GPU-based reactive-CFD-DEM (rCFD-DEM) modelling of multiphase reacting flows in an industrial-scale blast furnace: Enabling particle-scale high-speed simulations
Year: 2025
Type: article
Abstract: The blast furnace (BF) is the dominant ironmaking reactor with a complex internal state involving coupled multiphase flow, heat transfer, and chemical reactions, but its particle-scale understanding is limited at the industry scale due to the lack of high-speed particle-scale modelling techniques. This paper develops a graphics processing unit (GPU) − accelerated Reactive Computational Fluid Dynamics-Discrete Element Method (rCFD-DEM) model, capable of simulating the particle-scale flow and thermochemical behaviour in an industrial-scale BF at a high speed, especially capturing the particle-scale dynamics of the raceway, cohesive zone, and iron ore reduction simultaneously, for the first time. The model is validated against experimental data. The simulation results capture the cohesive zone with significantly higher particle volume fraction, altered gas percolation direction, and elevated reaction rates, resulting in strong heat absorption and high iron ore reduction degrees. A typical layered distribution of gas species and temperature is also captured, with higher temperatures and lower CO concentrations in the coke layers, and lower temperatures with higher CO in the iron ore layers. Iron ore reduction is quantified at the particle scale, high in the high-temperature central region of the furnace. In the cohesive zone, the reduction degree of particles increases linearly over time and stabilises after approximately 50 s. This GPU-accelerated rCFD-DEM BF model provides a revolutionary high-speed R&D platform for investigating particle-scale internal state of BFs, especially process optimisation and design to achieve low-cost ironmaking decarbonisation.
Source: Chemical Engineering Science
Authors: Dazhao Gou, Yansong Shen
Institution: UNSW Sydney
Language: English
Subfield: Mechanical Engineering
Field: Engineering
Domain: Physical Sciences
Open Access status: hybrid
Estimated APC paid: $4,050
OpenAlex ID: W4416412381