solution, the weighing coefficients are adjusted with as small a number of finite element analyses as possible. As a conclusion, the proposed procedure can reduce the computational cost remarkably compared to the conventional one, and can be applied practically to arbitrary shaped structures.