Critical Review on Heat Transfer Correlations and Numerical Modeling for Scrap Melting in Converter Steelmaking Process

Steel is an important product in many engineering sectors, however, the steelmaking industry is one of the largest CO2 emitters. Because of this, new governmental policies push the steelmaking industry toward a cleaner and more sustainable operation. Becoming carbon neutral, utilizing more scrap is one of the feasible solutions to achieve this goal. However, because of the heterogeneity of the scrap in size, shape, and composition, knowledge gaps need to be dealt with to assess the influence of increasing scrap ratio on the Basic Oxygen Furnace (BOF) steelmaking operation. In the context of steel production, understanding heat transfer and the melting behavior of the scrap in the BOF is crucial for two reasons. First, better heat transfer contributes to efficient energy utilization, reduces scrap melting time which potentially leads to more scrap content, prevents non-homogeneous temperature distribution, and improves the BOF operation and lifetime. Second, understanding the melting behavior of the scrap helps designing a better scrap charging process which will lead to a homogenized melt (if good mixing was not applied). The present study is divided into two parts. First, the difference in melting behavior with emphasis on the melting of a single metal element (e.g., Rod, Sphere, Prism) under the influence of natural and forced convection conditions and their corresponding correlations is highlighted. Also, the influence of solid fraction on the melting and the heat transfer process is emphasized. Secondly, a comparison between the utilized numerical modeling approaches is made. It shows that CFD-DEM (Computational Fluid Dynamics-Discrete Element Method) approach offer attractive features compared to traditional Computational Fluid Dynamics (CFD), especially modeling different scrap shapes and types and using of different contact models (e.g., linear - spring dashpot, Hertz-Mindlin). Furthermore, it is computationally cheaper compared to the Particle Resolved Direct Numerical Simulation (PR-DNS). However, to accurately model the scrap melting (and dissolution) a reliable heat (and mass) correlation model is needed to predict the melting (and dissolution) phenomena correctly. Even though DEM has the capability to model the swelling behavior (particle growth or shell formation), the current review shows that there is no model in CFD-DEM available in literature to capture both shell formation (growth) and melting of the scrap.