Extension And Application Of Nonequilibrium Solidification Models For Undercooled Alloy Melts

Nonequilibrium solidification in undercooled alloy melts has fuelled increasing attention with development of solidification technology. Microstructures determine mechanical and physical properties of materials. In order to control solidification process effectively and obtain materials with ideal microstructure, establishing solidification models which much more approach to reality and have more precise predictions and broader application range is very important.As a series of research, the planar interface response function, the morphological stability model and the free dendritic growth model, which are based on sharp interface assumption, and the solute trapping model, the solidification model with planar interface and the solute field profile during isothermal dendritic solidification, which adopt a diffuse interface, were developed. These extended models take into account the relaxation effect of nonequilibrium solute diffusion in bulk liquid and are applicable to non-dilute alloys. All of these models use Gibbs free energy functions of solid and liquid phases optimized by CALculation of PHAse Diagram method (CALPHAD) and are thus more general.Applying the planar interface response function with sharp interface assumption and the solute trapping model and the planar interface solidification model, which adopt a diffuse interface, to Si-9at.%As alloy, applying the morphological stability model to Si-Sn system and applying the free dendritic growth model to Cu70Ni30 alloy, satisfactory agreements between model predictions and the available experimental data were obtained.Comparative analysis of present model and existing typical models indicates that local equilibrium solute diffusion assumption at interface or in bulk liquid, assumptions of linear solidus and liquidus curves, Baker-Cahn relationship and dilute alloy assumption restrict application range of models in a certain extent, and may lead to remarkable deviations of model predictions from reality. Furthermore, the present model is also more precise and improved for dilute alloy.There is still a divergence of opinion as to whether the solute drag effect should be considered during solidification. The present model introduces an adjustable parameter to unify both forms with and without solute drag, and thus can consider so-called partial solute drag effect. After numerical calculation and comparison with the available experimental data for the present model and Aziz et al.'s continuous growth model (CGM), it is first indicated that the versions without solute drag of both models fails badly at fitting the data. In contrast, the present model (the generalized Hillert-Sundman model) provides a satisfactory agreement with the experimental data. Second, the CGM with partial solute drag also reproduces the data. Comparative analysis indicates that the CGM with partial drag can be regarded as the simplified version with sharp interface of the Hillert-Sundman model. Combined these results with the relevant conclusions from shrap interface models, it is finally demonstrated that considering the solute drag effect is required during solidification.