CFD–PBM simulation of two-phase flow in a pulsed disc and doughnut column with directly measured breakup kernel functions

Abstract

The pulsed disc and doughnut column (PDDC) is a widely used type of liquid–liquid extraction equipment. Both the research and design of this type of equipment require in-depth knowledge of the dispersed phase droplet size distribution (DSD). The coupling of computational fluid dynamics (CFD) and population balance model (PBM) has exhibited a significant potential for predicting the DSD by considering droplet breakage and coalescence. In the present work, the CFD–PBM simulation is adopted to predict the DSD, and to investigate local liquid–liquid flow behaviors in a square-sectioned PDDC. The empirical correlations of the breakup kernels, which are based on our previous work [Hao Zhou et al., Direct measurement of droplet breakage in a pulsed disc and doughnut column. AIChE Journal, 2017, 63(9), 4188–4200], are implemented in the CFD code. For comparison, the kernel functions obtained by Coulaloglou and Tavlarides [Coulaloglou and Tavlarides. Description of interaction processes in agitated liquid–liquid dispersions. Chemical Engineering Science, 1977, 32, 1289–1297] are also tested. The class method with the fixed pivot technique is utilized to solve the population balance equations. The droplet size distribution predicted by the CFD–PBM simulation combined with the kernel functions that we have previously determined experimentally are observed to be more precise than those of Coulaloglou and Tavlarides. Accordingly, the CFD–PBM simulation results with the kernel functions of our proposed correlations are used to investigate the liquid–liquid flow behaviors in the PDDC. The influences of droplet size range and initial droplet size distribution at the dispersed phase inlet in the predicted DSD are discussed. The hydraulic performance of simulation results agrees well with experimental data. Local flow behaviors are illustrated to depict the typical flow characteristics of PDDC. The simulation results suggest that CFD–PBM method with the proposed PBM kernel functions is a promising approach for the simulation of liquid–liquid flow in the PDDC.