Comparison between two-fluid model simulations and particle image analysis & velocimetry (PIV) results for a two-dimensional gas-solid fluidized bed

Abstract

This work compares simulation and experimental results of the hydrodynamics of a two-dimensional, bubbling air-fluidized bed. The simulation in this study has been conducted using an Eulerian–Eulerian two-fluid approach based on two different and well-known closure models for the gas–particle interaction: the drag models due to Gidaspow and Syamlal & O'Brien. The experimental results have been obtained by means of Digital Image Analysis (DIA) and Particle Image Velocimetry (PIV) techniques applied on a real bubbling fluidized bed of 0.005 m thickness to ensure its two-dimensional behaviour. Several results have been obtained in this work from both simulation and experiments and mutually compared. Previous studies in literature devoted to the comparison between two-fluid models and experiments are usually focused on bubble behaviour (i.e. bubble velocity and diameter) and dense-phase distribution. However, the present work examines and compares not only the bubble hydrodynamics and dense-phase probability within the bed, but also the time-averaged vertical and horizontal component of the dense-phase velocity, the air throughflow and the instantaneous interaction between bubbles and dense-phase. Besides, quantitative comparison of the time-averaged dense-phase probability as well as the velocity profiles at various distances from the distributor has been undertaken in this study by means of the definition of a discrepancy factor, which accounts for the quadratic difference between simulation and experiments The resulting comparison shows and acceptable resemblance between simulation and experiments for dense-phase probability, and good agreement for bubble diameter and velocity in two-dimensional beds, which is in harmony with other previous studies. However, regarding the time-averaged velocity of the dense-phase, the present study clearly reveals that simulation and experiments only agree qualitatively in the two-dimensional bed tested, the vertical component of the simulated dense-phase velocity being nearly an order of magnitude larger than the one obtained from the PIV experiments. This discrepancy increases with the height above the distributor of the two-dimensional bed, and it is even larger for the horizontal component of the time-averaged dense-phase velocity. In other words, the results presented in this work indicate that the fine agreement commonly encountered between simulated and real beds on bubble hydrodynamics is not a sufficient condition to ensure that the dense-phase velocity obtained with two-fluid models is similar to that from experimental measurements on two-dimensional beds