Applications Using Heat Pipes & Vapor Chambers
A transition from traditional heat sinks to ones using heat pipes or vapor chambers should be considered when the design is conduction limited and/or when non-thermal goals such as weight or size can’t be achieved with other materials such as solid aluminum or copper. As a general rule, if conduction losses in the base are more than 10oC, two-phase devices are likely a good candidate for applications across a wide range of industries. Check conduction loss for a specific application.
Regardless of industry segment, mechanical and electrical engineers must contend with cooling electronics that likely fall into one, or more, of three categories: space constrained, rugged, and high heat flux. These design challenges drive heat sink parameters like module thickness & weight, location of heat source relative to fin stack, fin volume based on available airflow, and heat transport / spreading capacity, among others.
Heat Sinks for Space Constrained Devices
When thinking about space constrained electronic enclosures, one probably envisions handheld devices almost to the exclusion of all else. We think of them as any device, regardless of size, where electronics are packed very tightly within a given space. Achieving thermal goals within cramped spaces often requires a fin stack (condenser) that is remote from the heat source and/or one that is capable of dissipating heat from multiple heat sources. Moreover, weight is often a key design goal.
Whether it be for mobile communications systems, embedded computers or handheld devices, heat pipes and vapor chambers can dramatically improve thermal performance, even in severely cramped enclosures. When compared to solid metal heat sinks, two-phase designs can offer temperature improvements at the semiconductor interface (T-case) as high as 30-50%, allowing engineers to increase processor computational speed while meeting system level thermal goals.
Although heat pipes can be made as thin as 0.6mm, performance requirements generally dictate a minimum thickness of about 1.5mm when flattened. Vapor chambers typically have thickness between 2-4mm, but with an associated width of between 50-100mm – allowing for improved multi-directional heat spreading.
The examples below illustrate how heat pipes and vapor chambers can be used in thermal assemblies for space constrained environments.
Fig. 1: Laptop PC
Fig. 2: SFF Desktop PC
Fig. 3: Test Equipment
Figure 1 illustrates a heat sink design for a specialty laptop PC. As in many space constrained applications, the fin stack needed to be located in an area with greater access to cool air and it needed to act as the condenser for multiple heat sources – in this case the CPU and chipset. Of particular interest in this design is the use of a dual diameter flattened primary heat pipe to which the CPU attaches.
Figure 2 shows a heatsink for a high-performance processor in a small form factor gaming PC. The original design used a solid copper baseplate, but a new Core i7 required better cooling. We replaced it with a vapor chamber, improving performance and reducing weight.
Figure 3 shows a heatsink used to cool six 80-watt ASICS. The design challenge here included weight restrictions as well as the requirement for each ASIC to remain within 2oC of each other. Here we used a two-piece vapor chamber with the center cut-out and a shared fin stack to meet weight requirements.
Heat Sinks for Rugged Electronics
Heat pipes and vapor chambers have proven effective for both military and industrial applications where electronics are exposed to harsh conditions – from extreme temperatures and wet or dusty environments, to situations where components are subject to repeated shock or exposure to corrosive substances. Success ranges from cooling environmentally exposed power electronics and energy systems to military and telecommunications applications.
Two-phase devices can withstand repeated freeze thaw cycles, are designed to perform well with changes to orientation and operate effectively in high ambient temperatures. Celsia works with both packaging and thermal engineering teams to ensure the highest level of efficiency and survivability for both commercial grade and rugged electronics in harsh environment applications.
Of concern, especially for mission-critical applications, is the durability of heat pipes and vapor chambers. These devices have useful lives in excess of 15 years of continuous operation with no change in performance. This statistic is based on both performance measurements of in-field devices as well as accelerated life testing from both academia and industry, including an ongoing study here at Celsia.
Example of one of the test sets Celsia is using as part of its long-term heat pipe reliability test program
If a two-phase device fails before predicted MTBF figures, it fails very early in its life. This ‘infant mortality’ can be mitigated with stringent production and testing procedures. Table 1 illustrates Celsia’s typical first article and mass production QA testing which is conducted on both the formed heat pipe or vapor chamber as well as the entire assembled heat sink.
|First Article Reliability||Test|
|Thermal Shock||2,000 cycles: -40 to 100 oC|
|Thermal Cycle||2,000 cycles: -40 to 100 oC|
|High Temperature||28 days @ 120 C, 48 hrs. @ 200 C|
|High Temp & High Humidity||96 hrs. @ 65 C and 85% humidity|
|Leak Check||High pressure helium leak check|
|Mass Production Testing||Test|
|100% of 2-phase devices & completed modules||Burn-in, helium leak check, performance|
The examples below illustrate how two-phase devices can be used in thermal assemblies for rugged environments.
Fig. 5: 100Gb Transponder
Fig. 6: Rugged Communications
Fig. 7: Outdoor Electronics
Figure 5 illustrates A high power transceiver used for long haul applications. High powers, high ambient, low airflow and tight Tjunction requirements required the use of both a large vapor chamber and stamped fins to hit the design goals.
Figure 6 shows A heat pipe assembly used in military battlefield logistics systems. This type of assembly allows the heat from the key hot components to be directly moved to the housing for cooling.
Figure 7 illustrates a single 6mm heat pipe soldered to a cast aluminum, finned housing body. The heat pipe cools three heat sources.
Heat Sinks for High Heat Flux Applications
Shrinking die sizes coupled with higher power output of modern electronics means not only increased total heat flux but also higher intensity localized hot spots. Heat pipes and vapor chambers can be the ideal solution to move and spread heat in applications using bare die computational ASICs, ultra-high brightness LEDs for medical & UV curing applications, laser diodes, and power amplifiers / transistors.
Many electronics engineers would consider heat fluxes of greater than 20 W/cm2 to be high. For power electronics engineers, the initial figure is probably above 75 W/cm2. When properly designed, and incorporated into a thermal module, both heat pipes and vapor chambers can be used in applications where power density approaches 500 W/cm2, although this is far from typical. At Celsia, most of the higher heat flux challenges we see are in the 35-80 W/cm2 range but with cooling challenges that are compounded by cramped space, low/no airflow, and/or the requirement to operate in high ambient temperatures.
Celsia designs are often required to work under high localized thermal loads using unique evaporator systems to reduce thermal resistance and increase power handling capabilities. These evaporator designs can sometimes compete with pumped liquid systems for cooling of high power devices.
Wick structures are the engine that run these devices. Like engines, wick structures can be designed to fit various applications. For very high heat fluxes the internal evaporator surface requires very high capillary forces to feed the liquid into the evaporator. Celsia engineers have been working on ultra-high power densities for decades. One very early project using lithium as the working fluid was estimated to be on the order of 187,000 W/cm2 at 1,600 oC setting a world record for power density in a heat pipe. Likewise, for other applications such as long distances or against gravity the wicks can be optimized for each application.
The examples below illustrate how two-phase devices can be used in thermal assemblies for high heat flux applications.
Fig. 8: LED Theater Spotlight
Fig. 9: Solar
Fig. 10: Laser Diode Array
Figure 8 illustrates a cooling solution for an LED spotlight used for professional theatrical lighting. It uses a 12mm diameter nickel plated heat pipe, which is squared off at one end for direct mounting of 4 LED light engines. Total power is 155 watts with a power density of 75 W/cm2.
Figure 9 shows a natural convection cooling system for a 500 sun concentrating photovoltaic system. Designed to work with an active solar tracking system this part needed to work in all positions and conditions relative to an outdoor environment.
Figure 10 combines a single fin array with 3 vapor chambers for cooling three different powered lasers (RGB). Each color has its own thermal requirements. The higher power lasers are on the upstream side assuring they meet the thermal specs while allowing the lower power lasers to be downstream.