Numerical experiments for inverse analysis of material properties and size in functionally graded materials using the Artificial Bee Colony algorithm

H. Sakurai
2014 High Performance and Optimum Design of Structures and Materials   unpublished
Functionally graded materials (FGMs) possess properties that vary gradually and are highly heat resistant. When incorporating these FGMs into a structure, the required distribution of material properties must be determined and produced to specification. Thus far, an inverse analysis method has been used for estimating a distribution of a Young's modulus of FGMs. However, minimizing an objective function to estimate a material property using the Davidon-Fletcher-Powell method did not converge
more » ... did not converge under some initial conditions. Therefore, convergence of a solution depends on initial conditions. On the other hand, the Artificial Bee Colony (ABC) algorithm has drawn considerable interest in global optimization of a multimodal function. The objective of the present paper is to propose a method of numerical experimentation to conduct inverse analysis in FGMs with the ABC algorithm. A numerical experiment estimating both size and a graded index of an FGM beam that is based on measured stress values is presented. Next, determining a distribution of thermal conductivity in twodimensional FGMs using measured steady state temperatures is carried out. The results of the numerical experiments demonstrate the effectiveness of the proposed method.
doi:10.2495/hpsm140111 fatcat:4i4jzgllmbbvjhlqnom2lufqde