August 3rd, 2016 – Day One Program
7:00 am – Registration Opens/Breakfast
8:00 am – Opening Remarks
8:05 am -
Chip-Embedded Pumped Two Phase Cooling
High-end server-class processors continue to push towards increased single thread and throughput performance as CMOS technology approaches fundamental scaling limits. Three dimensional (3D) chip stacking technology can enable improved computational performance and power efficiency through increased number of closely coupled cores and cache, and/or by combining heterogeneous technologies within a single chip-stack.
However, thermal and associated reliability issues cannot be resolved by traditional 2D chip cooling technology applied to one surface of a 3D chip stack. To achieve thermal performance targets of complex high power 3D chip stacks, inter- or intra-chip cooling will be required. Interlayer two-phase evaporative cooling with a chip-to-chip interconnect-compatible dielectric fluid is an enabling technology for high power 3D stacks. Development of this approach requires high fidelity conjugate thermal models which can predict the thermal performance of a 3D chip stack with spatially varying heat sources coupled to a two phase microfluidic convection network. The challenge is to achieve models which are both accurate and computationally manageable. In this talk, a state-of-the-art conjugate heat transfer model developed for simulating two-phase flow boiling through chip embedded micron-scale channels will be described in detail. Additionally, experimental results from, and model validation against, a large 3D compatible demonstration vehicle having a non-uniform power map will be presented.
Pritish R. Parida, PhD, Research Staff Member – IBM Research
8:45 am -
Thermal Properties of Freestanding Reduced Graphene Oxide Films
Graphene and few-layer graphene (FLG) has revealed excellent heat conduction properties. It has already been demonstrated that graphene and FLG can be used as efficient fillers for thermal interface materials (TIMs), while graphene laminate has potential as a thermal coating for plastics. Reduced graphene oxide (rGO) can be obtained via inexpensive scalable techniques. For practical applications it is important to understand how the reduction process affects the heat conduction in rGO films.
In this presentation we report the results of our investigation of thermal and electrical conductivity of chemically and thermally reduced freestanding rGO films. Both the in-plane and cross-plane thermal conductivities were measured using a combination of techniques: the “laser flash” and Raman optothermal. We show that the in-plane thermal conductivity of rGO films obtained via various processes can reach up to 40 W/mK at room temperature. This is significant enhancement as compared to the thermal conductivity of graphene oxide (GO), which are in the range from 0.5 W/mK to 1 W/mK. The electrical conductivity was also measured to correlate its evolution with the thermal conductivity. The phonon thermal conductivity dominates the heat transport. The reduction results in increased concentration of carbon atoms. We found that the thermal conductivity scales up nearly exponentially with the carbon concentration. Further reduction in oxygen and impurity content can lead to the thermal conductivity values above those in conventional semiconductors. The cross-plane thermal conductivity is defined by the film morphology. A strong anisotropy of thermal conductivity in freestanding rGO films can be useful for heat spreader applications in electronic and optoelectronic devices where high in-plane and low cross-plane thermal conductivities are required for component protection.
Hoda Malekpour, PhD Candidate in Electrical Engineering and Researcher, Professor Alexander A. Balandin’s Phonon Optimized Engineered Materials (POEM) Center – University of California – Riverside (UCR).
9:20 am -
Cooling Fan Design System with Automated Performance Prediction
A cooling fan is widely used as a cooling device in electronics. In CFD, modeling methods of cooling fans can be categorized into two methods depending on the requirements. One method is used to predict detailed flow by modeling the actual shape of the fan blades in detail. The other method is used to predict the flow rate as a simplified model defining the relation between the pressure and the flow rate (P-Q Curve). The former method is best suited for predicting performance of the fan itself and its noise while the latter method is suitable for quickly simulation of a system level simulation. While the latter method significantly simplifies the procedures and reduces the complexity of the simulation, it has been addressed by many engineers that a P-Q curve provided by a fan manufacture does not match with the actual performance especially when a fan is installed in the vicinity of other components which could completely change the flow behavior.
This presentation introduces an innovative tool that allows an engineer to quickly and easily design a fan in 3D and to predict its performance including P-Q curve, power curve, and efficiency curve. As a result, the designer can design an optimized axial fan specific to the configuration of the system design.
Ben Cook, Engineer – Cradle North America
10:00 am – Exhibit Hall Opens/Networking Break in the Hall
11:00 am -
Enhancements of Immersion Cooling of High Power Chips with Nucleate Boiling of Dielectric Liquids
Nucleate boiling of dielectric liquids of FC-72, HFE-7100, and PF-5060 has been recognized as an effective method for cooling of high power computer chips and CPUs. Much work has been reported, which investigated the enhancement of nucleate boiling heat removal rate in saturation and subcooled boiling. In this method, the entire circuit board is immersed in a pool of dielectric liquid and the dissipated heat is removed from the chip surface by nucleate boiling.
This paper presents results of t experiments performed in our laboratory that investigated saturation and subcooled boiling of FC-72, HFE-7100 and PF-5060 liquids on rough Cu, porous graphite, and micro-porous Cu (MPC) and dimpled Cu surfaces. These surfaces, measuring 10 x10 mm, are uniformly heated in the pool boiling experiments. The experiments also investigated the effect of surface inclination angle, from 0o (upward facing) to 180o (downward facing) on the heat removal by nucleate boiling and CHF. Results of flow field visualizations of bubble nucleation and growth determined both the bubble departure diameter and detachment frequency on the various surfaces.
Mohamed S. El-Genk, Regents’ Professor of Nuclear, Mechanical and Chemical Engineering, and Director Institute for Space and Nuclear Power Studies
- University of New Mexico
11:40 am -
Thermal Management and Reliability of Automotive Power Electronics and Electric Machines
Increasing the number of electric-drive vehicles (EDVs) on America’s roads has been identified as a strategy with near-term potential for dramatically decreasing the nation’s dependence on oil―by the US (DOE), the federal cross-agency EV-Everywhere Challenge, and the automotive industry. Mass-market deployment will rely on meeting aggressive technical targets, including improved efficiency and reduced size, weight, and cost.
Many of these advances will depend on optimization of thermal management. Effective thermal management is critical to improving the performance and ensuring the reliability of EDVs. Efficient heat removal makes higher power densities and lower operating temperatures possible, and in turn enables cost and size reductions. The National Renewable Energy Laboratory (NREL), along with DOE and industry partners is working to develop cost-effective thermal management solutions to increase device and component power densities. In this presentation, the activities in recent years related to thermal management and reliability of automotive power electronics and electric machines will be presented.
Sreekant Narumanchi, Ph.D, Power Electronics and Electric Machines Section Supervisor in the Transportation and Hydrogen Systems Center - NREL
12:20 pm – Lunch
1:20 pm -
Thermally Conductive Ceramics – Performance and Applications
The presentation will discuss and compare the performance of several different thermally conductive ceramic materials, including strength characteristics and other physical properties. The presentation will also include the thermal transfer characteristics of these materials at various operating temperatures, and will include a direct comparison video of the materials in a controlled test.
Wayne Boone, Senior Account Manager – Materion Ceramics
2:00 pm -
Thermal Management for Device Hot Spots and Improved Battery Safety
In the world of electrified mobility driven by rechargeable power systems, there is a need for faster and improved systems to allow for recharging of the battery systems. Simply forcing higher current into the device and battery pack can create excessively high temperatures which leads to a number of problems such as hot spots, high surface touch temperatures, high battery temperatures leading to battery degradation, potential battery thermal runaway, etc.
Possible solutions to alleviate these issues created by fast or rapid charging is to insulate or move the battery away from the surface, or develop a smart solution to control the charging process by incorporating sensors to monitor the temperatures and voltages. These smart processes will monitor the battery charge properties and ramp down the charge rate to prevent overheating and protect the battery. Depending on the power system where multiple batteries may be connected, these systems must monitor each battery and adjust the charging parameters and circuitry accordingly to prevent uneven battery charging and uneven pack heating.
This presentation will show how the proper design and implementation of thermal materials can provide lower surface touch temperatures due to hot spot heat absorption and spreading, improved device performance due to more stable running parameters and speeds, improved temperature homogeneity across the device and multiple battery systems, improved battery life due to temperature control and even improved safety due to thermal runway prevention properties. The use of proper tests can clearly elucidate the benefits and limitations of incorporating these systems.
Mark Hartmann, CTO/R&D - Outlast Technologies LLC, CoorsTek Polymers Group
2:40 pm – Networking Break in the Exhibit Hall
3:00 pm -
As Electronics Get Smaller, Infrared Plays a Greater Role in Thermal Management
Thermal management is a challenge for designers and producers of advanced electronics, especially when they don’t show a clear source of overheating. Contact measurement tools such as thermocouples may be too large to use, and even smaller probes could skew measurements by acting as heat sinks. In addition, contact measurement can be hit and miss if it’s unclear what’s causing the hot spot. An infrared image of an overheating circuit board can show not only the source of overheating, but also measure its intensity. We will explore how non-contact temperature measurement such as infrared thermography can help pinpoint the source of hot spots, reveal patterns of heat dissipation, and provide a thermal map of temperatures across the entire device. Through examples and demonstrations, we will show the most effective ways to use infrared inspections, from identifying thermal properties to detecting design flaws and manufacturing errors. We will also discuss the challenges of infrared imaging, such as finding ways to measure shiny surfaces with low emissivity. Finally we will review the cost of infrared thermography versus its potential benefits, including improvements to overall electronic design, the early discovery of design flaws, and the reduction in time-to-market.
Chris Bainter - FLIR R&D/Science Segment
3:40 pm -
Ultrathin Thermal Ground Planes
Vapor chambers are passive thermal management systems that employ the evaporation of an encapsulated liquid to lift heat from a hot spot, convection of the hot vapor to spread the heat, condensation of the vapor to reject heat at a condenser, and capillary pumping of the liquid to return it to the hot spot. Owing to the viscosity of the vapor, typical vapor chambers are limited in their design to thicknesses greater than 0.4 mm.
Two years ago at this forum, we introduced what was at the time the world’s thinnest vapor chamber, in the form of a thermal ground plane (TGP) 0.25 mm thick. Now, we can introduce a new world’s-thinnest vapor chamber, in the form of an all-polymer TGP 0.175 mm thick. Owing to its polymer construction, the TGP is flexible as well. In this presentation, we will discuss the thermal performance in terms of effective thermal conductivity and maximum power. Both these parameters are highly dependent on operating conditions (e.g. size of heat source, size and temperature of the condensation surface), and we will cover models that can be used to predict TGP performance, in order for TGPs to be used in thermal systems.
Dr. Ryan Lewis, Director of R&D – Kelvin Thermal Technologies
4:20 pm -
Ceramic Microchannel Devices for Thermal Management
Microchannel components in thermal management applications offer high effectiveness with low pressure drop, as well as compact designs. Since heat transfer in microchannel devices typically is dominated by heat transfer to and from microchannel walls, the effectiveness and overall heat transfer properties are less sensitive to the thermal conductivity of the material that components are made from.
In some cases, a relatively low thermal conductivity material improves the performance of components by inhibiting heat transfer in undesired directions. In addition to benefits offered by moderate thermal conductivity, the benefits of corrosion resistance and lower cost relative to super-alloys will be discussed. Therefore, the use of ceramics in microchannel devices can be beneficial, assuming that they are cost effective.
The presentation will discuss applications, principles of design, and fabrication methods for making practical microchannel devices using ceramic components. Applications span many thermal regimes from cryogenic to ambient to high temperature. Specific examples of gas-to-gas recuperation for microturbines, or small heat engines, will be described. Examples of cooling microelectronic and concentrated photovoltaic will also be presented. The impact of microchannel design on effectiveness, pressure drop, and reliability will be discussed. Commercially viable methods of fabrication also will be described.
Dr. Charles A. Lewinsohn, Director of Research - Ceramatec, Inc.
5:00 pm – Cocktail Reception in the Expo Hall
August 4th, 2016 – Day Two Program
8:00 am -
Heat Pipes and Vapor Chambers – Practical Usage Guidelines
Two-phase devices are incredible heat conductors and significantly boost heat sink performance. We’ll present a practical guide to using both heat pipes and vapor chambers: similarities, differences, operating parameters, mounting options, common mistakes, best uses, examples, and performance modeling.
George Meyer, CEO/CTO – Celsia, Inc.
8:40 am -
CFD Simulation of Power Conversion: Sensitivity to Model Assumptions
When analyzing electronics components utilizing convective cooling with heat sinks, typical simplifying assumptions such as ignoring passive components and using a bulk orthotropic conductivity for the PCB result in models with reasonable run-times and acceptable correlation to test results. In the case of power conversion circuits, the increased copper loading of the PCB and use of components with very high junction to case resistances and low junction to board resistances challenges these assumptions; when the heat sink is removed, the model no longer correlates to testing. A power conversion card CFD model is described with and without a heat sink, both versions solved with varying simplifying assumptions: insulated PCB, orthotropic PCB, imported ECAD, with passive components, and without passive components. A summary of the effects of each simplifying assumption on component temperature, and suggestions of best practices for modeling the part with and without a heat sink are proposed.
Levi A. Campbell, PPP&C IDT Chair, Master Inventor – IBM Advanced Thermal Lab
9:20 am -
Thermocouple Design Issues: How IC Manufactures are Addressing These Challenges
Thermocouples are one of the most commonly used temperature sensing elements, offering a variety of benefits including ruggedness, wide temperature range, fast transient response and simple interface. However, thermocouples can be problematic for system designers, as their implementation requires a range of design expertise including analog, mixed-signal, thermal and even firmware development. In order to simply the design process and speed time to market, Microchip Technology has researched an integrated thermocouple to degree Celsius solution. Learn about their research in these areas.
Kevin Tretter, principal product engineer, Analog Power and Interfaces Division – Microchip Technology
10:00 am – Networking Break in Exhibit Hall
10:20 am -
Advanced Materials and Process for High Performance Flip Chip
Next generation high speed network/communication packages require a much larger die size/more ball counts flip chip packages to meet high speed and high I/O functionality and better thermal performance due to very high power in the die. Thermal management in the flip chip package is becoming a major challenge now days in the industry.
IC Customers continue to push assembly suppliers to reduce theta jc on high power flip chip packages. Designing the package with right bill of material (BOM) is a must for high power flip chip package. Larger die size flip chip packages experience greater issues in the assembly process due to higher thermomechanical mismatch, resulting in high warpage and thicker TIMs (thermal interface material) BLT (bond line thickness) and hence, degrades thermal performance significantly.
Thermal performance of a high power package can be improved by using advanced higher conductive TIMs, and improved assembly process which ultimately minimizes the thermal resistance between die to lid. The study will focus on some key factors for selecting advanced BOM, and process for high thermal flip chip package. Very comprehensive experiments and simulation are being carried out to achieve the objective of the work. A test vehicle is designed using a flip chip package with various TIM materials and processes. More advanced TIMs are being selected for further investigate the scope of the project and possible future implementation.
Nokibul Islam, Product and Technology Marketing - STATS ChipPAC Inc.
11:00 am -
Thermal Simulation of Microchannel Two-Phase Liquid Cooling Cold Plates for Servers and Power Electronics
The passage from single-phase cooling (air and liquid) to two-phase flow boiling in micro channels of cold plates requires not only numerous well validated heat transfer, pressure drop and critical heat flux models, but also a simulator tool that implements these AND can handle both heat spreading and flow spreading, the latter arising from non-uniform heat fluxes dissipated by electronics, so-called hot spots. Furthermore, the micro-two-phase cooling circuit is usually designed for parallel cooling of multiple CPU’s or IGBT’s, which requires the modelling of the flow splitting to each when operating at different heat dissipation rates. In addition, cold startup and transient operation must be able to be simulated.
This lecture will give an overview of our codes for multi-microchannel evaporator cold plates and 3D-IC cooling geometries for these types of simulations and provide numerous experimental validations of the codes. Then, several parametric studies will be presented together with outputs (including graphical) of local temperature profiles in all layers (evaporator base, TIM and silicon die), pressure drops and gradients (including those of the piping and headers), critical heat flux, etc. for a variety of coolants.
John Thome, Professor – Ecole Polytechnique Fédérale de Lausanne, Switzerland
11:40 am -
Power Dissipation Limits for High-Performance Tablet Cooling
Some of the highest performance tablets currently available are using Intel I7 processors that generate considerable heat when pushed to their computational limits. These tablets, with the addition of a detachable keyboard, can compete performance-wise with laptop and desktop computers.
This presentation will contain strategies for cooling high performance tablets that exceed the power that can be cooled by natural convection. The presentation will include thermal modeling using Computational Fluid Dynamics (CFD) to determine the limits of both passive and active cooling of these devices along with experimental test results and IR images. The limiting factor for determining the maximum power dissipation is the surface touch temperature of the tablet when running software that puts the unit under both high computational and graphical stress.
Guy Wagner, Director – Electronic Cooling Solutions
12:20 pm – Networking Lunch
1:20 pm -
Practical Approaches for Heat Flux Enhancement in Multiphase Cooling
Pumped liquid multiphase cooling (PLMC) provides a heat transfer topology with field-demonstrated heat fluxes far greater than single phase solutions and can be an excellent upgrade from water based cooling. The Durbin Group participated in the implementation of PLMC upgrade for a land based High Power RF system and demonstrated a 3.7 factor reduction in system weight and 3.3 factor reduction in system size due to the SWaP enhancement enabled by PLMC . When retrofitting existing systems for conversion to PLMC, the designer can be required to carry-over components from the water based system due to integration constraints. Often, the passages in water-based heat sinks do not provide an optimal thermal interface for the multiphase system. Research is being conducted to evaluate how surface treatments, internal foaming, and bonded folded-fin can provide designers with a method to improve the performance of their existing hardware using PLMC technology. These heat flux enhancements enable the process fluid to more efficiently extract heat from the thermal load.
John Durbin, President – Thermal Form & Function Group David Talaiver, Director of Emerging Technology – Thermal Form and Function Group
2:00 pm -
Ceramic Design and Material Opportunities for Demanding Power Density Applications
The ongoing increase of power density within several applications needs more and more new design opportunities to solve the challenging technology roadmaps. New demanding design for LASER diode cooler or Two-Phase cooling systems for mobile applications can be fulfilled with unique fine structuring methods down to 0.15 mm. This in combination with the high performance material Aluminum Nitride and a thermal conductivity of 200 W/mK lead to a very good combination of thermal performance and wear resistance.
Joerg Mueller, Product Manager - CeramTec GmbH
2:40 pm -
Low Viscosity Nanocomposite Materials for Electronic Encapsulation
Epoxy resins commonly used for encapsulation of electronic materials are filled with various inorganic materials to reduce coefficient of thermal expansion (CTE), improve thermal conductivity and various mechanical properties. The increase in viscosity that accompanies increase in filler loading limits the amount of filler that can be added to a liquid encapsulation system. As a result, highly filled systems are difficult to use in applications that require better flow characteristics, such as flip chip under-fills, transfer molding masses and encapsulates. Moreover, trends of miniaturization require encapsulation systems with much lower viscosity and more efficient thermal management.
This presentation introduces nanosilica-epoxy composite materials that consist of surface treated spherical nanoparticles that allow equivalent filler loading at much lower viscosity compared to conventional fillers. Such particles can even penetrate small gaps between smallest microchips and PCB, or finest filaments like in ignition coils or glass fiber bundles inside PCBs and provide improvements in CTE and mechanical properties. An overview of the aforementioned nanocomposite materials and their mechanical, thermal and electrical properties is presented.
Dr. Kunal Kumar, Senior Technical Service Representative, Evonik Corp.
3:20 – Conference Conclusion