The Development of Improved Particle Drag Force Correlations for CFD Simulation

The particle drag force correlation is often the most important factor in a fluidized bed CFD simulation. While commonly-used drag models capture the basic functional dependencies on Reynolds number and voidage, the real-world particle dynamics may also be affected by particle characteristics such as the particle shape, the full-size distribution, and surface properties. These additional factors, which are often difficult or expensive to measure, can be a source of error in the particle drag force correlation and a CFD simulation as a whole. Therefore, in common practice, the accuracy of a particle drag force model is improved through iterative tuning based on experimental bed density, entrainment rates, or other fluidization characteristics.

The goal of this work is to reduce the computational time and cost of CFD tuning by developing particle drag force correlations which can be directly improved based on easily-obtained experimental pressure drop or minimum fluidization data. Model accuracy before and after tuning is demonstrated using the wide range of liquid-solid bed expansion and settling data that is available in literature. The same data sets are used for comparison with particle drag force models that commonly used in CFD.