Experimental investigations of pump‐driven closed‐loop thermosyphon system

Abstract

The thermal management of the compact electronic system is one of the major challenges in the present power electronics and computational industries. The present paper investigates the effects of a pump-driven flow on the thermal performance of a closed-loop thermosyphon (CLT) system. This paper discusses the effect of pump-driven flow on the thermal performance of CLT. In this study, the experimentation is carried out on the water-charged pump-driven closed-loop thermosyphon (PDLT) with different heat inputs, filling ratios (FR), and adiabatic lengths to understand the effects of these parameters on the thermal performance of the system. The results indicate that the heat transfer performance of CLT is improved using pump-driven flow in the PDLT and the minimum thermal resistance (0.035 k/W) is obtained at FR = 0.6 and 3 kW heat input. It is also noticed that the thermal resistance of PDLT is up to 35% higher than CLT at FR = 0.6, 0.5 kW heat input, and 500 mm adiabatic length. The unstable geyser boiling phenomenon is eliminated from the system at the adiabatic lengths of 200 and 500 mm. However, for long adiabatic length (800 mm), the geyser boiling occurs at a higher FR (FR = 0.6) and moderate heat flux (0.5 kW).

Machine learning approach to predict the thermal performance of closed‐loop thermosyphon

“…In linear regression, there is a straight inclined relation among the variables. The outputs are also named targets, which are represented by a linear combination of input variables or features 14 …”

“…The objective of this research is to predict CLT system performance parameters using different ML algorithms and compare their result. [10][11][12][13][14][15] 2 | METHODOLOGY…”

“…The outputs are also named targets, which are represented by a linear combination of input variables or features. 14 F I G U R E 2 Methodology for machine learning (ML) algorithm development.…”

“…In this paper, ML methods are used to predict the relationship between input parameters and output performance parameters. The objective of this research is to predict CLT system performance parameters using different ML algorithms and compare their result 10‐15 …”