(393a) Large Eddy Simulation of a Pipeline Rotor-Stator Mixer | AIChE

(393a) Large Eddy Simulation of a Pipeline Rotor-Stator Mixer


Minnick, B. A. - Presenter, University of Maryland
Calabrese, R., University of Maryland
Rotor-stator mixers provide high deformation rates to a relatively limited volume, resulting in a region in which intensive mixing, milling, and/or dispersion/emulsification operations can occur. Because they are geometrically complex, experimental measurement of important flow regions within rotor-stator mixers is often difficult, if not impossible. Computational fluid dynamics (CFD) simulations of mixers are a means to provide flow field information that are of benefit to both designers and end users.

This presentation focuses on our recent work in the fully transient (sliding mesh) three-dimensional large eddy simulation (LES) of a pipeline rotor-stator mixer, the Greerco High Shear Pipeline Mixer, using ANSYS FLUENT software. The modeled unit consists of two conical rotor-stator stages, each with a nominal diameter of 4 inches, aligned to produce an axial discharge flow. The mixer geometry consists largely of cylindrical and conical bodies; the impact of this on meshing will be discussed. The working fluid was water in turbulent flow. The wall-adapting local eddy-viscosity model is used to model subgrid scales and an equilibrium stress wall model describes the interaction of scales near no-slip surfaces.

Flow and turbulence quantities have been studied on a per stator slot and per rotor stage basis respectively. As with other rotor-stator mixers, fluid is shown to impinge on the trailing surface of the stator slot. Turbulent kinetic energy and its isotropic dissipation rate are shown in such regions. Comparisons are made between the LES filtered Navier-Stokes simulations and our previously reported Reynolds averaged Navier-Stokes (RANS) realizable k-ε model simulations. Turbulence scales important for mixing are approximated and compared for both of these turbulence models. The effect of mesh density to resolve the larger scales of turbulence are discussed by comparing results from a coarse and refined computational mesh. The effect of operating conditions on power draw, throughput and other quantities of practical utility are also discussed.