Graduation Semester and Year

2015

Language

English

Document Type

Dissertation

Degree Name

Doctor of Philosophy in Aerospace Engineering

Department

Mechanical and Aerospace Engineering

First Advisor

Dereje Agonafer

Abstract

Continually increasing demand for information technology (IT) applications and services has provided sustained growth and interest in data centers. The large amounts of energy consumed by data center facilities have placed a significant emphasis on the energy efficiency of their overall operation. One area of particular importance is the cooling energy required. Heat generation within a data center starts at the server level, specifically within the microelectronic devices that process digital information. Convective heat transfer is the primary driver for the removal of heat from an individual server. As such, cooling efficiency at the server level will be dictated by the pumping power required to move a cooling fluid through the system. Many methods are available for removing heat from the server, either with air or liquids as the cooling medium. This work evaluates new, efficient approaches for removing that heat and the pertinent design considerations that must be taken into account for successful implementation. In general, smaller fans operate at lower efficiencies than larger fans of proportional linear dimensions. The applicability of replacing smaller, 60mm fans from within the chassis of web servers with an array of either 80mm or 120mm fans consolidated to the back of a rack is experimentally tested. Initial characterization of the selected fans showed the larger 80mm and 120mm fans operate at double peak total efficiency of the smaller 60mm fans. A stack of four servers was used in a laboratory setting to represent a rack of servers. When all four servers were stressed at uniform computational loadings, the 80mm fan array resulted in between 50.1% to 52.6% reduction in total rack fan power compared to the baseline 60mm fans. The 120mm fan array showed similar reduction in rack fan power of 47.6% to 54.0% over the baseline 60mm fan configuration. Since actual data centers rarely operate at uniform computational loading across servers in a rack, a worst case scenario test was conceived. In this test, the arrays of larger fans were controlled by a single server operating at peak computational workload while the other three in the rack remained idle. Despite significant overcooling in the three idle servers, the 80mm and 120mm fan configurations still showed 35.3% and 33.8% reduction in total rack fan power compared to the best possible operation of the 60mm fans. The findings in this study strongly suggest that a rack-level fan scheme in which servers share airflow is more efficient alternative to fans contained within the server.Air flow management is a critical tool to maintain efficient operation of a data center cooling scheme. Provisioning of airflow from CRAC units and containment systems often lead to changes in the static pressure at the inlet to server racks. Through experimental testing on an Air Flow Bench it is observer that static pressure at the inlet to servers has a significant influence on the thermal performance and fan cooling energy consumption within the server itself. Reduction in server fan power or component temperatures can be achieved by increasing the static pressure at the server inlet. Complementary design and control at the room level with this information at the server level can lead to reduction in overall system fan power and more energy efficient data center operation.Complete immersion of servers in dielectric mineral oil has recently become a promising technique for minimizing cooling energy consumption in data centers. However, a lack of sufficient published data and long term documentation of oil immersion cooling performance makes most data center operators hesitant to apply these approaches to their mission critical facilities. In this study, a single server was fully submerged horizontally in mineral oil. Experiments were conducted to observe the effects of varying the volumetric flow rate and oil inlet temperature on the thermal performance and power consumption of the server. Specifically, temperature measurements of the CPUs, motherboard components, and bulk fluid were recorded at steady state conditions. These results provide an initial bounding envelope of environmental conditions suitable for an oil immersion data center. Comparing the results from baseline tests performed with traditional air cooing, the technology shows a 34.4% reduction in the thermal resistance of the system. The cooling loop was able to achieve partial power usage effectiveness (pPUECooling) values as low as 1.03. This server-level study provides a preview of possible facility energy savings by utilizing high temperature, low flow rate oil for cooling. Following this, visual observations, microscopic measurements, and testing of mechanical properties were taken. Evaluation of the technology’s impact on the mechanical reliability of components and operability of data centers is made.

Disciplines

Aerospace Engineering | Engineering | Mechanical Engineering

Comments

Degree granted by The University of Texas at Arlington

Download

Share

COinS