Effect of crossflow velocity on underwater bubble swarms

• Published in: International journal of multiphase flow 2018-08, Vol.105, p.60-73
• Main Authors: Xu, Yang, Aliyu, Aliyu Musa, Seo, Hyunduk, Wang, Jin-Jun, Kim, Kyung Chun
• Format: Article
• Language: English
• Subjects: Bubble swarms; Bubble trajectory; Bubbly flow; Crossflow; Visualization
• ISSN: 0301-9322; 1879-3533
• DOI: 10.1016/j.ijmultiphaseflow.2018.03.018

•Bubble swarm experiments performed in a channel with water crossflow.•High-speed visualization at various gas flow rates and liquid velocities.•Image processing used to extract trajectory and velocity of swarm.•Bubble motion is quasilinear and correlation developed to model trajectory.•Improved velocity correlation derived to capture our wide range of flow conditions. We investigate the effect of crossflow velocity on submerged bubble plumes or swarms by employing the use of high-speed photography and an image-processing method to measure bubble rise velocities. Particle image velocimetry (PIV) was used to accurately determine the crossflow freestream velocity as well as boundary layer information. We varied the gas flow rates from 2 to 25 L/min. This range exceeds those of previous studies we could find in the open literature which were mostly less than 5 L/min and involved isolated bubbles. Combined with the crossflow velocities, this resulted in the investigation of a wide range of flow conditions providing a database of 36 experimental data points and constitutes a substantial addition to the bubble swarm/crossflow literature. Because our experiments involved larger gas flow rates than previously reported, we had to develop a digital image-processing algorithm using standard functions in Matlab to measure swarm rise velocities, and angles of inclination under crossflow. Results were validated against reported data at similar experimental conditions. It was established that increasing freestream velocity strongly suppressed bubble rise velocities and resulted in bubble breakup. Relationships for predicting rise velocity and inclination angle were derived as non-dimensional functions of the crossflow velocity, fluid properties and inlet gas flow rates. These showed good agreement with the current as well as reported experimental data.