Cooling Technologies Research Center  




CTRC Breakthroughs

Patents and Invention Disclosures


Visualization of phase change from a sintered wick (.wmv file)

(presented as Supplementary Information with “Characterization of Evaporation and Boiling from Sintered-Powder Wicks Fed by Capillary Action,” by Weibel, Garimella, and North, International Journal of Heat and Mass Transfer, Vol. 53, pp. 4204-4215; due to the big file size, may take some time to start playing)


Anti-Noise Synthetic Jet Enhanced Heat Sink    Back to Top

The use of synthetic jet seeks to enhance the performance of heat sinks by adding a method of providing cool air to the heated area produced by interacting jets which is more efficient because of less power required to operate them as oppose to a fan which is noisy. The synthetic jet too, has an issue of noise but optimum phase angle between two jets mitigate this effect. Continued research seeks to find this optimum working angle. This project for the first time simultaneously maximizes heat transfer performance and minimizes the acoustic noise emission of a synthetic jet cooler. Thermal characterization of synthetic jets has received some attention in the past, but not much is known about the characteristics of two adjacent synthetic jets, which this project will be using to achieve noise suppression. Apparently, the acoustic noise emission of such a configuration has never been researched. Based on previous experience by the PI, noise suppression and heat transfer enhancement can be achieved using a pair of adjacent synthetic jets. This project investigates this configuration and provides guidelines for optimal design, and enable these systems to be used in addition to (or in place of) axial fan heat sinks, in order to cool electronics more efficiently.

In 2011, comparable stagnation Nusselt number results to previous experiments done with synthetic jets were obtained for round orifices, thus validating our laboratory setup. It is keen to note that the orientation used in this setup was in the horizontal manner, as oppose to the vertical one used elsewhere. This validation provides for us a basis for further experimentation with jets. These jets can be used as an alternate means of (or supplement for) cooling as opposed to axial fans. If these jets are designed correctly, this could result in more efficient heat transfer performances with less noise if this noise issue is looked at critically. This gives a detail synopsis of the impact of this research for the scientific community and that of the electronics cooling field. The figure is a one-dimensional representation of the heat transfer coefficient distribution, obtained when a 5-mm diameter round orifice jet acts upon a heated foil. Reynolds number of 1500 and a stroke length of 7 were other critical parameters

Performance and Reliability Impact of Compliant Thermal Interface Materials    Back to Top

Compliant TIMs are an important material set in many applications including automotive electronics, consumer electronics and LED industries. They are used to transmit heat and conform to local deformation mostly as TIM-II materials. While Gap pads are expected to conform to interface better and as a result, exhibit better thermal performance compared to greases, they are stiffer and therefore may impose stress on the rest of the assembly. These materials are also expected to exhibit a “solid-like” viscoelastic behavior. Commercially available compliant TIMs exhibit significant time-dependent deformation under compression limiting their utility. While these materials are expected to enhance thermal transport by filling gaps, their (often viscoelastic) mechanical behavior is not well understood. Present work is aimed at the development of multiple cross-validated test procedures to identify time-dependent viscoelastic behavior of compliant TIMs and to correlate the mechanical response of the materials to experimentally measured effective thermal conductivity.

During use-conditions, the thermal gap filler pad materials are subjected to extreme environments such as high temperature, mechanical and thermal cycles, high assembly pressures that result in changes in original properties of the pads. The present work aims at providing a fundamental understanding of mechanical behavior of these viscoelastic materials.  This study developed a novel micro-scale characterization procedure for such materials. The eventual utility of the study is to understand the materials’ response to varying assembly pressures as well as thermal cycling conditions. Such a response is critical to understanding the extent to which heat sink assembly causes solder joint failures. Exhaustive viscoelastic characterization of the Gap pad (compliant TIM) materials, and relating mechanical cycling to measured thermal conductivity degradation of these materials have been accomplished. This research also estimated the stress transmitted to the solder joints through the gap pads as a result of applied pressures during heatsink assembly. The top figure is a cross-section of the chosen thermal interface material as viewed under environmental mode in a scanning electron microscope showing a fiberglass pad on right side of the image. The lower two figures are the chosen material also view.

Thermal Cycling Fatigue Effects on the Contact Resistance of Organic Interface Materials on Si    Back to Top
High-conductance thermal interface materials will be crucial to the planned implementation of high-density integrated circuits into the future. The development of an accurate constitutive model of the effect of thermal cycling to polymer interface layers will be necessary for the reliable integration of organic materials into high performance packages. Environmental conditions highly influence the response of the thermal probe used in testing the thermal interface materials.  Accurate measurements of heat flow are dependent upon knowing and controlling the ambient operating conditions. Mapping the development of interface defects is important for improving the materials which aid in the removal of heat from electronic packages. Finding and documenting the evolution of the defects is dependent upon a full understanding of the thermal probes response to changing environmental conditions. Changes in test operating conditions can dominate changes to local interface properties.

Determination of Liquid Film Thickness in Slug and Annular Flow for Heat Transfer Model    Back to Top

By knowing the geometries of physical structures formed when liquid is boiling in microchannel heatsinks, it is possible to improve heat transfer models capable of predicting performance of these heatsinks.  Improved models can lead to improved designs; these heatsinks are critical to the reliable operation of high-power electronics and military avionics systems, as well as in renewable energy applications such as wind turbines, where power electronics find widespread use.

In 2011, this project developed a new measurement technique applicable to boiling of coolant in microchannel heatsinks.  The technique is capable of obtaining three-dimensional measurements of transparent liquid geometries in two-phase mixtures. The novel, non-intrusive technique developed will allow for investigations that were hitherto not possible in terms of in-situ measurements in liquid-vapor phase-change applications.  Ongoing research conducted as a part of this project using this measurement technique will benefit industry by providing improved heat transfer models to predict heat transfer performance in two-phase microchannel heatsinks. The ultimate goal of the project is to measure and analyze the complex liquid-vapor structures formed in two-phase microchannel heat sinks in order to improve predictive heat transfer performance models.  These improved models will aid industry in the design and implementation of microchannel heatsinks primarily in the cooling of high-power electronics.  Additionally, the accomplishments to date in this project have improved microfluidic measurement capabilities and techniques that may be used in future research projects for a wide variety of applications. The figure shows a 400 µm diameter capillary tube and a fluorescent image of a planar cross section within the tube.

Graphene-Based Thermal Interface Materials    Back to Top

Efficient heat dissipation is critical in many applications, such as in the integrated electronic circuits. Thermal interface materials (TIMs) are necessary for heat dissipation because the gaps/voids between heat transfer surfaces (e.g., between CPU and heat sinks) are unavoidable. The heat transfer efficiency can be dramatically enhanced by filling these gaps/voids with appropriate TIMs. Graphene, a single atomic layer of graphite with honeycomb lattice structure, exhibits very high thermal conductivity (~3000-5000 W/m-K), which renders it an outstanding candidate for TIM applications. In this project, graphene flakes/composite based materials are proposed as TIMs. Graphene flakes can be either vertically grown on many substrates by plasma enhanced chemical vapor deposition (PECVD) or chemically reduced from graphite oxide. Mixture of commercial thermal paste or polymer with graphene flakes is another graphene-based TIM option.

Graphene composite shows significant increase of thermal conductivity when the volume percent of graphene is increased. Vertically grown graphene shows superior performance, with a thermal resistance in the order of 1 mm2 K/W, which is better than available commercial thermal interface materials. The latter is of potential importance for industry relating to heat dissipation, such as IC technologies in which heat generated is necessarily required to be removed. This figure is the synthesis of graphene composite, which includes the preparation of graphene solution, the drying of the powders, and the hot pressing to form nanocomposite. Shown in the right lower corner is the vertical graphene petals prepared using chemical vapor deposition (CVD).

Organic Rankine Cycle with Solution Circuit for Low Grade Heat Recovery    Back to Top

The Organic Rankine Cycle with Solution Circuit (ORCSC) is an attractive technology for harnessing low grade heat as an energy source for power generation. Low grade heat sources include alternative sources such as geothermal, solar and biomass, as well as waste heat otherwise rejected to the environment by various industrial processes. These sources provide energy at a far lower temperature than found in conventional fossil fuels, and require new technologies to make them efficient and cost effective to utilize. This project aims to be the first to demonstrate the use of ORCSC technology. Preliminary numerical modeling has shown that the ORCSC offers significant efficiency improvements over current technology for comparable systems.  It also allows the use of high pressure, environmentally neutral working fluids such as carbon dioxide at far lower pressures than found in other systems. This eliminates the need for capital-intensive pressure sealed units. Furthermore, the model has also proved that the system capacity can be easily adjusted by simply changing the intrinsic properties of the working fluid used in the system. This offers a simple, cost-effective solution for adapting the system for peak and non-peak loads that the system is likely to encounter in practice.

The fundamental objective for this system is to utilize low grade heat in a power generating application. Existing power generating technologies rely primarily on steam based cycles, which are not feasible for low-grade heat sources. Utilizing low grade heat is important because several “green energy” sources including geothermal, solar, biomass etc. fall under the gambit of low-grade heat sources. To the best of the PI’s knowledge, no experimental investigations have been conducted with source temperatures as low as 60°C – 120°C for large scale power generating technologies. As such, integrating two different bodies of knowledge, namely mechanical fluid flow machines (expansion turbines) and heat and mass transfer (absorption and desorption) in the heat exchangers, into one technology is the broad technical challenge for this system. Simulation and preliminary experimental results have shown that this technology is particularly effective for low source temperatures where achieving cycle efficiencies as close to the theoretical (Carnot) limit as possible is important for ensuring the economic feasibility of the technology. The figure shows the experimental test facility for Organic Rankine Cycle with Solution Circuit showing all major system components

Optimal Porous Microstructures for Enhanced Thermal Management    Back to Top

The goal of this project is to quantify the uncertainties in engineering designs, especially in the field of electronics cooling, and thereby optimize them in a robust fashion. This optimization methodology, widely employed in other disciplines such as the design of MEMS equipment, will aid the industry in identifying critical input parameters to which the outputs are most sensitive. By considering uncertainties in the input parameters, the dependence of the output uncertainties on the input uncertainties may be identified, along with the expected range of the output quantities of interest. This dependence may then be used for finding the allowable uncertainty (range) in the input parameters given an acceptable uncertainty (range) in the output quantities of interest. In summary, the developed methodology aids the industry in arriving at failure-safe designs that are insensitive to inherent variability in input quantities. The size and cost of the equipment is also significantly reduced by this means, satisfying an important challenge in the electronics cooling industry.

A computationally inexpensive methodology for robust optimization is developed, with specific application to electronics cooling device designs. The methodology, capable of producing robust and reliable optimal designs, is demonstrated by optimizing conventional air-cooled pin-fin heat sinks. The methodology developed in this work has many benefits and may be extended to any optimization problem. Using this approach, the uncertainty or variations in the outputs, corresponding to the variable set of inputs, can be accurately quantified.  The input parameters to which the outputs are most sensitive can also be identified. The methodology produces truly optimal solutions that minimize product weight and size. This figure is a 3-dimensional view of the pin-fin heat sink problem solved using the developed methodology, shown along with flow pathlines and temperature contours. The arrows indicating the flow configuration are colored by temperature.

Surface Treatment for Boiling Heat Transfer Enhancement    Back to Top

Contradictory results have been offered in the literature on effect of surface treatment techniques on boiling heat transfer.  Most previous studies dealt with fabrication methods for fixed porous structures for boiling enhancement. The concept of a free-particle technique in which metal particles are placed on a heated surface for improvement in boiling heat transfer does not appear to have been previously proposed and systematically studied for practical applications in the heat transfer industry.  The experimental results from this project, obtained in two different working fluids, DI water and FC-72, with very different wetting characteristics, offer design guidelines for treating surfaces for boiling heat transfer enhancement.

The concept of a free-particle technique is demonstrated to improve pool boiling heat transfer via an experimental study.  In this technique, properly sized metal particles are placed on a heated surface and facilitate liquid-to-vapor phase change heat transfer by providing bubble nucleation sites on the surface. Due to the simple heat transfer enhancement mechanism, the free-particle technique can be applied to phase change heat transfer systems and enhance thermal performance with no added fabrication costs.  The results of the free-particle experiment also offer guidelines to fabricate fixed porous structures for boiling heat transfer optimization. Shows is a series of high speed camera images, showing vapor bubble nucleation and detachment from a narrow corner cavity between a 13 mm copper ball and a copper heated surface.

Optimization of Manifold Microchannel Heatsinks    Back to Top

In engineering problems that involve design uncertainties due to fabrication or other input uncertainties, a deterministic optimization approach may not provide a truly optimized design.  Optimization under uncertainty is a powerful method that ensures safe design of systems, taking the input uncertainties into account.  This tool is very flexible and can be used for design of any system that involves uncertainty in inputs.  The current project shows its application to manifold microchannel heat sinks, which enhance the heat transfer performance of conventional microchannels and reduce the pumping power required for operation.

The computational model has been developed and the effect of input uncertainties on the output parameters has been demonstrated for this application.  Deterministic optimization of design parameters has been performed. The developed methodology aids the industry in arriving at failure safe designs, which are fairly insensitive to uncertainty in design parameters.  Such robust design ensures improved size and cost efficiency which is an important factor in electronics cooling industry. The current study investigates the application of this method to the design of manifold microchannels to distribute the flow through multiple alternating inlet and exit pairs, thereby reducing the pressure drop while improving heat transfer performance.   The figure shows  (a) Unit cell computational domain  and (b) Probabilistic distribution of outputs with uncertainties in inputs.

Characterization of Composite Heat Spreaders    Back to Top

The characterization of composite heat spreaders is being explored by Drs. Timothy Fisher and Suresh Garimella. The 3-omega method has been used widely to provide accurate thermal conductivity measurement of materials in the past several decades. In this work, effort is made to establish and benchmark the method as a resource to the thermal science community at CTRC and Purdue.  The total thermal resistance of a cooling package can be reduced by replacing traditional Al/Cu heat spreaders with suitable low cost new materials. Furthermore, carbon nanotubes (CNTs) are synthesized on heat spreaders to improve heat transfer at interface. The results indicate that, using an Al/Ti/Fe tri-layer catalyst, a high-density CNT array can be directly and firmly bonded to heat spreader surface. The figure shows CNT array synthesis on silicon substrate.

Thermal and Electrical Characterization of CNT Vias    Back to Top

Drs. Timothy Fisher and Timothy Sands are conducting the project of thermal and electrical characterization of CNT vias. Today’s conventional copper interconnects face intrinsic problems such as electromigration at high current density (J>106 A/cm2) and resistivity changes due to grain and surface scattering effects. CNTs with their unique combination of high current carrying capacity, high thermal conductivity, mechanical integrity and electron mean free paths of several microns may be ideal candidates for next-generation interconnects. In this project, stable and dense CNTs were grown from electrodeposited Pd in the bottom of porous anodic alumina (PAA) pores, using buckypaper as a plasma screen. CNT density was calculated to be up to 1.0×1010 cm-2, and the minimum current density transported in CNT vias was calculated to be 106 A/cm2, which is comparable to or higher than Cu interconnects. Modifications to CNTs by electron donating and electron withdrawing chemicals (e.g., TTF and TCNE) have been made to improve the electrical properties of CNT vias. The first figure shows a side view of as-grown CNT’s using Pd as catalysts. The second figure has CNT via patterns where the patterns were fabricated with shadow masks on a PAA template. Ti&Au were evaporated on the top of CNT vias as contact metals for electrical measurements.

Miniature Scale Linear Compressor for Electronics Cooling    Back to Top

Drs. Eckhard Groll and Suresh Garimella are researching miniature scale linear compressors. The need for better active cooling solutions in electronics has been challenging due to the lack of components commercially available. A linear compressor is considered a promising new development in component technology due to the potential for high performance and high scalability. This work has developed a comprehensive model of a linear compressor based on first principals. This model has shown good agreement when compared with experimental data and can be used for future development of linear compressor technology. This addition to component technology will further the overall development of vapor compression refrigeration for electronics cooling. The figure is a schematic of a linear compressor.

Acoustically Enhanced Heat Transfer    Back to Top

Drs. Eckhard Groll and Stuart Bolton are exploring acoustically enhanced heat transfer. The aim of the project is the enhancement of heat transfer with an acoustic sound field with, ideally, a smaller pressure drop in comparison to other surface enhancements. This project involved experimental investigation of acoustically enhancing the heat transfer from a fluid flow to the solid wall of a plain tube and determining the influence of an acoustic field on the rate of heat transfer. A prototype test stand had been constructed in which air was the working fluid. The experiments have shown that the heat transfer can be enhanced significantly with an acoustic field. The effect of the sound field depends on frequency and sound level, whereas further parameters cannot be excluded. The figure plots dimensionless heat transfer enhancement (Q/Q0) against the frequency at various sound levels (top) showing that heat transfer can be enhanced with an acoustic field by up to 30%. Q/Q0 is defined as the ratio of the heat transfer with acoustic enhancement to the base-case heat transfer Q0. Q/Q0 increases as the sound level increases. Differential pressure inside of the test section plotted against the frequency (below) shows that with a particular frequency, the differential pressure is negative. It can be considered as a pumping effect.

Electromechanical Transport of Nanofluids: Microfluidic Pumping and Heat Transfer Enhancement    Back to Top

The electromechanical transport of nanofluids is being investigated by Drs. Dong Liu and Suresh Garimella. Nanofluids are considered a promising candidate for advanced heat transfer fluids that can be used in electronics cooling. It was expected that the dispersed nanoparticles lead to significant increase in thermal conductivity and convective heat transfer. However, the reported results are wide spread and often in discrepancy. In addition, the increased viscosity poses a great challenge for circulating nanofluids with traditional pumping methods. In this work, a thorough investigation was conducted to study the fundamental mechanisms of single-phase forced convection of Al2O3-water nanofluids in a minichannel. Friction factor and convection heat transfer coefficient were measured for nanofluids of various volume fractions. The Reynolds number covers a wide range of flow conditions to provide a complete description of flow physics from the laminar to fully developed turbulent flow. Several salient features were found, which had not been reported in the literature before. Nanofluids exhibit a pronounced entrance region behavior in laminar flows. The onset of transition to turbulence was delayed. Convective heat transfer of nanofluids is mildly enhanced in the laminar region, however, deteriorates in the transition and turbulent regions. It suggests that nanofluids should be used in laminar region to yield heat transfer enhancement. A novel pumping method was also developed in this work which exploits dielectrophoresis-induced particle-fluid interaction to actuate nanofluids without needing external pumps. With this approach, the flow is generated by inducing strong electromechanical effects in the fluid using integrated microelectrodes. The fluidic driving mechanisms due to the particle-fluid and particle-particle interactions under twDEP were analyzed and the induced flow field was obtained from numerical simulations. Experimental measurements of the flow velocity in a prototype DEP micropumping device show satisfactory agreement with the numerical predications. Results from this work indicate that the DEP-induced micropumping scheme holds great promise for devising versatile, self-contained microscale fluidic delivery systems. The figure shows a comparison of the average Nusselt number measured for nanofluids and the base fluid over the entire range of Reynolds number studied. While nanofluids provide modest heat transfer enhancement in laminar flow, they cause significant heat transfer degradation in the transition region when compared to the heat transfer data of water, and the degradation gets worse as the nanoparticle concentration increases. The delayed transition from laminar to turbulent heat transfer in nanofluids can be clearly identified. Once the flow becomes fully turbulent, convective heat transfer of nanofluids becomes less dependent on the nanoparticle concentration and gradually converges to that of water.

Electrical Control of Liquid Droplet Motion and Morphology on Smooth and Microstructured Superhydrophobic Surfaces    Back to Top

Dr. Suresh Garimella is leading the electrowetting project. A fundamental study of electrical and gravitational actuation of liquid droplets on smooth and superhydrophobic surfaces was undertaken.A generic modeling framework was developed to predict the electrical actuation of low electrical conductivity and insulating liquid droplets using AC and DC voltages.This modeling framework can be used to study electrical actuation of liquid droplets across the entire spectrum of electromechanical actuation. Detailed and careful measurements of droplet velocities (of different types of liquids) under electrical actuation were carried out in a specially designed setup.The work led to the first reported velocity measurements of insulating liquid droplets under electrical actuation. Electrowetting-induced Cassie (nonwetting) to Wenzel (wetting) transition on superhydrophobic surfaces was studied analytically and experimentally. The microscale surface design enabled the first reported direct visual confirmation of the electrowetting-induced Cassie-Wenzel transition. Conceptualization, development and characterization of superhydrophobic surfaces with a robust Cassie state (using non-communicating roughness elements) to prevent the Cassie-Wenzel transition was reported. A careful experimental study of heat transfer due to droplet motion on a hot inclined substrate was carried out as part of the FRS project. The temperature drop of the substrate due to the droplet motion was measured using infrared thermography. The effect of droplet volume and droplet velocity on the temperature drop was identified. A numerical model was developed to calculate heat transfer coefficient due to droplet motion on a single plate and droplet motion sandwiched between two plates. The model was validated with the experimental results. The model is used to predict heat dissipation rates using multiple electrically actuated droplets to cool electronic devices. A detailed experimental study of the transient dynamics of droplet oscillation using electrical actuation was carried out. The droplet spreading under a wide range of DC voltage was studied. Dynamic contact angle and contact radius were measured experimentally for a wide range of frequencies at different AC voltages. A numerical modeling technique was developed to predict droplet retention under gravitational forces. The model is verified against in-house experimental measurements. Advancing and receding contact angle measurements were made for different droplet volumes at different angle of inclination of the surface. Correlations are proposed for the measured contact angles as a function of Bond number. The droplet shapes show dynamic contact angle on inclined surfaces for different volume (β is angle of inclination).

Thermoelectric Power Generation from Waste Heat in Electronic Systems    Back to Top

Drs. Timothy Sands and Timothy Fisher are conducting a project about thermoelectric power generation from waste heat in electronic systems. A large fraction of the electrical energy supplied to operate a data center or to power a laptop is dissipated as heat at the chip level. If a small fraction of this waste heat flux were converted to electrical power, it is conceivable that the overall system performance would be improved and the cost of running the data center or portable device would be reduced accordingly. One promising technology for converting waste heat to electric power in internal combustion vehicles and industrial processes is the thermoelectric generator (TEG). Recent improvements in the thermoelectric figure-of-merit, ZT, bode well for the implementation of these solid-state waste heat recovery systems. For example, the laptop application would utilize thin profile generators with relatively high heat transfer coefficients and the possibility of dynamic switching between Peltier and Seebeck modes. The data center application would utilize waste heat from water-cooled chips, boards or racks. This project not only involves designing the fabrication process of a low grade waste heat generator but also developing a modeling tool to optimize the element configuration. The model completed recently allows assessment of the effect of element design, contact resistance, leg length, and thermoelectric properties on power density and efficiency. The figure shows 350 micron thick self-supporting branched porous anodic alumina template for microcoolers and waste heat energy converters.

Alternative Waste Heat Recovery Opportunities    Back to Top

Dr. Srinivas Garimella at Georgia Tech is studying an assessment of alternative waste heat recovery opportunities. Thermally activated systems based on sorption cycles, as well as mechanical systems based on vapor compression/expansion are assessed in this study for waste heat recovery applications. Two cycles are considered to provide refrigeration or air-conditioning, two cycles are considered for power generation, and one is considered to upgrade low-grade heat to higher temperatures. The size and performance of each system is estimated for two general cases. One for smaller-scale and low-temperature waste heat availabilities (waste heat at 60°C,) and the other is for larger-scale and higher-temperature applications with waste heat at 120°C. Comparative assessments of these cycles on the basis of efficiencies or COP, as appropriate, and system size, guide the selection of waste heat recovery and upgrade systems for different applications and waste heat availabilities. It is found that even with such low-grade waste heat sources, work produced is approximately 10% of the available waste. Cooling output can be as high as 75% of the available waste heat, and temperature boosting of the waste heat can be achieved with efficiencies as high as 47%. Specific illustrative case studies have been conducted on waste heat recovery for electronics devices, data centers, vehicles, and process plants, yielding performance as stated above within small system footprints. Passive means of implementing such cycles with few moving parts are also being explored.

Flow Regime-Based Predictive Model for Flow Boiling in Microchannels    Back to Top

Dr. Suresh Garimella continues to study flow properties in microchannels with results in flow regime-basepredictive model for flow boiling. A comprehensive understanding of flow boiling heat transfer and pressure drop in microchannels for various liquid/surface combinations, across a wide range of operating conditions, has been developed. A large database of boiling flow pattern visualizations is obtained for a wide range of channel dimensions and flow parameters, leading to a good understanding of microscale flow regimes. This gallery of visualizations has been widely disseminated through the Center website for researchers and practitioners worldwide to use. The effects of microchannel geometry on flow boiling have been elucidated; the cross-sectional area of the microchannels is found to play a determining role in boiling mechanisms and heat transfer. The effect of surface roughness on boiling heat transfer enhancement has also been established. Vapor confinement and emergence of microscale effects are shown to depend not only on channel size and fluid properties, but also on the flow rate. Based on the experimental results, a new transition criterion is developed which predicts the conditions under which microscale confinement effects are exhibited in flow boiling. A comprehensive flow regime map for flow boiling of FC-77 is developed, along with quantitative regime-transition criteria, based on approximately 390 data points encompassing a wide range of microchannel dimensions, mass fluxes, and heat fluxes. The figure shows a high-speed flow visualization of bubbly flow and slug flow in 400 µm-wide microchannels

This project involved developing an electrical impedance-based void fraction meter to directly measure the void fraction and identify flow regimes in a two-phase microchannel. This project has successfully developed a calibration curve relating electrical impedance to void fraction in two-phase flow in a microchannel and developed a procedure for determining flow regimes quantitatively using the statistical characteristics of the void fraction signal. The capability to measure void fraction and flow regimes will be used to enhance physics based models for heat transfer in two-phase microchannels. This is applicable in electronics cooling as well as microscale heat exchangers in fuel cells and companies have shown interest in incorporating his type of technology in their products.

Investigation of Nucleate Boiling Mechanisms in Forced Convection Using Laser-Induced Flourescence Thermography    Back to Top

Dr. Suresh Garimella is investigating nucleate boiling mechanisms in forced convection using laser-induced fluorescence thermography (LIFT). LIFT is used to obtain the fluid temperature field around growing and collapsing vapor bubbles generated during subcooled nucleate boiling. Limited data on the liquid temperature field during nucleate boiling currently exists in the literature and, as far as the investigators are aware, this study represents the first time LIFT has been employed to study such a phenomenon. LIFT relies on the temperature-dependent fluorescence intensity exhibited by certain dye molecules. Preliminary LIFT measurements during flow boiling indicate noticeable changes in fluorescence intensity, which is believed to be the result of a spatially and temporally-varying temperature field. The figures show the fluorescence intensity field following a vapor bubble collapse during subcooled nucleate boiling: a temperature-sensitive fluorescent dye (upper image) and a temperature-insensitive fluorescent dye (lower image). The circled region in both images indicates the region where the bubble collapsed. Changes in the fluorescence intensity field, which are apparent in the upper image but not the lower image, are believed to indicate changes in the liquid temperature field.

Enhanced Electrohydrodynamic Micropump    Back to Top

The enhanced electrohydrodynamic micropump project is led by Dr. Suresh Garimella. The design consists of a traveling-wave electrode array in contact with a fluid that moves ions induced in the fluid domain with the traveling electric field. A numerical model has been completed to study optimized designs for improvements in fluid motion. Higher phase representations of the traveling-wave (for a given wavelength), in addition to an increase in electric field strength by electrode design modifications have been shown to increase flow rate by approximately 75%. A microfabricated pump has also been completed to benchmark modeling results. A peak in the velocity has been observed at a frequency that corresponds to the charge relaxation time of an ion in the fluid. Fluid motion increases linearly with the square of the applied peak voltage to the traveling-wave electrodes. Additionally, the effect of externally applied heat to increase the temperature gradient (necessary for ion induction) has been demonstrated experimentally. A velocity increase of approximately 50% was observed with an additional 0.3 W of heat flux applied on one boundary. Using the validated numerical model, the capability of removing heat with electrohydrodynamic micropumping has been evaluated. The figures show device layout and instrumentation of an electrohydrodynamic micropump.

Control of Cooling Fan Noise by Radiation Efficiency Control    Back to Top

Dr. Stuart Bolton is studying the control of cooling fan noise by radiation efficiency control. The noise level caused by axial cooling fans can be controlled by careful consideration of installation conditions as well as improving the aerodynamic design of fans. Such an installation effect is due to the dipole-like acoustic features of axial fans and this was experimentally proven by confirming the out-of-phase sound wave propagation between the front and back fields of an axial fan. Utilizing this result, an axial fan was able to be numerically modeled as a simple dipole source so that the installation effect of an axial fan could be simulated with very low computing cost. The red points in the figure indicate measurement points for directivity pattern of an axial fan mounted on ISO 10302 plenum.

Boiling Heat Transfer Enhancement in Microchannels Using Surface Treatment    Back to Top

Dr. Suresh Garimella is continuing a project focusing on heat transfer in microchannels. Power electronics and high power density integrated circuits demand novel cooling strategies which are capable of removing large amounts of heat and maintaining device temperatures in the desired operational range. One such strategy is the utilization of the highly effective heat transfer mechanisms associated with the liquid-vapor phase change process. The present work seeks to characterize and predict the enhancement to boiling heat transfer provided by arbitrarily rough surfaces in microchannel heat sinks. Results indicate that increasing the surface roughness by a factor of 3 yields a 30% enhancement in the amount of heat that can be removed while keeping the heat sink temperature constant. Further increases in surface roughness appear to be of little additional benefit. Boiling from a flat, polished surface in a pool of water is shown in the figure.

Enhanced Thermal Contact Conductance Using Carbon Nanotube Interfaces    Back to Top

A project led by Dr. Timothy Fisher of Purdue University investigates the thermal contact conductance enhancement with directly synthesized carbon nanotube arrays. A tri-layer catalyst configuration has been developed for well anchored and vertically oriented CNT arrays direct synthesized on various substrates (e.g., bare silicon, metals, dielectric materials) with plasma-enhanced chemical vapor deposition and different catalyst metals. A test rig for thermal contact resistance measurement (adapted from ASTM D 5470) has been established in a high-vacuum environment, and temperature measurements are accomplished with an infrared imaging system. The figure above summarizes thermal interfacial resistance results for a bare copper-silicon interface, a single-sided CNT array, and a double-sided array. The results demonstrate that the CNT arrays can exhibit resistances below 10 mm2K/W that are comparable to the resistance of an ultra-thin soldered joint. Important and relevant capabilities that have been developed include: (a) anchoring of CNT arrays to substrates, (b) control of CNT wall type, diameter, density, length, and alignment, (c) synthesis at moderate and low temperatures using catalyst-containing dendrimers, and (d) conductances of dry interfaces that are comparable to soldered interfaces and that exhibit good mechanical robustness. Six technical papers (in journals and conferences) have been published or are in press, and a patent application has been submitted. In addition, parameter variation studies on dry CNT arrays, CNT-wax composites, and diamond-CNT composites are ongoing for possible further conductance enhancement.

Coordinated Miniature Piezofan Arrays    Back to Top

Fluidic coupling between neighboring piezofans (see figure) and its impact on the heat transfer performance is being explored in this project conducted by Drs. Suresh Garimella and Arvind Raman. Experimental results suggest that coupling is almost entirely due to fluid-structure interaction, and is seen as a decrease in viscous drag on each piezoelectric fan. This coupling phenomenon causes an increase in vibration amplitude of as large as 40 percent compared to an isolated single fan, which causes a further increase in heat transfer performance. There appears to be an optimum condition for separation distance between these two fans (pitch) where the performance is greatest (see Kimber, Garimella, and Raman, Procs. ITHERM 2006). The smallest fan pitch does not necessarily yield the greatest heat transfer. More detailed heat transfer experiments are underway to better quantify the local and area-averaged convection coefficients on a larger heated surface subject to a controlled heat flux. Additional experiments are also being conducted for piezoelectric fans of different geometries and materials. However, the work completed in the last year shows that fan coupling can indeed be exploited to improve the thermal performance obtained when multiple fans operate in close proximity.

Thin Film Evaporation    Back to Top

An evaporating meniscus is at the core of most two-phase heat transfer devices; a comprehensive understanding of the underlying physical mechanisms in thin film evaporation is critical for miniaturization of heat pipes, cold-plates, and other thermal solutions. Evaporation taking placeat the very thin solid-liquid-vapor junction is claimed to be the dominant mode of heat transfer in a liquid-vapor phase change process. This project conducted by Drs. Suresh V Garimella and Jayathi Y Murthy of Purdue University is aimed at experimental investigation of the fluidic and thermal transport in an evaporating meniscus, and the development of an experimentally validated theoretical model. Experiments are characterizing the effect of thermocapillary convection and thin-film evaporation on the heat transfer. Micro-PIV experiments are being performed to study the flow pattern near the evaporating meniscus. The figure below shows the measured flow field induced by thermocapillary convection due to the differential evaporation observed in a horizontal diametrical section of a volatile liquid in a 400μm ID capillary. The differential evaporation sets up a temperature gradient between the corners and the center of meniscus which generates the thermocapillary convection seen in the form of two counter-rotating vortices. This strong convection is expected to increase the total heat transfer by providing better mixing of the fluid. A comprehensive model of transport from an evaporating meniscus is currently being generated.

Transport in Porous Structures and Metal Foams    Back to Top

A novel computational methodology for direct numerical simulation of open-cell foams and heat pipe wicking structures has been developed in a project conducted by Drs. Jayathi Murthy and Suresh Garimella of Purdue University . This comprehensive model predicts the thermal and flow characteristics of foams and particle beds, which are in excellent agreement with published measurements. The results show the capability of the developed methodology and presents opportunities to explore other problems such as thermal dispersion, particulate fouling and foam structure optimization.


Two-Phase Transport in Microchannels    Back to Top

A project led by Dr. Suresh Garimella of Purdue University is exploring boiling and two-phase flow in microchannels. Transport through microchannels that range in width from 100 to 400 micrometers in copper and silicon substrates is experimentally characterized. Experiments include high-speed flow visualization, and local and global temperature and pressure measurements. A predictive model is also being formulated to aid in the design and optimization of microchannel heat sinks.


Validated Models for Particulate Thermal Interface Materials    Back to Top

Thermal Interface Materials (TIMs) continue to be a bottleneck for developing the next generation of microprocessors with smaller chip sizes and increased power. Development of better TIMs is imperative to ensure efficient heat removal from microelectronic systems, which in turn improves the system reliability and performance. This is the goal of researchers Dr. Ganesh Subbarayan-Shastri and Dr. Thomas Siegmund at Purdue University.  Accurate modeling of thermal interface materials requires either complex 3-D computational simulations or improved analytical models. Analytical models that calculate the effective conductivity of particle-laden matrices range from the classical Maxwell’s model (1891) to its more recent size-dependent generalization by Nan et al. (1997). However, most existing models do not consider particle-particle interactions, and therefore fail when volume loading exceeds 30%.


Images: 1.) 3-D Random Microstructure. 2.) Temperature field at y=0.5 3.) Heat Flux field at y=0.5.

Numerical modeling of realistic 3D microstructures (at high filler volume loadings) considering inter-particle interactions was performed using full-field meshless simulations and random particle network simulations. The developed models are validated with experiments on representative systems. The models can be efficiently used to accurately predict the effect of varying: i) the filler particle conductivity, ii) the base polymer matrix conductivity, and iii) size-distribution and arrangement of the filler particles on the composite thermal conductivity of TIMs. These models are expected to provide critical help in the design of high performance TIMs.


Experimental Investigation of the Transport Properties of Wicks    Back to Top

Heat pipes are commonly used in electronics cooling applications due to their ability to move high amounts of heat over reasonable distances with only a small drop in temperature. Porous wick structures imbedded inside heat pipes provide the interfacial tension necessary to drive the working fluid. For most electronics cooling applications, a balance between capillary pressure head and the corresponding permeability determines whether a heat pipe will meet the temperature rise and heat flux requirements in a given application. An experimental investigation of wick operating limits is the focus this project undertaken by Dr. Suresh Garimella at Purdue University. Discrepancies in operating limits reported in the literature due to inadequate test procedures are being resolved. Improved measurements and observation of transport processes are also aiding in the design of miniaturized heat pipes. Two dedicated heat pipe wick-testing facilities have been established. One measures mass transport in a variety of flat wick structures, including sintered copper and aluminum grooved wicks. The second facility is an instrumented thermosyphon test bed to measure conductivity values for wick structures under varying degrees of wick saturation.


Miniature-Scale Refrigeration System (MSRS) for Electronics Cooling    Back to Top

Drs. Eckhard A. Groll and Suresh V. Garimella are investigating the performance and feasibility of meso- and micro-scale refrigeration systems for electronics cooling. A comprehensive theoretical analysis of miniature-scale refrigeration systems (MSRS) is being conducted, in conjunction with design and fabrication of innovative refrigeration-based solutions to cool microchips. The simulation model is based on thermodynamics, heat transfer and fluid mechanics concepts applied to miniature scale components of a micro vapor compression system, including the compressor, mini-channel condenser, expansion device, and integrated cold plate evaporator and heat spreader.


The system model shown in the figure simulates a refrigeration system with an active expansion device. Hence, only two parameters, suction and discharge pressures, need to be estimated to completely define the operating conditions and performance of the vapor compression system. The simulation model can be used as an engineering design tool to evaluate the feasibility of MSRS to cool high heat dissipation microprocessors. The model predicts all state points of the micro-refrigeration system and the power consumption and the efficiency of the cooling system.


Miniature-Scale Diaphragm Compressor for Electronics Cooling    Back to Top

A novel analytical model of a Miniature Scale Diaphragm Compressor (MSDC) for electronics cooling has been developed through a CTRC project conducted by Drs Eckhard Groll and Suresh Garimella. This comprehensive physics-based model predicts the performance of the MSDC and can be used to design a miniature compressor. Results from the model indicate that an MSDC has excellent thermodynamic performance and good potential for miniaturization and integration with micro-refrigeration systems to cool high-heat-dissipation microprocessors.

The diaphragm compressor consists mainly of two contoured conductive planes that serve as electrodes. These are separated by dielectric insulation layers and a gas/refrigerant gap. If a voltage differential is applied between the electrodes, the electrostatic force deforms the diaphragm and pulls it towards the electrode on the chamber wall. The particular contour of the compression chamber promotes a progressive and continuous zipping action of the diaphragm until the membrane mates with the entire chamber wall. At the end of the compression strokes, the gas has almost zero dead space and the flexible diaphragm provides perfect rectification. Thus, the pressure of the refrigerant inside the chamber rises. Suction and discharge flapper mini-valves control the refrigerant flow in and out of the compressor chamber. Target parameters for the diaphragm compressor include a heat removal of 200 W, pressure head of 750 kPa, pressure ratio of 2, and flow rate of 3000 ml/min, accomplished with a diaphragm compressor of 80 mm in diameter and 20 mm in height.

Direct Cooling by Electron Field Emmission    Back to Top

Emission of electrons from a cathode has long been known to produce a heating or cooling effect at the emission site depending on the conditions of the emission, such as the emitter work function and temperature and the applied electric field. Many refrigeration devices have been proposed based on this phenomenon; however, a successful device has not been demonstrated due to several challenges, one of which is that an emission gap of a few nanometers is needed for efficient cooling. The present concept provides a method by which an emission gap distance in the nanometer range can be maintained over areas as large as several square millimeters using a carbon nanotube array emitter and a liquid metal anode.

Liquid indium and other liquid metals wet carbon nanotubes very poorly. Because of this characteristic, a narrow gap may be maintained between a densely packed array of carbon nanotubes and a liquid metal by placing thin dielectric spacers at strategic locations between the carbon nanotube-liquid metal interface (see the accompanying figure). The dielectric spacers “float” on the liquid metal surface due to surface tension and establish an emission gap between the carbon nanotubes and the liquid metal. One advantage of this approach is that the emission gap distance is not affected by thermal expansion, and another is that the gap distance may be adjusted by varying the thickness of the dielectric spacers.


  ►The NSF SEE website has highlighted three breakthroughs from CTRC research: Detecting Fatigue in Thermal Materials, Helping Electronics Beat the Heat and Graphene Flakes Take the Heat.

Cooling Technologies Research Center

CTRC Home | Purdue University | ME Home | Facebook | Linkedin
Electronics Cooling Lab | Solidification Heat Transfer Lab | Renewable and Sustainable Energy Technology and Policy

Copyright © 2013, Purdue University, all rights reserved.
An equal access/equal opportunity university.
If you have trouble accessing this page because of a disability, please contact the consortium administrator at