A hybrid Potts Monte Carlo model with realistic time scaling to simulate prior austenite structures in additively manufactured high-strength steel

Computationally efficient models are required to understand the relationship between additive manufacturing (AM) processes, such as directed energy deposition (DED), and the microstructures they form. Such models provide a means of mapping the large design spaces that AM provides. The current work presents a mesoscale hybrid Potts Monte Carlo (PMC) and cellular automata (CA) model that builds on recent methods in the literature to simulate the DED processing of a martensitic high-strength steel. The model simulates the evolution of the prior austenite grain structure in this material, capturing multiple physical phenomena occurring near the solidification front. The formulation of the current hybrid model implements a method of scaling the PMC time step relative to the CA algorithm that ensures that grain coarsening and solidification processes occur in realistic proportion to one another. The model is validated by comparison to the size and morphology of the austenite structure in a DED printed thin wall, revealed by electron backscatter diffraction imaging and the computational reconstruction of prior austenite. Model predictions of austenite grain size and aspect ratio agree very well with experimental observations. Quantification of the model’s behavior demonstrates the relative roles played by different modes of grain growth and nucleation during DED processing of this material. The hybridization of PMC and CA methods allows the model to capture relevant growth behaviors in a computationally efficient manner, spanning volumes large enough to capture the effect of changing processing conditions on microstructural features. [Display omitted] •Efficient mesoscale models can capture effects of scan strategy on AM microstructure.•Proper scaling of time for simulated thermally activated processes is essential.•Hybrid model balances epitaxial growth, coarsening, and nucleation in DED simulation.•Varied conditions at DED solidification front promote mixed growth behaviors.•Prior austenite grain morphology in DED processed martensitic steel can be predicted.