Speaker
Description
The phase-field method has provided a powerful tool for studying grain growth over the last decades. Thanks to its diffuse character, it has been efficiently applied to the modelling of microstructure evolutions involving complicated topological transformations, without explicitly tracking the interface positions. However, in order to numerically resolve the diffuse interfaces, an interface width sufficiently greater than the grid spacing is mandatory. Depending on the required accuracy, the interfaces are resolved with 6 to 10 grid points. Hence, for representative grain growth simulations, where microstructures of a statistically significant number of grains are required, simulations must be carried out on a very large numerical grid, making the method computationally very expensive.
Recently, the S-PFM approach, developed by Finel et al [Phys.Rev.Lett. 121(2), (2018) 025501], has provided a new inherently discrete formulation, where interfaces can be resolved with only one grid point, thus drastically improving the numerical performances of the method. Such numerical improvement was fully exploited by Dimokrati et al [Acta Materialia, 201, (2020), 147], where the S-PFM approach was successively extended to a multi-phase-field model and applied to ideal grain growth.
In the present work, the S-PFM multi-phase field model for grain growth is further extended to the case of non-uniform grain boundary properties where interface energies and mobilities depend on the misorientation between adjacent grains. We
show that the S-PFM approach is perfectly adapted for simulations with non-uniform interfaces, since both the energy and the mobility can be adjusted independently of the interface width. The model is then used to investigate abnormal grain growth in highly textured materials, where we demonstrate that, for grain boundaries with low non-uniformity degrees, a combined effect of both energy and mobility advantages is sufficient to trigger abnormal growth.
| Speaker Country | Morocco |
|---|