Growth and breakup of a wet agglomerate in a dry gas–solid fluidized bed

Abstract

SignificanceUsing CFD‐DEM simulations, a wet agglomerate of particles was placed in a void region of a dry vigorously fluidized bed to understand how wet agglomerates grow or breakup and how liquid spreads when agglomerates interact with dry fluidized particles. In the CFD‐DEM model, cohesive and viscous forces arising from liquid bridges between particles were modeled, as well as a finite rate of liquid bridge filling. The liquid properties were varied between different simulations to vary Bond number (surface tension forces/gravitational forces) and Capillary number (viscous forces/surface tension forces) in the system. Resulting agglomerate behavior was divided into regimes of (i) the agglomerate breaking up, (ii) the agglomerate retaining its initial form, but not growing, and (iii) the agglomerate retaining its initial form and growing. Regimes were mapped based on Bo and Ca. Implications of agglomerate behavior on spreading of liquid to initially dry particles were investigated. This article identifies a new way to map agglomerate growth and breakup behavior based on Bo and Ca. In modeling both liquid forces and a finite rate of liquid transfer, it identifies the complex influence viscosity has on agglomeration by strengthening liquid bridges while slowing their formation. Viewing Ca as the ratio of bridge formation time to particle collision and separation time capture why agglomerates with high Ca struggle to grow. © 2017 American Institute of Chemical Engineers AIChE J, 63: 2520–2527, 2017

The effect of liquid bridge model details on the dynamics of wet fluidized beds

“…This previous research concluded that an “agglomerate breakup” regime exists, in line with our findings. The key difference between the work of Boyce et al is that they were able to also identify an “agglomerate retention region”, which is absent in our simulations since we assume that liquid is homogeneously spread initially. It is clear that an initially non‐uniform liquid distribution will support arguments for using model C. Clearly, this due to the fact that the liquid spreading rate (which is strongly affected by the bridge filling rate) affects the particles' liquid content, and hence their sedimentation behaviour.…”

“…Also, we consider regions characterized by very high Bo and BoCa as “false fluidization,” since they result in the formation of a single, unphysically‐large agglomerate in our simulation domain see (Figure , panel f). Finally, we have indicated a region denoted as “agglomerate retention region” in Figure , which is based on the findings of Boyce et al Specifically, in this region a dynamic bridge filling model (i.e., model C) is necessary to picture the spreading of liquid initially not homogeneously distributed in the particle bed.…”

“…For an initially non‐uniform liquid distribution, however, model C is recommended even for large values of BoCa , since liquid spreading becomes rate limiting. Recalling that the capillary number can be interpreted as the ratio of bridge filling time and a characteristic time between collisions (Boyce et al), it appears that a simple criterion for judging the effect of liquid spreading is the capillary number. We would like to view our results also in the light of the recent findings of Boyce et al, who studied partially wetted fluidized beds of particles over a range of Bond and capillary numbers (i.e., up to 100 and 1000, respectively). This previous research concluded that an “agglomerate breakup” regime exists, in line with our findings.…”

Towards filtered drag force model for non-cohesive and cohesive particle-gas flows

“…In contrast, numerous Euler-Lagrange models, where the locally averaged equations of motion for the fluid phase are solved in an Eulerian framework and the particles are tracked in a Lagrangian fashion by solving Newton's equations of motion, have been successfully used to simulate gas-solid flows with such complex interactions. [27][28][29][30][31][32][33][34][35] Instead of incorporating complex particle physics into Euler-Euler models via theoretical derivations, it is more straightforward to perform Euler-Lagrange simulations including complex particle physics and use these simulation results to directly propose constitutive relations for filtered Euler-Euler models. As shown in our previous study, 36 one can use the drag corrections deduced from Euler-Lagrange simulations for filtered Euler-Euler simulations.…”

Magnetic resonance imaging of gas–solid fluidization with liquid bridging

“…Previous studies have shown that the addition of small amounts of liquid alters the oscillations in pressure drop across the bed, slows the speed of particles, and increases the minimum fluidization velocity . Studies have also sought to map the behavior of wet fluidized beds into regimes based on the amount of liquid added as well as the surface tension and viscosity of the liquid; these regimes have included shifts in the Geldart grouping of the particles and the growth or breakup of agglomerates . Despite decades of studies, key questions remain open in the field, including the relative importance of liquid loading, surface tension, and viscosity of the liquid, the non‐dimensionalization of liquid bridge behavior based on force‐ or energy‐based analyses and the validity of approximating a wet and dry bed of behaving similarly if the ratio of superficial velocity to minimum fluidization velocity ( U/U mf ) is kept constant .…”