CFD simulations of dense solid-liquid suspensions in baffled stirred tanks: Prediction of suspension curves

Mixing of solid particles into liquids within contactors mechanically agitated by stirrers is a topic of primary importance for several industrial applications. A great research effort has been devoted to the assessment of the minimum impeller speed (N-js) able to guarantee the suspension of all particles. Conversely, only little attention has been paid so far to the evaluation of the amount of solid particles that are suspended at impeller speeds lower than N-js. In some cases the loss in available interfacial area between particles and liquid could be reasonably counterbalanced by a decreased mechanical power, making it of interest to evaluate the percentage of suspended solids at different impeller speeds <N-js in order to quantify the possible economical advantage of adopting an incomplete suspension (filleting) regime inside the stirred tank. The present work deals with computational fluid dynamics simulations of dense solid-liquid partial suspensions in baffled stirred tanks and particularly focuses on the prediction of the amount of suspended particles at agitation speeds encompassing both the filleting and the complete suspension regime. An Eulerian-Eulerian Multi Fluid Model coupled with a standard k-epsilon turbulence model for the continuous (liquid) phase only was adopted for CFD simulations. Both the Sliding Grid (SG) and the Multiple Reference Frame (MRF) approaches were employed to simulate the impeller-tank relative rotation. Different turbulence corrections to the fluid-particle drag correlation were considered. Experimental evidence was used for validation purposes: data collected by using the Pressure Gauge Technique (PGT) [1] and snapshots of the simulated tank. Comparison between CFD predictions and all experimental data showed a satisfactory agreement; for large particle diameters the drag coefficient correlation due to Pinelli et al. [2] gave the best results. The influence of the impeller motion treatment was found to be negligible. (C) 2011 Elsevier B.V. All rights reserved.