CFD simulation of gas–solid fluidized bed hydrodynamics; prediction accuracy study

Abstract The present study aims at increasing the prediction accuracy of simulating gas–solid fluidized bed hydrodynamics. Two simulation packages, Fluent and MFIX, were used to predict the pressure drop, voidage, and solid-phase velocities by solving mass, momentum, and energy balance equations. A 2D multi-fluid Eulerian model with the kinetic theory of granular flow (KTGF) was applied to simulate the process by considering two different drag models. The same comparative criterion of average absolute relative deviation (AARD%) was considered to compare the present simulation with the previous works. Compared to the prior works, the minimum decrease in error (AARD% of 5.91%) was 3.17% related to the estimation of the time-averaged voidage by applying the Gidaspow model, while the maximum reduction in error (ARRD% of 5.88%) was 17.35% attributed to the prediction of pressure drop by employing the Syamlal-O’Brien model, both in Fluent software. However, MFIX software was the best CFD tool in predicting time-averaged voidage by AARD% values less than 9% under all conditions. Furthermore, similar patterns in contours were observed for solid-phase volume fraction and gas/solid phase velocities in both simulation tools, which are compatible with results from the literature without any significant difference.