The expansion of physical models integrated into Computational Fluid Dynamics (CFD) codes is broadening the range of problems being addressed. Multi-phase flow modeling techniques are becoming computationally stable and industrial applications are expanding. This paper focuses on modeling gas/liquid mixing in a stirred tank. This type of multi-phase reactor is widely used in the chemical and pharmaceutical industries. The performance of these reactors can be dependent upon the interfacial surface area. The objective of this work is to evaluate the current capabilities of various multiphase models for predicting the gas/liquid contacting. Accurate predictive models facilitate the screening of unconventional gas/liquid reactor designs, where full-scale trials can be cost prohibitive. Modeling results are compared against experimental data reported in 1977 by Smith and others. These experiments demonstrated the effect of impeller speed and gassing rate on the gas-liquid mass transfer rate, kLa. A standard baffled stirred tank configuration was used (H/T = 1, single Rushton impeller, D/T = 1/3). The data was compared to predictions for three different multiphase approaches. The first model evaluated was the steady Algebraic Slip Model (ASM) implemented in Fluent 5. The second model was a steady Euler/Euler formulation using CFX4. The third approach was a transient Euler/Euler method in Fluent 4. These algorithms employ a single bubble size in the dispersed-phase momentum equations. The effect of variation of the dispersed-phase bubble diameter is shown for the shear-dependent mean-size (SDMS) and the Multiple Size Group (MUSIG) models. The interphase interaction term for all models is commonly calculated using the assumption of rigid spherical bubbles. The effect of bubble distortion is illustrated by using an empirical correlation for bubble rise velocity. Both qualitative and quantitative comparisons are made between the experimental and numerical results. The qualitative comparisons examine the various flow regimes. The quantitative evaluations involve the prediction of interfacial area, overall gas hold-up, and the ratio of gassed impeller power to ungassed impeller power. These comparisons show the bubble size modeling have a significant positive impact on the prediction of gas-liquid contacting.

This content is only available via PDF.
You do not currently have access to this content.