COMPUTATIONAL FLUID DYNAMICS OF A FLUID BED EMPLOYING TUNED GAS-SOLID DRAG MODELS

Abstract

Fluidized beds are devices in which a fluid flows from the bottom through a bed of particles, keeping them under suspension. Fluidized beds find many applications as reactors for combustion and gasification of solid fuels. For a given fluid-particulate combination, there is a minimum fluidization velocity (U mf) which exerts a drag force that equals the weight of the bed, fluidizing the system. Therefore, it is possible to calculate gas-solid drag forces parameters from a minimum fluidization velocity (Umf) obtained experimentally. In the present work, the objective was to tune gas-solid drag correlations to be used in the Computational Fluid Dynamics (CFD) of a fluidized bed employing the Umf and to analyze the improvement of CFD results. The particles employed were one of Geldart-B (sand-like) and two of Geldart-D (spoutable) types, fluidized in a cylindrical riser with 0.114 m internal diameter. The CFD multiphase model employed was the Two-Fluid-Model (TFM). In this model both gas and solid phases are assumed interpenetrating continua, mapped along the domain via its volume fraction, and the Kinetic Theory of Granular Flows (KTGF) is used to model solids phase viscosity term. The force interactions between phases are modeled using gas-solid drag correlations, which in this work were based on Syamlal-O'Brien and Di Felice models. A finite volume method CFD code was used to perform the simulations. The simulations for superficial velocity of 1.5 Umf was performed in order to confront experimental and numerical results of pressure drop and bed height. So far tuned models were better than the original ones in the prediction of fluidization curves (pressure drop versus superficial velocity), and in the prediction of bed expansion and bubble formation.