Characterization of Drag Force on Agglomerates in Fluidized Beds

Agglomeration of bed materials is an important phenomenon in several applications that use fluidized bed technologies such as granulation, compacting, drying, combustion and gasification, etc. Agglomerates complicate the flow structure, thus influence the hydrodynamics as well as mass and heat transfer in fluidized beds. For instant, fluidization of Geldart C/A particles is expected to be particularly difficult due to agglomeration behaviour by cohesive forces existing between particles, which become more and more prominent as the particle size decreases. On the other hand, in fluidization drag force is one of the dominant particle-fluid interactions to control particle movements, circulation and concentration distribution. It is well understood that formation of agglomerates (similar to particle clusters) significantly reduces the drag force comparing to single particles. Therefore, understanding the mechanisms underlying particle agglomeration and their effect on the hydrodynamics (mainly gas-solid drag force) is crucial for the design and control of the desired process performance. Unfortunately, currently available drag correlations that are applied in modelling of fluidization do not truly take into account the agglomeration phenomenon.

Computationally, direct numerical simulations (DNS) is proven effective to study in detail the gas-solid interactions in multiphase flows. Such simulations employ no assumptions for solving the fluid motions, and resolve the flow field from the boundary layer of the particle surface. The hydrodynamic forces acting on the solid particles are directly computed in DNS from the fluid flow via enforcing a no-slip boundary condition at particle surface.

In this work, a detailed understanding and quantification of agglomeration behaviour of fine particles as well as drag resistance of the formed agglomerates under fluidization conditions is provide using DNS. The sedimentation of hundreds of primary particles of same diameters (orders of micrometers) in a fluidizing fluid is considered under different flow conditions (Reynolds numbers upto 200). The particles experience gravity, drag as well as van der Waals forces. During the sedimentation process, the particles can stick together and form agglomerates, which can then also be broken up. The obtained agglomerate morphologies are characterized by various structural parameters such as sphericity, a convex hull-based porosity and the radius of gyration. The time-dependent growth of agglomerates size is simultaneously tracked. The formed agglomerates with different morphologies are then analysed with respect to the drag force in comparison with the drag on single primary particles. Finally, by comparing all the simulation results to the existing drag correlations for non-spherical particles, a new drag correlation is formulated. This new correlation is a correction to the Schiller-Naumann [2] correlation for spheres but includes the dependency on the agglomerates morphologies. The improved drag correlation can be used to improve the accuracy of prediction of CFD-DEM and TFM models for fluidized beds with agglomeration phenomenon.