8th International Congress on Science and Technology of Ironmaking - ICSTI 2018

The campaign length of a blast furnace is limited by the hearth inner lining lifetime. In order to maximize the campaign length and ensure a good draining of hot metal and slag, a good understanding of the flow in the hearth is essential. Challenges in modeling the flow involve several continuous phases (hot metal, slag and hot blast) as well as the presence of the deadman, a dense bed of coke particles. The shape and position of the deadman is highly dynamic and depend on several conditions, e.g. the weight of the burden, the deadman porosity as well as the liquid iron and slag levels in the hearth. We present a numerical coupled CFD (Computational Fluid Dynamics) – DEM (Discrete Element Method) model to account for the transient behavior of the deadman. A VOF (Volume of Fluid) method is used to model the multiple continuous phases and the DEM method to model the discrete particles. The VOF and DEM models are coupled together in a 2-way manner, resulting in a complete 4-way coupled CFD-DEM model. Experimental validation was performed on a lab-scale particle filled tank and a demonstration of a small-scale blast furnace hearth is shown. Finally, we present our latest advancements towards modeling industrial scale blast furnaces. Due to the big dimensions, difficulties with long computational times arise as the particle number becomes infeasible. Commonly a parcel approach (coarse graining) is used to handle this problem, where several particles are treated as one large particle. However, in combination with VOF, the amount of coarse graining is limited as the increased particle size impairs the resolution of the fluid interfaces. Additionally, the wide range of time scales (ranging from particle collisions of fractions of a second to several hours of operational times) present in the blast furnace remains a major challenge. With information gathered from several short-term coarse grained DEM simulations for various liquid levels, we performed a long-term fully Eulerian simulation that used the data to create a dynamic porosity field. By not resolving any particle collisions in the long term simulation, we could simulate a longer process time while still accounting for the dynamic behavior of the deadman.