Development of the AI-Assisted Thermal Hydraulic Analysis Method for Condensing Bubbles in Vertical Subcooled Flow Boiling

Abstract Subcooled flow boiling plays a pivotal role in various industrial applications, including nuclear reactors and thermal management systems. However, the rapid and complex changes of condensing bubbles experience from their inception to collapse in subcooled flow boiling present substantial obstacles for conventional bubble detection methods, especially when it comes to condensation bubbles. In light of this, a state-of-the-art AI method is developed and validated for the efficient detection and tracking of condensation bubbles in subcooled flow boiling, thereby enabling the effective execution of thermal hydraulic analyses. This study initially employs computer vision technology to efficiently construct a bubble dataset. Subsequently, a bubble detection model is trained based on this dataset and the YOLOv8 with attention mechanism. Utilizing the results of bubble detection, a multi-object tracking algorithm is then applied to track all bubbles in each frame. The developed AI-based method exhibits a strong ability to detect 95% of condensation bubbles in subcooled flow boiling, additionally, it streamlines the extraction of critical thermal hydraulic parameters, including void fraction, bubble lifetime, and nucleation site density. The model’s accuracy and consistency is demonstrated compared to empirical correlations, affirming its reliability in analyzing the intricate dynamics of subcooled flow boiling.