Abstract
Data centers are becoming more powerful and more integrated with the continuous development of smart cities, which brings us more technological convenience, but also generates a large amount of waste heat. At present, the efficient and green cooling scheme is one of the key researches and development points to ensure the stable and safe operation of power electronic devices and achieve energy saving and consumption reduction. As a branch of the heat pipe, the pulsating heat pipe is one of most promising passive cooling techniques among many candidates for its unique advantages such as small size, simple and compact structure, and high heat dissipation efficiency, but its application in data centers just begins, and there are few reports on research and implementation. Based on the introduction of the basic structure, working mechanism and outstanding advantages of pulsating heat pipes, this paper reviews in detail the researches on the factors affecting its performance, so as to evaluate the possibility of using pulsating heat pipes in data centers. Finally, the latest application and development of pulsating heat pipes applied to heat dissipation of high-power CPUs are summarized, which can provide a guidance for subsequent research and engineering application.
Highlights
With the fast development of the Artificial Intelligence, internet communication, high performance computing, and big data analysis, especially in the context of the global fight against the COVID-19 epidemic in 2020, data center with a large amount of information processing, storage and computing is becoming a hot spot for a new round of investment and construction
The pressure difference on both ends will be reduced, and the pulsating frequency and amplitude will be correspondingly reduced, resulting in the difficulty of startup; 2) Larger saturation pressure change value with saturation temperature sat. This will cause the working fluid to produce larger pressure fluctuations under smaller temperature changes, which is conducive to oscillating heat transfer; 3) Smaller viscosity can reduce the shear stress along the tube wall, thereby reducing the pressure drop in the pipe, which will reduce the input heat load for maintaining the pulsating flow
The results show that the surfactant solution is favorable to the operation of pulsating heat pipe (PHP) at the concentration of 10 ppm under all experimental input heat loads
Summary
With the fast development of the Artificial Intelligence, internet communication, high performance computing, and big data analysis, especially in the context of the global fight against the COVID-19 epidemic in 2020, data center with a large amount of information processing, storage and computing is becoming a hot spot for a new round of investment and construction. The study found (Zhang and Faghri, 2008) that only when Bo ≤ 2, that is, when the gravity is less than the surface tension, stable vapor slugs and liquid plugs will be formed in PHPs. The smaller pipe diameter results in smaller liquid plugs amplitude, higher frequency, greater heat transfer per unit area, and higher thermal conductivity. The pressure difference on both ends will be reduced, and the pulsating frequency and amplitude will be correspondingly reduced, resulting in the difficulty of startup; 2) Larger saturation pressure change value with saturation temperature (dP/dT) sat This will cause the working fluid to produce larger pressure fluctuations under smaller temperature changes, which is conducive to oscillating heat transfer; 3) Smaller viscosity can reduce the shear stress along the tube wall, thereby reducing the pressure drop in the pipe, which will reduce the input heat load for maintaining the pulsating flow. Rahman et al (2015b) studied the effect of open-loop PHP performance at different filling rates using deionized water and acetone as working fluids, and found that the performance reaches the best when the filling rate of acetone and water is 0.7 and 0.5 respectively. Verma et al (2013) concluded that the optimal filling rates of deionized water and methanol are 0.5 and 0.4, respectively, and methanol has a shorter startup time and a lower startup temperature. Shi and Pan (2017) found that when the TABLE 2 | Summary of experimental investigations on factors affecting the performance of PHPs
Sign-in/Register to view highlights & summary Sign-in/Register to access full text options 
Free) 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call
