Abstract
The ability of traditional room-conditioning systems to accommodate expanding information technology (IT) loads is limited in contemporary data centers, where the storage, storing, and processing of data have grown quickly as a result of evolving technological trends and rising demand for online services, which has also led to an increase in the amount of waste heat generated by IT equipment. Through the implementation of hybrid air and liquid cooling technologies, targeted, on-demand cooling is made possible by employing a variety of techniques, which include but are not limited to in-row, overhead, and rear door heat exchanger cooling systems. The ongoing trends of rising power densities of microprocessors and the presence of hot spots present a considerable challenge to the air-cooling system, which pushes it to its limits. As a result of the significant rise in IT power densities, single-phase cold plate cooling is becoming increasingly popular. However, the limited availability of a source of chilled water or constrained air distribution routes and the unavailability of raised floors are the main obstacles to the adoption of single-phase cooling technologies in many old data centers, which led to the use of liquid-to-air cooling distribution units (CDUs). These types of CDUs effectively cool servers by distributing liquid coolant into cooling loops that are installed on top of each server. The CDU pumps cold liquid to cooling loops attached to servers or chips, where they absorb heat. After returning the heated fluid to the CDU, the heat exchanger (HX) uses the cold air to eject the by the use of fans.
One of the most common liquid cooling techniques will be examined in the current study based on different conditions for high power density racks (+50 kW). This paper investigates the cooling performance of a liquid-to-air in-row CDU in a test rack containing seven thermal test vehicles (TTVs) under various operational conditions, focusing on cooling capacity, HX effectiveness, and TTV case temperatures under different supply air temperatures (SAT). This test rig has the necessary instruments to monitor and analyze the experiments on both the liquid coolant and the air sides. Moreover, another experiment is conducted to assess the performance of the CDU that runs under different control fan schemes, as well as how the change of the control type will affect the supply fluid temperature and the TTV case temperatures at 10%, 50%, and 100% of the total delivered power. Finally, suggestions for the best control fan scheme to use for these systems and units are provided at the conclusion of the study.
