Microstructure based fatigue life prediction of polycrystalline materials using SFEM and CDM

Abstract

Accurate fatigue life prediction of polycrystalline materials is crucial for many engineering applications. In polycrystalline materials, a significant portion of life is spent in the crack nucleation phase at the microstructural scale. Hence, the total fatigue life shows high sensitivity to the local microstructure. To predict fatigue life accurately, the microstructure models of polycrystalline material i.e., titanium alloy are virtually generated with the help of the Voronoi tessellation technique. These models incorporate critical microstructural features such as grain size, grain shape, and the volume fraction of different phases within the material. To efficiently predict microstructure sensitive fatigue life, the smooth finite element method (SFEM) is coupled with continuum damage mechanics (CDM). The SFEM provides flexibility in the meshing of complex microstructure geometries as it alleviates the need to use only triangular and quadrilateral elements. Moreover, there is no need of isoparametric mapping and explicit form of shape function derivatives in SFEM, hence it requires less computation time. To obtain the fatigue life (in number of cycles), jump in cycles algorithm is implemented using SFEM-CDM. The numerical results of fatigue life data obtained from simulations are compared with experimental data, which reveals the validity of the present approach. This approach is useful to find out the scatter in fatigue life data of polycrystalline materials along with the source of scatter.