(554g) Numerical Simulation of Droplet Size Distribution in Pulsed Disc and Doughnut Column with Wettable Internals

In the present work, droplet size distributions of the liquid–liquid two-phase flow in the pulsed disc and doughnut columns (PDDC) are numerically investigated. This is achieved by coupling the computational fluid dynamics (CFD) and population balance model (PBM). The key points of the CFD–PBM simulation method is to determine the kernel functions of the PBM. The kernel functions are used to describe the birth and death events of the dispersed droplets in the liquid–liquid two-phase flow. Because the internals of the PDDC, the liquid-liquid extraction equipment, are wetted by the dispersed phase, the traditional PBM kernel functions are unable to describe the droplet behaviors adequately.

In view of the drawbacks of most present kernel functions, the birth and death events of droplets in a laboratory-scale PDDC of wettable internals are observed firstly by virtue of the online high-speed camera. It is found that droplets will break up and daughter droplets will be created after droplet breakage. To describe this breakage event, the droplet breakage frequency function, b(d), and the daughter size distribution function, β (d|d_m), should be properly set. In addition, because the internals are wetted by the dispersed phase, liquid layers are found on the internal surfaces. It is observed that droplets will coalesce into the liquid layer and a residual droplet will form after droplet-layer coalescence . Moreover, droplets will also be created by the dispersion effect of the liquid layers. To describe these events properly, the droplet–liquid layer coalescence rate, G(d), the residual droplets size distribution function, H(d|d_m), and the droplet dispersion frequency function, Λ(d), are required to be determined. Then, based on tremendous experiment data, the kernel functions describing these droplets birth and death events are correlated with operational and physical parameters.

These kernel functions are then imbedded into the CFD code to calculate the droplet size distributions in the laboratory-scale PDDC of wettable internals. The simulation results are found to accord well with the experimental results. Finally, the established CFD–PBM simulation method is applied to predict the droplet size distributions in a pilot-scale PDDC of wettable internals. The simulation results suggest that the pulsation intensity is the main factor to affect the droplet size distributions.