A numerical model for dendritic growth in binary alloys

Srinivasan Raghavan, Purdue University


A Cellular Automaton (CA) model to track the solid-liquid interface movement combined with finite volume computation of solute diffusion has been developed to simulate dendritic structures in binary alloys. One of the significant problems in the previous CA formulations was the presence of artificial anisotropy in growth kinetics at zero and 45° introduced by a square CA grid. A new technique to track the interface movement is proposed which makes it easier to model dendritic growth in different crystallographic orientations and also model coarsening and remelting of arms. The model stability with respect to the numerical parameters (grid size, Δx and time step, Δt) for varying operating conditions is also studied. An operating window process window in Δt and Δx was identified, in which the model gives a consistent set of results in dendrite tip radius and tip undercooling. Finally, model validation with experimental results from literature and analytical results for both directional and equiaxed growth conditions are presented.




Johnson, Purdue University.

Subject Area

Materials science|Metallurgy

Off-Campus Purdue Users:
To access this dissertation, please log in to our
proxy server