Evaluation of a filtered model for the simulation of large scale bubbling and turbulent fluidized beds

Abstract

Full 3D flow simulations of lab and industrial scale dense fluidized beds were carried out using a filtered Eulerian–Eulerian approach. Filtered closures for interphase momentum exchange, solids stresses and additional wall corrections were implemented in the standard equations of motion. These closures had a very large effect on the overall model performance when solved on the large cell sizes required for computationally affordable 3D fluidized bed simulations. Numerical experiments conducted under different fluidization conditions showed that the current model formulation performs well over a wide range of operating conditions. It was found that additional modelling accounting for flow non-uniformity is essential under certain fluidization conditions. The current method for dealing with flow non-uniformity by means of wall corrections yielded good results under vigorous fluidization, but caused a slight inaccuracy at low fluidization velocities. In general, comparisons to a wide range of experimental data showed good quantitative agreement, suggesting that the formulation of the filtered model is highly generic. The filtered approach was also successfully verified in a large scale bubbling fluidized bed reactor by comparisons with a highly computationally expensive, well resolved, non-filtered flow simulation.