Gas dynamics in fluidized beds heavily influence chemical reactions, heat and mass transfer and overall hydrodynamics. Despite this fact, there have been few direct measurements of gas dynamics in fluidized beds (1, 2), due to the difficulties in obtaining reliable experimental data on gas motion in 3D beds filled with opaque particles. Subsequently, computational models of fluidized beds, which model the flow of gas and particles as well as their interaction, have been largely validated against experimental measurements of particle dynamics (3–5), leaving uncertainties in the accuracy of gas dynamics predicted by computational models.

Recently, Boyce et al. (6) presented results of an MRI study measuring gas dynamics in fluidized beds. These spatially-resolved measurements of time-averaged gas velocity and velocity distribution both in the bed of particles and in the freeboard revealed the nature of gas flow through bubbling and particulate regions of fluidized beds. The measurements also showed time-averaged particle velocity and void fraction in the same fluidized bed to provide insights on how gas dynamics relate to particle dynamics. These measurements were previously compared against classical analytical theories for gas dynamics in fluidized beds (6), such as the two-phase theory of fluidization (7).

Here, we compare the MRI measurements with simulation predictions using the computational fluid dynamics – discrete element method (CFD-DEM) (8). This simulation technique is commonly used for detailed simulations of laboratory-sized fluidized beds because it resolves the motion of each individual particle using a Lagrangian method, while resolving gas flow on Eulerian grids coarser than the particle diameter and accounting for gas-particle interaction using a drag law. The accuracy of this method in predicting gas and particle dynamics in bubbling and homogeneously fluidized beds is assessed, while varying important parameters such as drag law, fluid grid sizing and gas distribution. Additionally, since only time-averaged results could be provided experimentally, instantaneous predictions from computer simulations are compared with classical analytical theory for gas flow in fluidized beds, such as bubble rise velocity (9) and gas flow through bubbles (10).

Using CFD-DEM, we extend this analysis to “wet” fluidized bed in which a small amount of liquid allows for cohesive pendular bridges to form between particles which can drastically alter gas and particle dynamics. In CFD-DEM simulations of wet fluidized beds, liquid loading, viscosity and surface tension are accounted for to track the amount of liquid on each particle and in each pendular bridge, as well as the cohesive force provided by liquid bridges (11) and the finite rate of liquid transfer between particles and bridges. In these simulations, the cohesive force from liquid bridges can lead to the formation and growth of agglomerates of particles, while the shearing force from liquid and particle dynamics can lead to agglomerate breakup. We use these simulations to start to identify parameter spaces for agglomerate dynamics based on liquid, particle and gas properties and instruct future experiments, using MRI or other techniques, to validate models and provide further insights on flow behavior in wet fluidized beds.

References:

1.         T. Pavlin et al., Noninvasive Measurements of Gas Exchange in a Three-Dimensional Fluidized Bed by Hyperpolarized 129Xe NMR. Appl. Magn. Reson. 32, 93–112 (2007).

2.         P. N. Rowe, P. F. Wace, Gas-Flow Patterns in Fluidized Beds. Nature. 188, 737–738 (1960).

3.         C. M. Boyce, D. J. Holland, S. A. Scott, J. S. Dennis, Adapting Data Processing To Compare Model and Experiment Accurately: A Discrete Element Model and Magnetic Resonance Measurements of a 3D Cylindrical Fluidized Bed. Ind. Eng. Chem. Res. 52, 18085–18094 (2013).

4.         X. Lu, C. M. Boyce, S. A. Scott, J. S. Dennis, D. J. Holland, Investigation of Two-fluid Models of Fluidisation Using Magnetic Resonance and Discrete Element Simulations. Procedia Eng. 102, 1436–1445 (2015).

5.         C. R. Müller et al., Granular temperature: Comparison of Magnetic Resonance measurements with Discrete Element Model simulations. Powder Technol. 184, 241–253 (2008).

6.         C. M. Boyce, J. S. Dennis, D. J. Holland, (Atlanta, GA, 2014).

7.         R. D. Toomey, H. F. Johnstone, Gaseous fluidization of solid particles. Chem. Eng. Prog. 48, 220–226 (1952).

8.         Y. Tsuji, T. Kawaguchi, T. Tanaka, Discrete particle simulation of two-dimensional fluidized bed. Powder Technol. 77, 79–87 (1993).

9.         R. M. Davies, G. Taylor, The Mechanics of Large Bubbles Rising through Extended Liquids and through Liquids in Tubes. Proc. R. Soc. Lond. Ser. Math. Phys. Sci. 200, 375–390.

10.       D. Harrison, J. F. Davidson, de Kock, J.W., On the nature of aggregative and particulate fluidisation. Trans Inst Chem Eng. 39, 202–212 (1961).

11.       T. Mikami, H. Kamiya, M. Horio, Numerical simulation of cohesive powder behavior in a fluidized bed. Chem. Eng. Sci. 53, 1927–1940 (1998).