Viscosity of Iron Oxide Aluminosilicate Melts

Zhiming Yan,Ramana G. Reddy,Xuewei Lv,Zhengde Pang,Chenguang Bai
DOI: https://doi.org/10.1007/s11663-018-1460-z
2018-01-01
Metallurgical and Materials Transactions B
Abstract:A structure-based viscosity model has been developed for the CaO-MgO-FeO-SiO2-Al2O3 system and its subsystems. A critical analysis of the experimental data for iron oxide aluminosilicate systems has been conducted and used in the optimization of a model, which provides a link between the viscosity and the structures of the melts. The structural characteristics of aluminosilicate melt can be described by four types of oxygen: bridging oxygen \( {\text{O}}_{i}^{0} \), non-bridging oxygen \( {\text{O}}_{j}^{ - } \), free oxygen O2−, and excess bridging oxygen O*. The present model is capable of predicting the viscosities in the CaO-MgO-FeO-SiO2-Al2O3 system and its subsystems over wide ranges of composition and temperature above the liquidus within experimental uncertainties, and the average of relative errors for this model was found to be 19.94 pct. This model includes FeO-SiO2, CaO-SiO2-FeO, Al2O3-SiO2-FeO, MgO-SiO2-FeO, CaO-SiO2-MgO-FeO, CaO-SiO2-Al2O3-FeO, and CaO-MgO-SiO2-Al2O3-FeO. It was found that the effects of different metal cations on the viscosity can be determined, and the ability to reduce viscosity in the silicate melts follows the order: FeO > CaO > MgO. In the present model, the predicted viscosity decreased when CaO was initially initial added to the Al2O3-MgO-FeO-SiO2 system at a fixed SiO2 content because Ca2+ has the highest priority for charge compensation and FeO is the strongest modifier for this slag.
What problem does this paper attempt to address?