From the time a fledgling IT maven first installed multiple servers close to one another, it became clear that thermal management, and the choice between air and liquid cooling, would be the ongoing challenge. Initially, air-based cooling systems did an adequate job of dissipating heat, but these proved to be inadequate in the age of high-density data centers, some extending beyond a whopping 10mW in size.
Today’s server racks and supercomputers require a cooling system that exceeds efficiency, effectiveness, and safety standards. The closed-loop liquid cooling system meets those requirements.
For optimal performance, a closed-loop system must remain free of contaminants. The introduction of such contaminants such as dust, dirt, debris, or even air can cause system malfunctions.
Contaminates can be introduced into the system in a variety of ways. When specifying the components of the liquid cooling system, the highest quality construction, materials, and installation procedures must be considered.
To lessen the risks of introducing contaminates into your cooling system, here are the four main areas to consider.
Installation is one time that poses a risk of contaminants being introduced into the liquid cooling system. Without precise adherence to a safe cooling line installation process, contaminants, such as those from elastomeric components, can be introduced into a closed-loop system. Developing an installation process and training staff on the procedures is key to a successful installation.
When a server goes down, a 'hot-swap' replacement is required. This procedure requires disconnecting the cooling line coupling without disconnecting additional servers—a delicate operation mandating well-planned, meticulously executed protocols.
For this change-out to go smoothly (i.e., no leaks, spills, or contaminants introduced in the system), a plan of action must be in place. Of course, advanced training of personnel in the methods and precautions are paramount. Ongoing 'dry-run' sessions should be a standard component of this training, allowing technicians to sharpen their skills for the next 'real thing,' whether it be of the scheduled or on-demand variety.
Reducing the risk of line contamination begins before the system installation date has been circled on the calendar; this means choosing the appropriate components for the job.
First and foremost, a flush-faced valve design is a must. Its construction must be durable and robust to reduce air inclusion and function reliably across various temperatures. Additionally, it must be designed specifically for low-pressure applications.
Of particular note is that many coupling designs are not intended for use in cooling applications. Therefore, the importance of selecting metal couplings designed specifically for liquid cooling cannot be underestimated.
The materials used in the coupler's assembly should be examined closely as well before making a purchase decision. To prevent the chipping associated with plastic components, high-quality stainless steel is the way to go to prevent debris from breaking off and entering the system.
Evaluating couplings based on robustness, valves, seal compounds, and materials compatibility can help ensure trouble-free serviceability and long-term performance.
Liquid cooling system manufacturers and data center operators understand that a closed-loop rack cooling system's integrity is only as reliable as the system's components. With the vital role these systems play in protecting servers and supercomputers, the manufacturer must meet the highest production and testing standards available.
Parker's long-standing commitment to producing the cleanest components possible is apparent along every step of the manufacturing process:
• Materials: It all begins with material selection. Parker uses stainless steel that is less likely to chip and burr. Only high-integrity elastomers are used in the manufacturing of seals and O-rings. These materials, if of lower quality, could potentially cause problems, causing debris to potentially get into the cooling system.
• Cleaning: After machining, every piece of every component is thoroughly cleaned to remove particles and debris.
• Assembly: Parker's assembly process takes place in a 'clean-room' setting, ensuring that each part is protected from manufacturing residue or other environmental contaminants.
• Testing: Parker employs one of the most rigorous leak-testing protocols in the industry. 100% of the completed couplings are helium-leak checked to ensure integrity. Why helium? Because a helium molecule is significantly smaller than a water molecule.
• Packaging: Once tested, the finished components remain in Parker's sanitary packaging room to be prepared for shipping. Even our product caps and shipping bags are specially designed to prevent impurities from tainting the couplers' integrity.
Offering a variety of couplings and seals, Parker's liquid cooling couplings combine a compact design with robust components built to resist mechanical stresses and avoid any fluid loss. Each design features non-spill valving, higher flow rates, and low-pressure drop capabilities to ensure reliable thermal management cooling.
For the critical needs of data centers, compromise is not an option. Parker understands that these facilities must continuously stay ahead of the game to deal with the ever-rising temperatures that new technologies generate effectively. Parker is committed to meeting these challenges by delivering liquid cooling solutions that protect a company's significant investment in hardware and contribute significantly to its profitability.
This article was contributed by Todd Lambert, market sales manager, Parker Hannifin's Quick Coupling Division.