Numerical simulation of the gas-solid flow in a fluidized bed by combining discrete particle method with computational fluid dynamics

Abstract

The gas-solid flow in a fluidized bed is modelled by a combined approach of discrete particle method and computational fluid dynamics (DPM-CFD), in which the motion of individual particles is obtained by solving Newton's second law of motion and gas flow by the Navier-Stokes equation based on the concept of local average. The coupling between DPM and CFD is achieved directly by applying the principle of Newton's third law of motion to the discrete particle and continuum gas which are modelled at different length and time scales. The equations of motion for a system of particles are solved by a collision dynamic model developed in this work which, in conjunction with the predictor-corrector method, allows stiff particles ( κ = 50,000 Nm −1) to be used with a reasonable computational time step (1.5 × 10 −5 s) while conserving the energy and momentum. The gas-phase equations are solved by the conventional SIMPLE method facilitated with the Crank-Nicolson scheme to give the second order accuracy in the time discretization. The proposed model shows its capacity of simulating the gas fluidization process realistically from a fixed to fully fluidized bed via an incipient fluidization stage. This is done by a series of numerical tests to reproduce the experimental procedures in determining the minimum fluidization velocity of 2400 particles ( ϱ p = 2700 kgm −3, D = 4 × 10 −3 m) in a pseudo-three-dimensional central jet fluidized bed of dimensions 0.9 × 0.15 × 0.004 m. The hysteretic feature of bed pressure drop vs superficial gas velocity curve is obtained for the first time realistically from first principles, with the predicted minimum fluidization velocity in good agreement with experiment. It is demonstrated that the proposed model is able to capture the gas-solid flow features in a fluidized bed from the largest length and time scales relevant to the processing equipment down to the smallest ones relevant to the individual particles.