(554e) Immiscible Liquid-Liquid Mixing Strategy in an Agitated Chemical Reactor – a Computational Fluid Dynamics Study
AIChE Annual Meeting
2019
2019 AIChE Annual Meeting
North American Mixing Forum
The Use of CFD in Simulation of Multiphase Mixing Processes
Wednesday, November 13, 2019 - 1:54pm to 2:15pm
Immiscible
liquid-liquid mixing strategy in an agitated chemical reactor A Computational
Fluid Dynamics study
Rui
C. Silva, Maria Inês Lopes, Filipe Neves
Immiscible liquid-liquid mixing performed in agitated chemical
reactors is a fundamental unit operation in batch pharmaceutical and chemical
production. The efficiency of said unit operation is affected by a number of operational
parameters such as impeller geometry, number of impellers, number of baffles,
reactor geometry, power number and dispersed phase volume fraction as well as physical
properties of phases such as viscosity and density.
CFD simulations are a cost-effective tool to calculate mixing
efficiency particularly for complex impeller geometries. A significant number
of advantages of a numerical CFD approach to processes involving complex impeller
geometries, when compared with traditional experimental measurement techniques:
overall enhancement of process knowledge on the influence of several operational
parameters is attained for every point of the study domain in a reduced
timespan, minimized burdening the process train with scale testing and easy
troubleshoot.
In this work a numerical study using an open-source CFD package,
OpenFOAM, was performed to characterize several impeller geometries as well as
assessing their impact on the mixing efficiency of immiscible liquids. The
efficiency of mixing (i. e. mixing index) was assessed using the numerical
velocity field, energy consumption and the amount of dispersed phase
transferred. Different impeller geometries, number of baffles and immiscible
liquids addition strategy were tested using Computational Fluid Dynamics (CFD).
Steady state and transient flow fields were calculated for given flow
conditions using solvers based on simpleFoam (Semi-Implicit Method for
Pressure-Linked Equations) and pimpleFoam (merged PISO-SIMPLE) respectively. Turbulence
was modelled by the two-equation standard high-Re k-epsilon turbulence model,
as implemented in OpenFOAM, in tandem with a Moving Reference Frame (MRF).
Figure Velocity vector and streamlines on bottom impeller with
one baffle.