Investigation of coalescence and breakup of bubble in packed-bed

Daeseong Jo, Purdue University

Abstract

Model of bubble interactions of breakup and coalescence in a two-dimensional packed-bed reactor has been developed for dispersed two-phase flow conditions. Image processing techniques are used to study dominant bubble mechanisms at pore level under the bubbly flow regime. Bubble breakup and coalescence are identified as dominant mechanisms from analysis using a large number of image samples. Two types of coalescence mechanisms are identified that occur due to compression and deceleration associated with the bubbles and three breakup mechanisms are identified that are result of liquid shear force, bubble acceleration, and bubble-to-bubble impact. The two-dimensional packed-bed system is designed and built with two different inlets: one is for uncontrolled sized bubbles, and the other is for controlled sized bubbles. For uncontrolled sized bubbles, bubble images are taken at a single axial location in order to study changes of two-phase parameters as well as bubble size PDF (population density function) distribution with different gas and liquid superficial velocities. For controlled sized bubbles, bubble images are taken at multiple axial locations in order to study changes of bubble size PDF distribution from inlet to the far downstream. Bubble breakup and coalescence dominated flow regimes near the inlet have been simulated by adjusting appropriate flow rates of air and water. Data on various two-phase parameters, such as local void fraction, bubble velocity, size, number, and shape are obtained from assessment of the bubble images. Results indicate that when a flow regime changed from bubbly to either trickling or pulsing flow, the number of average sized bubbles significantly decreases and the shape of the majority of the bubbles is no longer spherical. Although a mean bubble velocity of all sized bubbles is uniform for given gas and liquid superficial velocities, individual bubble velocities are quite different depending on the bubble location in the pore. The bubble size PDF distributions taken for the uncontrolled sized bubbles are compared with previous studies and the results on bubble size are in general agreement. Bubble size PDF distributions under either bubble breakup or coalescence dominated flow regimes change rapidly near the inlet and slowly far away from the inlet. The median as a function of axial locations show how the bubble size distributions propagate along the axial direction. The different initial medians of the bubble breakup and coalescence dominated flows reach the same final median far downstream. Based on the experimental observations, bubble interaction models to predict bubble size PDF distribution along the axial direction are developed by modifying existing models as well as adding new geometry effects due to the complex and narrow channels. For breakup, an additional pressure term is added and surface tension stress is modified due to bubble-to-solid collision. For coalescence, probability of bubble collision, and cross sectional area of bubbles are modified due to restricted flow channels. Pressure ratios enhance the breakup and the coalescence near packing. A velocity ratio enhances the coalescence at vertical narrow channels. The developed mechanistic bubble interaction models are implemented in a commercial CFD code, CFX-4.0, which solves population balance equations for dispersed gas-liquid flows. To validate the present bubble interaction models of breakup and coalescence, bubble breakup and coalescence dominated flow regimes near the inlet are separately simulated and the bubble size PDF distributions estimated by CFD analyses are compared with experimental data. As a result, the estimated bubble size distributions for the bubble breakup and coalescence dominated flows are comparable with each other and their median variations along the axial direction agree with the experimental data. With the adjusted coalescence coefficients, the estimated medians fall within the uncertainties.

Degree

Ph.D.

Advisors

Revankar, Purdue University.

Subject Area

Chemical engineering|Nuclear engineering

COinS