Experimental Study of a Novel Prototype of Dual-Cycle Two-Phase Cooling System for a Data Cabinet

As Information Technology continues to rapidly evolve, the scale and energy consumption of data centers have seen a significant surge. Traditional air-cooling systems in data centers are notably energy intensive. This study proposes an innovative dual-cycle two-phase cooling system (DTCS) for retrof...

Full description

Saved in:
Bibliographic Details
Main Authors: Hao Cheng, Tongzhi Yang, Yifan Zhao, Leixin Wang, Weixing Yuan, Kexian Ren
Format: Article
Language:English
Published: MDPI AG 2025-02-01
Series:Applied Sciences
Subjects:
Online Access:https://www.mdpi.com/2076-3417/15/5/2386
Tags: Add Tag
No Tags, Be the first to tag this record!
Description
Summary:As Information Technology continues to rapidly evolve, the scale and energy consumption of data centers have seen a significant surge. Traditional air-cooling systems in data centers are notably energy intensive. This study proposes an innovative dual-cycle two-phase cooling system (DTCS) for retrofitting existing server cabinets. The system integrates the two synergistic subsystems of a pump-driven chip-level two-phase cooling system (PCTCS) and a pump-driven backplane-level two-phase cooling system (PBTCS). The PCTCS provides year-round natural cooling of high-power chips, and experimental results indicate that even under extreme outdoor conditions of up to 42 °C, the PCTCS can maintain the chip temperature at 76 °C. At the same time, the PBTCS effectively cools all components in the cabinet except the high-power chips. By efficiently controlling the chip temperature through the PCTCS, the novel DTCS ensures stable operation at data room temperatures up to 33 °C, thereby significantly reducing the energy consumption of the cooling system. The average annual pPUE of the cooling system is 1.078 at the test site in Xiangyang, Hubei province in China. The DTCS could perform well across various Chinese cities even under severe hot and moist conditions, with the average annual pPUE consistently below 1.1.
ISSN:2076-3417