A tracking algorithm of vapor-liquid interfaces for pulsating heat pipe: Theoretical and experimental validation at liquid nitrogen temperature

Abstract

Pulsating heat pipes (PHPs) have a wide range of applications in electronics cooling, space exploration, energy conversion, and many other fields due to their efficient heat transfer capability. Studying the phase change process, flow pattern distribution, and flow velocity of PHPs is essential to reveal the heat transfer mechanism. The visualization of PHPs helps the observation of flow patterns, but cannot measure the flow velocity quantitatively. In light of this, a multiple object tracking algorithm based on the visualization of a liquid nitrogen PHP is proposed to calculate all vapor-liquid interface velocity from the PHP working video. The tracking algorithm is realized through constrained orbit tracking and is designed to track all the vapor-liquid interfaces inside the PHP simultaneously. The averaged flow velocities of all the interfaces inside the PHP are calculated. The algorithm has high accuracy and good real-time performance. The accuracy of the algorithm is 99.5% through testing compared with Tracker, which helps to further reveal the heat transfer mechanism of PHPs. The principle of the tracking algorithm is also applicable to processing other two-phase flow videos for vapor-liquid interface tracking, which helps to further obtain information about the flow intensity of two-phase flow.