DEM-CFD investigation on the blast furnace raceway characteristics: Heat/ mass transfer, dynamics and mechanisms

Exploring the kinetic characteristics and internal thermochemical behavior of the raceway is essential to provide in-depth insights and valuable references for the actual blast furnace (BF) production. In this paper, the flow, heat and mass transfer behaviors and kinetic characteristics of coke particles in the BF raceway were modelled using Discrete Element Method-Computational Fluid Dynamics (DEM-CFD) method from particulate scale. The effects of process parameters (blast velocity/temperature, and oxygen concentration) on gas species distribution, reaction kinetic rate, particle temperature, particle size, dynamic characteristics (raceway size, kinetic energy) and microscopic properties (coordination number, contact force, total force, and pore distribution) were systematically investigated. On this basis, the back propagation neural network (BPNN) model was established to predict raceway size for any combination of process parameters. The results show that blast velocity/temperature and oxygen concentration have strong influences on the reaction kinetic rate, gas species distributions and microscopic properties of the coke bed. Higher blast velocity or temperature makes the higher temperature and carbon loss of individual coke particles around the raceway, and the mass transfer between gas and coke particles is not diminished due to the excessive oxygen at a large oxygen concentration. The width and height of the raceway increase with the increase of blast velocity and oxygen concentration. The BPNN model can well predict the raceway size, which provides a quick and easy way to establish a more accurate and effective strategy for size control.