Experimental study of bubble breakup in water and solid suspension by using the image‐based method

Abstract

This work investigates the bubble breakup process with and without particles in turbulent conditions using the image‐based method. A binocular high‐speed camera was employed to capture breakup events. A deep learning‐based image identification software (Large Deformation Dispersed Phase Analysis in Multiphase Flows) and a highly deformed bubble volume/surface area quantification method (Dense Adaptive Segmentation Method) are proposed. An energy barrier is found during the bubble breakup process, with the maximum increase in surface area (ΔSmax) being two to three times the final increase after breakup (ΔSfinal). This indicates that the critical energy required for bubble breakup is underestimated in most breakup models. The presence of suspended particles raises this energy barrier, thus reducing the breakup probability. The daughter bubble size distribution follows an M‐type distribution in water, while the addition of particles leads to a tendency towards equal‐size breakup. This work provides a reliable technology and the experimental data for further clarifying the bubble breakup mechanism.