Parametric study of the time-averaged gas–solid drag force in circulating fluidized bed conditions

• Time-averaging of transient Eulerian–Eulerian simulations of 2D CFB risers done
• Time-averaged drag force was calculated in each mesh point in 17 CFB simulations.
• Voidage, wall distance, slip velocity and phase properties affect the drag force.
• Solid density, gas viscosity and particle size also have clear effects on the drag.

Steady state modeling based on time-averaged transport equations is a computationally efficient method for CFD simulation of large industrial-scale circulating fluidized beds. The feasible alternative for steady state modeling is to carry out transient simulations in a coarse mesh, which would require mesh resolution dependent closure laws and lead to long simulations in order to characterize the average process behavior. These closures could be avoided by simulations with adequate spatial and temporal resolutions, but the required computational effort renders such simulations unfeasible in industrial scale applications in near future. Equation closures for steady state modeling can be developed by time-averaging results from transient simulations. One of the largest terms to be modeled is the time-averaged drag force. In the present work, parameters that need to be accounted for in a model for the time-averaged drag force were studied by analyzing time-averaged results from a number of transient 2D simulations carried out with fairly fine spatial resolution. The analysis was limited to Geldart B particles. In the literature, the solid volume fraction, distance to the wall and the gas–solid slip velocity have been included as parameters in drag correction functions developed for transient coarse-mesh simulations. In the present work, the same parameters are found to have significant effects on the time-averaged gas–solid drag force. Additionally, solid density, particle size and gas phase laminar viscosity are shown to have significant effects on the average drag force. Thus process conditions, which significantly vary inside a CFB, need to be accounted for in the model.

Figure optionsDownload as PowerPoint slide