F Vakili Tahami - Associate professor, University of Tabriz
R Hassannejad - Assistant professor, University of Tabriz
A Hossainzadeh Foroughi - M.Sc. student, University of Tabriz
A Mohammad Alizadeh Fard - M.Sc. student, University of Tabriz

The performance of FGMs is dependent on optimal utilization of their material constituents. Although there are several research studies in optimizing FGM components based on their elastic behavior, there are few references which address the optimum design of FGM components based on their creep behavior. In this study, bi-objective optimization of material distribution through the thickness of FGM sphere is presented considering the stress redistribution due to creep. The two objective functions are factor of safety and weight of the spherical vessel. Two methods of Sum of Weighted Cost Functions (SWCF) and Nondominated Sorted Genetic Algorithm (NSGA) methods are used and their results are compared. Since the selected objective functions are conflicting, Pareto optimal solutions for the optimum distribution of materials are obtained. The results show that stress redistribution due to the creep has major role and it has to be taken into account. Also, it is shown that for creep solutions, the objective space is non-convex and NSGA method is more efficient in comparison with SWCF.

FGM, bi-objective optimization, Pareto optimal, SWCF, NSGA, creep behavior