A multi-fluid Eulerian model incorporating the kinetic theory of granular flow is used for the simulation of bubbling fluidized beds containing a binary mixture of Geldart B particles at low gas velocities. The cases of density, size and combined density/size segregation are investigated using computational fluid dynamic simulations. Various expressions for the drag force are evaluated for predicting different segregations. The simulation results show that summation of the particle-particle drag force, i.e., the "hindrance effect" term, and the Stokes drag of particles, which is modified based on the Wen-Yu drag model can be used for accurate simulation of a binary mixture of particles differing in size, density, or both. Bed expansion and dimensionless axial segregation profiles of CFD results are compared with the experimental data and good agreement is found.
IntroductionMixtures of solid particles of different sizes and/or densities tend to separate during fluidization. The segregation behavior of a mixture of particles is important since particle distributions in the fluidized bed influence chemical reactions, bed expansion, and various mass and heat transfer properties in the fluidized beds. In fluidized bed reactors in the food, chemical, pharmaceutical and metallurgical industries, the processes of particle mixing and segregation change the distribution of the mixture components in the bed, and therefore, play a very important role in the overall processes [1]. Segregation, as a transient process, is strongly affected by gas velocity [1]. Bubbles, which are known to play an intricate and ambiguous role in fluidized bed reactors, can strongly influence the different mixing and segregation phenomena. Both the chemical reaction and mass/heat transfer depend on the local particle size distribution (PSD) in the bed. Therefore, a better understanding of the spatial distribution of different solid components for a given particle size distribution is required to improve the design, operation and scale-up of different gas-fluidized bed processes.While there are several experimental procedures for measuring particle size distribution and the physical properties of fluidized beds, the use of computational fluid dynamic (CFD) tools for providing valuable information for the design of the reactor, scale up and optimization has attracted considerable attention in recent years. Two methods have been typically used for CFD modeling of gas-solid flows, i.e., the EulerianLagrangian method and the Eulerian-Eulerian approach. In the Eulerian-Lagrangian approach, the computational demand rises sharply with the number of traced particles, which constrains its applicability to high concentration flows. In the Eulerian-Eulerian method, which is used in the current study, two phases are mathematically treated as interpenetrating continua. The success of a multi-fluid model (MFM) depends on the proper description of the interfacial forces and the solid stresses. The interfacial forces are used to describe the momentum transfer be...