Modeling of Grain Refinement and Nucleation Behavior of Mg-4Y-0.5Zr (wt.%) Alloy via Cellular Automaton Model

Grain refinement during solidification of Mg-4Y-0.5Zr (wt.%) alloy was both experimentally and numerically studied. A two-dimensional cellular automaton model is adopted. A LGK (Lipton–Glicksman–Kurz) analytical model is solved to predict the variation of dendrite tip velocity with undercooling under the conditions with and without Zr. With addition of 0.5 Zr, the growth velocity of dendrite tip is approximately one sixth of that without Zr. Experimental characterizations showed that the addition of Zr resulted in a significant grain refinement. The effects of cooling rate, initial nucleation density and Zr addition were numerically examined. A high cooling rate is important to the grain refinement, because it effectively increases the degree of effective undercooling and nullifies the effect of latent heat on stopping the heterogeneous nucleation. For Mg-4Y alloy, solute suppressed nucleation is magnified with the increase in initial nuclei density. Since a large portion of nuclei is depressed due to the fast overlap of solute diffusion layers, the grain refinement is not as desirable as expected. The mechanism of grain refinement with addition of Zr is that a reduced growth rate is accompanied with a slowed release rate of latent heat. Without much temperature interference from the neighboring developing dendrites, more nuclei can achieve enough undercooling for nucleation and become grains. A high cooling rate promotes multi-step nucleation and significantly refines the grain structure, which cannot be achieved at a low cooling rate.