Molecular-dynamics simulations of molten Ni-based superalloys

Fundamental parameters of liquid metal alloys are calculated using a first principles approach, based on Density Functional Theory (DFT), with the goal of informing models of defect formation during solidification processing. Ab-initio molecular dynamic simulations (AIMD) are applied to liquid metal alloys ranging from simple metals to Rene-N4 and CMSX-4 to predict molar volumes (density), diffusion rates and local ordering. These include elemental Ni, Ni-5.4X, Ni-20X, and Ni-10A1-2.8X (X= W, Re, and Ta) (at%) alloys at 1750 and 1830 K. Calculated kinetics and the atomic distribution in the liquids indicate that simulations of 500 atoms run for approximately 7 ps converge the time-averaged properties, including molar volume. Overall diffusion rates and molar volumes are in good agreement with available experimental measurements, though the AIMD predictions appear to systematically underestimate thermal expansion. The method is then used to predict density inversion in three additional alloys, Ni-14A1-3W, CMSX-4 and Rene-N4 for temperatures and compositions expected in the mushy zone during directional solidification. Density inversion is predicted for these three alloys based on density and density contrast in the mushy zone, the prediction and ranking of the effect is consistent with previous studies. Predictions for molar volume are compared with the recent parameterization of molar volumes using extensive measurements of binary alloys. AIMD calculations validate the underlying assumptions of these models and illustrate the bounds in alloy chemistry for applying such techniques.