• Energy Research

The heat generation from recent advanced computer chips is increasing rapidly. This creates a challenge in cooling the chips while maintaining their temperatures below the threshold values. Another challenge is that the heat generation in the chip is not uniform where some chip components generate more heat than other components. This would create a large temperature gradient across the chip, resulting in inducing thermal stresses inside the chip that may lead to a high probability to damage the chip. The locations in the chip with heat rates that correspond to high heat fluxes are known as hotspots. This research study focuses on using thermoelectric modules (TEMs) for cooling chip hotspots of different heat fluxes. When a TEM is used for cooling a chip hotspot, it is called a thermoelectric cooler (TEC), which requires electrical power. Additionally, when a TEM is used for converting a chip’s wasted heat to electrical power, it is called a thermoelectric generator (TEG). In this study, the TEMs are used for cooling the hotspots of computer chips, and a TEC is attached to the hotspot to reduce its temperature to an acceptable value. On the other hand, the other cold surfaces of the chip are attached to TEGs for harvesting electrical power from the chip’s wasted heat. Thereafter, this harvested electrical power (HEP) is then used to run the TEC attached to the hotspot. Since no external electrical power is needed for cooling the hotspot to an acceptable temperature, this technique is called a sustainable self-cooling framework (SSCF). In this paper, the operation principles of the SSCF to cool the hotspot, subjected to different operating conditions, are discussed. As well, considerations are given to investigate the effect of the TEM geometrical parameters, such as the P-/N-leg height and spacing between the legs in both operations of the TEC mode and TEG mode on the SSCF performance.

Enclosed airspaces to reduce heat flow have been recognized for well over 100 years. Airspaces with one or more reflective surfaces define reflective insulation (RI) assemblies, a product type used in walls, roofs, windows with multiple panes, curtain walls and skylights. The thermal resistance (R value) of airspaces depends on the emittance of all surfaces, airspace dimensions and orientation, heat flow direction and surfaces temperatures. The modeling of RI now includes CFD coupled with radiation to quantify the total heat transfer. This study compares a validated model for airspace R values with existing methods such as ISO 6946 and hot-box results that provide the R values in the ASHRAE Handbook of Fundamentals. The existing methods do not include an airspace aspect ratio. This study showed that the aspect ratio can impact the R value by a factor of two. The impact of aspect ratio was calculated for double airspaces variation such as that for single airspaces. The present calculations are two-dimensional and also consider all the bounding airspace surfaces, while previous methods are one-dimensional and do not include surface temperature variations or detailed radiative transport.

Abstract When solar radiation hits a roof surface, a part of solar energy is reflected and part is absorbed. The absorbed part of solar energy results in an increase of the surface temperature of the roof. Cool reflective (white) roofs use bright surfaces to reflect a significant portion of the incident short-wave solar radiation, which lowers the surface temperature compared to conventional (black) roofs with bituminous membrane. As such, white roofs help reduce the urban heat island effect during the summer. The question is “do white roofs lead to moisture-related problems in northern and southern climates?” To help answer this question, numerical simulations were conducted to compare the hygrothermal performance of a single kind of white and black roofs under different outdoor and indoor conditions. The outdoor conditions are obtained from the weather database of the National Research Council of Canada, Institute for Research in Construction (NRC–IRC). The indoor conditions are taken based on the European standard (EN 15026) and ASHRAE recommendations for conditioned space. The type of roofs considered in this study is Modified-Bitumen (MOD-BIT) roofing systems. The numerical simulations were conducted for the outdoor climate of Toronto (ON), Montreal (QC), St John’s (NL), Saskatoon (SK), Seattle (WA), Wilmington (NC) and Phoenix (AZ). Results showed that for the outdoor climates of St John’s and Saskatoon, the white roofs could lead to longer-term moisture-related problems. However, for the outdoor climates of Toronto, Montreal, Seattle, Wilmington and Phoenix, buildings with white roofs were shown to have a low risk of experiencing moisture damage. Also, buildings with white roofs in these locations were predicted to show a net yearly energy savings compared to buildings with black roofs.

Abstract Reflective insulations are being used in home attics, flat roofs, sloped roofs and wall systems of building envelopes. The present model, hygIRC-C, was used to investigate the contribution of the reflective insulations to the thermal resistance of specimens. The predictions of the present model were compared with test data of different sample stacks with different types of reflective insulations. In a previous study, the present model was benchmarked using test data obtained from a Guarded Hot Box (GHB) in accordance with the ASTM C-1363 test method. In this study, the test data was obtained from a different test method based on the heat flow meter in accordance of ASTM C-518 in the case of horizontal sample stacks with reflective insulations. Results showed that the predicted heat fluxes on the same area and same location of Heat Flux Transducers (HFTs) on the top and bottom surfaces of the sample stacks are in good agreement with the measured heat fluxes (within ±1%). The derived R-values using these heat fluxes are also in good agreements. Due to the combined effect of heat transfer by convection and radiation in the airspace (facing the reflective surface), these predicted and measured heat fluxes are greater than the area-weighted average heat flux of whole sample stack, which is needed to determine the effective R-value of the sample. As such, the derived R-value from the test data resulted in underestimation of the effective R-value of the sample stack. After gaining confidence in the present model, it was used to conduct parametric study in order to quantify the contribution of reflective insulations to the effective R-value for a sample stack with different inclination angles, different directions of heat flow (upward and downward) and for a wide range of foil emissivity. Furthermore, the present model was used to compare the predicted R-values with the listed R-values in the 2009 ASHRAE Handbook [22] for enclosed air cavity (20 mm thick) of different effective emittance, inclinations and directions of heat flow.

Abstract Thermoelectric devices are currently being used in the applications of cooling and generating electricity. This study mainly focuses on using these devices for both applications toward cooling down computer chips. An important aspect in designing the cooling system is to minimize the non-uniformity of the temperature distribution in the computer chip so as to reduce the thermal stresses in it. Another aspect in designing the cooling system is to minimize its power requirements. To investigate these two aspects, the temperatures of the cold chip areas can be allowed to increase, but not to exceed a certain temperature threshold, by installing Thermoelectric Generators (TEGs) on these areas that can harvest electrical power from the chip wasted heat. Thereafter, the chip hotspot areas can be cooled down by installing Thermoelectric Coolers (TECs) on these areas that can be powered by the harvested electrical power from the TEGs in order to maintain the temperatures of these hotspots to be less than or equal a certain temperature threshold. This cooling technique is called “sustainable self-cooling framework” for cooling chip hotspots. However, the question is: can the harvested electrical power by the TEGs be enough to power the TECs for cooling chip hotspots? In this study, a 3D model is developed to optimize the performance of both TEGs and TECs. Thereafter, this model is validated against experimental data of TEC and TEG. The results showed that the model predictions were in good agreements with the experimental data to within ±4%. Also, considerations are given in this study to optimize the performance of cascaded and non-cascaded TEGs and TECs for future use them to develop sustainable self-cooling frameworks for cooling chip hotspots at different operating conditions. Finally, a case study is conducted in this paper for a sustainable self-cooling framework in order to address the question above. The results showed that the self-cooling framework can successfully cool down the hotspot at an acceptable temperature with not only no need for additional electrical power requirements but also for reducing the non-uniformity in the chip temperature distribution.

Abstract Skutterudite based thermoelectric unicouples are being considered for use in Advanced Radioisotope Power Systems (ARPSs) to support NASA’s planetary exploration missions. For these systems, which would be much lighter than state of the art Radioisotope Thermoelectric Generators (RTGs), it is important to ensure minimal degradation in the performance of unicouples that may be caused by material sublimation. In this work, two unicouples, JAN-04 with a thin metallic coating on the legs near the hot junction to suppress antimony sublimation and SEP-03 without coating, are tested for >1000 and 3600 h, respectively. The legs in the two unicouples are of almost the same dimensions and compositions; the p-legs are made of CeFe 3.5 Co 0.5 Sb 12 and Bi 0.4 Sb 1.6 Te 3 segments and the n-legs are made of CoSb 3 and Bi 2 Te 2.95 Se 0.05 segments. SEP-03 is tested at average hot and cold junction temperatures of 961.5 ± 22.0 and 296.3 ± 5.7 K, respectively, in argon gas at ∼0.068 MPa, and JAN-04 is tested at 962.8 ± 20.5 and 294.5 ± 3.3 K, respectively, initially in argon gas at the same pressure for ∼26.5 h then in vacuum ∼9.0 × 10 −7 Torr for >973.5 h. The measured open circuit voltage V oc (240 mV) and peak electrical power (1.64 W e ) for SEP-03 at the beginning of test (BOT) are higher than those for JAN-04 (188 mV and 0.84 W e , respectively). Although the argon gas effectively decreased the antimony loss from the legs of SEP-03, marked degradations in performance occurred. The estimated peak efficiency for SEP-03 decreased from 13.8% at BOT to 5.8% at end of test (EOT), and the peak power decreased from 1.64 W e at BOT to 0.48 W e at EOT, however, V oc decreased by ∼14%. The latter for JAN-04 decreased only by ∼3%, the estimated peak efficiency (∼12%) changed very little and the peak power decreased by ∼20%. Unlike SEP-03, the measured total and contact resistances of the legs in JAN-04 changed very little.

Abstract A one-dimensional, steady-state analytical model was developed to predict the CCFL in GATPTs, which treats the shear stress at the liquid-vapor interface as the sum of two terms: (a) adiabatic shear stress; and (b) dynamic shear stress. The latter accounts for the effect of evaporation/condensation at the liquid-vapor interface. The model predictions were in good agreement (within ±10%) with the data of other investigators for water and methanol. The results showed that neglecting the dynamic shear stress at intermediate and high liquid film flows underestimates the film Reynolds number at CCFL by more than 20%. The model was used to develop operation maps for R-113, acetone, methanol, heptane, water and Dowtherm-A working fluids, which give the film Reynolds number at the CCFL (or maximum power throughput) as a function of the vapor temperature in the range from 250 to 700 K. The effects of the thermosyphon inner diameter and length of the evaporator section on the film Reynolds number at CCFL were also investigated.

Abstract Heat transfer data of numerous investigators for uniformly-heated liquid pools of water, ethanol, methanol, Dowtherm-A, R-11 and R-113 in small cylindrical enclosures were compiled, sorted, and correlated in the following heat transfer regimes: (a) natural convection; (b) nucleate boiling; and (c) combined convection. In the combined convection, where both natural convection and nucleate boiling contribute to the heat transfer, the data were correlated by superimposing the natural convection and nucleate boiling heat transfer correlations using a power law approach as: Nu CC =(Nu 4 NC +Nu 4 NB ) 0.25 All correlations were within ±15% of most experimental data. The data covered a wide range of pool diameters (6–37 mm), heated pool heights (50–800 mm), working fluid filling ratios (0.1–3.25), and wall heat fluxes (0.7–383 kW m-2).

Heat transfer correlations were developed for the liquid film region, in the evaporator section of closed, two-phase, gravity-assisted thermosyphons in the following regimes: (a) laminar convection, at low heat fluxes, (b) combined convection, at intermediate heat fluxes, and (c) nucleate boiling, at high heat fluxes. These correlations were based on a data set consisting of a total of 305 points for ethanol, acetone, R-11, and R-113 working fluids, wall heat fluxes of 0.99–52.62 kW/m2, working fluid filling ratios of 0.01–0.62, inner diameters of 6–37 mm, evaporator section lengths of 50–609.6 mm, and vapor temperatures of 261–352 K. The combined convention data were correlated by superimposing the correlations of laminar convention and nucleate boiling using a power law approach, to ensure smooth transition among the three heat transfer regimes. The three heat transfer correlations developed in this work are within ±15 percent of experimental data.