Quantitative Evaluation of Slag Corrosion on MgO-C Refractory by Experimental and Numerical Simulation

To predict slag corrosion numerically, a transient 3D mathematical model that considers the fluid flow, heat, and mass transfer was developed. A dynamic corrosion experiment using the rotating immersion approach was carried out to assess the overall corrosion activation of the MgO-C refractory. An expression for the corrosion rate was determined based on the wall shear stress, slag viscosity, sample size, overall corrosion activation, and difference in MgO content. A greater corrosion rate was observed at the bottom corner of the refractory sample compared to other parts, and it was concluded that higher temperatures and speeds encourage slag corrosion. The averaged corrosion rates at the sample wall with different rotating speeds and holding periods were compared. The relative error varied from 1.92% to 12.19%, which is within the acceptable range. It is expected that the proposed computational framework can be potentially extended to other refractory corrosion scenarios in metallurgical reactors.