Advanced Thermal Management for LEDs
Billy: Welcome and thank you for joining us for today’s webcast, Advanced Thermal Management for LEDs – Practical Solutions When Heat Sinks Aren’t Good Enough, sponsored by Advanced Cooling Technologies and Tech Briefs Media Group. I’m Billy Hurley, Associate Editor with Tech Briefs Media Group, and I will be your moderator today.
Our webcast will last approximately 30 minutes, and there will be a question and answer period at the end of the presentation. If you have a question, you may submit it at any time during the presentation by entering it in the box at the bottom of your screen. Our presenters will answer as many questions as possible at the conclusion of the presentation, and those questions not addressed during the live event will be answered after the webcast.
In order to view the presentation properly, please disable any pop-up blockers you may have on your browser. Twice during the presentation, we will also present you with poll questions, and we invite you to answer those at the appropriate time. At this time, I would like to introduce our speakers. Pete Ritt is Vice President of Technical Services at ACT. He has successfully managed numerous advanced technology introductions in a variety of industries including the display, consumer electronics, and household goods industries. He has been granted over 20 patents and holds degrees in Chemical Engineering from the University of Notre Dame and an MBA from Shippensburg University.
Also on the line for our Q&A part of the presentation is Dr. Richard Bonner. Dr. Bonner, manager of custom products has been with ACT for over eight years. He has served as principal investigator on multiple government and industrial R&D programs involving advanced thermal management. The thermal topics have included two-face heat transfer, nanoscale coatings, thermal storage, junction level cooling, and jet impingement. He’s also served as a technical lead on many technical services programs including aerospace defense, medical and commercial electronics cooling programs. So now at this time, I would like to hand the webinar over to our first speaker, Pete Ritt. Pete?
Pete: Thank you, Billy. I’m Pete Ritt along with Dr. Richard Bonner. We’re delighted to be with you. For today’s webinar, we’d like to talk about some advanced passive thermal management technologies, heat pipes and vapor chambers that are being used more frequently by LED device designers to solve challenging thermal problems. We’ll explain the fundamentals of how these passive technologies work, and then offer some suggestions as to when they should be considered for implementation. Specifically, we’ll investigate three types of passive technologies. Heat pipes, conduction plates with heat pipes embedded in them which we call HiK plates, and finally vapor chambers. We will then discuss five LED product examples and how they benefit from these passive technologies. We’ll conclude with a quick wrap up and answer some questions that you might have.
LEDs, as we all know, are experiencing increasing acceptance and growth across many market segments. Part of this growth is due to new, higher power LEDs that can be packaged in increasingly small areas. These new, brighter more powerful products meet challenging customer requirements, and satisfy evolving consumer preferences. However, increased power places greater importance on thermal management, because of the negative effect that heat has on LED performance. For typical LED devices, 70% to 80% of the input electrical power becomes waste heat. The waste heat if not properly managed can have a significant impact on LED device performance. Increasing device temperature is directly correlated to device life.
A rule of thumb is that for every 10 degrees heat increase in operating temperature over the max operating condition, device life can decrease by 50%. Additionally, many products are wave length specific. Increase in temperature can cause a wave length output reducing the efficacy of these devices. So it is clear that LEDs require thermal management solutions, and as input power increase and package size decrease, these thermal management solutions would be of greater importance to device life and performance.
Today, we would be discussing passive two phase heat transport technologies. We will be discussing heat pipes, heat pipe embedded conduction plates, and vapor chambers. By passive, we mean the devices do not require any external power to operate. By two phase, we’re refereeing to simultaneously having the working fluid be in two phases, liquid and vapor inside the device. Inherent advantages of these devices are that they take benefit of the power transport capability of the latent heat of vaporization. They require no input power to operate. They provide years of reliable operation, and they operate with virtually no noise. We will now examine these technologies in a little more detail.
First, we’ll look at heat pipes. Heat pipes are sealed vacuum devices. They’re housed in a metal tube. Inside the tube is a wick structure and a small amount of a working fluid. Most applications are copper tube, copper wick, and water as the working fluid. But there’re several other envelop materials, wick structures and working fluid combinations. Heat pipes are passive two phase heat transfer devices that operate in a close system. To work, a heat pipe must be connected to a hot end or evaporator, and a cold end or a condenser as can be seen in this schematic on the left. The difference in temperature between the hot evaporator end and the cold condenser end is the driving force. The heat from the evaporator causes the working fluid to vaporize. Pressure pushes the vapor to the cooler end where it condenses to a liquid and is absorbed by the wick structure. The condensed liquid then returns to the evaporator by capillary force of the wick structure.
In the schematic on the right, you can see a cut away showing the heat pipe structure including the envelop material, the wick structure, and the central vapor path. Fluid inventory control is an important design and processing parameter. Typically, only a small amount of liquid is used in heat pipes. So now that we know a little about heat pipe operation, let’s look at how this can be put to use. First of all, with different combinations of envelop materials, wick structures and working fluid, heat pipes can be used over a wide temperature range from cryogenic to liquid metal, from -150 to 1,500 degrees C. Because they’re sealed vacuum devices, some working fluids can operate well beyond their nominal boiling point.
Water for example, can be an effective working fluid from 20 to 250 degrees C. By controlling the amount of working fluid and selecting the appropriate wick structures, heat pipes can restart operation after freezing. They’re freeze-thaw tolerant. Also, with the right wick structure, heat pipes can operate against gravity, that is, heat pipes can transfer heat downwards. Typically, heat pipes can transfer heat around 8 inches against gravity, although gravity aided operation, that is, when the heat source is below the heat sink is generally preferred. In terms of [inaudible 00:07:36] since boiling and condensing are occurring at the same temperature, heat pipes can have an effective thermal conductivity of 10,000 to 20,000 watts per meter kelvin. We can contrast that with copper and often used conductor which has a thermal conductivity of around 400 watts per meter K.
Other advantages include continuous passive operation, excellent isothermal characteristics, and quiet operation. In terms of heat flux capabilities, heat pipes can operate with heat fluxes up to 50 to 75 watts per centimeter square, with custom wicks up to 500 watt per centimeter square.
As mentioned, one of the key benefits of heat pipes is heat transport. Heat pipes can be used to transfer heat to external sinks. They can be used, for example, to move heat away from PCBs, chips or other hot components to heat sinks which then be dissipated by either natural conduction or with air flow from fans. Heat pipes are capable of transferring heat over long distances with minimal temperature differential, delta-T, within the device. The typical delta-T is about 2 to 5 degrees C over the length of the pipe.
We mentioned typical heat pipes can transfer heat around 8 inches against gravity. Gravity aided and other specialized heat pipes such as loopy pipes can transfer heat over longer distances. Bending and flattening enable increased geometric flexibility to design. You can see in the picture on the upper right, a heat pipe which is bent in three dimensions, transports heat away from an electronic component to an external heat sink. The picture below that shows a curve heat pipe removing heat from the center of an electrical assembly to a cold [inaudible 00:09:25].
Additionally, heat pipes can be used to move heat away from the inside of an enclosure to the exterior without subjecting the components to the outside environment. This solution can be implemented to dissipate heat from LEDs in sealed packages. The other major transport benefit of heat pipes is heat spreading. One specific embodiment of that is the HiK plate. HiK plates are aluminium conduction plates with heat pipes embedded in them. Aluminium has a thermal conductivity of around 200 watts per meter K. Typical thermal conductivity for aluminium HiK plates is 500 to 1,200 watts per meter K depending on the number, diameter and location of the heat pipes.
HiK plates, termed that because of their high thermal conductivity can be used to reduce hot spot temperatures by moving or spreading heat to less critical areas. They are fully compatible with either liquid of air cool chassis that can be used to improve fin efficiency and lower fin weight. Pictured on the lower right is an application where all of the high powered electrical components were located on one side of the heat sink. Heat pipes were embedded into the plate to move and spread the heat to the other side of the heat sink and get improved fin efficiency. Plate thickness is typically 4 millimeters, but can be thinner. We have made plate as thin as 1.8 millimeters. The structural strength and weight of a HiK plate is the same as aluminium. Aluminium silicon carbide can be used instead of aluminium or direct die attachment application. A sample [inaudible 00:11:07] HiK plate is pictured on the upper right. And for reduced weight applications, magnesium can be used in place of aluminium.
Shown here, is the thermal analysis of an aluminium plate containing several high powered electrical components with and without embedded heat pipes. Pictured on the far left is a solid aluminium plate without any heat pipes. You can clearly see the three hot spots, two at the top, and one at the bottom. Pictured in the middle, is that same aluminium plate, but now, with heat pipes embedded in them. You can now see that the max temperature has dropped about 20 degree C, and the temperature uniformity has improved greatly. Pictured on the right is the actual plate with the heat pipes embedded in it. The heat pipes can be seen as silver lines. This technology has many heat transport and heat spreading applications and certainly could be considered for improving thermal management for LED arrays.
Next, we’ll talk about vapor chambers. Vapor chambers work similarly to heat pipes, but due to its shape, can move heat in two dimensions as can be seen in this schematic on the right. In vapor chamber operation, high heat flux input causes the working fluid to vaporize. The vapor then spreads in two dimensions. There’s wick material around the inside structure of the chamber. When the vapor reaches the other low flux side, it condenses and returns to the high heat flux area via the wick structure. One of the benefits of vapor chamber is that the vapor heat transport yields nearly perfect isothermal heat spreading as is illustrated on the low flux side in the right side of the schematic. Here, the temperature across that side of the vapor chamber is highly uniform. So in a sense, a vapor chamber is a heat flux transformer.
Vapor chambers are often used in conjunction with heat sinks, taking advantage of their nearly isothermal output, and making the heat sink more efficient. Typical components of a vapor chamber heat sink assembly are seen in the lower left. They consist of the vapor chamber wick structure in the middle, a copper envelop seen below, and a heat sink seen at the top. Vapor chambers are used to dissipate highly concentrated heat flux regions. They can handle higher input heat fluxes than a typical heat pipe, and they can dissipate that heat more uniform. As we’ll see, it is a great solution for high powered, wave length sensitive LED devices. We will now pause for our first polling question.
Billy: Thanks Pete. At this time, we would like to present you with the first of two polling questions. It should appear on your screen now. The question is, “Who does your company use to resolve critical thermal management issues?” And your choices are A, in-house resources, B, external experts, C, a combination of both in-house and external resources, or D, not sure. Again, “Who does your company use to resolve critical thermal management issues?” In-house resources, external experts, a combination, or not sure. And as you answer that question, I will hand the presentation back over to Pete Ritt. Pete?
Pete: Thanks Billy. Now that we’ve learned a little bit about heat pipes, HiK plate and vapor chambers, it could be useful to examine exactly how these technologies could be used as part of a thermal management solution for LEDs. In the following, we’ll look at five different LED applications where they can be beneficially implemented.
The first is the remote sink. In many lighting applications, the LED device mush fit in a big space to accommodate a variety of customer requirements, which usually exclude thermal management considerations. A common example is a Luminaire design where the ceiling or wall fixtures are based on a pre-existing design using non-LED technology. These designs commonly have both restricted space or heat dissipation through conduction, and limited air flow to remove heat via conduction. In cases where there’s space to remotely dissipate heat, heat pipes can be used to transport the heat from the device to a heat sink located elsewhere. This is called the “Remote Sink.”
The remote sink solution has a heat pipe in direct contact with the LED device, a PCB or a similar component at one end, which serves as the evaporator. At the other end, the heat pipe is connected to the heat sink, the condenser. A sketch of a conceptual design can be seen on the lower left. Here, two heat pipes are in direct contact with both the heat generating component at the bottom, and the heat dissipating fins at the top. A wall or other enclosure can be place in between the LED and heat sink to separate the two. In the example on the right, a single heat pipe is used to transfer heat from the source on the bottom to the heat sink above it. A thermal image on the right shows heat being transferred from the hot yellow LED device to the blue green yellow heat sink above it. The presence of green and yellow within the heat sink shows that the heat pipe is effectively transferring heat from the device to the radial heat sink fins. Heat pipes can efficiently transfer heat approximately 8 inches against gravity with no minimal thermal gradient, and even greater distances when gravity assistance. Note that the number, size, shape and location of heat pipes would be specific to the design.
Case number two, extrusion isothermalization. In some cases, it is not possible or desirable to have a heat sink in a remote location, say and/or outdoor lighting products for example. Frequently large extrusions, typically made of aluminium are utilized. Heat dissipation occurs through conduction resulting in large temperature gradients across the heat sink structure with the high temperatures found closest to the LED heat generating source. As we have mentioned, the excessive temperature has a negative effect on LED device life and performance. What can be done to mitigate this?
We talked earlier about how embedding heat pipes in an aluminium extrusion can be used to more effectively spread heat in that structure. The same benefit can be realized by inserting heat pipes into these large heat sink extrusions. The heat pipes themselves will isothermalize. By doing this, they will spread heat through the heat sink. This heat spreading reduces the thermal gradient and likewise reduces the max temperature at the LED source. Benefits can be taken in different ways, such as, one can run the device at a desire higher temperature and reduce the risk of overheating or one can reduce the overall size of the heat sink.
Our testing and analysis has confirmed that the longer the extrusion, the more benefit from the presence of heat pipes principally because of the larger available area. As seen in the graph on the upper right, the percentage improvement in thermal resistance with heat pipes increases approximately linearly with increasing heat sink length. For example, one can expect to see a 5% improvement in thermal resistance for a 5 centimeter long heat sink. That improvement can increase to 30% for a 30 centimeter long heat sink. Additionally, one can expect the benefit will be more noticeable in the natural conduction heat sinks as the fin operation plays a major role in forced induction performance.
Let’s look at a specific example seen in the upper right. In this case, heat pipes were embedded in a 200 millimeter long radial heat sink that was dissipating 100 watts of heat. As seen in section B in the center of the thermal images at the bottom of the page, addition of the heat pipes decreased the maximum temperature to 10 degrees C, which as mentioned previously could increase device life by 50%. So we see here an example of how using heat pipes to isothermalize a heat sink to get better performance.
The third case, heat sink size, weight and power, again, uses the embedded heat pipe or HiK plate as we all it, but takes the benefit in a different way. It’s well known that placing a discrete heat source on a large metal heat sink will produce thermal gradients as the heat slowly conducts and dissipates heat to the external fins. We have also discussed in this presentation that embedding heat pipes can increase thermal conductivity from around 200 watts meter K to 500 to 1,200 watts per meter K, offering the opportunity to reduce heat sink plate thickness and fin area. This approach can be implemented in a variety of LED applications including large arrays, outdoor lighting as well as some down lighting applications.
Let’s now look at what kind of benefit we can expect. Here’s the heat sink size and weight analysis with and without heat pipes. Total heat dissipation is 150 watts in both cases. The conventional metal heat sink is 12 inches long, weighs 9.6 pounds, and has a base thickness of six tenth of an inch. Introduction of five heat pipes, three in close proximity to the heat source, and another two a little further out for heat spreading can have a dramatic effect. The overall length can be reduced to 10 inches and the weight can be reduced to 6.3 pounds, an overall material reduction of 35%.
Here are some actual thermal images that demonstrate the improvement. The HiK heat sink seen on the right, more effectively spreads heat as can be seen in the yellow areas surrounding the source even though the heat sink is shorter, lighter, and thinner. The improvement is directly attributable to the addition of heat pipes, which can be seen as red lines in the picture on the right.
The next example of passive heat transfer technologies for LEDs is directed close to the source of heat. Obviously it is advantageous to dissipate heat as close to the source as possible. This can be difficult as electrical isolation requirements must be satisfied, otherwise the device will not function. Unfortunately in many cases, electrical isolation is only achieved using materials that are thermally insulated such as with FR-4 boards. Recent work at advanced cooling technologies has explored adding heat pipes to the structure of metal core printed circuit boards to help spread heat right at the source. You can see in the picture on the right an example of heat pipes embedded into an MCPCB. Heat pipes are seen on the left in close proximity to the circuit tree. The circuit side is seen in the image to the right of the embedded heat pipes. Let’s next look at some results.
As we have seen, embedding heat pipes can provide much more heat spreading, enabling lower operating temperatures and/or smaller heat sinks. Here we see a three LED structure on the lower right. A thermal image of that structure is seen on its left. The temperature scale is only 10 degrees, from 58 to 68 degrees C, to demonstrate the heat spreading capability of this concept. Measurements have shown that the embedded heat pipes can reduce the heat spreading resistance by 45% over standard aluminium MCPCB, and 15% over copper MCPCBs. A valuable improvement particularly so close to the heat source.
Our final example is for a growing number of high power LED applications such as UV curing devices. These products typical use high power, requiring dissipation of hundreds of watts, but must maintain tight temperature range, so that the optical output wave length remains constant. This problem is further aggravated because an additional thermal interface layer is needed between the heat source typically made of some low thermal expansion semi-conductor material, and the heat dissipating device, usually copper or aluminium which have much higher coefficient of thermal expansions, CTEs.
Typical interface materials include thermal gap pads and thermal paste. Without such a thermal interface material, the different materials coefficient of thermal expansion will result in the heat source device cracking or severing from the heat dissipating device. Unfortunately the presence of this thermal interface layer increases the overall thermal resistance and likewise increases the temperature on the LED device. The thermal management issue is analogous to some laser diode application, and fortunately the solution can be the same as well. Let’s look at this in more detail.
With both heat pipes and vapor chambers, it is most advantages to have water as a working fluid for both cost and performance reasons. Copper has been shown to have excellent compatibility with water, but copper has relatively high CTE, and would be thermally mismatched with common high heat load LED devices, and would require a thermal interface layer. Additionally, conventional wick structures have limited performance capabilities which may not always satisfy requirements for high heat load devices. The desired solution is to have a water compatible vapor chamber that is CTE matched to eliminate unnecessary thermal barriers, and has a wick structure design for heat flux devices.
Such a device is available and has been called the “CTE Match Vapor Chamber.” This unique device has been demonstrated to dissipate heat fluxes as high as 700 watts per centimeter square, and 2 kilowatt overall. The overall envelop structure is aluminium nitrite with a direct bond copper exterior. The copper on the inside of the vapor chamber ensures the well-known water-copper performance is maintained. In areas where the heat source is to be attached, the copper layer is removed exposing the aluminium nitrite. Aluminium nitrite has a CTE of around 5.5 PPM per degree C, much closer to many common semi-conductor materials. Devices can be directly attached to the vapor chamber, eliminating the need for a thermal interface layer.
Additionally, there’s a unique wick structure located at the heat source. It’s a combination of very thin and very think wick material collocated at the evaporator. The thick wick collects the condensate returning from the condenser portion of the vapor chamber, and delivers it to the thin wick area. The thin wick quickly evaporates the fluid, achieving very low thermal resistance values. An example of the thin, thick wick structure can be seen at the bottom of the page termed lateral liquid delivery. This structure has been shown to have a very high critical heat flux, CHF, with evaporator resistance of only 0.05 K centimeters square per watt. The CTE match vapor chamber is an excellent solution for high powered LED devices. We will now take you back to Billy for another polling question.
Billy: Thanks Pete. At this time, we like to present you with our second and final poll question for today. It should appear on your screen now. The question is, “Are you facing any thermal management issues in the next A, 0 to 6 months, B, 6 to 24 months, or C, you have no immediate issues?” Again, are you facing any thermal management issues in the next 0 to 6 months, 6 to 24 months, or you’re having no immediate issues? So as you answer that question, I will hand the presentation back over to Pete Ritt again. Pete?
Pete: Thanks Billy. In summary, as LED devices become more accepted and more prevalent, customers are requiring higher performing products in smaller packaging size. This puts an increased burden of thermal management on LED device designers. The conventional metal heat sink solutions will continue to be implemented wherever practicable, but are more frequently being found to be insufficient for a variety of performance and size reasons. Fortunately, there exists excellent passive, two phased, heat transfer technology, heat pipes, HiK plates, and vapor chambers that have provided many successful examples of thermal management solutions in the broader electronics industry. Those same technologies can be applied to many LED devices as well.
We presented today five examples of how heat pipes, HiK plates and vapor chambers can be implemented to solve real LED thermal management issues. The remote heat sink demonstrated the heat transport capability of heat pipes. Heat can be effectively moved 8 inches from the source against gravity, and even more when gravity aided. The extrusion isothermalization showed how heat pipes can reduce the peak temperature, 10 degrees in the example shown by spreading the heat more efficiently across a heat sink. Extrusion, size, weight and power showed that heat sink size can be decreased 35% in the example shown through implementation of heat pipe without degrading heat sink performance.
PCB level spreading offers thermal management solutions, very close to the heat source, reducing the heat spreading resistance by 45% over aluminium, metal or printed circuit boards and 15% over copper ones and finally, for very high heat flux requirements such as some LED UV curing applications. The CTE match vapor chamber offers heat flux capabilities over 500 watts per centimeter square with low evaporative resistance. All of these solutions are based on technologies that have proven track records for long, stable, reliable operations, and can be expected to find many new applications with the proliferation of LEDs. We thank you for your attention. And now, we throw it back to Billy and try to answer some questions that you might have.
Billy: Thanks Pete. At this time, we would like to begin out Q&A, and for this I’d like to welcome Dr. Richard Bonner to the line. Dr. Bonner is manager of custom products at ACT. Richard, welcome. We got our time for one or two questions here. We had a couple of questions actually about working fluids. So what are the options and considerations when selecting working fluids? And then more specifically someone asked, “Do you recommend a working fluid for an operating temperature of 3 degree Celsius?”
Dr. Bonner: Sure. Those are two great questions. Now, when we consider the selection of the working fluid, the first thing we normally look at is the performance over the required temperature range. We have quick ways to evaluate that looking at what we call “figures of merit,” look at all physical properties associated with heat pipe performance. So we can very quickly tell you what the working fluid candidates are in that temperature range. After that the next main issue is reliability. So for every working fluid, there’s usually exist a set of candidate envelop materials. Maybe if I can use this 3 degree C temperature as an example. I think in that range, we would be looking at using either ammonia, methanol or acetone as a working fluid.
Now, with ammonia, aluminium is a very good candidate that shows long life. With methanol and acetone, stainless steel might be better options. From that point, it really depends on the application, which material set combination is going to work best.
Billy: Can copper water heat pipes be used in applications where freezing is a concern?
Dr. Bonner: Yeah, absolutely. That’s a great question, very relevant to LEDs, and lots of people want to try and install their fixtures outdoors or in unheated warehouses, so it’s just something we commonly see. And actually the answer is yes. As a matter of fact, we use copper water heat pipes in a number of military applications where the temperature ranges -40 to +85 degree C. Really the key here is to make sure that your working fluid is charged such that when it goes through a freeze cycle, it doesn’t cause any bulging of the heat pipe envelop. That’s something we have good control over here in our manufacturing processes. Just to make it clear, the heat pipe isn’t going to work as you freeze, but it will survive that freeze, reheat out and start working as you do rise above the freezing point.
Billy: Richard, this will have to be our last question. How does orientation affect heat pipe and vapor chamber performance?
Dr. Bonner: It certainly affects the amount of power that your heat pipe or vapor chamber consume. So whenever you’re working against gravity that put additional stresses on the wick that must be overcome, that ultimately leads to degradation to the amount of power that you can transfer. However, as long as you’re operating within that limit, operating against gravity doesn’t affect your thermal resistance. So in other words, the delta-T, or again, the thermal resistance required to move that power doesn’t change the function of orientation, but there is a smaller power limit as you try to move heat against gravity.
Billy: Thanks Richard. We will end it there. That concludes today’s webcast. Again, if we did not get a chance to answer your question today, our sponsors will do their best to address them after today’s presentation. So our thanks to Pete Ritt and Dr. Richard Bonner, and thanks to everyone for joining us. Just a reminder, this webcast will be available on demand at www.TechBriefs.com for the next 12 months. Have a great day.