Simulation of Flocculation in Stirred Vessels Lagrangian Versus Eulerian

A study has been performed to evaluate Lagrangian and Eulerian approaches for simulating flocculation in stirred vessels. The prediction of the transient floc size evolution was performed using the quadrature method of moments (QMOM) while flow field characteristics within the turbulent stirred vessel were obtained using computational fluid dynamics (CFD). The Eulerian and Lagrangian CFD/QMOM models were applied to a 28 l square tank using either a Rushton turbine or a fluid foil impeller. Simulations were performed with an initial concentration of 100 mg L−1 of 1 μm nominal clay particles for several average characteristic velocity gradients (40-, 70-, 90-, 150-s−1). For the Lagrangian approach, the results showed that the average floc size transient evolution curve does not predict a peak followed by a lower steady-state size as observed for higher shear rates with the Eulerian approach. However, the overall good agreement between the Eulerian and Lagrangian CFD/QMOM models, indicates that a Lagrangian approach combined with a QMOM model would be an efficient method to quantify the impact of non-fluid flow experimental conditions on the flocculation process. In addition, the Lagrangian CFD/QMOM approach may be a useful tool to study the dynamics of flocculation and determine appropriate coalescence/breakup kernels when performing an inverse problem technique.