Category Archives: Analysis

Fin Optimization in Heat Sinks and Heat Exchangers

Posted on May 10, 2017 by | 1 comment

(This article was featured in an issue of Qpedia Thermal e-Magazine, an online publication produced by Advanced Thermal Solutions, Inc. (ATS) 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.)

In electronics cooling, often separately managed Thermal/Mechanical (TM) and Software/Electrical (SE) engineering teams are finding themselves facing common challenges, as they are being driven towards similar business goals, such as product differentiation, company growth and profitability.

More so than ever today, these teams are being directed to find ways to increase component performance, particularly on highly populated boards within complex systems, at an acceptable cost of manufacturing. They are also discovering that their goals are being held back by governing specifications, environmental conditions, mechanical limitations and budget restrictions.

Heat Exchangers

Closeup of fin array on an ATS tube-to-fin heat exchanger. (Advanced Thermal Solutions, Inc.)

TM’s design thermal solutions based on airflow, envelope size, power dissipation, etc. and migrate (as expected) to the lower cost “standard solutions” whenever possible. If adequate margin is not met, reliability implications are more apparent as engineers will have to optimize solutions. This is because, in most cases, the form factor, layout, boundary conditions, etc. are set.

Thermal solutions become the gatekeeper, and in some cases, the determining factor in product deployment.

Many leading companies design their products by using technologies that will sustain long product life cycles for increased market share and brand awareness. As products are refined through the design cycle, thermal solutions may have to be optimized and this requires many investigations to be undertaken.

As the electronics industry continues to use components dissipating more and more power, new heat sink solutions must be able to accommodate large heat fluxes while keeping the same spatial dimensions [1]. Finned heat sinks and heat exchangers are largely employed in many engineering fields, and this demand spurs researchers into devising and testing new geometries for the heat sinks.

Engineers constantly try to develop new designs to enhance the performance of heat exchangers. One such effort is the design of the wavy fins to enhance the surface area.

Figure 1 shows a close up view of an extrusion type thermal solution where the profile has a feature of undulated fins. In general, a wavy fin heat sink should perform better under natural and forced convection due to the increased surface area created by the fins. This feature can easily be manufactured with a die. The “waviness” can be adjusted to increase surface area resulting in a positive impact on thermal performance.

Heat Exchangers

Figure 1. Close-Up View of Simply Wavy Fin Geometry [1]

Theoretical models have been devised to find the pressure drop and the heat transfer from wavy fin geometries. Figure 2 shows the schematic of a wavy fin.

Heat Exchangers

Figure 2. Schematic of a Wavy Fin Geometry [2]

In this figure, the fins are assumed to have a sinusoidal geometry where

λ = Wave length (m)
H = channel width (m)
S = channel height
2A = twice the amplitude of the wave

The shape of the curve is assumed to be:

The length of the curve can be found from the following equation:

Shah and London [3] came up with the following equation for the friction and Nusselt number in channels:

Where,
F = fanning friction factor
aspect ratio

The same equation applies for a wavy fin based on the correct length:

The Nusselt number for the straight fins and wavy fins is the same as long as the correct surface area is used:

The above equations are for the low Reynolds number.

For high Reynolds number Shapiro et. al [4] derived the following equations:

Where,
Dh = hydraulic diameter (m)
Reynolds number based on hydraulic diameter
L = half length of the channel (Le/2)
Pr = prandtl number
Dh = 2SH/(S+H)

The combined asymptotic for the friction and Nusselt number is as follows:

Figure 3 compares the results of the above analytical equations with the results from Kays and London [5]. In the graph, the Colburn j factor is shown and is defined as:

The results show that the experimental values of Shah and London are within 20% band of the values obtained from the above relations. The data is for the fin type 11.44-3/8W.

Heat Exchangers

Figure 3. f and j Values as a Function of Reynolds Number.[2]

Marthinuss et al. [6] reviewed published data for air-cooled heat sinks, primarily from Compact Heat Exchangers by Kays et al [5] and concluded that for identical fin arrays consisting of circular and rectangular passages, including circular tubes, tube banks, straight fins, louvered fins, strip or lanced offset fins, wavy fins and pin fins, the optimum heat sink is a compromise among heat transfer, pressure drop, volume, weight and cost.

Figure 4 shows that if the goal is to get a higher value of heat transfer per unit of pressure drop, the straight fin is the best. Figure 5 shows that when heat transfer per unit height is of concern pin fin is the best.

Heat Exchangers

Figure 4. Profile Comparisons Based on Heat Transfer/Pressure Drop. [6]

Figure 5. Profile Comparisons Based on Heat Transfer/Volume. [6]

Sikka et al. [7] performed experiments on heat sinks with different fin geometries. Figure 6 shows 3 different categories of heat sinks tested. The conventional fins, such as straight and pin fins, are shown in (a); (b) shows the fluted fins and (c) shows the wavy fin design. The tests were done for both horizontal and vertical direction of air flow at natural convection and low Reynolds number forced flow. Table 1 shows the dimensional values of each of these heat sinks.

The last column shows the values of At/Ab (total surface area/base surface area).

Figure 6. (a) Traditional Fins, (b) Fluted Fins, (c) Wavy Fins. [7]

Table 1. Geometries and Dimensions of the Heat Sinks. [7]

The values of the Nusselt number were reported based on the following relation:

Figure 7 shows that for natural convection in the horizontal direction, the pin fin has the best performance. The fluted fins have, in general, a better performance compared to longitudinal fins. The lower graph in figure 7 shows that the wavy fins are essentially the same as the longitudinal fins.

Figure 7. Nusselt Number As a Function of Rayleigh Number for Natural Convection-Horizontal Direction. [7]

Figure 8 shows the natural convection cases for the vertical direction. The figure shows that heat transfer decreases for the pin fin, but increases for the plate fin. The pin fin still is better than the plate fin, but the difference is only 4-6%. Figure 8 also shows that the cross cut heat sink has the best performance. The bottom figure in 8 confirms that the wavy fins do not have much better heat transfer compared to plate fins.

Figure 8. Nusselt Number as a Function of Rayleigh Number for Natural Convection-Vertical Direction. [7]

Figure 9 shows the Nusselt number for forced convection over a horizontal plate as a function of Reynolds number. This figure indicates that, for very low Reynolds numbers, the cross fin is better than the pin fin; but, around Re = 2000, the situation reverses and the pin fin gets better than the cross cut heat sink. For low Reynolds numbers, the longitudinal pins are better than the wavy fins; but, at higher Reynolds numbers, the performance of the wavy fins gets better by almost 12-18%.

Figure 9. Nusselt Number as a Function of Reynolds Number for Forced Convection-Horizontal Direction. [7]

Figure 10 provides the Nusselt numbers for the vertical direction for forced flow. In comparing the results with the horizontal direction, the results are almost the same, with the difference being that the wavy fin heat sinks perform better than the plate fin heat sinks, by about 14-20%.

Figure 10. Nusselt Number as a Function of Reynolds Number for Forced Convection-Vertical Direction.[7]

The results presented in this article strengthen our understanding about how heat exchangers and heat sinks can be made more compact and efficient. The results show that the design of the fin field is still an issue and much remains to be investigated for optimization, depending on the conditions and application.

Further empirical testing is warranted for the evaluation of the effects of wavy fin heat sinks, as fine meshing and a high degree of confidence is not easily obtained through simulating these profiles using commercial CFD tools.

References:

1. Lorenzini, M., “Performance Evaluation of a Wavy-Fin Heat Sink for Power Electronics” Applied Thermal Engineering, 2007.
2. Awad, M., Muzychka, S., “Models for pressure drop and heat transfer in air cooled compact wavy fin heat exchangers”, Journal of Enhanced Heat Transfer, 18(3):191-207(2011).
3. Shah, R., London, A., “Advances in heat transfer, suppl. 1, laminar forced flow convection in ducts”, New York, Academic press, 1978
4. Shapiro, A., Sigel, R., Kline, S., “Friction factor in the laminar entry region of a smooth tube,” Proc., 2nd V.S.Nat. Congress of applied mechanics, PP. 733-741, 1954.
5. Kays, M., London,L., “Compact Heat Exchangers”, Third Edition, McGraw-Hill, 1984.
6. Marthinuss, E., Hall, G., “Air cooled compact heat exchanger design for electronics cooling”, Electronics cooling magazine, Feb 1st, 2004
7. Sikka, K., Torrance, K., Scholler, U., Salanova, I., “Heat sinks with fluted and wavy fins in natural and low-velocity forced convection”, IEEE, Intersoceity Conference, 2000.

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

1 Comment

Posted in Analysis, Heat Exchangers, Heat Sinks, heat transfer, Thermal Analysis

Tagged , , , , , ,

How Do Heat Sink Materials Impact Performance

Posted on April 19, 2017 by | 3 comments

By Michael Haskell, Thermal Engineer
and Norman Quesnel, Senior Member of Marketing Staff
Advanced Thermal Solutions, Inc.

(This article was featured in an issue of Qpedia Thermal e-Magazine, an online publication produced by Advanced Thermal Solutions, Inc. (ATS) 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.)

Heat Sink Materials

This article examines the difference in thermal performance between copper, aluminum, and graphite foam heat sinks. (Advanced Thermal Solutions, Inc.)

Introduction

As thermal solutions for today’s electronics grow more challenging, demand rises for novel cooling ideas or materials. As a result, the proven methods of analytical calculations, modeling and laboratory testing are sometimes bypassed for a quick “cure-all” solution. Evolutionary progress is required of the thermal industry, of course. But, despite the urgency to introduce new ideas and materials, thorough testing should be performed in determining the thermal performance of a solution before it is implemented.

This article addresses the impact of material choice on heat sink performance. First, an evaluation of different materials is made in a laboratory setting, using mechanical samples and a research quality wind tunnel. This testing compares a constant heat sink geometry made from copper, aluminum, and graphite foam. Next, an application-specific heat sink study is presented using computational fluid dynamics (CFD) software.

In this study, a heat sink was designed in 3D CAD to cool a dual core host processor. The performance of both an aluminum and copper design was then evaluated using CFD.

Laboratory Tests of Copper, Aluminum, and Graphite Foam

The stated thermal properties of engineered graphite foams have enhanced their consideration as heat sink materials. Yet, the literature is void of a true comparison of these materials with copper and aluminum. To evaluate graphite foam as a viable material for heat sinks, a series of tests were conducted to compare the thermal performance of geometrically identical heat sinks made of copper, aluminum, and graphite foam respectively.

Testing was conducted in a research quality laboratory wind tunnel where the unducted air flow was consistent with typical applications.

(The results for ducted and jet impingement flows, though similar to the unducted case, will be presented in a future article along with a secondary graphite foam material.)

Test Procedure

Earlier foam experiments by Coursey et al. [1] used solder brazing to affix a foam heat sink to a heated component. The solder method reduced the problematic interfacial resistance when using foams, due to their porous nature. Directly bonding the heat sink to a component has two potential drawbacks. First, the high temperatures common in brazing could damage the electrical component itself.

The other issue concerns the complicated replacement or rework of the component. Due to the low tensile strength of foam (Table 1) a greater potential for heat sink damage occurs than with aluminum or copper [2]. If the heat sink is damaged or the attached component needs to be serviced, direct bonding increases the cost of rework.

Table 1. Thermal and Mechanical Properties of the Heat Sink Materials. (Advanced Thermal Solutions, Inc.)

To avoid these problems, the foam heat sink can be soldered to an aluminum or copper carrier plate. This foam-and-plate assembly can then be mounted to a component in a standard fashion. The carrier plate allows sufficient pressure to be applied to the interface material, ensuring low contact resistance.

In this study, the heat sinks were clamped directly to the test component without a carrier plate as a baseline for all three materials. Shin-Etsu X23 thermal grease was used as an interface material to fill the porous surface of the foam and reduce interfacial resistance. Five J-type thermocouples were placed in the following locations: upstream of the heat sink to record ambient air temperatures, in the heater block, in the center of the heat sink base, at the edge of the heat sink base, and in the tip of the outermost fin.

Heat Sink Material

Figure 1. Test Heat Sink Drawing. (Advanced Thermal Solutions, Inc.)

A thin film heater was set at 10 watts during all testing, and the heat source area was 25 mm x 25 mm, or one quarter of the overall sink base area, as shown in Figure 1. Both cardboard and FR-4 board were used to insulate the bottom of the heater, The estimated value of Ψjb is 62.5°C/W. Throughout testing, the value of Ψjb was 36–92 times greater than that of Ψja.

Results

As expected, the traditional copper and aluminum heat sinks performed similarly. The main difference was due to the higher thermal conductivity of copper, which reduced spreading resistance. During slow velocity flow conditions, the lower heat transfer rate means that convection thermal resistance makes up a large portion of the overall Θja.

Heat Sink Materials

Table 2. Test Heat Sink Geometry. (Advanced Thermal Solutions, Inc.)

Heat Sink Materials

Figure 2. Experimental Heater and Measurement Setup. (Advanced Thermal Solutions, Inc.)

As flow speed increases, the convection resistance decreases, and the internal heat sink conduction resistance is more of a factor in the overall Θja value. This behavior is evident in the table below, and when comparing the different heat sink materials. The graphite heat sink’s thermal performance was only 12% lower than aluminum at low flow rates. However, the performance difference increased to 25-30% as the flow rate increased (Table 3).

Heat Sink Materials

Table 3. Specific Thermal Test Results. (Advanced Thermal Solutions, Inc.)

Due to the lack of a solder joint, the foam heat sink experienced a larger interfacial resistance when compared to the solid heat sinks. This difference can be seen when comparing ΨHEATER-BASE in Table 3. To decouple the effect of interfacial resistance ΨBASE-AIR can be calculated. When ignoring interfacial resistance in this manner foam performs within 1% of aluminum at 1.5 m/s, and within 15% at 3.5 m/s.

Heat Sink Materials

Figure 3. Heat Sink Thermal Resistance as a Function of Velocity. (Advanced Thermal Solutions, Inc.)

Graphite foam-derived heat sinks show promise in specific applications, but exhibit several drawbacks in mainstream electronics cooling. Due to the frail nature of graphite foam, unique precautions must be taken during the handling and use of these heat sinks. When coupled to a copper base plate, graphite foam can perform with acceptably small spreading resistances.

However, the foam’s lower thermal conductivity reduces thermal performance at high flow velocities compared to a traditional copper heat sink.

The mechanical attachment needed to ensure acceptable thermal interface performance without soldering or brazing also hinders foam-based heat sinks from being explored in mainstream applications. Despite these challenges, the thermal performance-to-weight ratio of foam is very attractive and well-suited to the aerospace and military industries, where cost and ease of use come second to weight and performance.

Thermal Software Comparison of Aluminum and Copper Heat Sinks

A challenging thermal application was considered. This involved the use of a dual core host processor on a board with limited footprint area for a heat sink of sufficient size. A heat sink with a stepped base was designed to clear onboard components. It provided sufficient surface area to dissipate heat (Figure 4).

Due to the complexity of the heat sink, machining a test sample from each material was not practical. Instead, CFD was used to predict the performance difference between the two materials and determine if the additional cost of copper was warranted.

Heat Sink Materials

Figure 4. Stepped Base maxiFLOW™ Heat Sink (ATS). (Advanced Thermal Solutions, Inc.)

Because of the stepped base and a long heat conduction path, spreading resistance was a major factor in the overall thermal resistance. The effect of copper in place of aluminum due to its higher thermal conductivity (400 and 180 W/m*K respectively) is shown in Table 4. The CFD software predicted a 21% improvement using copper in place of aluminum. More importantly, it reduced the processor case temperature below the required goal of 95°C.

The performance improvement with copper is due to the reduced spreading resistance from the processor die to the heat sink fins. This effect is shown in Figure 5, where the base temperatures of both heat sinks are obtained from the CFD analysis and plotted together. The aluminum heat sink shows a hotter center base temperature and a more pronounced drop off in temperature along the outer fins. The copper heat sink spreads the heat to all fins in a more even fashion, increasing the overall efficiency of the design. This temperature distribution can be seen in Figures 6 and 7, which were created using CFDesign software.

Heat Sink Materials

Figure 5. Effect of Heat Sink Material on Temperature Distribution. (Advanced Thermal Solutions, Inc.)

Heat Sink Materials

Figure 6. Aluminum Stepped Base maxiFLOW™ Heat Sink Simulation. (Advanced Thermal Solutions, Inc.)

Heat Sink Materials

Figure 7. Copper Stepped Base maxiFLOW™ Heat Sink Simulation. (Advanced Thermal Solutions, Inc.)

Conclusion

Design engineers have many materials at their disposal to meet the challenging thermal needs of modern components. Classic materials such as aluminum and copper are joined by new technologies that bring improvements in cost, weight, or conductivity. The choice between a metallic, foam or plastic heat sink can be difficult because thermal conductivity provides the only available information to predict performance.

The first method for determining material selection is a classic thermodynamics problem: what effect does conductivity have on the overall thermal resistance in my system? Only once this is answered can the benefits of cost, weight, and manufacture be addressed.

References

1. Coursey, J., Jungho, K., and Boudreaux, P. Performance of Graphite Foam Evaporator for Use in Thermal Management, Journal of Electronics Packaging, June 2005.
2. Klett, J., High Conductivity Graphitic Foams, Oak Ridge National Laboratory, 2003.

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

3 Comments

Posted in Analysis, heat sink, Heat Sink Material, Heat Sinks, Thermal Analysis

Tagged , , ,

How Did Thermal Performance of Aluminum Heat Sink Compare to Copper?

Posted on February 2, 2017 by | 4 comments

Advanced Thermal Solutions, Inc. (ATS) was recently tasked with creating a more cost-effective and lighter solution for a customer that was looking to replace a relatively large heat sink, which was dissipating the heat from four components on a printed circuit board (PCB). The customer did not want a skived heat sink, so ATS engineers created a custom aluminum heat sink embedded with copper heat pipes to draw the heat away from the components.

ATS engineers worked on a comparison of a copper heat sink with an aluminum heat sink that had embedded heat pipes running above the components. Analysis showed that the aluminum heat sink nearly matched the thermal performance of the copper and was within the margin required by the client. (Advanced Thermal Solutions, Inc.)

ATS engineers used analytical modeling and CFD simulations to examine the thermal performance of two aluminum heat sink designs: one with heat pipes that stopped at the edge of the components and the other with heat pipes that ran above the components. Analysis demonstrated that the design with heat pipes running above the components kept junction temperatures within 2°C of the original copper heat sink and an average difference of less than 1°C.

Peter Konstatilakis, a Field Application Engineer at ATS who worked with the client on this analysis, sat down with Marketing Communications Specialist Josh Perry to discuss the technical details behind the thermal analysis and the results that were presented to the customer.

JP: Thanks for taking the time to talk about this project Peter. What was it that they approached you with? What was the problem or the challenge?
PK: There was a long lead time with sourcing this copper; it’s a relatively large and heavy part.  This size bar of copper isn’t typically stocked. So, we were having sourcing issues with this non-standard copper stock and they were having weight and cost issues. They had to cut this heat sink in half for testing because they were overweight on the board. Through shock and vibe testing, if the heat sink is too heavy then it can actually rip out of the board.

An alternative was to make the heat sink through a manufacturing process called skiving. Skived heat sinks have a fin count tolerance, so you may have more fins than are specified or you might have less fins, and some of the fins may be curved, which poses cosmetic issues with skived heat sinks; the fins aren’t perfectly straight. It’s not really an issue thermally, especially if companies don’t see the heat sinks too often, but this client’s customers see the boards, see the heat sinks, and they wanted them to look perfect.

Instead of having to get this copper, we thought, why don’t we make an aluminum heat sink with heat pipes? That’s sort of where this came from.

JP: So the problem with skiving a heat sink was mostly an issue with aesthetics?
PK: Yeah, exactly. The tolerance on the fin spacing was +/- three fins, due to the high number of fins. I did a quick analytical analysis with our heat sink calculation tool and the difference in thermal resistance was maybe 1%. That was because the heat sink has such a large surface area and losing a fin or two only changes the performance by a percent or less. On a smaller heat sink, you will see a greater difference. I told the customer but they said that they still didn’t want to go with skived for aesthetic reasons. Instead, we extruded aluminum and then we put heat pipes in the base.

JP: Why was it necessary to add heat pipes to the heat sink?
PK: The big thing, in this case, is the spreading. You can see the locations of the components and then how large the heat sink is. There’s definitely a lot of spreading resistance in the base because there’s so much distance between the heat sink and all the components, so that’s the main issue that we were trying to take care of with the heat pipes. An aluminum heat sink with heat pipes is definitely a lot lighter than a copper heat sink, about three times lighter. Overall it’s much easier to source and also much cheaper. I think it’s again about three times as much for copper.

JP: When this challenge came across your desk, what was the first thing that you looked at? How did you approach the challenge?
PK: What I did was look at our analytical tool again and I modeled this heat sink in all copper. Since there are four components it’s a little complicated, but I modeled them as one component in the middle of the heat sink with gap pad and everything and got the performance of that heat sink. Once I did that, I ran CFD simulations on the copper heat sink with the components placed as they are in the application and the performance values were within 15%. So, doing that, we knew that we had a good CFD model.

After running the baseline simulations on the copper, I moved onto the aluminum heat sink knowing that we had a good CFD model and that we could trust the results. I used the aluminum heat sink and put heat pipes in the base. I started with heat pipes out in front of the components and then the next simulation was with heat pipes above the components. Obviously, if the heat pipes are above the component then you’ll get a little better spreading resistance and the heat will flow better.

Aluminum Heat Sinks

The first of two aluminum heat sink designs had heat pipes that stopped at the components. This design was not as effective as when the heat pipes ran above the components. (Advanced Thermal Solutions, Inc.)

JP: How significant of a difference was it?
PK: From the base line of the copper heat sink, it was around a 1-2°C difference, on average.

After looking at these two simulations, I met with Dr. Kaveh Azar (founder, CEO and President of ATS) to discuss the results. With the heat pipes above the components, we were seeing an average difference of less than 1%. It performs worse by less than 1% and I’m currently doing a couple of other simulations to see if we can improve that by adding more heat pipes, making the heat pipes wider, or even running less conservative heat pipes since the conductivity I’m running with is 2000 W/m-K axially and 400 W/m-K through the cross section. Really, the axial conductivity should be around 20,000-50,000 W/m-K, and the copper wall and wick effective conductivity is around 100-200 W/m-K due to the low conductivity of the porous copper sintered wick. The conservative values I used were to get the simulation up and running, while I’ll end up analytically determining the respective heat pipe conductivity.

I’m also doing an all-aluminum simulation just so we can see what that looks like and so we can see how much better the copper heat sink is in general.

This turned into just looking at the heat sink and trying to put heat pipes in them to seeing if we could also vary the length and see if we could get better performance. Your pressure drop increases as the length increases, so the higher the pressure drop then the lower the air flow is going to be in the system, the lower the airflow then the lower the performance. There is sweet spot for the length. I’m looking at that with our analytical calculator. And then the base thickness as well, we’re looking at that too.

Aluminum Heat Sinks

The results of the CFD analysis showed that the average temperature difference between the copper and the second aluminum heat sink design was less than one degree. (Advanced Thermal Solutions, Inc.)

JP: With the aluminum heat sink within 1% of the copper, did that make switching from copper worth it for the customer?
PK: It definitely did. If you’re within 1% and the customer has a little margin already, then it’s worth it because it’s three times lower cost, lower weight, and it will look better because it’s extruded rather than skived.

JP: Just to clarify, what is the difference between skiving and extruding?
PK: Extruding, basically, is pushing a hot piece of metal through a die that is in the shape of a heat sink, so it’s like putting play-doh through a die. You get a heat sink with the fin pitch and everything, where skiving uses a copper block and they come in with a blade and peel the fin out. The blade comes in and pushes a layer up. You can skive aluminum as well and they’re about the same cost, but you can’t extrude copper for a heat sink.

This showed our thermal capability to the customer. It showed that we can design custom heat sinks. We can make them more cost-effective, better performing, whatever they need.

JP: When you’re working through these types of challenges, how much of it becomes a foundation of knowledge that you can then take to another customer’s project?
PK: The more experience that you have, the better. Like any field, the more experience you have then you can look at something and know right off the bat if it’s going to work or not. It also helps in terms of understanding how to model certain applications and where to start with the design.

JP: Did we run these simulations here or did we have (ATS engineer) Sridevi Iyengar run the simulations in India?
PK: We did it here. Sri does a lot, but she uses FloTHERM and I’m quicker with Autodesk CFDesign. FloTHERM can be used for bigger systems because it takes less of a mesh. Generally, FloTHERM only works in rectangular coordinates, where CFDesign works with tetrahedrons, allowing the simulation of angled objects. Since it works with tetrahedrons though, it takes longer to mesh and run than FloTHERM. You can’t really have anything angled in FloTHERM and obtain accurate results. We ended up having to angle the heat pipes in order to contact the components, which are in a different plane than the rest of the heat sink.

JP: I know it is a priority at ATS, but why was it important to have an analytical component, not just CFD, in finding a solution?
PK: Analytical modeling is used to ensure that the CFD results make sense. When you see the graphs from CFD, it looks appealing to the eye and you get drawn to it. It’s science and engineering that is made visible, whereas heat transfer and fluid dynamics (for air) are invisible to the naked eye. Another method of ‘seeing’ heat transfer is using an infrared thermal camera or liquid crystal thermography, while a water tunnel or inducing smoke into the flow can be used to see fluid flow. The analytical also gives us a good first judgement and solid design direction.

Optimization for the length of the heat sink and the base thickness, I did with our analytical tool. CFD simulations take a lot of time, so I can narrow down the number of designs and determine what we want to simulate. Rather than doing 10 different simulations, when each takes on average three or four hours, I can get instant results and say, okay, a 5 mm base is the sweet spot, so let me try in CFD 4 mm thickness, 5 mm, and 6 mm; narrowing it down to three simulations.

Analytical modeling gives us quick what-if scenarios, which we say a lot, and it definitely helps give you an understanding of what to expect. If the numbers are way off then I know something is wrong in the CFD model and I check to see if my mesh and other parameters are correct. It humbles you almost and it helps you understand the application and what you’re simulating. The last thing you want to do is give a customer incorrect data.

It gives you two independent solutions. We say analytically this solution is validated, so we have faith in the model. Now, here is the model and it shows better what we want to do.

To learn more about Advanced Thermal Solutions, Inc., visit www.qats.com or contact ATS at 781.769.2800 or ats-hq@qats.com.

4 Comments

Posted in Analysis, CFD, Consulting, heat sink design, Thermal Analysis

Tagged , , , , , , ,

Discussion of Thermal Solution for Stratix 10 FPGA

Posted on October 28, 2016 by | Leave a comment

An Advanced Thermal Solutions, Inc. (ATS) client was planning on upgrading an existing board by adding Altera’s high-powered Stratix 10 FPGAs, with estimates of as many as 90 watts of power being dissipated by two of the components and 40 watts from a third. The client was using ATS heat sinks on the original iteration of the board and wanted ATS to test whether or not the same heat sinks would work with higher power demands.

In the end, the original heat sinks proved to be effective and lowered the case temperature below the required maximum. Through a combination of analytical modeling and CFD simulations, ATS was able to demonstrate that the heat sinks would be able to cool the new, more powerful components.

ATS Field Application Engineer Vineet Barot recently spoke with Marketing Director John O’Day and Marketing Communications Specialist Josh Perry about the process he undertook to meet the requirements of the client and to test the heat sinks under these new conditions.

JP: Thanks again for sitting down with us to talk about the project Vineet. What was the challenge that this client presented to us?
VB: They had a previous-generation PCB on which they were using ATS heat sinks, ATS 1634-C2-R1, and they wanted to know if they switched to the next-gen design with three Altera Stratix 10 FPGAs, two of them being relatively high-powered, could they still use the same heat sinks?

Stratix 10 FPGA

The board that was given to ATS engineers to determine whether the original ATS heat sinks would be effective with new, high-powered Stratix 10 FPGA from Altera. (Advanced Thermal Solutions, Inc.)

They don’t even know what the power of the FPGAs is exactly, but they gave us these parameters: 40°C ambient with the junction temperatures to be no more than 100°C. Even though the initial package is capable of going higher, they wanted this limit. That translates to a 90°C case temperature. You have the silicon chip, the actual component with the gates and everything, and you have a package that puts all that together and there’s typically a thermal path that it follows to the lid that has either metal or plastic. So, there’s some amount of temperature lost from the junction to the case.

The resistance is constant so you know for any given power what the max will be. The power that they wanted for FPGAs 1 and 2, which are down at the bottom, was 90 watts, again this is an estimate, and the third one was 40 watts.

JP: How did you get started working towards a solution?
VB: Immediately we tried to identify the worst-case scenario. Overall the board lay-out is pretty well done because you have nice, linear flow. The fans are relatively powerful, lots of good flow going through there. It’s a well-designed board and they wanted to know what we could do with it.

I said, let’s start with the heat sinks that you’re already using, which are the 1634s, and then go from there. Here are the fan specs. They wanted to use the most powerful fan here in this top curve here. This is flow rate versus pressure. The more pressure you have in front of a fan, the slower it can pump out the air and this is the curve that determines that.

Stratix 10 FPGA

Fan operating points on the board, determined by CFD simulations. (Advanced Thermal Solutions, Inc.)

This little area here is sometime called the knee of the fan curve. Let’s say we’re in this area, the flow rate and pressure is relatively linear, so if I increase my pressure, if I put my hand in front of the fan, the flow rate goes down. If I have no pressure, I have my maximum flow rate. If I increase my pressure then the flow rate goes down. What happens in this part, the same thing. In the knee, a slight increase in pressure, so from .59 to .63, reduces the flow rate quite a bit.

Stratix 10 FPGA

CFD simulations showed that the fans were operating in the “knee” where it is hard to judge the impact of pressure changes on flow rate and vice versa. (Advanced Thermal Solutions, Inc.)

So, for a 0.1 difference in flow rate (in cubic meters per second) it took 0.4 inches of water pressure difference, whereas here for a 0.1 difference in flow rate it only took a .04 increase in pressure. That’s why there’s a circle there. It’s a danger area because if you’re in that range it gets harder to predict what the flow will be because any pressure-change, any dust build-up, any change in estimated open area might change your flow rate.

The 1634 is what they were using previously. It’s a copper heat pipe, straight-fin, mounted with a hardware kit and a backing plate that they have. It’s a custom heat sink that we made for them and actually the next –gen, C2-R1, we also made for them for the previous-gen of their board, they originally wanted us to add heat pipes to this copper heat sink, but I took the latest version and said, let’s see what this one will do. For the third heat sink, I went and did some analytical modeling to see what kind of requirement would be needed and I chose one of our off-the-shelf pushPIN™ heat sinks to work because it was 40 watts.

JO: Is the push pin heat sink down flow from the 1634, so it’s getting preheated air?
VB: Yes. This is a pull system, so the air is going out towards the fans.

Stratix 10 FPGA

CFD simulations done with FloTherm, which uses a recto-linear grid. (Advanced Thermal Solutions, Inc.)

This is the CFD modeling that ATS thermal engineer Sridevi Iyengar did in FloTherm. This is a big board. There are a lot of different nodes, a lot of different cells and FloTherm uses recto-linear grids to avoid waviness. You can change the shape of the lines depending on where you need to be. Sri’s also really good at modeling. She was able to turn it around in a day.

Stratix 10 FPGA

Flow vectors at the cut plane, as determined by CFD simulations. (Advanced Thermal Solutions, Inc.)

These are the different fans and she pointed out what the different fan operating curves. Within this curve, she’s able to point out where the different fans are and she’s pointing out that fan 5 is operating around the knee. If you look at all the different fans they all operate around this are, which is not the best area to operate around. You want to operate down here so that you have a lot of flow. If you look at the case temperatures, remember the max was 90°C, we’re at 75°C. We’re 15°C below, 15° margin of error. This was a push pin heat sink on this one up here and 1634s on the high-powered FPGAs down here.

Stratix 10 FPGA

JP: Was there more analysis that you did before deciding the original heat sinks were the solution?
VB: I calculated analytical models using the flow and the fan operating curves from CFD because it’s relatively hard to predict what the flow is going to be. Using that flow and doing a thermal analysis using HSM (heat sink modeling tool), we were within five percent. What Sri simulated with FloTherm was if a copper heat sink with the heat pipe was working super well, let’s try copper without the heat pipe and you can see the temperature increased from 74° to 76°C here, still way under the case temperature. Aluminum with the heat pipe was 77°; aluminum without the heat pipe was 81°, so you’re still under.

Basically there were enough margins for error, so you could go to smaller fans because there’s some concern about operating in the knee region, or you can downgrade the heat sink if the customer wanted. We presented this and they were very happy with the results. They weren’t super worried about operating in the knee region because there’s going to be some other things that might shift the curve a little bit and they didn’t want to downgrade the heat sink because of the power being dissipated.

Stratix 10 FPGA

Final case temperatures determined by CFD simulations and backed up by analytical modeling. (Advanced Thermal Solutions, Inc.)

JO: What were some of the challenges in this design work that surprised you?
VB: The biggest challenges were translating their board into a board that’s workable for CFD. It’s tricky to simplify it without really removing all of the details. We had to decide what are the details that are important that we need to simulate. The single board computer and power supply, this relatively complex looking piece here with the heat sink, and we simplified that into one dummy heat sink to sort of see if it’s going to get too hot. It all comes with it, so we didn’t have to work on it.

The power supply is even harder, so I didn’t put it in there because I didn’t know what power it would be, didn’t know how hot it would be. I put a dummy component in there to make sure it doesn’t affect the air flow too much but that it does have some effect so you can see the pressure drop from it but thermally it’s not going to affect anything.

JO: It really shows that we know how to cool Stratix FPGAs from Altera, we have clear solutions for that both custom and off-the-shelf and that we understand how to model them in two different ways. We can model them with CFD and analytical modeling. We have pretty much a full complement of capabilities when dealing with this technology.

JP: Are there times when we want to create a TLB (thermal load board) or prototype and test this in a wind tunnel or in our lab?
VB: For the most part, customers will do that part themselves. They have the capability, they have the rack and if it’s a thing where they have the fans built into the rack then they can just test it. On a single individual heat sink basis, it’s not necessary because CFD and analytical modeling are so established. You want two independent solutions to make sure you’re in the right ballpark but it’s not something you’re too concerned that the result will be too far off of the theoretical. For another client, for example, we had to make load boards, but even then they did all the testing.

To learn more about Advanced Thermal Solutions, Inc. consulting services, visit https://www.qats.com/consulting or contact ATS at 781.769.2800 or ats-hq@qats.com.

Leave a comment

Posted in Analysis, CFD, Chassis, Consulting, Heat Sinks, Modeling, PCB, Simulation, Thermal Analysis

Tagged , , , , , , , , , , , , , ,

Technical Discussion of ATS Telecom PCB solution

Posted on October 28, 2016 by | 1 comment

Last year, Advanced Thermal Solutions, Inc. (ATS) was brought in to assist a customer with finding a thermal solution for a PCB that was included in a data center rack being used in the telecommunications industry. The engineers needed to keep in consideration that the board’s two power-dissipating components were on opposite ends and the airflow across the board could be from either side.

Telecom PCB

The PCB layout that ATS engineer Vineet Barot was asked to design a thermal solution for included two components on opposite ends and airflow that could be coming from either direction. (Advanced Thermal Solutions, Inc.)

The original solution had been to use heat sinks embedded with heat pipes, but the client was looking for a more cost-effective and a more reliable solution. The client approached ATS and Field Application Engineer Vineet Barot examined the problem to find the best answer. Using analytical and CFD modeling, he was able to determine that ATS’ patented maxiFLOW™ heat sinks would provide the solution.

Vineet sat down with Marketing Director John O’Day and Marketing Communications Specialist Josh Perry to discuss the challenges that he faced in this project and the importance of using analytical modeling to back up the results of the CFD (computational fluid dynamics).

JP: Thanks for sitting down with us Vineet. How was the project presented to you by the client?
VB: They had a board that was unique – where it would be inserted into a rack, but it could be inserted in either direction. So, we faced a unique challenge because airflow could be from either side of the board. There were two components on either side of the board, so if airflow was coming from one side then component ‘A’ would get hot and from the other side then component ‘B’ would get hot. The other thing was that the customer, who is a very smart thermal engineer, had already set up everything and he was planning on using these heat sinks that had heat pipes embedded in them. The goal was to try and come up with a heat sink that would do the same thing, hopefully without requiring the heat pipes.

JO: Can we talk for a second about the application? You mentioned that airflow was from either side, the board was going to be used in a data center or a telecom node?
VB: It was for a telecom company.

JP: Was there a reason he didn’t want to use a heat pipe?
VB: I think probably cost and reliability. We use heat pipes embedded in the heat sinks too, so it’s not a something we never want to use, but the client wanted to throw that at us and see if we had alternatives.

JP: Can you take us through the board and the challenges that you saw?
VB: As you can see from this slide, there are four main components and two of the hottest ones are on the edge. Airflow can be from right to left or left to right, so which one would be the worst-case scenario?

Telecom PCB

JO: From right to left, I think?
VB: Correct. This one is a straightforward one to figure out because not only is the component smaller but the power is also higher. Even though [air] can go both ways, there’s a worst-case scenario.

This was the customer’s idea – a straight-fin heat sink with a heat pipe and he put one block of heat pipe in there instead of two or three heat pipes that would normally be embedded in there. You can clearly see what the goal was. You have a small component in here, you want to put a large heat sink over the top and you want to spread the heat throughout the base of the heat sink. All the other components are also occupied by straight-fin heat sinks.

JO: Okay, at this point in the analysis, this is the rough estimate of the problem that you face?
VB: This is a straightforward project in terms of problem definition, which can be a big issue sometimes. This time problem definition was clear because the customer had defined the exact heat sink that they wanted to use. It’s not a bad heat sink they just wanted an improvement, cost-wise, reliability-wise.

This is the G600, which is the air going from left to right. The two main components are represented here and we want to make sure that the junction temperatures that the CFD calculated is lower than the maximum junction temperatures allowed, which they were. These heat sinks work. As we always like to do at ATS, we like to have two, independent solutions to verify any problem. That was the CFD result but we also did the analytical modeling to see what these heat sinks are capable of and the percent difference from CFD was less than 10 percent. Twenty percent is the goal typically. If it’s less than 20 percent then you know you’re in the ballpark.

(Advanced Thermal Solutions, Inc.)

(Advanced Thermal Solutions, Inc.)

JO: Do you use a spreadsheet to do these analytical modeling?
VB: HSM, which is our heat sink modeling tool, and then for determining what velocity you have through the fins, the correct way of doing this is to come up with the flow pattern on your own. You go through all the formulas in the book and determine what the flow will be everywhere or figure out what CFD is giving you for the fan curve and check it with analytical modeling. You can look at pressure drop in there, look at the fan curve and see if you’re in the ballpark. You can also check other things in CFD, for example flow balance. Input the flow data into HSM and it will spit out what the thermal performance is for any given heat sink. HSM calculations are based on its internal formulas.

JO: We effectively have a proprietary internal tool. We’ve made a conscious decision to use it.
VB: To actually use it is unique. Not everybody would use it. A lot of people would skip this step and go straight to CFD. We use CFD too but we want to make sure that it’s on the right path.

JP: What do you see as the benefit of doing both analytical and CFD modeling?
VB: CFD, because it’s so easy to use, can be a tool that will lead you astray if you don’t check it because it’s very easy to use and the software can’t tell you if your results are accurate. If you do any calculation, you use a calculator. The calculator is never going to give you a wrong answer but just because you’re using a calculator doesn’t mean that you’re doing the math right. You want to have a secondary answer to verify that what you did is correct.

JP: What was the solution that you came up with for this particular challenge?
VB: We replaced these heat sinks with the heat pipe with maxiFLOW™, no heat pipe needed. One of the little tricks that I used was to off-set the heat sinks a little bit so that these fins are out here and so the airflow here would be kind of unobstructed. And I set this one a little lower so it would have some fins over here, not much, that would be unobstructed. The G600 configurations worked out with the junction temperatures being below what the maximum requirement was without having to use any heat pipes for the main components. There is also a note showing that one of the ancillary components was also below the max. Analytical modeling of that was within 10-11 percent.

The final PCB layout with maxiFLOW heat sinks covering the hottest components on both ends of the board. (Advanced Thermal Solutions, Inc.)

The final PCB layout with maxiFLOW heat sinks covering the hottest components on both ends of the board. (Advanced Thermal Solutions, Inc.)

As you noted, this was the worst-case scenario, going from right to left and you can see because it’s the worst-case scenario this tiny little component here that’s 14 watts that’s having all this pre-heated air going into it, it’s junction temperature was exactly at the maximum allowed. That’s not entirely great. We want to build in a little bit of margin but it was below what was needed.

The conclusion here was that maxiFLOW™ was able to provide enough cooling without needing to use the heat pipes and analytical calculation agreed to less than 20 percent. We would need to explore some alternate designs and strategies if we want to reduce the junction temperature even further because that close to the maximum temperature is uncomfortable. The other idea that we had was to switch the remaining heat sinks, the ones in the middle, which are straight fin, also to maxiFLOW™ to reduce pressure drop and to get more flow through this final component.

(Advanced Thermal Solutions, Inc.)

(Advanced Thermal Solutions, Inc.)

JP: If you have an idea like that, is it something that you broach with the customer?
VB: They really liked the result. If this was a project where the customer said, ‘Yep, we need this,’ then we would have said here’s the initial result and we have an additional strategy. At that point the customer would have said, ‘Yeah this is making us uncomfortable and we need to explore further’ or they would have said, ‘You know what? Fourteen watts is a max and I don’t know if we’ll ever go to 14 watts or the ambient we’re saying is 50°C but we don’t know that it will ever get to 50°C so the fact that you’re at max junction temperature at the worst-case scenario is okay by us.’

JP: Do you always test for the worst-case scenario?
VB: It’s always at the worst-case scenario. It’s always at the max power and maximum ambient temperature.

JP: Was this the first option that we came up with, using maxiFLOW™? Were there other options that we explored?
VB: Pretty much. The way that I approached it was doing the analytical first. You can generate 50 results from analytical modeling in an hour whereas it takes a day and a half for every CFD model – or longer. These numbers here were arrived at with analytical modeling; the height, the width, the top width, were all from analytical modeling, base thickness to measure spreading resistance, all of that was done on HSM and spreadsheets to say this will work.

JP: Do you find that people outside ATS aren’t doing analytical?
VB: No one is doing it, which is really bad because it’s very useful. It gives you a quick idea if it’s acceptable, if this solution is feasible.

To learn more about Advanced Thermal Solutions, Inc. (ATS) consulting services, visit https://www.qats.com/consulting or contact ATS at 781.769.2800 or ats-hq@qats.com.

1 Comment

Posted in Analysis, CFD, Consulting, heat sink, PCB, Thermal Analysis

Tagged , , , , , , , , , , , , , ,