Numerical investigation of the interphase forces and turbulence closure in 3D square bubble columns

Abstract

3D time dependent numerical study has been performed to predict the flow hydrodynamics in bubble columns by employing explicit algebraic Reynolds stress (EARSM), re-normalization group (RNG) and RNG bubble induced turbulence (BIT) k–ε models and the computational fluid dynamics (CFD) results are compared with experimental work of Deen (2001). This work shows the comparison between different drag force models (Ishii and Zuber, 1979; Schiller and Nauman, 1935; Grace et al., 1976) and drag coefficient). Moreover, the effects of the lift, virtual mass and suchlike other imperative non drag forces inclusive of wall lubrication as well as the turbulent dispersion have been examined. All the results show good quantitative agreement with experiments in case of turbulent kinetic energy, axial gas and liquid velocity profiles. EARSM k–ε has found to perform better than both the RNG k–ε modalities in estimating the turbulent kinetic energy profiles, but gave poor predictions of axial velocity profiles. Comparatively, RNG and RNG BIT models have successfully predicted the averaged flow field. Whereas the turbulent kinetic energy profiles at some locations, have been slightly under estimated by the RNG models. The inclusion of bubble induced turbulence led to slight improvement in estimation of the turbulent kinetic energy profiles. The advanced performance of RNG models stems in the better estimation of rate of turbulent dissipation and turbulent viscosity. Thus, both RNG and RNG BIT models can be readily utilized for the analysis of average flow fields and turbulent flow structures.