Flow Visualization In Closed Loop Pulsating Heat Pipe (CLPHP) Using Deep Learning Techniques

¹ Department of Mechanical Engineering, Matrusri Engineering College, Hyderabad, India
² Department of Mechanical Engineering, Jawaharlal Nehru Technological University Hyderabad, Telanagana - 500085, India
³ Department of Computer Science of Engineering, Pydah College of Engineering and Technology, Visakhapatnam, Andhra Pradesh - 531163, India
⁴ Department of Computer Science Engineering, Vignan’s Institute of Engineering for Women, Visakhapatnam, Andhra Pradesh - 530049, India

^* Corresponding author: santhinerella@matrusri.edu.in

Abstract

The present work describes an alternative method for recognizing and tracking the flow in PHP which is based on visualization to address the issue of examining individual flows graphically. For this an experimental analysis which was already carried out on an 8-turn Closed Loop Pulsating Heat Pipe (CLPHP) with copper tube capillary dimensions with 50% fill ratio in vertical mode changing the range of heat inputs by using water as the working fluid is taken as reference. The time and cost required for experimentation and simulation is significantly reduced by using deep learning method. YOLOv5 based Convolutional Neural Network (CNN) is used to identify and classify the flow in CLPHP. The results obtained by CNN are compared with existing experimental and CFD results.