Study on Bubble Entrapment Mechanism in a Laminar Rotational Mixing System

Abstract

Industrial mixing is a fundamental operation across diverse engineering sectors, essential for achieving spatial homogeneity and enhancing heat and mass transfer. While turbulent mixing efficiently drives homogenization in low-viscosity fluids, processing highly viscous media dampens inertial forces, strictly confining the hydrodynamics to the laminar regime. Under these conditions, radial impellers frequently generate Isolated Mixing Regions (IMRs), which are stable, non-mixing toroidal structures that topologically isolate fluid elements from the active, chaotic mixing regions. Achieving efficient homogenization is severely hindered by these IMRs, which resist traditional suppression techniques without incurring prohibitive mechanical and energetic costs. Furthermore, the introduction of dispersed phases (bubbles and particles) significantly complicates the flow field. Driven by a non-linear interplay of buoyancy, drag, lift, and add-mass force, bubbles exhibit predictable cross-streamline migration directly into these invariant boundaries. Despite extensive qualitative study, current entrapment models predominantly focus on the inertial clustering of neutrally buoyant particles. A critical gap remains in understanding and harnessing buoyancy-driven entrapment for large, non-passive dispersed phases in viscous environments. This dissertation aims to quantitatively characterise IMRs and shift the stereotype from viewing them as destructive barriers to utilising them as highly specific, positive control mechanisms for multiphase distribution.

First, we demonstrate stable, buoyancy-dominant bubble entrapment in a conventional radial mixing tank. Diverging from prior inertial clustering models that rely on negligible external forces (gL/U²≪1), we establish that accumulation is not restricted to sub-millimetre, neutrally buoyant particles. By actively controlling the injection position and impeller speed, we achieved reproducible entrapment of macro-scale (millimetre to centimetre) bubbles, demonstrating their prolonged residence time and stability within the invariant boundaries. Through experimental tracking of the bubble trajectory across the entrapment process, we elucidate the underlying commonalities between IMRs, static vortices, Taylor-Couette flow and solid-body-rotating flows. Ultimately, a critical mapping between bubble size and impeller speed is developed, decoding the macroscopic trapping mechanisms to directly advise future industrial parameters.

Second, we employ Direct Numerical Simulation (DNS) to bypass the spatial limitations of experimental observation, comprehensively resolving the full three-dimensional laminar flow field, and providing a quantitative analysis of the IMR velocity scales across varying impeller speeds. To address the experimental limitations of conventional 2D Particle Image Velocimetry (PIV) in capturing obscured multiphase interfaces, we utilise advanced front-tracking level-set methods. These multiphase simulations capture highly localised, previously unrecorded dynamics, including bubble deformation and interfacial velocity distribution. By coupling these decomposed, quantified velocity scales with a bubble force balance, we develop a theoretical trapping criterion to accurately predict critical bubble size thresholds, which is subsequently validated against our experimental data.

In summary, this dissertation establishes a comprehensive framework for understanding and harnessing the interaction between IMRs and dispersed gas bubbles in highly viscous laminar flows. Through a tightly integrated approach utilising experimental tracking, mathematical modelling, and advanced computational fluid dynamics, this work identifies the direct correlations between hydrodynamic topology and practical engineering design parameters. By demonstrating that IMRs can be precisely predicted and actively leveraged, our study opens a new avenue for optimising multiphase reactor designs, pipeline junctions, and continuous separation units.

Date of Award	10 Jul 2026
Original language	English
Awarding Institution	City University of Hong Kong
Supervisor	Steven WANG (Supervisor)