Passive, gravity-driven thermosyphons represent a step-change in technology towards the goal of greatly reducing PUE (Power Usage Effectiveness) of datacenters by replacing energy hungry fans of air-cooling approach with a highly-reliable solution able to dissipate the rising heat loads demanded in a cost-effective manner. The European Union has launched a zero carbon-footprint target for datacenters by the timeline of 2030, which would include new standards for implementing green solutions. In the present study, a newly updated version of the general thermosyphon simulation code previously presented at InterPACK 2019 and InterPACK 2020 is considered. To facilitate the industrial transition to thermosyphon cooling technology, with its intrinsic complex flow phenomena, the availability of a general-use, widely validated design tool that handles both air-cooled and liquid-cooled types of thermosyphons is of paramount importance. The solver must be able to analyze and design thermosyphon-based cooling systems with high accuracy and handle the numerous geometric singularities in the working fluid’s flow path, besides that of the secondary coolant. Therefore, a new extensive validation of the thermosyphon simulation solver is performed and presented here versus experimental data gathered for a compact liquid-cooled thermosyphon design, which is being considered for the cooling of high-performance servers. The new experimental database has been gathered to be able to characterize the effect of filling ratio, heat load, secondary coolant temperature and mass flow rate on the cooling performance, using R1234ze(E) as a low GWP (Global Warming Potential) working fluid. This compact design has experimentally demonstrated high performance, maintaining the pseudo chip’s temperature lower than 45°C for evaporator footprint heat fluxes up to 18W/cm 2 . The comparison shows that the solver is able to accurately predict thermosyphon thermal-hydraulic performance, and based on this prediction, characterize the internal flow rate generated by the thermosyphon, which is key to correctly estimate the maximum heat removal capability.

The trade-off between efficient cooling and low power consumption is a goal that has always been very desirable in <b>electronics</b> <b>cooling</b>, especially nowadays that power densities of processing units are increasing. Conventional cooling solutions do not have the necessary cooling capacities for these power densities or require significant power consumption. In this study, a novel air-cooled thermosyphon cooling system for desktop computers is presented and experimentally tested. The thermosyphon comprises a vertical micro-channel cold plate as the evaporator and a horizontal air-cooled multiport coil as the condenser. The thermosyphon has a total height of 12 cm and operates with a fan speed of 1700 RPM. The working fluid selected for the thermosyphon loop is R1234ze(E), chosen for its advantageous thermophysical properties and nearly zero-GWP (Global Warming Potential). The test results presented in this paper aim to analyze thermosyphon’s thermal and hydraulic performance by studying the trends of thermal resistance and mass flow rate as a function of different operating conditions. The maximum heat rejection under safe conditions is 250 W, corresponding to a heat flux of about 18 W&#x2F;cm 2 .

Two-phase cooling systems based on the thermosyphon operating principle exhibit excellent heat transfer performance, reliability, and flexibility, therefore can be applied to overcome thermal challenges in a wide range of <b>electronic</b> <b>cooling</b> applications and deployment scenarios. However, extremely complex nature of two-phase flow physics involving flow patterns and phase transitions has been the major challenge for technology adoption in industry. This paper demonstrates a machine learning (ML) based model for evaluating the thermal performance and refrigerant mass flow rate, of a thermosyphon cooling system for telecom equipment. Unlike conventional laboratory approach that requires numerous sensors attached to a cooling system to capture their thermal performance, the new model requires a minimum number of sensors to monitor the health of a thermal management solution. Using the proposed model, a system control module can be further developed which could identify optimal operating parameters in real-time under dynamically changing heat load conditions and actively maintain safety and thermal requirements.

Processor energy density is exceeding the capabilities of conventional air-cooling technology, but two-phase cooling has the potential to manage these resulting heat fluxes at reliable temperatures and higher electrical efficiency. When two-phase cooling is used in tandem with overclocking, data center footprints are reduced as individual chip processing power can be set at limits well beyond the manufacturer’s Thermal Design Power (TDP) or nominal operating condition. This study examines how Liquid Nitrogen (LN2) can be used with Additive Manufacturing (AM) and overclocking to increase the computational performance of a commercially available GPU. The power consumption and frequency relationship were established for both the cryogenically cooled solution and a comparative air-cooled solution. The cryogenic solution saw up to a 17.4% increase in compute efficiency and an 18.1% improvement in compute speed with comparable power efficiency at an equivalent performance level to the air-cooled solution. This study considers the computational performance and efficiency gains that can be acquired through cryogenic cooling on an individual graphics card, which can be replicated on a larger scale in data center applications.

Data centers housing high performance computing equipment have large and growing rack densities, which pushes the limits of traditional air cooling technologies because of limited heat transfer coefficients. Therefore, on-chip cooling using so-called cold plates is emerging as a necessary cooling option for high-density electronics. The use of mini-channels or pins fins to enhance internal heat transfer area inside cold plates requires extensive micro-machining that is relatively time consuming and expensive for mass production. As an alternative approach, inserting and bonding pre-manufactured metal foams into hollow bodies are explored as a potentially inexpensive means to enhance the interior heat transfer area of cold plates. One key aspect of the performance of metal foams in cold plates is the thermal contact resistance in the bonding between the foam and the substrate. This project predicts the contact resistance using measurements of different foam types (pure Cu and Cu with oxide), porosities (63%, 80%, 93%, and 95%) and thicknesses (4 mm, 8 mm, and 10 mm). These measurements are carried out with and without the use of thermal interface material (TIM) pads. A theory is proposed and implemented to estimate the contact and foam thermal resistances, but further work is needed to gain confidence in the results. Observations suggest that different thermal behavior is seen for the Cu foams compared to the Cu with oxide foams, and that the use of TIM pads can achieve 10x to 40x reduction in overall thermal resistance for highly porous foams bonded on Cu substrates.

With increased focus on miniature high power density electronic packages, there is a need for the development of new interface materials with lower thermal resistance. To this end, high conductivity thermal paste or similar thermal interface materials (TIMs), reinforced with superior thermal conductivity materials such as multi-walled carbon nanotubes (MWCNTs), graphene nanoplatelets (GNPs), graphite-derived multilayer graphene (g-MLG) offer an effective strategy to provide efficient paths for heat dissipation from heat source to heat sink. In an earlier paper, we had demonstrated that multilayer graphene derived from coal (coal-MLG) synthesized using our in-house developed one-pot process, has increased presence of phenolic groups on its surfaces, which translates into better dispersion of coal-MLG in silicone thermal paste. In this paper, we first compare the thermal conductance of a high conductivity thermal paste (k = 8.9 W/mK) using coal-MLG as an additive with that realized with other nano additives — MWCNTs, GNPs, and g-MLG. The data shows that coal-MLG as an additive outperforms all the other investigated nano additives in enhancing the thermal performance of the paste. With the coal-MLG as an additive, ∼70% increase in thermal performance was observed as compared to the base thermal paste used. This increase is about 2.5 times higher than that obtained using g-MLG as an additive. We also measured the thermal performance of coal-MLG-based TIM with its different wt.% fractions. The data confirmed our hypothesis that the optimum level of the loading fraction of the additive that can be dispersed in the matrix (paste in this case) before the onset of agglomeration is higher for the coal-MLG (3%) than for the other additives (2%). The implication is further improvement thermal performance with coal-MLG. The data shows the additional thermal enhancement to ∼2X. Finally, since coal-MLG produced by our in-house process is relatively cheaper and more environmentally friendly, we believe that these results would pave the path for enhanced thermal performance with non-silicone thermal pastes at a significantly lower cost. We also expect similar benefits for the silicone-based thermal pastes.

An in-rack cooling system connected to an external vapor recompression loop can be an economical solution to harness waste heat recovery in data centers. Validated subsystem-level models of the thermosyphon cooling and recompression loops (evaporator, heat exchangers, compressor, etc.) are needed to predict overall system performance and to perform design optimization based on the operating conditions. This paper specifically focuses on the model of the evaporator, which is a finned-tube heat exchanger incorporated in a thermosyphon cooling loop. The fin-pack is divided into individual segments to analyze the refrigerant and air side heat transfer characteristics. Refrigerant flow in the tubes is modeled as 1-D flow scheme with transport equations solved on a staggered grid. The air side is modeled using differential equations to represent the air temperature and humidity ratio and to predict if moisture removal will occur, in which case the airside heat transfer coefficient is suitably reduced. The louver fins are modeled as individual hexagons and are treated in conjunction with the tube walls. A segment-by-segment approach is utilized for each tube and the heat exchanger geometry is subsequently evaluated from one end to the other, with air property changes considered for each subsequent row of tubes. Model predictions of stream outlet temperature and pressure, refrigerant outlet vapor quality and heat exchanger duty show good agreement when compared against a commercial software.

This paper is focused on the modeling of a brazed plate heat exchanger (BPHE) for a novel in-rack cooling loop coupled with heat recovery capability for enhanced thermal management of datacenters. In the proposed technology, the BPHE is acting as a condenser, and the model presented in this study can be applied in either the cooling loop or vapor recompression loop. Thus, the primary fluid enters as either superheated (in the vapor recompression loop) or saturated vapor (in the cooling loop), while the secondary fluid enters as a sub-cooled liquid. The model augments an existing technique from the open literature and is applied to condensation of a low-pressure refrigerant R245fa. The model assumes a two-fluid heat exchanger with R245fa and water as the primary and secondary fluids, respectively, flowing in counterflow configuration; however, the model can also handle parallel flow configuration. The 2-D model divides the heat exchanger geometry into a discrete number of slices to analyze heat transfer and pressure drops (including static, momentum and frictional losses) of both fluids, which are used to predict the exit temperature and pressure of both fluids. The model predicts the exchanger duty based on the local energy balance. The predicted values of fluid output properties (secondary fluid temperature and pressure, and primary fluid vapor quality and pressure) along with heat exchanger duty show good agreement when compared against a commercial software.

This paper introduces a novel thermal management solution coupling in-rack cooling and heat recovery system. System-level modeling capabilities are the key to design and analyze thermal performance for different applications. In this study, a semi-empirical model for a hermetically sealed scroll compressor is developed and applied to different scroll geometries. The model parameters are tuned and validated such that the model is applicable to a variety of working fluids. The identified parameters are split into two groups: one group is dependent on the compressor geometry and independent of working fluid, whereas the other group is fluid dependent. By modifying the fluid-dependent parameters using the specific heat ratios of two refrigerants, the model shows promise in predicting the refrigerant mass flow rate, discharge temperature and compressor shaft power of a third refrigerant. Here, the approach has been applied using data for two refrigerants (R22 and R134a) to achieve predictions for a third refrigerant’s (R407c) mass flow rate, discharge temperature, and compressor shaft power, with normalized root mean square errors of 0.01, 0.04 and 0.020, respectively. The normalization is performed based on the minimum and maximum values of the measured variable data. The technique thus presented in this study can be used to accurately predict the primary variables of interest for a scroll compressor running on a given refrigerant for which data may be limited, enabling component-level design or analysis for different operating conditions and system requirements.

This paper advances the state-of-the-art in novel passive two-phase systems for more efficient cooling of datacenters and telecom central offices compared to the traditional air-based cooling solutions (e.g. aisle-based containment systems). The proposed passive two-phase technology uses numerous server-level thermosyphons to dissipate the heat generated by critical components, such as central processing units, accelerators, etc., with the flexibility of selecting the rack-level and room-level cooling elements depending on the deployment scenarios. The main goal of this paper is to experimentally investigate the thermal performance and maximum heat removal capability of a server-level thermosyphon for cooling compact servers. The experimental apparatus, built at Nokia Bell Labs, incorporates a single 7-cm high liquid-cooled thermosyphon that fits within a 2U server (smaller form factors can be achieved by a proper design that would further reduce the thermosyphon’s height). The heat source is represented by a pseudo-chip, composed of six parallel cartridge heaters installed in a copper block that incorporates local temperature measurements and is able of dissipating a total power of ≈ 500 W over a footprint area of 3.5 cm × 3.5 cm (corresponding heat flux of ≈ 41 W/cm 2 ). Steady-state experiments were carried out at various heat loads up to 240 W (corresponding heat flux of ≈ 20 W/cm 2 ), filling ratios and secondary side inlet conditions (coolant temperatures and mass flow rates), using R1234ze(E) and deionized water as the working fluids on the primary and secondary side, respectively. Test results demonstrate high heat transfer performance of the server-level thermosyphon over a wide range of conditions, and operating points are identified and classified into an operational map. Thermosyphon-based cooling systems across multiple length scales can significantly improve operation in terms of lowering energy consumption, allowing for higher hardware density, increased processing speed and reliability.

Frost formation on evaporators negatively affects the cooling performance of refrigerators. It increases the thermal resistance between the refrigerant and air leading to a reduction in the system cooling capacity. In this study, the effect of frost accumulation over a bare and finned surface on the convective thermal resistances has been experimentally investigated under impinging flow conditions. The surfaces are vertically positioned in a horizontal wind tunnel. The convective resistances have been measured with an in-house developed heat flux measurement system. Finally, the effectiveness of the finned surface was derived from the measurements for dry, condensing flow and as well as for frosting conditions. Under frosting conditions, the effectiveness of the finned surface is measured as 1.4 that is by a factor of 2X lower compared to the effectiveness of the same finned surface operating under dry conditions. It has been observed that the frost accumulation initially takes place at the tip of the fins and leads to a 45% drop in the heat transfer rate when the fin tips are completely covered with frost. Further frost accumulation on the fin base does not result in an additional drop in the heat transfer rate. In this regard, the study emphasizes the importance of the fin tip design for the heat sinks operating under frosting conditions.

With the recent advances in wide band gap device technology, solid-state lighting (SSL) has become favorable for many lighting applications due to energy savings, long life, green nature for environment, and exceptional color performance. Light emitting diodes (LED) as SSL devices have recently offered unique advantages for a wide range of commercial and residential applications. However, LED operation is strictly limited by temperature as its preferred chip junction temperature is below 100 °C. This is very similar to advanced electronics components with continuously increasing heat fluxes due to the expanding microprocessor power dissipation coupled with reduction in feature sizes. While in some of the applications standard cooling techniques cannot achieve an effective cooling performance due to physical limitations or poor heat transfer capabilities, development of novel cooling techniques is necessary. The emergence of LED hot spots has also turned attention to the cooling with dielectric liquids intimately in contact with the heat and photon dissipating surfaces, where elevated LED temperatures will adversely affect light extraction and reliability. In the interest of highly effective heat removal from LEDs with direct liquid cooling, the current paper starts with explaining the increasing thermal problems in electronics and also in lighting technologies followed by a brief overview of the state of the art for liquid cooling technologies. Then, attention will be turned into thermal consideration of approximately a 60W replacement LED light engine. A conjugate CFD model is deployed to determine local hot spots and to optimize the thermal resistance by varying multiple design parameters, boundary conditions, and the type of fluid. Detailed system level simulations also point out possible abatement techniques for local hot spots while keeping light extraction at maximum.

The movement to LED lighting systems worldwide is accelerating quickly as energy savings and reduction in hazardous materials increase in importance. Government regulations and rapidly lowering prices help to further this trend. Today’s strong drive is to replace light bulbs of common outputs (60W, 75W and 100W) without resorting to Compact Fluorescent (CFL) bulbs containing mercury while maintaining the standard industry bulb size and shape referred to as A19. For many bulb designs, this A19 size and shape restriction forces a small heat sink which is barely capable of dissipating heat for 60W equivalent LED bulbs with natural convection for today’s LED efficacies. 75W and 100W equivalent bulbs require larger sizes, some method of forced cooling, or some unusual liquid cooling system; generally none of these approaches are desirable for light bulbs from a consumer point of view. Thus, there is interest in developing natural convection cooled A19 light bulb designs for LEDs that cool far more effectively than today’s current designs. Current A19 size heat sink designs typically have thermal resistances of 5–7°C/W. This paper presents designs utilizing the effects of chimney cooling, well developed for other fields that reduce heat sink resistances by significant amounts while meeting all other requirements for bulb system design. Numerical studies and test data show performance of 3–4°C/W for various orientations including methods for keeping the chimney partially active in horizontal orientations. Significant parameters are also studied with effects upon performance. The simulations are in good agreement with the experimental data. Such chimney-based designs are shown to enable 75W and 100W equivalent LED light bulb designs critical for faster penetration of LED systems into general lighting applications.

As the power dissipated by advanced microelectronic devices continues to increase, the demand for reliability also increases. This increases the requirements on the thermal performance of every part of the system, including the heat sink. One of the objectives of this study is to examine the effect of the shape of the heat sink fins on the thermal performance of the system. The pressure gradient at the center of the heat sink, near the base, tends to be high. This significantly reduces the airflow at that location and, hence, decreases transport in that region. Different fin shapes and airflow rates have been studied with the objective of searching for an optimal heat sink design that would improve the thermal performance without increasing the pressure drop across the heat sink. Parallel plate fins have been investigated by removing fin material from the region near the center of the heat sink along the length and height of the fins. The study also examines the impact of uniform and non-uniform heat sources in the device upon the overall system thermal performance. Twenty one heat sink designs with different cuts were simulated and compared and an improved heat sink design was proposed by eliminating the fin material at the center of the heat sink, thereby enhancing its thermal performance.

This paper discusses active cooling devices used to maintain acceptable temperature levels for electronics packages in a high temperature environment. The temperature control is provided by a free-piston Stirling cooler (FPSC). The FPSC is a compact, quiet, low-vibration, environmentally benign cooling device based on the Stirling cycle. In principle, the cold side temperature of the FPSC, that thermally regulates the electronics package, is controllable from a low end, at cryogenic temperatures, upwards to the hot side temperature, which is limited only to the maximum temperature survivable by the materials used to construct the driving motor. General characteristics of the FPSC are discussed from the standpoint of <b>electronics</b> <b>cooling</b> and various options for heat transport are presented. Two prototype FPSCs have been developed for cooling electronics packages in applications where it has not been possible to accomplish even close to the requirements with existing technology. The first prototype is being used to cool an electronics package to below 100°C in a 200°C operating environment. The hot side temperature of the FPSC was designed to operate at 230°C in order to reject heat to the 200°C ambient, while limiting the motor temperature to less than 250°C. A cooling capacity of 120 W was achieved. The second prototype is used to provide conditioned air for a sealed, waterproof, electronics enclosure that is exposed to a 60°C environment. The FPSC operating temperatures for this application are 30°C on the cold side and 80°C on the hot side, with a cooling capacity of 500 W.

This paper presents an experimental study and theoretical interpretation of two-phase flow in a closed loop. The objective of this work is to find the optimum flow rate with respect to the thermal design power (amount of heat to be rejected). We assume that forced-convection boiling characteristics are explained based on mass and energy conservation, and claim that our proposed coefficient ( C ≡ Q L / Q : a ratio of amount of evaporated liquid to the flow rate) indicates the optimum flow rate for wide variation of evaporator-shapes and working fluids. In order to verify our model, we have measured the thermal resistance of evaporator with respect to heater input power for various flow rates. Hydrofluoro ether (HFE) and Fluorinert™ refrigerants were used as the working fluid in the experiment. Here flow rate of 40∼120ml/min and thermal design power of 50∼200W were controlled by the pump and by the heater, respectively. We observed that the coefficient resulted in the optimum flow rate is almost the same regardless of working fluids and evaporator shapes. The data which indicates the optimum flow rate were quite well reproduced by our proposed theory when the value of this coefficient is C ≈ 0.7∼0.95. For the demonstration, we designed the assembled-type two-phase cooling module with the optimum flow rate based on our model, and we observed that the evaporator had a relatively small thermal resistance of 0.1K/W.

A detailed evaluation of modeling techniques used to incorporate a complex component level steady-state thermal model of a 32-pin SSOP package in a system level simulation is described. The device is an Application Specific IC designed for the present system, intended for use in a demanding automotive environment. The paper discusses the development of a compact model, used to enable the incorporation into a system level model with reasonable computational detail. The methodology involves a preliminary system level thermal analysis to estimate the ambient conditions in the vicinity of the device, followed by a detailed model of the device analyzed for various boundary conditions. A simplified compact model of the device is developed and validated against the detailed model. The compact model is shown to be accurate in the prediction of peak temperature in the package and boundary condition independent within realistic extremes of the operating environment. The fact that a compact model may not match the more detailed model at all points in the package is scrutinized in the context of correlation of the model to experimental results. The study highlights the significance of board-level conductivity enhancement and thermal vias under the package in reduction of package temperature.

The cooling performance of piezoelectric actuators is evaluated for low-form-factor electronics in this work. A piezoelectric actuator is a cantilever made from metal or plastic with a piezoelectric material bonded to it. Under an alternating electrical current, the piezo actuator oscillates back and forth, generating airflow. Compared to conventional fans, these actuators have the advantages of low power consumption, low noise, and smaller dimensions. The parameters investigated in the experiments are actuator orientation, actuator-to-heat source distance, and actuator amplitude. For an actuator power consumption of 31 mW, the heat source temperature was lowered by more than 25°C compared to natural convection conditions (for a 2.45 W heater power dissipation). Performance comparisons against axial fans and natural convection heat sinks show that the piezo actuators perform significantly better in terms of power consumption and cooling volume. This paper was also originally published as part of the Proceedings of the ASME 2005 Heat Transfer Summer Conference .

Plate fin heat sinks are commonly used in <b>electronics</b> <b>cooling</b> including high end processors. A number of empirical and analytical methods are available to predict their performance but most of the models are valid for fin pitch larger than 3 mm heat sinks in laminar flow. The present work is to investigate high dense plate fin heat sink in both laminar and turbulent regimes. Thermal and hydraulic performance of several dense plate-fin heat sinks were characterized for high end processors in a fully-ducted wind tunnel. All the three heat sinks tested have the same dimensions of 89 mm (L) × 56 mm (W) × 50 mm (H), and fin number varied between 23 and 33. Heat sink base for all heat sinks was made of solid copper, while different fin materials of Aluminum and Copper are used. Several analytical methods for laminar flow from literature were reviewed in this study. A new heat transfer analytical method was proposed for both laminar and turbulent flows. The characterization data from these three parallel plate heat sinks were compared with the analytical methods. Finally, empirical heat transfer correlations were developed for both laminar and turbulent flows.

Loop heat pipe (LHP) is a very versatile heat transfer device that uses capillary forces developed in the wick structure and latent heat of evaporation of the working fluid to carry high heat loads over considerable distances. Robust behaviour and temperature control capabilities of this device has enable it to score an edge over the traditional heat pipes. In the past, LHPs has been invariably assessed for <b>electronic</b> <b>cooling</b> at large scale. As the size of the thermal footprint and available space is going down drastically, miniature size of the LHP has to be developed. In this paper, results of the investigation on the miniature LHP (mLHP) for thermal control of electronic devices with heat dissipation capacity of up to 70 W have been discussed. Copper mLHP with disk-shaped flat evaporator 30 mm in diameter and 10 mm thickness was developed. Flat evaporators are easy to attach to the heat source without any need of cylinder-plane-reducer saddle that creates additional thermal resistance in the case of cylindrical evaporators. Wick structure made from sintered nickel powder with pore size of 3–5 μm was able to provide adequate capillary forces for the continuos circulation of the working fluid, and successfully transport heat load at the required distance of 60 mm. Heat was transferred using 3 mm ID copper tube with vapour and liquid lines of 60 mm and 200 mm length respectively. mLHP showed very reliable start up at different heat loads and was able to achieve steady state without any symptoms of wick dry-out. Tests were conducted on the mLHP with evaporator and condenser at the same level. Total thermal resistance, R total of the mLHP came out to be in the range of 1–4°C&#x2F;W. It is concluded from the outcomes of the investigation that mLHP with flat evaporator can be effectively used for the thermal control of the electronic equipments with restricted space and high heat flux chipsets.