Simulation of bubbly flow in non-Newtonian fluids using computational fluid dynamics and population balance model based on the entire turbulent energy spectrum

Abstract

To investigate the effects of interfacial force models and the coalescence and breakup models for bubbly flow simulation in non-Newtonian fluids, visualization experiments were conducted in a rectangle bubble column. Particle image velocimetry and shadowgraphy were employed to measure bubbly flow in shear-thinning fluids. By extending the kernel functions to the entire turbulent energy spectrum, the simulation accurately captured bubble coalescence and breakup behaviors, leading to improved predictions of bubble size distribution. The results indicate that the Liao and Lehr models underestimated bubble coalescence frequency, while the Luo model overestimated it. The Prince–Blanch model provided the most accurate description of bubble size and distribution. The primary mechanisms driving bubble coalescence in the system were identified as buoyancy-driven motion, wake entrainment, and viscous shear. A comparison of drag force models developed for non-Newtonian fluids in previous literature revealed that, at low superficial gas velocities, the differences in predicted velocity fields and bubble sizes among various drag models were minimal. However, the HosenB and Li models predicted larger liquid velocity. As the superficial gas velocity increased, these discrepancies became more pronounced, with all models except HosenB underestimating the velocity. Additionally, other interfacial forces, including lift force, turbulent dispersion force, and wall lubrication force, were compared, revealing that lift force and turbulent dispersion force significantly influenced the simulation outcomes.