Thermodynamic Modeling of Isotherms of Oxygen Solubility in Liquid Metal of the Fe–Mg–Al–O System

Abstract

In this paper, we studied the interaction of magnesium and aluminum dissolved in liquid iron with oxygen, which is an important problem for choosing the optimal parameters of steel refining and casting. The relevance of the study is determined by the possibility and conditions for the formation of adverse refractory particles of magnesium oxide and magnesian spinel in the melt. We performed the thermodynamic modeling of phase equilibria realized in the liquid metal of the Fe–Mg–O, Fe–Al–O, and Fe–Mg–Al–O systems in the temperature range of 1550–1650°C. The calculation was carried out using a technique for constructing the surface of the component solubility in a metal, which associates quantitative changes in the liquid metal composition with changes in the composition of the interaction products of molten metal components. The modeling method was based both on the use of equilibrium constants of reactions occurring between the components of the studied systems in the selected temperature range and on the account of the values ​​of the first-order interaction parameters (according to Wagner) of elements in liquid iron. To simulate the activities of an oxide melt conjugated with a metal melt, the approximation of the theory of subregular ionic solutions was used. The approximation of the theory of regular ionic solutions was used to model the activity of a solid solution of oxides while the theory of perfect ionic solutions was used to model the activity of a solid solution of spinels. The isotherms of oxygen solubility in the liquid metal of the Fe–Mg–O, Fe–Al–O, and Fe–Mg–Al–O systems were plotted, and the regions of thermodynamic stability of oxide phases conjugated with the metal melt were determined. In particular, for the Fe–Mg–Al–O system, the region of liquid metal compositions, in equilibrium with which there will be a solid solution of spinels |FeAl₂O₄, MgAl₂O₄|_ss, was determined. The obtained results of thermodynamic modeling are compared with experimental data.