The growing demand for data center facilities has made intelligently managed data center operations necessary. For temperature measurement and thermal management, a common practice is to install a limited number of temperature sensors evenly distributed throughout the room. However, data center operators rarely fully equip facilities with temperature sensors due to their cost, complexity, and maintenance requirements, creating vacancies in the data center temperature and cooling picture. The local nature of sensor data can also be misinterpreted and misused. Without novel methods to interpret and visualize temperatures obtained by prediction or measurement, data center operators cannot easily identify urgent local cooling issues or quickly examine the temperature at other location. This paper presents methods to predict a full three-dimensional temperature field in data centers from a limited number of measurement points. Several different statistical interpolating schemes are discussed. We also validate the interpolated temperature fields against benchmark data from Computation Fluid Dynamics (CFD) and show good agreement.
- Electronic and Photonic Packaging Division
Measured-Temperature Interpolation and Visualization for Data Centers
- Views Icon Views
- Share Icon Share
- Search Site
Zhang, X(, Healey, CM, Sheffer, ZR, & VanGilder, JW. "Measured-Temperature Interpolation and Visualization for Data Centers." Proceedings of the ASME 2013 International Technical Conference and Exhibition on Packaging and Integration of Electronic and Photonic Microsystems. Volume 2: Thermal Management; Data Centers and Energy Efficient Electronic Systems. Burlingame, California, USA. July 16–18, 2013. V002T09A006. ASME. https://doi.org/10.1115/IPACK2013-73081
Download citation file: