Date of Award
Master of Applied Science (MASc)
Materials Science and Engineering
Jeffrey J. Hoyt
Crystal-melt interface properties and their associated anisotropies playa crucial role during solidification in controlling the nucleation, crystallization rate and growth morphology. There are two solid-liquid interfacial (SLI) properties affecting the dendritic microstructures that form in the crystallization process and the SLI properties are interfacial free energy (γ) and kinetic coefficient (μ) . In this research work, atomic scale simulation techniques, such as Monte Carlo (MC) and Molecular Dynamics (MD), have been applied to compute the crystal-melt interface properties and their anisotropies of the AI-Mg system.
An inter-atomic potential is utilized for describing the pair interactions of binary AI-Mg system during atomistic simulations. Actually the potential was developed particularly for the simulation of solid-liquid interface properties of AI-Mg alloys. Optimization of the potential is conducted by determining the equilibrium phase diagram employing Monte Carlo (MC) simulation techniques and comparing with the experimental results. A method is discussed for fitting the potential into phase diagram by varying the data for liquid solution energies. A good agreement of the AI-rich side of AI-Mg phase diagram, determined from this potential, is found with the experimental phase diagram. The inter-atomic potential is also optimized by comparing the liquid enthalpy of mixing of AI-Mg alloys with that of experiments.
The crystal-melt interfacial energy (γ) and its anisotropies in AI-Mg binary alloys are computed utilizing a combination of MC and MD simulations in association with the analysis of capillary fluctuation method (CFM). The orientation averaged surface energy γo is observed to increase with increasing temperature which is consistent with other computational results of Lennard Jonnes (LJ) and Hard Sphere (HS) system. The anisotropy of y is found to follow the ordering of γ100> γ110> γ111. Superimposition of the y anisotropy parameters on the orientation selection map, proposed by Haximali et al., predicts the primary dendrite growth in <100> direction for pure Al and the growth is examined to be stabilized in the same orientation with the addition of Mg atoms to Al i.e. in the concentrated alloys as well.
Kinetic coefficient (μ) of pure Al is determined from the free solidification method utilizing Molecular Dynamics (MD) simulations employing multiple thermostats in the system which avoids the underestimation of μ due to slow dissipation of the generated latent heat at the solid-liquid interface. Kinetic coefficient is also extracted from equilibrium fluctuation analysis with a correction due to the contribution of thermally controlled interfacial kinetics. μ is computed for both (100) and (110) orientations of the crystal-melt interfaces and the values from both techniques are found to be equivalent. The magnitudes of interface mobility for pure Al is determined as μ100 = 163 cm/s/K and μ110 = 129 cm/s/K.
Rahman, Jahidur, "Atomistic Simulations for computing solid liquid interface properties of the Al-Mg system" (2009). Open Access Dissertations and Theses. Paper 4315.
McMaster University Library