Scale-bridging atomistic simulations to density-based phase-field (DPF) modelling of grain boundaries
Quelle: BAM
Conventional phase-field approaches have successfully captured microstructural evolution for decades, yet they often lose the story told by individual atoms, particularly at the interface. For complex and technologically important defects such as grain boundaries (GBs), which govern a wide range of mechanical and functional properties in polycrystalline materials, this simplification can lead to significant inaccuracies in predicting interfacial properties and segregation-related phenomena. What if we could give these mesoscale models a “memory” of the underlying atomistic physics in predicting defects’ properties?
In this two-part series, we extend and enhance the density-based phase-field (DPF) framework to link and integrate with atomistic simulations. To benchmark, we analyse a large dataset of atomistically computed GB structures and corresponding energies for BCC-Fe and -Mo in the context of the DPF model.
In Part I of this series, we leverage atomistic simulations to explore the significance of the atomic density field parameter, and particularly, the average GB atomic density used in the DPF model. Here, we introduce a coarse-graining approach and show how discrete ground-state atomistic GB structures are converted into smooth, physically meaningful continuum atomic density fields. We found that, the GB excess free volume, which is an integral property of the GB, and a physically measurable property is directly related to the GB atomic density, that is the average atomic density at the GB central plane. A linear correlation between the excess free volume and GB density is revealed. In addition, clear correlations between the GB density and both GB energy and misorientation angle is also revealed.
Building on these results, in part II of this series, we further our systematic study by directly incorporate (atomistic) interatomic potentials into the DPF free energy functional. By employing GB energies calculated from atomistic simulations, the first order density gradient energy coefficient is accessed through coarse-graining. Furthermore, we uncover a direct correlation between the GB atomic density, and the GB energy contribution exits.
This new framework ensures that mesoscale simulations accurately reflect the physical reality of the material under study, providing more reliable and insightful results.
Linking atomistic and phase-field modeling of grain boundaries I: Coarse-graining atomistic structures
Theophilus Wallis, Sutatch Ratanaphan, Reza Darvishi Kamachali
Acta Materialia, 2025
Linking atomistic and phase-field modeling of grain boundaries II: Incorporating atomistic potentials into free energy functional
Theophilus Wallis, Reza Darvishi Kamachali
Acta Materialia, 2025