Multidevice Cooling With Flow Boiling in a Variable Microgap

Author(s):  
S. J. Young ◽  
D. Janssen ◽  
E. A. Wenzel ◽  
B. M. Shadakofsky ◽  
F. A. Kulacki

Abstract Flow boiling in an onboard variable microgap is demonstrated as a viable cooling method for multidevice electronics. The microgap is created by a bonded conformal encapsulation that delivers uniform subcooled inlet coolant flow across a multidevice layout comprising a processor and two in-line, symmetrically placed memory devices. Each device is simulated with a ceramic resistance heater on a 1:1 scale, and the heights of the devices create the variable microgap under the roof line of the encapsulation. The gap height for the processor is 0.5 mm and 1 mm for the memory devices. Parameters investigated are pressure drop, average device temperature, processor power, and coefficient of performance (COP). For inlet coolant flow first over the memory devices, the average device temperature exceeds the 95 °C limit when processor power is ∼50 W or less. For inlet flow over the processor, memory device temperatures are approximately the same over all the levels of processor and memory chip power. For processor power <30 W and an inlet coolant temperature of 25 °C, single-phase heat transfer is the dominant cooling mechanism. When processor power is >40 W, two-phase heat transfer dominates, and processor power of 120 W is reached within the 95 °C threshold. Volumetric power density across the data set is 134 to 1209 W/cm3.

Author(s):  
S. J. Young ◽  
D. Janssen ◽  
E. A. Wenzel ◽  
B. M. Shadakofsky ◽  
F. A. Kulacki

Onboard liquid cooling of electronic devices is demonstrated with liquid delivered externally to the point of heat removal through a conformal encapsulation. The encapsulation creates a flat microgap above the integrated circuit (IC) and delivers a uniform inlet coolant flow over the device. The coolant is Novec™ 7200, and the electronics are simulated with a resistance heater on a 1:1 scale. Thermal performance is demonstrated at power densities of ∼1 kW/cm3 in the microgap. Parameters investigated are pressure drop, average device temperature, heat transfer coefficient, and coefficient of performance (COP). Nusselt numbers for gap sizes of 0.25, 0.5, and 0.75 mm are reduced to a dimensionless correlation. With low coolant inlet subcooling, two-phase heat transfer is seen at all mass flows. Device temperatures reach 95 °C for power dissipation of 50–80 W (0.67–1.08 kW/cm3) depending on coolant flow for a gap of 0.5 mm. Coefficients of performance of ∼100 to 70,000 are determined via measured pressure drop and demonstrate a low pumping penalty at the device level within the range of power and coolant flow considered. The encapsulation with microgap flow boiling provides a means for use of higher power central processing unit and graphics processing unit devices and thereby enables higher computing performance, for example, in embedded airborne computers.


2015 ◽  
Vol 137 (11) ◽  
Author(s):  
D. Janssen ◽  
J. M. Dixon ◽  
S. J. Young ◽  
F. A. Kulacki

Heat transfer coefficients in a set of three symmetrically heated narrow gap channels arranged in line are reported at power densities of 1 kW/cm3 and wall heat flux of 3–40 W/cm2. This configuration emulates an electronics system wherein power dissipation can vary across an array of processors, memory chips, or other components. Three pairs of parallel ceramic resistance heaters in a nearly adiabatic housing form the flow passage, and length-to-gap ratios for each pair of heaters are 34 at a gap of 0.36 mm. Novec™ 7200 and 7300 are used as the heat transfer fluids. Nonuniform longitudinal power distributions are designed with the center heater pair at 1.5X and 2X the level of the first and third heater pairs. At all levels of inlet subcooling, single-phase heat transfer dominates over the first two heater pairs, while the third pair exhibits significant increases because of the presence of flow boiling. Reynolds numbers range from 250 to 1200, Weber numbers from 2 to 14, and boiling numbers from O(10−4) to O(10−3). Exit quality can reach 30% in some cases. Overall heat transfer coefficients of 40 kW/m2K are obtained. Pressure drops for both Novec™ heat transfer fluids are approximately equal at a given mass flux, and a high ratio of heat transfer to pumping power (coefficient of performance (COP)) is obtained. With a mass flux of 250 kg/m2s, heater temperatures can exceed 95 °C, which is the acceptable limit of steady operation for contemporary high performance electronics. Thus, an optimal operating point involving power density, power distribution, mass flux, and inlet subcooling is suggested by the data set for this benchmark multiheater configuration.


Author(s):  
D. D. Janssen ◽  
J. M. Dixon ◽  
S. J. Young ◽  
F. A. Kulacki

Heat transfer coefficients in a set of three symmetrically heated narrow gap channels arranged in line are reported at watt densities of 1 kW/cm3. This experimental configuration emulates an electronics system wherein power dissipation can vary across an array of processors, memory chips, or other components. Three pairs of parallel ceramic resistance heaters in a nearly adiabatic housing form the flow passage, and length-to-gap ratios for each pair of heaters is 34.1 at a gap of 0.36 mm. Novec™ 7200 and 7300 are used as the heat transfer fluids. Non-uniform longitudinal power distributions are designed with the center heater pair at 1.5X and 2X the level of the first and third heater pairs. At all levels of inlet sub-cooling, single-phase heat transfer dominates heat transfer over the first two heater pairs, while the third pair exhibits significant increases because of the presence of flow boiling. Reynolds numbers range from 250 to 1200, Weber numbers from 2 to 15, and boiling numbers from O(10−4) to O(10−3). Exit quality can reach 30 percent in some cases. Overall heat transfer coefficients of 40 kW/m2K are obtained. Pressure drops for both Novec™ heat transfer fluids are approximately equal at a given mass flux, and a high coefficient of performance is obtained. With a mass flux of 250 kg/m2s, heater temperatures can exceed 95 °C, which is the acceptable limit of steady operation for contemporary high performance electronics. Thus, an optimal operating point involving power density, power distribution, mass flux, and inlet sub-cooling is suggested by the data set for this benchmark multi-heater configuration.


Materials ◽  
2021 ◽  
Vol 14 (16) ◽  
pp. 4617
Author(s):  
Sanghyun Nam ◽  
Dae Yeon Kim ◽  
Youngwoo Kim ◽  
Kyung Chun Kim

Heat transfer under flow boiling is better in a rectangular channel filled with open-cell metal foam than in an empty channel, but the high pressure drop is a drawback of the empty channel method. In this study, various types of metal foam insert configurations were tested to reduce the pressure drop while maintaining high heat transfer. Specifically, we measured the boiling heat transfer and pressure drop of a two-phase vertical upward flow of R245fa inside a channel. To measure the pressure and temperature differences of the metal foam, differential pressure transducers and T-type thermocouples were used at both ends of the test section. While the saturation pressure was kept constant at 5.9 bar, the steam quality at the inlet of the test section was changed from 0.05 to 0.99. The channel height, moreover, was 3 mm, and the mass flux ranged from 133 to 300 kg/m2s. The two-phase flow characteristics were observed through a high-speed visualization experiment. Heat transfer tended to increase with the mean vapor quality, and, as expected, the fully filled metal foam channel offered the highest thermal performance. The streamwise insert pattern model had the lowest heat transfer at a low mass flux. However, at a higher mass flux, the three different insert models presented almost the same heat transfer coefficients. We found that the streamwise pattern model had a very low pressure drop compared to that of the spanwise pattern models. The goodness factors of the flow area and the core volume of the streamwise patterned model were higher than those of the full-filled metal foam channel.


Author(s):  
Wenhai Li ◽  
Ken Alabi ◽  
Foluso Ladeinde

Over the years, empirical correlations have been developed for predicting saturated flow boiling [1–15] and condensation [16–30] heat transfer coefficients inside horizontal/vertical tubes or micro-channels. In the present work, we have examined 30 of these models, and modified many of them for use in compact plate-fin heat exchangers. However, the various correlations, which have been developed for pipes and ducts, have been modified in our work to make them applicable to extended fin surfaces. The various correlations have been used in a low-order, one-dimensional, finite-volume type numerical integration of the flow and heat transfer equations in heat exchangers. The NIST’s REFPROP database [31] is used to account for the large variations in the fluid thermo-physical properties during phase change. The numerical results are compared with Yara’s experimental data [32]. The validity of the various boiling and condensation models for a real plate-fin heat exchanger design is discussed. The results show that some of the modified boiling and condensation correlations can provide acceptable prediction of heat transfer coefficient for two-phase flows in compact plate-fin heat exchangers.


2021 ◽  
Author(s):  
Matt Harrison ◽  
Joshua Gess

Abstract Using Particle Image Velocimetry (PIV), the amount of fluid required to sustain nucleate boiling was quantified to a microstructured copper circular disk. Having prepared the disk with preferential nucleation sites, an analytical model of the net coolant flow rate requirements to a single site has been produced and validated against experimental data. The model assumes that there are three primary phenomena contributing to the coolant flow rate requirements at the boiling surface; radial growth of vapor throughout incipience to departure, bubble rise, and natural convection around the periphery. The total mass flowrate is the sum of these contributing portions. The model accurately predicts the quenching fluid flow rate at low and high heat fluxes with 4% and 30% error of the measured value respectively. For the microstructured surface examined in this study, coolant flow rate requirements ranged from 0.1 to 0.16 kg/sec for a range of heat fluxes from 5.5 to 11.0 W/cm2. Under subcooled conditions, the coolant flow rate requirements plummeted to a nearly negligible value due to domination of transient conduction as the primary heat transfer mechanism at the liquid/vapor/surface interface. PIV and the validated analytical model could be used as a test standard where the amount of coolant the surface needs in relation to its heat transfer coefficient or thermal resistance is a benchmark for the efficacy of a standard surface or boiling enhancement coating/surface structure.


2019 ◽  
Vol 32 (2) ◽  
pp. 1905117 ◽  
Author(s):  
Wenming Li ◽  
Zuankai Wang ◽  
Fanghao Yang ◽  
Tamanna Alam ◽  
Mengnan Jiang ◽  
...  

2008 ◽  
Vol 131 (1) ◽  
Author(s):  
Jong Chull Jo ◽  
Woong Sik Kim ◽  
Chang-Yong Choi ◽  
Yong Kab Lee

This paper addresses the numerical simulation of two-phase flow heat transfer in the helically coiled tubes of an integral type pressurized water reactor steam generator under normal operation using a computational fluid dynamics code. The shell-side flow field where a single-phase fluid flows in the downward direction is also calculated in conjunction with the tube-side two-phase flow characteristics. For the calculation of tube-side two-phase flow, the inhomogeneous two-fluid model is used. Both the Rensselaer Polytechnic Institute wall boiling model and the bulk boiling model are implemented for the numerical simulations of boiling-induced two-phase flow in a vertical straight pipe and channel, and the computed results are compared with the available measured data. The conjugate heat transfer analysis method is employed to calculate the conduction in the tube wall with finite thickness and the convections in the internal and external fluids simultaneously so as to match the fluid-wall-fluid interface conditions properly. Both the internal and external turbulent flows are simulated using the standard k-ε model. From the results of the present numerical simulation, it is shown that the bulk boiling model can be applied to the simulation of two-phase flow in the helically coiled steam generator tubes. In addition, the present simulation method is considered to be physically plausible in the light of discussions on the computed results.


Sign in / Sign up

Export Citation Format

Share Document