Technology Review: Data Center Cooling

(This article was featured in an issue of Qpedia Thermal e-Magazine, an online publication dedicated to the thermal management of electronics. To get the current issue or to look through the archives, visit http://www.qats.com/Qpedia-Thermal-eMagazine.)

Qpedia continues its review of technologies developed for electronics cooling applications. We are presenting selected patents that were awarded to developers around the world to address cooling challenges. After reading the series, you will be more aware of both the historic developments and the latest breakthroughs in both product design and applications.

Data Center Cooling

This article explores recent patents and technical advancements from the data center cooling industry. (Wikimedia Commons)

We are specifically focusing on patented technologies to show the breadth of development in thermal management product sectors. Please note that there are many patents within these areas. Limited by article space, we are presenting a small number to offer a representation of the entire field. You are encouraged to do your own patent investigation.

Further, if you have been awarded a patent and would like to have it included in these reviews, please send us your patent number or patent application.

In this issue our spotlight is on data center cooling solutions.

There are many U.S. patents in this area of technology, and those presented here are among the more recent. These patents show some of the salient features that are the focus of different inventors.

Hybrid Immersion Cooled Server with Integral Spot and Bath Cooling

US 7724524 B1, Campbell, L., Chu, R., Ellsworth, M., Iyengar, M. and Simons, R.

An immersion cooling apparatus and method is provided for cooling of electronic components housed in a computing environment. The components are divided into primary and secondary heat generating components and are housed in a liquid sealed enclosure. The primary heat generating components are cooled by indirect liquid cooling provided by at least one cold plate having fins. The cold plate is coupled to a first coolant conduit that circulates a first coolant in the enclosure and supplies the cold plate. Immersion cooling is provided for secondary heat generating components through a second coolant that will be disposed inside the enclosure such as to partially submerge the cold plate and the first coolant conduit as swell as the heat generating components.

The shortcomings of the prior art are overcome and additional advantages are provided through the provision of a method and associated hybrid immersion cooling apparatus for cooling of electronic components housed in a computing environment. The components are categorized as primary and secondary heat generating components and are housed in a liquid tight enclosure. The primary heat generating components are cooled by indirect liquid cooling provided by at least one cold plate having fins.

The cold plate is coupled to a first coolant conduit that circulates a first coolant in the enclosure and supplies the cold plate. Immersion or direct liquid cooling is provided for secondary heat generating components through a second coolant that will be disposed inside the enclosure such as to partially submerge the cold plate and the first coolant conduit as Well as the heat generating components.

In one embodiment, the cold plate and the coolant conduit each comprise extended external surfaces used to transfer heat from the second coolant to the first coolant. In one embodiment the second coolant cools the electronics via free convection or a combination of free convection and boiling while an alternate embodiment provides a sub merged pump to aid circulation. In yet another embodiment, the electronics are housed on an electronics board that is tilted at an angle to also aid circulation of the second coolant.

Thermal Caching For Liquid Cooled Computer Systems

US 7436666 B1, Konshak, M.

The present invention involves supplementing the liquid cooling path within a given rack of electronic components in order to reduce the temperature rise of critical components before damage or data loss can occur in the event of a primary cooling system failure. Liquid cooled racks generally have piping and heat exchangers that contain a certain volume of liquid or vapor that is being constantly replaced by a data center pump system. Upon a data center power or pump failure, the How stops or is diminished. The coolant, however, is still present within the rack components. Upon detection of a lack of How, an unexpected rise in the coolant temperature, or a significant reduction in How pressure, and before any liquid can be drained away from the racks, an electrically or hydraulically operated valve bypasses the supply and return of the data center cooling path, creating an independent closed loop supplemental cooling system within the rack itself.

One embodiment of a system for mitigating a failure of a data center’s liquid cooling system is shown. A high level block diagram of one embodiment of a system for thermal caching in a liquid cooled computer system is presented. Under normal circumstances, coolant is delivered to a rack housing a plurality of liquid cooled components by a data center liquid cooling system supply and return lines. The supply and return lines are coupled to the rack by quick disconnecting and the coolant is pumped through the conduits by one or more data center pumps. Within the rack a secondary cooling loop is established that is in fluid communication with the data center cooling loop.

Upon entering the rack the data center supply line enters a monitoring device. In one embodiment of the present invention the device monitors fluid flow, pressure, and/or temperature. Other characteristics of the fluid that can also be used to identify a failure of the data center cooling system as imposed on the supply line. Thereafter, and before allowing the coolant to access any of the electronic components housed within the rack, a one way valve is placed on the supply lid entering the rock. The valve allows fluid to pass from the data center cooling supply line into the rack but prevents flow from regressing toward the supply line should the flow stop and/or an adverse pressure gradient is experienced.

Similarly, a two-way valve is placed on the coolant return line that returns heated coolant from the electronic components to the data center coolant return line. The two way valve is in communication with the monitoring device and capable of receiving a signal that indicates a failure in the primary or data center liquid cooling system.

Upon receiving such a signal from the monitoring device, the two-way valve closes either electrically or hydraulically and diverts the return coolant from the electronic devices to the supplemental or secondary liquid cooling loop. The supplemental liquid cooling loop, which is housed entirely within the rack, thereafter operates independent of the data center liquid cooling system.

Modular High-Density Computer System

US 7688578 B2, Mann, R., Landrum, G. and Hintz, R.

A modular high-density computer system has an infrastructure that includes a framework component forming a plurality of bays and has one or more cooling components. The computer system also has one or more computational components that include a rack assembly and a plurality of servers installed in the rack assembly. Each of the one or more computational components is assembled and shipped to an installation site separately from the infrastructure and then installed within one of the plurality of bays after the infrastructure is installed at the site.

Historically, a room oriented cooling infrastructure was sufficient to handle this cooling problem. Such an infrastructure was typically made up of one or more bulk air conditioning units designed to cool the room to some average temperature. This type of cooling infrastructure evolved based on the assumption that the computer equipment is relatively homogeneous and that the power density is on the order 1 -2 Kilowatts per rack. The high density racks mentioned above, however, can be on the order of 20 Kilowatts per rack and above. This increase in power density, and the fact that the equipment can be quite heterogeneous leads to the creation of hot spots that can no longer be sufficiently cooled with simple room-oriented cooling systems.

By separating the computational components of a high density computer system from the requisite cooling, power distribution, data l/O distribution, and housing infrastructure components, significant advantages can be achieved by embodiments of the invention over current configurations of such computer systems. The computational components can be assembled and tested for proper operation while the infrastructure components are shipped separately in advance of the computational components. Thus, the infrastructure cooling components, data and power distribution components, and framework components can be delivered and their time-intensive installation completed more easily and prior to receiving the tested computational components.

Moreover, the shipping of the computational components separately from the infrastructure components achieves less laborious and less costly delivery and handling of the components. Finally, it makes it easier for the data center to upgrade or add additional infrastructure and infrastructure components to a current installation.

For more information about Advanced Thermal Solutions, Inc. (ATS) thermal management consulting and design services, visit https://www.qats.com/consulting or contact ATS at 781.769.2800 or ats-hq@qats.com.

Leave a Reply

Your email address will not be published. Required fields are marked *

This site uses Akismet to reduce spam. Learn how your comment data is processed.