SLAG CORROSION OF STEELMAKING REFRACTORIES – THE USE OF THERMODYNAMIC SIMULATION MODELS

Reference Presenter Authors
(Institution)
Abstract
14-067 Analía Gladys Tomba Martinez Tomba Martinez, A.G.(Instituto de investigación en Ciencia y Tecnología de Materiales); Refractories are ceramics for high temperature by excellence. In the case of steelmaking refractories, thermal conditions vary from room temperature up to 1700°C depending on the vessel and the specific operation. At high temperature, liquid steelmaking slags usually attack the refractories in contact with this melt, by penetration (through open pores) and dissolution. Ideally, the slag-refractory system will tend to the thermodynamic equilibrium: produced the saturation of the liquid, solid phases will precipitate and then the chemical attack will stop. Slag corrosion frequently limits the refractory lifetime. However, its approach has difficulties related to the multiple variables determining the material behaviour. The current practice consists in the experimental evaluation of corrosion by different lab tests which reproduce more or less closely the service conditions. In this context, the thermodynamic simulation of the slag-refractory system (using commercial softwares) has demonstrated to be a powerful tool to explain and predict the mineralogical composition at the end of the process and even corrosion wear. Moreover, it can be used to anticipate the final equilibrium state for different experimental conditions, such as changes in temperature, slag or refractory compositions, for instance, thus reducing the number of expensive and time-consuming lab tests. In this talk, examples of the successful use of thermodynamic simulation models in the slag corrosion study of MgO.Al2O3 spinel forming refractories at high temperatures are presented. MgO-Al2O3 castables of different compositions and Al2O3-MgO-C commercial bricks of different qualities were tested by cup test, in contact with industrial slags. Phases formed in slag-refractory interfaces were identified by SEM/EDS and compared with those predicted by the simulation. Moreover, quantitative indicators of corrosion such as penetration and/or wear could be also linked to a simulation model parameter.
<< Back
Copyright © 2019 Metallum. All rights reserved.
Site produced by: SITESP.NET