Modeling the effect of grain clustering on the intergranular crack-propagation behaviour in polycrystalline materials

The effect of grain clustering on the intergranular crack-propagation behaviour of polycrystalline materials has been simulated using a novel integrated modeling approach combining Voronoi microstructure, Markov Chain theory and Monte Carlo simulations. The model takes into account the size, number, and spatial positions of clusters as well as the grain boundary character distribution (GBCD) of the cluster and non-cluster areas. Besides, the effect of grain boundary alignment with respect to the applied stress direction is considered. It has been found that the overall GBCD, often used as the input parameter to assess the crack-propagation extent is not a suitable mean for this purpose when the microstructures to be evaluated contain clustered regions. For the same overall GBCD, the crack-length, percolation frequency and percolation threshold have been found to be significantly different for the grain boundary network containing clustered zones than those of the non-clustered microstructure. In addition, the overall GBCD method was not found to be appropriate even when two clustered microstructures were compared. It was observed that, once the low-energy boundary frequency in the clustered areas reaches a value that can effectively blunt all the cracks that enter into them, the GBCD in the non-cluster region controls the cracking behaviour and percolation frequency. Besides, the spatial distribution of clusters has been found to have momentous effect on the crack-propagation and arrest behaviour. The percolation thresholds obtained for different cluster containing microstructures clearly demonstrate that this parameter strongly depends on the descriptions of the clusters present in the microstructure.