摘要

The properties of material strongly depend on the microstructure, and the development of microstructure is closely related to the phase transition with the temperature-dependent spatial correlation. To consider more realistic microstructures, we have proposed an efficient and simple algorithm for generating the spatially-correlated random field, which is obtained by the weighted average of random fields without spatial correlation according to the spatially-correlated length and anisotropy parameter. By using a mesoscale finite element model with the microstructures generated by our algorithm, an application study on the effective elastic behavior of Al2O3-NiAl composite materials is given. Our numerical results are in agreement with the experimental measurements. The proposed method is general and robust, which can be extended to the multi-phase materials.