New insights into ternary geometrical models for material design

Geometrical model is an important part of the CALPHAD method and has been widely used for the prediction of physicochemical properties of multicomponent systems. However, most of the researches are focused on the comparison of the predicted results by various geometrical models, and very few studies have analyzed the underlying causes of these differences. In this paper, we will discuss the factors affecting the predicted results of geometrical models by analyzing the whole calculation procedures. When the experimental data is constant, it is found that the predicted results are mainly affected by two factors, which can be concluded as the fitting order of Redlich-Kister equation and the selection rules of three representative points. The influences of the two factors are analyzed in the viscosity prediction of Au-Cu-Ag ternary system. Geometrical models are essential parts of the CALPHAD method and have been widely used for the prediction of physicochemical properties of multicomponent systems. However, what are the underlying reasons for these differences of results predicted by various geometrical models? How can we make the prediction more accurate? Very few studies have analyzed this question which is of great help for material design. In this paper, we will discuss about it by analyzing the whole calculation procedures of geometrical models and new insights will be proposed. [Display omitted] •New insights into the calculation procedures of geometrical model are proposed.•The affecting factors are discussed in the viscosity calculation of Au-Cu-Ag alloy.•R-K fitting orders and selection rules of representative points are the main factors.•It provides new theoretical basis for the analysis of the prediction results.