Abstract
In this study, we deal with an optimization problem of material compositions for a hollow sphere of functionally graded material (FGM) under heated and cooled process. Using the analytical procedure of a laminated hollow sphere model, the analytical solution for the FGM hollow sphere is derived approximately. As the optimization problem of minimizing the thermal stress distribution, the numerical calculations for functionally graded material are carried out making use of hierarchical neural network, and the optimum material composition is determined under arbitrary radius ratio and arbitrary heat transfer coefficient. Furthermore, the results obtained by neural network and ordinary non-linear programming are compared.
