Evaluation of large Eddy simulation and Euler‐Euler CFD models for solids flow dynamics in a stirred tank reactor

Abstract

Mechanically agitated reactors find wide range of applications for solid suspension and mixing in the chemical, biochemical, and mineral processing industries. Understanding the solids dynamics in these reactors is necessary to improve the design and operation of such reactors. Computational fluid dynamic (CFD) models are often useful in this regard, as it can provide significant insights into the flow and mixing of the phases involved. However, the model predictions need extensive evaluation with experimental results before they can be confidently used for the scale‐up and optimization of large scale reactors. Recently, Guha et al. carried out a systematic experimental investigation of the solids hydrodynamics in dense solid–liquid suspensions (2.5–19% solids loading w/w) in a stirred tank using the Computer Automated Radioactive Particle Tracking (CARPT) technique, which provided extensive information to efficiently assess the ability of the existing CFD models in predicting the solids dynamics in slurry reactors. This work presents such an evaluation by comparing the averaged solids velocities, turbulent kinetic energy, and solids sojourn time distributions predicted by CFD models with those obtained from the CARPT experiment for overall solids holdup of 1% (v/v) (2.5% w/w) at Reynolds number of 74,000. The Large Eddy Simulation (LES) and the Euler‐Euler model are the models chosen for evaluation in the current study. © 2008 American Institute of Chemical Engineers AIChE J, 2008

Numerical simulation of solid suspension via mechanical agitation: effect of the modelling approach, turbulence model and hindered settling drag law

“…Recently, Guha et al (2008) have compared large eddy simulation (LES) coupled with a Lagrangian particle model that includes collisions, and a two-phase Eulerian model with experimental data for a tank agitated by a Rushton turbine with solids volume fraction of 1%. For their Eulerian two-phase simulations they used turbulent diffusion in the volume fraction equation and the homogeneous k-e model.…”

Computational Fluid Mixing

“…The liquid flow was obtained using LES on a grid of some 14 million cells; Lagrangian tracking of almost 7 million particles was carried out over 10-20 impeller revolutions, with 2800 time steps per revolution and both two-way coupling and particle-particle interactions were included; importantly the latter were required to keep the particles in suspension, even though the impeller speed was above the just-suspended condition. Guha et al (2008) Thus although solid-liquid simulations can be conducted, there are still a number of uncertainties related to the physics of these flows, particularly with respect to particle-particle and particle-impeller interactions. There are even issues about the calculation of the drag coefficient for such flows: Montante and Magelli (2005) recommended correction of the drag coefficient to allow for local turbulence levels (Brucato et al, 1998b), but not for local solids volume fraction (Crowe et al, 1998), whereas Derksen (2003) makes the opposite recommendation, which seems to be the more conventional thinking.…”

Models and Applications for Simulating Turbulent Solid–Liquid Suspensions in Stirred Tanks

“…Guha et al (2008) compared the solids velocities and turbulent kinetic energy predicted by the E-L and E-E models with those measured by using the computer automated radioactive particle tracking (CARPT) technique. LES and the standard k-ε model with mixture properties were respectively used in the E-L and E-E models.…”