Optimum Design of Composite Laminated Plate Using Genetic Algorithm and RSM


The paper is focused on the application of the response surface method (RSM)in structural optimization. Applications of the response surface method in thedesign of composite laminated plate have been discussed. The response surfacemethod consists of two stages. In the first stage, the random variables is selected inorder to perform a deterministic computer simulation (finite element solution) inthe sample points. In the second stage, the approximation of the function (whichrepresent the buckling load) is performed in order to obtain response surfacesusing PDS module included in the ANSYS Program. This response surface isincorporated into a genetic algorithm (GAs) for optimization of random inputvariables to obtain maximum buckling load for composite laminated platesubjected to both mechanical and thermal loading. GAs are stochastic optimizationalgorithms based on natural selection and genetics. In contrast to traditionalgradient-based methods, GAs work on populations of solutions which evolvetypically over hundreds of generations. Four and five different variableformulations are examined. It was found that for SSSS boundary condition and twolayer laminate the optimum values of buckling load for all thermal loading occur atq1=33o, q2=59o, t1=1.23 mm and t2= 1.25 mm, also it can observe that thesignificant random variable are t1 and t3 (in the case of five independent variables)since the value of buckling load effected with t1 and t3 more than for t2.