CFD-DEM with Dynamic Meshing – a Novel Approach to Predict Particle Dispersion in an Agitated Tubular Reactor

This study focuses on particle dispersion within an Agitated Tubular Reactor (ATR) by coupling large eddy simulation (LES) with the discrete element method (DEM). The operating unit of the ATR system is essentially a tubular reactor containing a loose internal agitator. The reactor system is subject to an applied oscillation, consequently the motion of the internal agitator is determined by the hydrodynamic forces of the fluid, forces due to particle collisions and the forces due to collision between the internal agitator and the outer tube. The co-existence of a large internal structure and small particles presents a significant Fluid Structure Interaction (FSI) challenge which has to be overcome to investigate the particle behaviour in the ATR system. In the present study, a new numerical approach is proposed to solve the fluid flow induced by the free-moving internal agitator, in which a body-fitted dynamic meshing approach is used to account for the motion of the agitator. For the large agitator, the forces due to the fluid flow are calculated by integrating pressure and viscous stress over the agitator surface while the forces due to collisions with the smaller particles and the outer tube are calculated using a soft-sphere model. For the small particles, the influence of the flow structure on the particle motion is considered by coupling LES with DEM such that the impact of the fluctuating fluid velocity on the particle motion can be considered. The developed model has been fully implemented on the ANSYS/Fluent platform due to its efficient handling of dynamic meshing and complex and/or free-moving boundary conditions. The model has been calibrated and validated against experimental measurements. First, the collision parameters between the agitator and the outer tube are calibrated by comparing the predicted motion of the agitator, both translational and rotational, with the experimental data obtained by high-speed camera. Then, the mean and the fluctuating fluid velocity are compared with experimental measurements using acoustic backscatter technology. With the validated model, the effects of operational conditions on the particle dispersion are investigated.