Simulation of the Effect of Steel Flow Velocity on Slag Droplet Distribution and Interfacial Area Between Steel and Slag

• Published in: Steel research international 2015-03, Vol.86 (3), p.212-222
• Main Authors: Sulasalmi, Petri, Visuri, Ville-Valtteri, Kärnä, Aki, Fabritius, Timo
• Format: Article
• Language: English
• Subjects: Alloys; CAS-OB; CFD model; Computational fluid dynamics; Computer simulation; droplet distribution; Droplets; Flow velocity; interfacial area; Iron and steel making; Mathematical analysis; Mathematical models; Simulation; slag emulsification; Slags; Steel; steel flow velocity; Steels
• ISSN: 1611-3683; 1869-344X
• DOI: 10.1002/srin.201400026

Computational fluid dynamics (CFD) simulations have been carried out to study interfacial velocity and its effect on slag emulsification at the steel–slag interface. The multiphase volume of fluid (VOF) method was used to track the interface between slag and steel and a user defined function (UDF) was applied to calculate the interfacial velocity and to track separate droplets. Slag entrainment based on composition adjustment by sealed argon bubbling‐oxygen blowing (CAS‐OB) process was simulated using different inlet velocities of steel. The main objective of the present work was to study how steel flow velocity effects droplet distributions. Continuous Rosin–Rammler–Sperling (RRS) distribution function was fitted to computational results in order to provide a quantitative description of the droplet size distribution. The generation rates of the interfacial area between slag and steel were estimated based on obtained RRS‐distributions. The Sauter diameter was found to vary between 4 and 6 mm and the generation rate of the interfacial area between 0.2 and 0.6 m2 s−1. The results agree well with previous studies presented in the literature. A novel approach, which combines CFD‐modeling and regression analysis is applied to examine slag droplet distribution and generation rate with different steel flow velocities. Three phase CFD calculations are carried out for determining the shape of the distributions and then Rossin–Rammler–Sperling distribution function is fitted to the data, which enable estimating the birthrates of the emulsified droplets.