Multiscale evolutionary optimization of functionally graded porous materials

This paper deals with the optimization of the microstructure of selected porous functionally graded materials (FGMs). Porous materials are inhomogeneous structural materials whose properties at macroscale strongly depend on their microstructure. The designing of porous materials as FGMs allows obtaining structures well-tailored to their operation conditions, which require the use of optimization methods. As it is assumed that multiple contradictory optimization criteria are defined, the multiobjective optimization is performed. Multiobjective Genetic Algorithm (MOGA) included in ANSYS Workbench software is employed in the work as the optimization tool. The application of the global optimization algorithms allows avoiding problems with multimodal objective functions and the calculation of the objective function gradient algorithm. Microstructural material parameters are optimization design variables. Numerical homogenization with the use of representative volume element (RVE) is applied to obtain equivalent homogeneous properties of inhomogeneous structures. The finite element method software ANSYS Workbench is used to solve the boundary-value problem in both scales. The use of FEM software and optimization algorithm included in a single software package significantly reduces the time needed to exchange data between independent systems. The numerical example presenting the optimization results in the form of Pareto frontiers of non-dominated solutions shows an efficiency of the proposed attitude.