Bubble Trapping in Two-Phase Wakes from a Liquid-Gas Flow around a Cylinder
Open Access
- Author:
- Kim, Dohwan
- Graduate Program:
- Mechanical Engineering (PHD)
- Degree:
- Doctor of Philosophy
- Document Type:
- Dissertation
- Date of Defense:
- October 10, 2022
- Committee Members:
- Alexander Rattner, Major Field Member
Ali Borhan, Outside Unit, Field & Minor Member
Fan-Bill Cheung, Major Field Member
Matthew Rau, Chair & Dissertation Advisor
Stephen Lynch, Major Field Member
Daniel Haworth, Professor in Charge/Director of Graduate Studies - Keywords:
- Two-Phase Flow
Particle Tracking Velocimetry
Particle Image Velocimetry
Bubbly Flow
Bubble Trapping
High-Speed Visualization
Flow around a Cylinder - Abstract:
- Small tubes and fins have long been used as methods to increase surface area for convective heat transfer in single-phase flow applications. As demands for high heat transfer effectiveness have increased, implementing evaporative phase-change heat transfer in conjunction with these methods to increase surface area in advanced heat exchanger and heat sink designs has become increasingly attractive. However, the complex two-phase flow that results from these configurations is poorly understood, particularly in how the gas phase interacts with the flow structure of the wake created by these bluff bodies. This experimental research was conducted to understand the bubbles in a liquid-gas around a cylinder interact with the flow structure. In Chapter 2, a vertical water channel was developed to conduct an experimental study of liquid-gas flow around the cylinder. The bubbly flow movement around the cylinder was visualized with a high-speed camera and varying liquid Reynolds numbers from Re = 99 and Re = 2,956 and air superficial velocities varied from jg = 0.06 m/s to jg = 0.60 m/s. The mean bubble diameter observed during the experiment varied from 0.5 mm to 3.5 mm. Time-averaged images were examined to calculate the local void fraction values in the two-phase wake. A liquid-phase region with characteristically low void fractions and a bubble-trapping region with characteristically high void fractions could be easily determined by calculating the ratio between local void fraction values in the wakes and the freestream void fraction values. The liquid-phase region occurred throughout the experimental investigations when the Reynolds number was varied from Re = 99 to Re = 2,956. The overall length of the liquid-phase region decreased with increasing Reynolds number as the bubble-trapping region appeared when the Reynolds number was greater than 2,000. However, the bubble-trapping region also occurred at a lower Reynolds number of Re = 1,182 when the mean bubble diameter was reduced from 3.5 mm to 0.5 mm. In Chapter 3, a force balance model was developed to explain the occurrence of the bubble-trapping region. The high-speed images collected from the experimental facility were processed to detect the air bubbles and calculate their trajectories using particle tracking velocimetry and particle image velocimetry algorithms. In addition to the bubble velocities, the liquid velocities were also calculated by tracking the movement of bubbles and neutrally-buoyant flow tracers from the high-speed images using particle image shadow velocimetry. The phase-resolved velocities confirmed the formation of the bubble-trapping region behind the cylinder, where air bubbles were concentrated in the cylinder wake. To explain the bubble movement in the two-phase wakes, a time-averaged force balance equation was derived. The force balance equation and the phase-resolved velocity results revealed that strong inertial and lift forces were present when the Reynolds number was greater than 2,000. In addition, the reduced-order force balance analysis suggested that the bubble-trapping occurred at a lower Reynolds number of Re = 1,182 with a mean diameter of 0.5 mm due to a strong lift force acting on the small air bubbles despite the relatively weak inertial forces. In Chapter 4, further visualization of bubbly flow movement with the flow tracers was conducted by focusing Reynolds number from Re = 493 to Re = 2.463 at an air superficial velocity of jg = 0.36 m/s to explain the importance of the lift force on the bubble-trapping region at a transitional Reynolds number. The result from the flow measurements revealed that the bubble-trapping forces are composted in horizontal and vertical directions. The horizontal component of the inertial and lift forces attracted bubbles toward the centerline of the water channel; the vertical component of the inertial and lift forces pulled the air bubbles downward to hold them in the bubble-trapping region. The analysis discovered that the bubble-trapping forces act in two distinct locations based on their direction. Bubbles traveling around the cylinder faced strong horizontal forces at a downstream location of y/D = 1.0, which induced the bubble movement towards the centerline. Then, the bubbles faced strong vertical forces at a downstream location of y/D = 1.5, slowing down their vertical velocity and remaining them in the bubble-trapping region momentarily. In addition, the force balance analysis showed that lift forces also played a significant role in the formation of the bubble-trapping region. The lift force on the bubbles was also present in the horizontal and vertical directions; however, it was more important for explaining the bubble motion in the horizontal direction. Additionally, the comparison of liquid velocities in the two-phase and single-phase crossflows was made at Re = 2,946 to determine the effects of air bubbles in the flow. The transient analysis showed that the flow around the cylinder could be periodic; however, the velocity measurements are suitable for the time-average force balance analysis. It also revealed that the injection of air bubbles increased the mean and fluctuating liquid velocities in the wakes.