Experimental investigation of flow boiling heat transfer enhancement under ultrasound fields in a minichannel heat sink.

Ultrasound is considered to be an effective active heat transfer enhancement method, which is widely used in various fields. But there is no clear understanding of flow boiling heat transfer characteristics in micro/mini-channels under ultrasonic field since the studies related are limited up to now. In this paper, a novel minichannel heat exchanger with two ultrasonic transducers inside the inlet and outlet plenum respectively is designed to experimentally investigate the impacts of ultrasound on flow boiling heat transfer enhancement in a minichannel heat sink. Flow visualization analyses reveal that ultrasound can promote rapid bubble motion, bubble detachment from heating wall surface and thereby new bubble generation, and decrease the length of confined bubble. Furthermore, the flow boiling experiments are initiated employing working fluid R141b at different ultrasonic parameters (e.g., frequency, power, angle of radiation) and heat flux under three types of ultrasound excitations: no ultrasound (NU), single inlet ultrasound (IU), inlet and outlet ultrasound (IOU). The results indicate that ultrasound has obvious augmentation effects on flow boiling heat transfer even though the intensification effects will be limited with the heat flux increases. The higher ultrasonic power, the lower ultrasonic frequency and the higher ultrasonic radiation angle, the better intensification efficiency. The maximum enhancement ratio of have in the saturated boiling section reaches 1.88 at 50 W, 23 kHz and 45° under the experimental conditions. This study will be beneficial for future applications of ultrasound on flow boiling heat transfer in micro/mini-channels.

pressure amplitude of ultrasound, Pa

specific heat capacity, J/kg K

ultrasonic frequency, kHz

the secondary Bjerknes force, N

mass velocity of single minichannel, kg/m2 s

height of minichannel, m

distance between measurement point and minichannel bottom wall, m

local heat transfer coefficient, W/m2 K

average saturated heat transfer coefficient, W/m2 K

the numbers of thermocouples positioned in saturated boiling heat transfer

length of minichannel heat sink, m

fin parameter

mass flux of fluid, kg/s

total number of minichannels

outlet pressure, kPa

ultrasonic power, W

effective heat flux, W/m2

heating power absorbed by the working fluid, W

total heating power of heater plate, W

latent heat of vaporization, J/kg

distance between the centers of two bubbles, m

equilibrium radii of bubble j, m

top planform area of heat sink, m2

local bulk liquid temperature, °C

fluid temperature at inlet, °C

fluid temperature at outlet, °C

temperature of minichannel bottom wall, °C

local saturation temperature, °C

temperature of the upper thermocouple, °C

inlet subcooling, °C

width of minichannel, m

width of minichannel heat sink, m

width of the fin between minichannels, m

distance from the inlet, m

heat loss ratio

fin efficiency

radiation angle of ultrasound, °

thermal conductivity, W/m K

density of the liquid, kg/m3

vapor quality

angular driving frequency of ultrasound

natural frequency of bubble j

average

minichannel

minichannel inlet

liquid phase

serial number of thermocouple

without ultrasound

outlet of minichannel

with ultrasound

Introduction

With the rapid pace at which device functionalities in micro/nano electronic applications are jumping for advances in many electric technologies, processor chips are expected to contain tens of billions of nanoscale transistors [1] which causes the generation of extremely high heat fluxes. That poses a severe challenge to the traditional heat transfer technology. One of the most effective solutions accepted widely by researchers and applicants is flow boiling heat transfer technology in micro/mini-channel heat sinks with high heat transfer efficiency and low power consumption [2]. However, with the continuing rapid expansion in micro/nano electronic technologies, which gradually induce that existing technologies are unbale to meet demands of higher heat dissipation, more innovative and effective heat transfer intensification technologies are required to overcome the challenges. Recently, studies are mainly focused on two types of heat transfer enhancement technologies in flow boiling heat transfer [3]: passive technologies, e.g. surface roughness [4], [5], surface nanostructures [6], [7], microchannel structure [8], [9], fluid additives (such as nanoparticles [10], [11], etc.); and active technologies, e.g. electric field [12], [13], magnetic field [14], [15], ultrasonic field [16], [17] etc.

Ultrasound is regarded as an active heat transfer enhancement technology by receiving increasing interest from researchers since 1960s. In 1965, Bergles and Newell [18] experimentally studied the effects of ultrasound on heat transfer to water flowing in annuli. The phenomenon that applying 20 kHz ultrasound was able to enhance heat transfer was observed for the first time. The results showed that up to 40% local increase of heat transfer coefficient was achieved but only 10% in the global coefficient by ultrasound, which was in part due to the attenuation of sound effect, or to a bad contact between the ultrasonic emitter and the tube containing water. Until 1969, Wong and Chon [19] investigated the effects of ultrasonic vibrations on heat transfer by natural convection and by boiling. They found ultrasound had remarkable influence on boiling heat transfer. The research subject, the effects of ultrasound on boiling heat transfer, was studied for the first time. Despite the significant reinforcement effect of ultrasound on heat transfer, the researches on ultrasound as a possible thermal enhancement technique has not attracted much attention until the 2000, because the early studies did not obtain promising results to lead to deeper research [20]. It was in 2010 that Qu and Qiu [16], [17] firstly investigated the influence of acoustic field on flow boiling heat transfer in a microscale.

Boiling is one of the most studied modes of heat transfer enhancement in the presence of an ultrasonic field [20]. The subject has been favored by more and more researchers in recent 20 years. Some people found that acoustic cavitation plays an important role in boiling heat transfer enhancement. Laborde et al. [21] conducted the experiment of pool boiling to understand the complex physical mechanism induced by the propagation an ultrasound wave (from 20 to 800 kHz) through a liquid. They observed that bubbles oscillate, often non-linearly, around some equilibrium size during many cycles of acoustic pressure. Zhou and Liu [22] experimentally explored and showed that acoustic cavitation enhanced remarkably the pool boiling heat transfer and decreased the incipient boiling superheat. Also, they found the distance between cavitation bubbles enable to affect the heat transfer efficiency. Moehrle and Chung [23] visually studied and found that the acoustic force could promote the departure of bubbles from the heating surface, and therefore played an important role in enhancement of pool boiling heat transfer. Chen et al. [24] performed a test of a stainless steel circular heater rod for heat transfer performance with and without ultrasonic vibration in a thermostat water tank. The maximum ratio of heat transfer enhancement (the definition of enhancement ratio is same for all mentioned publications, as seen in Eq. (12)) under the action of ultrasound can be about 3.01. And, they found the enhancement ratios of ultrasound on boiling varied from 0.18 to 4.17 with different experimental conditions over the survey of the past work. However, Mandroyan and Hihn et al. [25], [26] initiated the study and found that excessive acoustic cavitation bubbles inducing void fraction increasing would attenuate the acoustic intensity and then the acoustic driven bubbles phenomenon. Hegedűs and Mettin et al. [27] numerically studied that the active cavitation threshold of a dual-frequency driven single spherical gas bubble for a given relative expansion.

The acoustic streaming is regarded as mainly another reason for the heat transfer enhancement of boiling in recent studies. Kim et al. [28] proposed that the violent motion of cavitation bubbles caused by ultrasound and acoustic streaming were the major reason for the enhancement in pool boiling. Chouvellon et al. [29] and Kumar et al. [30] revealed that the acoustic streaming can drive the fluid with a velocity of 0.0066 m/s ~ 1.6 m/s depending on the type, frequency, and power of the ultrasonic transducer. Tang et al. [31] presented that the acoustic streaming induced by high amplitude ultrasound has an obvious enhancement effect on nucleate pool boiling at low heat flux, but the augmentation will be restricted with increase in heat flux and decrease in ultrasonic amplitude. In addition, Li et al. [32] reported that the influence of different surface characteristics (smooth tube, screwed tube and finned tube) on heat transfer enhancement due to ultrasonic vibration on horizontal copper tubes in a sub-cooled boiling regime. Zheng et al. [33] investigated that the effects of ultrasonic frequency and power on the boiling heat transfer of the three different structural tubes and found visually that the acoustic streaming could help bubbles escape from tube surface.

The mutual interaction force between bubbles, so-called secondary Bjerknes force caused by radiation pressure exerted by other bubble [34], in the presence of ultrasound is another interest to researchers. The secondary Bjerknes force is attractive or repulsive, which leads to agglomeration or dispersion of bubbles [35], [36]. Therefore, it is of great significance to fully understand the type of action force for the application of ultrasound. Ida. [37], [38] found that the distance between the bubbles could affect the direction of the secondary Bjerknes force. Zhang et al. [39] investigated numerically the basic features of the secondary Bjerknes force under dual frequency excitation, and found that change of the sign of the secondary Bjerknes force occurred near the resonance bubble radii corresponding to the driving frequencies. Barbat and Ashgriz [40] studied theoretically that planar motion (relative 2-D) of two spherical gas bubbles under the secondary Bjerknes forces.

The literatures reviewed above show that the studies on boiling heat transfer enhancement by ultrasound are mainly focused on pool boiling or convection heat transfer, while the researches of flow boiling heat transfer intensification in micro/mini-channels is extremely limit so that there is no obvious understanding of heat transfer mechanism in micro/mini-channels under ultrasonic field. In the present work, a novel minichannel heat exchanger with two ultrasonic transducers inside the inlet and outlet plenum respectively is designed to experimentally explore the impact of direct-contact ultrasound on flow boiling heat transfer enhancement in a minichannel heat sink. The bubble motion behaviors with and without ultrasound are analyzed by performing flow visualization study. Additionally, the effects of different ultrasonic parameters (e.g., frequency, power, angle of radiation) and ultrasonic excitations (NU, IU, IOU) on heat transfer enhancement are investigated. Furthermore, for constant ultrasonic parameters, the influence of outlet vapor quality and inlet subcooling on flow boiling heat transfer is studied.

Experimental apparatus and method

Test loop

The schematic diagram of the experimental test loop used for the flow boiling is illustrated in the Fig. 1. The test loop is made up of a closed flow loop 1 (including minichannel heat exchanger, reservoir, magnetic pump, filter, float flowmeter, etc.) with several attached parts involving ultrasonic generator, heating unit, measuring instrument, high speed camera, data acquisition module, air-cooled circulation device with a flow cooling loop 2 (including air chiller and condenser), etc. R-141b, the working fluid employed in this experimental study, is injected by working fluid charge device to reservoir, where it is circulated through the loop by a variable frequency magnetic pump. The fluid flowing out of the pump is spilt into two streams. One stream flow through a bypass valve, and then goes back into the reservoir, while the other stream passes through a filter to prevent clogging in the flow loop. Afterwards, a float flowmeter with a range of 0 ~ 250 L/h is used to measure the volumetric flow rate before entering a pipeline immersed in a constant temperature water bath to obtain the required temperature at the inlet of minichannel heat exchanger. Furthermore, R-141b is converted into a two-phase mixture due to heating in the minichannel heat exchanger, which is wrapped with thermal insulation material to prevent heat loss. The two-phase working fluid from the heat exchanger is cooled by the flow loop 2, and then flows back to the tank after returning to the liquid state.

Fig. 1

Schematic diagram of the experimental flow loop.

Temperature of the working fluid at the heat exchanger inlet & outlet and along the streamwise are measured by employing K-type thermocouples ranging from 0 ~ 200 °C. Pressure measurements of the working fluid are obtained using pressure transducer in the range of 0 ~ 700 kPa. An adjustable AC power with a power range of 0 ~ 3 kW is used to supply heating power to the heat sink. Ultrasonic transducers at the heat exchanger inlet and outlet plenum are connected to an ultrasonic generator (KMD-K1), which can produce ultrasonic field with frequency of 20 kHz ~ 40 kHz and power level of 0 W ~ 50 W. Additionally, A high speed camera (SVSi GigaView) fitted with a microscope of Nikon Micro-Nikkor 60 mm f/2.8D is used for flow visualization at an image capture rate of 4261 frames/s and resolution of 1280 × 128 pixels. The measurement data from the thermocouples, pressure transducers are collected and recorded by data acquisition module (Agilent 34970).

Minichannel heat exchanger

The schematic of the minichannel heat exchanger with exploded view is presented in Fig. 2(a). It consists of an aluminum alloy minichannel heat sink, two ultrasonic transducers (inlet and outlet), two ultrasound transducer fixing blocks (inlet and outlet), a cover plate, an aluminum alloy bottom housing, a heater plate, a cover brace and two O-rings. The minichannel heat exchanger components are sequentially fastened tightly through the 26 bolts. Two fluorous O-rings (one is placed between the cover brace and cover plate, the other is located between the cover brace and the bottom housing) are applied to prevent the R141b leakage. The cover plate positioned between the cover brace and the minichannel heat sink is made by Pyrex glass in order to perform flow visualization studies. The visual window occupies a part of cover plate.

Fig. 2

Schematic diagram of the minichannel heat exchanger: (a) Exploded view, (b) Assembly view, (c) Heat sink view, (d) View of mounting angle of ultrasonic transducer.

The bottom housing assembly view is shown in Fig. 2(b). The minichannel heat sink is connected to the top center of bottom housing using thermal grease, while the heater plate acting as the heat source is attached to the bottom housing base symmetrical to the heat sink. Two ultrasonic transducers are placed at the inlet and outlet plenum of bottom housing by fixing blocks, which are bolted to the bottom housing. The inlet and outlet plenum are designed to be large enough to place transducers. Additionally, the ultrasonic transducers are connected to an ultrasonic generator through the connection holes of ultrasonic transducer, which can create three types of ultrasound excitation: NU, IU, IOU. NU denotes that there is no ultrasonic excitation in the minichannel heat exchanger. IU and IOU refer to the application of ultrasound in the inlet and inlet & outlet of the minichannel heat exchanger respectively. The transducers at the inlet and outlet by IOU have same ultrasonic frequency and power. Holes are drilled into the front of bottom housing as well as into the inlet and outlet plenums for inserting thermocouple probes. The six holes behind of bottom housing are employed to mount pressure transducers. Four of them along the heat sink are corresponding to the positions of thermocouples (as seen in the Fig. 2(a)).

A view of minichannel heat sink is depicted in Fig. 2(c). The minichannel heat sink occupying a footprint of 220 mm × 100 mm (L × Wt) comprises of fourteen parallel rectangular minichannels with a single channel size of 2 mm × 2 mm (Wch × Hch). The width between adjacent minichannels is 5 mm (Ww).

Fig. 2(d) shows the schematic diagram of mounting angle of ultrasonic transducer. The mounting angle is the angle between the axis of ultrasonic transducer and the minichannel heating surface. The radiation angle of ultrasonic transducers can be set from 0° to 45° by adjusting the mounting angle. The three types of radiation angle (0°, 25°, 45°) are chosen to study the effects of the angle of ultrasonic radiation on heat transfer.

Fig. 3 shows that the layout of thermocouple installation locations on the bottom housing of the minichannel heat exchanger (as shown in Fig. 2). The upper row of these thermocouple holes (T1 ~ T4, Tin, Tout) is located at the 5 mm below the bottom wall of minichannels, while the lower (T5 ~ T8) is situated at 19 mm below the upper row. Two thermocouples (Tin, Tout) whose center is 115 mm from the sides of heat sink are respectively inserted into the inlet and outlet plenum for measuring the inlet and outlet temperature of the working fluid. Besides, the upper row of thermocouple holes (T1 ~ T4) is used to measure the temperature of the minichannel heat sink, while the lower (T5 ~ T8) is employed to collect the temperature data of the heating part of bottom housing in order to figure out the mode of heat transfer from the heating block to the bottom of minichannel heat sink.

Fig. 3

Thermocouple hole distribution view.

Experimental procedure

Experiments are performed at the operation conditions of mass velocity from 90.8 to 181.7 kg/m2 s; heat flux from 9.7 to 21.2 kW/m2; inlet temperature from 35 to 35.5 °C; outlet pressure 152 ± 1 kPa; ultrasonic frequency of 23 kHz, 28 kHz, 32 kHz and 40 kHz; ultrasonic power of 12.5 W, 25 W, 37.5 W and 50 W; the ultrasonic radiation angle of 0°, 25°, 45°. The experimental run for this study starts with vacuumizing by a vacuum pump after setting the ultrasonic radiation angle. The non-condensable gases are discharged from the test system to prevent the influence of non-condensable gases and steam on experiments. Then, as shown in the Fig. 1, opening the valve 1 and closing the valve 2 and 3, the working fluid (R141b) is charged into the reservoir by the charge device. Further, the circulation in the flow loop 1 starts after the magnetic pump turning on. The flow rate is adjusted by regulating the bypass valve or the valve 4 of the flowmeter. Furthermore, adjusting the valve 6 located at the downstream of the minichannel heat exchanger, the pressure at the outlet of the minichannel heat exchanger is maintained at around 152 kPa in order to ensure the saturation temperature of R141b constant. Next, when both the flow rate and outlet pressure are gradually adjusted to the desired operating conditions, the inlet temperature is regulated to the required value by setting the water temperature of thermostatic bath. When the flow rate, inlet temperature and outlet pressure reach the desired values, the electrical power for the heater plate is turned on to adjust heat flux. For each required heat flux, the ultrasonic generator is applied to produce three types of ultrasound excitation (NU, IU, IOU), and then to set the corresponding ultrasonic frequency and power level. Finally, the temperature and pressure measurements are collected by DAQ after the system attains a steady state (the variations of all temperature and pressure measurements are within ±0.3 °C for at least 5 min), while flow visualization videos are recorded by using a SVSi GigaView high-speed video camera with the FPS 4261 frames/s.

Heat loss estimation

In order to ensure the validity of the experimental study, it is necessary to take into account the heat loss ratio in the experimental section. The heat loss ratio () in the present is obtained by conducting single-phase heat transfer balance experiments [41], [42] under the conditions of outlet pressure around 152 kPa, initial heat power 60 W. Heat loss ratio () is obtained fromwhere is the total heating power of heater plate, is heating power absorbed by the working fluid.where , , , are the mass flux, the specific heat capacity at constant pressure, the outlet temperature and the inlet temperature of working fluid, respectively.

Fig. 4 depicts variation of heat loss ratio () with heat flux at the mass flux 121.1 kg/m2 s and 181.7 kg/m2 s. It can be seen that the heat loss ratio decreases with the increase of heat flux, and finally stabilizes near the average value 0.18. Also, the heat loss ratio is in the range of 0.1 ~ 0.27, which is similar to Deng et al. [41] 0.1 ~ 0.3, Tang et al. [42] 0.1 ~ 0.25 and Xu et al. [43] 0.1 ~ 0.23. According to those studies, it is valid that the mean value of is applied to the flow boiling tests even though there is a certain deviation from the actual value under specific conditions.

Fig. 4

Variation of heat loss ratio with heat flux.

Data reduction

The effective heat flux () is calculated fromwhere denotes the heat flux transmitted to the working fluid, is the top planform area of the heat sink, , is the width of the minichannel heat sink, is the length of the minichannel.

The mass velocity of single minichannel () is given bywhere represents the mass flux of fluid, represents the number of minichannels; represents the width of a single minichannel; represents the height of a single minichannel.

The outlet vapor quality () in the present study is expressed by [12], [43],where denotes the latent heat of vaporization, Tsat,out is the outlet saturated temperature, and is set to the value of the saturated temperature based on the outlet pressure. is controlled by regulating the heat flux and increases with the rising in the heat flux.

According to the calculation method in ref [44], applying a fin model to the minichannels, the local heat transfer coefficient () at the temperature measuring point can be obtained byin which is local temperature of the minichannel bottom surface corresponding to the upper thermocouple. is the fluid temperature at the thermocouple location, is the height of minichannel, is the fin efficiency which is given byin which is the fin parameter defined asin which is the thermal conductivity of aluminum alloy.

There are mainly three stages after the working fluid entering the heat sink: single-phase flow, subcooling boiling and saturated boiling. At the beginning, the fluid is pure liquid flow. Then, the first boiling bubble generates at the onset of nucleate boiling (ONB) with the working fluid heating, and goes into subcooling boiling. As the fluid flows, it is further heated and enters saturation boiling. The heat transfer efficiency in saturated boiling stage will sharply increases. In the whole process, the working fluid experiences brief subcooled stage and then is heated up to saturated state. The fluid boiling in the minichannel is divided into two regions: subcooled region and saturated region.

For the subcooled region, the temperature of working fluid () in Eq. (6) is expressed asand for the saturated region, is given byin which is the local saturated temperature at the thermocouple position and the corresponding local pressure at this position, is the distance from the inlet temperature measuring point.

In Eq. (6) is expressed as follows:in which (namely T1 ~ T4) is the temperature of the upper thermocouple, is the distance between the temperature measurement point and the bottom of the minichannel, . It is worthy to note that the heat transfer model of the minichannel bottom housing is assumed to be one-dimensional steady-state heat conduction, and is considered to be effective and accurate in previous work [12], [45], [46].

After and are determined by Eqs. (9), (10), (11), the local heat transfer coefficients () can be obtained by Eqs. (6), (7), (8) with the iteration method as seen in Fig. 5.

Fig. 5

The computation process of the iteration method.

In order to understand more intuitively the effect of ultrasound on flow boiling heat transfer enhancement in minichannels, the formula of ultrasound enhancement ratio (Uh) is defined as follows:where and denote the average saturated boiling heat transfer coefficient with and without ultrasound, respectively. is given bywhere j denotes the numbers of thermocouples positioned in saturated boiling heat transfer in minichannel heat sink.

Experimental uncertainties

The uncertainties of measuring instruments employed in present study and derived parameters are tabulated in Table 1. All the thermocouples and pressure transducers have been calibrated before being mounted in the heat exchanger. The uncertainties of derived parameters involving the effective heat flux, average saturated boiling heat transfer coefficient, local heat transfer coefficient and ultrasound enhancement ratio are determined by principle of error transfer in ref. [12], [47], [48].

Table 1

Uncertainties of primary experimental measurements.

Parameters	Maximum Uncertainty
Temperature	0.2%
Pressure	0.5%
Volumetric flow rate	0.5%
Heat sink geometric dimensions	0.2%
Effective heat flux	4.28%
Average saturated boiling heat transfer coefficient	7.41%
Local heat transfer coefficient	10.12%
Ultrasound enhancement ratio	8.39%

Results and discussion

Flow visualization analysis

Subcooled boiling section

Fig. 6 (a)-(c) illustrate the schematics of bubble motion in the subcooled boiling stage of a single minichannel at the mass velocity 121.1 kg/m2·s, the heat flux 10.9 kW/m2, ultrasonic frequency 23 kHz, ultrasonic power 50 W for each ultrasonic transducer, radiation angle 25° under three ultrasonic actuations of NU, IU, IOU respectively. The images captured by the high-speed camera show the location at different time of the bubble marked out by the red circle in the single minichannel of the visual window under each ultrasonic excitation when the bubble passes through the subcooled boiling passage. The T0, T1, T2 on the Fig. 6 are the time the marked bubble entering the single minichannel for NU, IU, IOU respectively. The red arrow indicates flow direction. The length of the subcooled boiling passage is about 20 mm. The time interval between adjacent frames elected is 4.694 ms. The rightmost graphs in Fig. 6(a)-(c) are the reconstructed diagram of bubble distribution when the bubble marked exits the minichannel. The diagrams show that a bubble experiences around 66 ms, 28 ms, 56 ms to pass through this subcooled passage for NU, IU, IOU respectively. The present work applies the time to represent the speed of bubble motion. The velocity of bubble for NU, IU, IOU was around 0.304, 0.710, 0.355 m/s respectively. It can be concluded that the speed of bubble motion increases to 2.3 times under the action of IU, while that under IOU rises to 1.2 times. The reason why the presence of IU can speed up bubble motion is mainly the action of acoustic streaming inducing pressure gradient and then accelerating fluid movement [29], [30]. However, the effects of acoustic streaming under IOU could be weakened by the superposition of inlet and outlet acoustic waves. The acoustic streaming generated by outlet ultrasound is weaker than that of inlet ultrasound due to excessive bubbles in the outlet plenum preventing the ultrasound propagation [25], [26]. The acoustic cavitation in the minichannel could be so rare that there is neglect of effects of acoustic cavitation on bubble motion because the mode of bubble generation (producing in the downstream area) is closely similar to that without ultrasound. Another possibility could be that the secondary Bjerknes forces between the bubbles which push bubbles and the fluid.

Fig. 6

Schematics of bubbles motion in the subcooled boiling passage under three ultrasonic excitations for heat flux 10.9 kW/m2, mass velocity 121.1 kg/m2·s, ultrasonic frequency 23 kHz, ultrasonic power 50 W, θ = 25°.

Furthermore, the numbers of bubble in the presence of ultrasound is a little more than that in silent regime; it can be count from the rightmost reconstructed diagrams of Fig. 6(a)-(c) that there are about 10 bubbles, 12 bubbles and 13 bubbles under NU, IU and IOU respectively. The reason why the number of bubbles increase with ultrasound is that ultrasonic forces can promote the bubbles detachment from the heating wall surface [17], which can incur the new bubble formation in the heating wall surface under experimental conditions and therefore increase the bubble detachment frequency from heating surface. That phenomenon can also be seen in the Fig. 6 that more bubbles with ultrasound left the heating surface while most of bubbles without ultrasound attached to the heating wall .

Fig. 7 show that the motion trajectories of the bubbles smaller than the size of minichannel in the subcooling boiling section with and without ultrasound. The upper first ones in Fig. 7(a) and (b) are reconstructed based on those position images to present the bubbles motion trajectories; The lower four figures in the Fig. 7(a) and (b) are pictured by high-speed video camera and show the position of the bubbles at different moments. Note that there is a shadow on the left side of each bubble in the Fig. 7(b) due to the oblique lighting of dysprosium lamp configuring for high-speed camera. It can be seen that the bubbles move along a straight line against the wall under NU, while in the presence of ultrasound the motion trajectories show that the bubbles move with up and down vibration perpendicular to the fins between two minichannel fins along the flow direction. It cannot be observed that bubbles vibrate along other directions due to the limitation of flow visualization apparatus. In the process of bubble motion along the trajectories in the ultrasound, the bubble moving in the minichannel can disturb the bulk liquid so that enhances heat transfer performance, while the bubble contacting to the heating wall is observed to sweep away those small bubbles in the wall, which leads to the new bubble generation. In conclusion, those bubble motion behaviors in the ultrasound will be beneficial to enhance flow boiling heat transfer.

Fig. 7

Schematics of bubble motion trajectory with and without ultrasound: (a) under the ultrasound condition, (b) under the silent condition.

Saturated boiling section

Fig. 8(a)-(c) depict the schematics of bubbles motion in the saturated boiling stage of a single minichannel for three ultrasonic excitations of NU, IU, IOU respectively. The T0, T1, T2 on the Fig. 8 are the time the marked bubble entering the single minichannel for NU, IU, IOU respectively. Each image presents the location of the bubble in the single minichannel of visual window at different time pictured by high-speed camera. The length of the saturated boiling passage is also about 20 mm. It can be seen from the Fig. 8(a)-(c) that the bubble marked experiences approximately 80 ms for NU, 33 ms for IU, 52 ms for IOU respectively to pass through this saturated stage in the visual window. The velocity of bubble for NU, IU, IOU was around 0.251, 0.609, 0.387 m/s respectively. It can be concluded that the speed of the bubble motion is increased to 2.4 times and 1.5 times by IU and IOU respectively. Applying the time to represent the speed of bubble motion and comparing to the phenomenon in Fig. 6, it can be found that in the absence of ultrasound, the speed (0.251 m/s) of the bubble in the saturated boiling stage is slower than that (0.304 m/s) of subcooled boiling stage as the bubbles grow and coalesce to form elongating confined bubbles. In the presence of ultrasound, the bubble motion (0.609 m/s) for IU in the saturated boiling is little slower than that (0.710 m/s) in the subcooling boiling, while the speed (0.387 m/s) of bubble for IOU in the saturated boiling is slightly faster than that (0.355 m/s) in the subcooling boiling.

Fig. 8

Schematics of bubbles motion in the saturated boiling passage under three ultrasonic excitations for heat flux 10.9 kW/m2, mass velocity 121.1 kg/m2·s, ultrasonic frequency 23 kHz, ultrasonic power 50 W for each ultrasonic transducer, θ = 25°.

Additionally, it can be observed from Fig. 8 that the length of elongating confined bubble (whose the size is larger than the cross section of the microchannels) with ultrasound is less than that without ultrasound under same experimental conditions. That could be mainly attributed to the repulsive action of the secondary Bjerknes force preventing large bubbles coalescing to confined bubble [39]. Under the acoustic actuation, the force between bubbles, the secondary Bjerknes force, can prevent small bubbles approaching each other and merge [35], [36], [37], [38]. The secondary Bjerknes force is expressed by ref. [39], [49],where indicates repulsive action, indicates attactive action; is the complex pressure amplitude of ultrasound, is the angular driving frequency of ultrasound, and are the equilibrium radii of the bubbles, is the density of the liquid, separation distance between the centers of two bubbles, and are the natural frequency of bubbles corresponding to the equilibrium radii. It can be seen from Eq. (14) that when the driving frequency lies between the two natural frequencies corresponding to the equilibrium radii (namely or ,), indicates the mutual repulsion of the bubbles; otherwise the bubbles attract each other, which is partly responsible for enlarging the bubbles in the subcooled boiling the section.

On the other hand, according to the images (Fig. 8) with and without ultrasound by high-speed photography, it can be seen obviously that the number of bubbles without ultrasound (around 4 bubbles) is smaller than that with ultrasound (around 6 bubbles for IU, 7 bubbles for IOU), but the size of confined bubble is larger than that with ultrasound. This indicates that ultrasound can reduce the size of the dominant confined bubble and increase the number of boiling bubbles under the same heat flux, which benefits the heat transfer enhancement [50]. From the perspective of flow patterns [51], it can be seen from Fig. 8 that ultrasound can prevent to some extents liquid in the saturated region changing from bubbly flow to elongating confined bubble flow, which is able to intensify the flow boiling heat transfer [50].

Effects of ultrasonic field on flow boiling heat transfer

Fig. 9(a)-(b) present variations of the local heat transfer coefficient (h) with the axial distance (Z) along the stream-wise direction of the minichannel at different heat fluxes and three ultrasonic actuations (NU, IU, IOU) for two mass velocities of 151.4 kg/m2·s and 181.7 kg/m2·s, respectively. When the mass velocity is 151.4 kg/m2·s in the absence of ultrasound, Fig. 9(a) shows that the heat transfer coefficient at the heat flux 10.7 kW/m2 rises slightly along the flow direction in the upstream region, and then remains nearly stable from the third thermocouple location in the downstream region. The reason for this phenomenon is that the upstream of minichannel is subcooled region, while downstream region is saturated boiling region. With the heat flux increasing from 10.7 kW/m2 to 16.5 kW/m2, the local heat transfer coefficient generally increases. It is worthy to note that the working fluid at high heat flux (16.5 kW/m2) enters the saturated boiling ahead of time compared with low heat flux (10.7 kW/m2), followed by a rather flat decrease, which is caused by the change of flow regime in the downstream from bubbly flow to confined bubble flow.

Fig. 9

Variations of local heat transfer coefficient (h) along the stream-wise direction of the minichannel at f = 23 kHz, θ = 25°, Pus = 50 W, Tin = 35 ~ 35.5 °C, P = 152 kPa for IU, IOU and mass velocity of (a) 151.4 kg/m2·s and (b) 181.7 kg/m2·s.

When it comes to the mass velocity 151.4 kg/m2·s for two ultrasound excitations (IU, IOU), it can be seen from Fig. 9 that the local heat transfer coefficient (h) at each thermocouple location in the presence of ultrasound is much larger than that without ultrasound for both of heat flux, and h by IOU is larger than that by IU. The results indicate that ultrasonic field can obviously intensify the flow boiling heat transfer in minichannel heat sink, and the enhancement effect by IOU is much stronger than that by IU. At the same ultrasonic frequency and power for each transducer, IOU has a better heat transfer enhancement effect than IU due to the higher power (to be discussed in below section, as seen in Fig. 11 (a)-(b)), even though the working fluid is less accelerated for IOU than for IU configuration. It can be inferred in our experiment that the ultrasound power has more effects on heat transfer than that of the speed of bubble motion. When the heat flux is 10.7 kW/m2 under the ultrasonic actuation, h increases slightly along the flow direction in the upstream region, followed by a relatively rapid rise from the second thermocouple location in the middle region, and then keeps stable from the third thermocouple location in the downstream. The strongest enhancement efficiency (84% by IOU) occurs at the fourth thermocouple location, increasing from 2490.95 W/m2·K without ultrasound to 4581.65 W/m2·K with IOU. That is because the upstream, middle of minichannel and downstream region are single phase, subcooled boiling and saturated boiling region, respectively. In the single-phase region, the heat transfer intensification is mainly caused by intensified effects of ultrasound on convective heat transfer by accelerating bulk fluid movement, while in the two-phase region, the heat transfer enhancement is mainly attributed to the ultrasonic effect on bubble dynamics [12], [13]. When the heat flux increases to 16.5 kW/m2 under ultrasound, h increases sharper than that without ultrasound in the upstream, and then presents slightly drop. The maximum enhancement ratio (64% by IOU) is obtained at the second thermocouple location, increasing from 2978.44 W/m2·K without ultrasound to 4875.27 W/m2·K with IOU. That suggests that the enhancement effects weaken with the heat flux increasing.

Fig. 11

Variations of average heat transfer coefficient (have) and heat transfer enhancement ratio (Uh) in saturated boiling region with heat flux at different ultrasonic power and ultrasonic actuations.

When it comes to the mass velocity 181.7 kg/m2·s with and without ultrasonic field, Fig. 9(b) shows that the trends of the heat transfer coefficients with the stream-wise distance are similar to that for the mass velocity 151.4 kg/m2·s. It is worth noting that in the presence of ultrasound, the differences of maximum heat transfer coefficient at two heat flux narrows a little with the increasing mass velocity. That indicates that the effects of heat flux in the presence of ultrasound decreases with the mass velocity rising, which is also drawn by ref. [52].

Effects of ultrasonic parameters on heat transfer augmentation

In order to further explore the influence of ultrasound on flow boiling heat transfer in minichannel, the effects of different ultrasonic parameters (frequency, power, radiation angle of ultrasound) on saturated boiling heat transfer characteristic are studied in this section.

Ultrasonic frequency

Fig. 10 (a) and (b) depict variations of average heat transfer coefficient (have) in saturated boiling region with heat flux at four ultrasonic frequency (23 kHz, 28 kHz, 32 kHz, 40 kHz) and three ultrasonic actuations (NU, IU, IOU) for θ = 25°, Pus = 50 W, G = 121.1 kg/m2·s. It can be seen from Fig. 10(a) that the average heat transfer coefficient in silent regime presents monotonic increasing with the heat flux increasing, while in the presence of IU, the changes of have at different ultrasonic frequency shows similar trend, a slow increase to the maximum at around 15 kW/m2, followed by a slightly decrease. The heat transfer enhancement by ultrasound is prominent when the heat flux is low, and the effects of ultrasound on flow boiling heat transfer weakens with the increase of heat flux, which is because the flow regime changes from bubbly flow to confined bubbly flow with the heat flux increasing. After the flow regime is becoming the elongating confined bubbly flow, the heat transfer will be deteriorated if the heat flux continues to increase [50]. When the heat flux is approximately greater than 18 kW/m2, have in the presence of ultrasound is basically same. That can be interpreted by the flow regime transition during the saturated boiling region. The flow regime at low heat flux is mainly bubbly flow, while at the high heat flux, the flow regime is dominated by elongating confined bubble flow. Additionally, have decreases with the increasing ultrasound frequency at the same the heat flux. That may be explained by higher ultrasound frequency promoting the bubbles coalescences more drastically due to the pressure changer from the positive to the negative pressure rapidly [53], which indicates that bubble grows more by coalescing than by absorbing heat with the rising ultrasound frequency. Therefore, the heat absorbed under higher ultrasound frequency is less than that under lower ultrasound frequency in the process of bubble growth.

Fig. 10

Variations of average heat transfer coefficient (have) and heat transfer enhancement ratio (Uh) in saturated boiling region with heat flux at different ultrasonic frequency and ultrasonic actuations.

For the ultrasonic actuation IOU, have presents an obvious rise at each heat flux comparing to IU. The increment of have caused by different ultrasonic actuations decreases with the increasing heat flux and ultrasonic frequency from 824.60 W/m2·K at 23 kHz and 9.72 kW/m2 to 67.24 W/m2·K at 40 kHz and 21 kW/m2. That implies that the effects of different ultrasonic actuation on heat transfer is weakened with the increasing heat flux and ultrasonic frequency. Furthermore, although the variation of have for IOU presents the similar trend with that for IU, the ultrasonic actuation IOU can achieve higher heat transfer efficiency (3709.29 ~ 4157.97 W/m2·K for IOU, 2991.80 ~ 3508.61 W/m2·K for IU) at low heat flux (qeff < 15 kW/m2). Also, when qeff greater than 15 kW/m2, the descending in have for IOU with the rising heat flux is much sharper comparing to the IU, which is mainly attributed to more bubble generation and growth to form large elongating confined bubbles under IOU so as to deteriorate the heat transfer.

Fig. 10(c) and (d) depict variations of heat transfer enhancement ratios (Uh) in saturated boiling region at four ultrasonic frequency (23 kHz, 28 kHz, 32 kHz, 40 kHz) and two ultrasonic actuations (IU, IOU). It can be seen that the enhancement efficiency by ultrasound is inversely proportional to ultrasonic frequency and heat flux. At the same heat flux, the higher ultrasonic frequency is, the lower Uh will be. For the frequency 23 kHz, the enhancement ratios by IU range from 0.05 to 0.45, while the ratios by IOU are in the range of 0.09 ~ 0.82. However, when the frequency is 40 kHz, the heat transfer enhancement ratios by IU and IOU range from 0.02 to 0.35, 0.04 to 0.68 respectively.

Ultrasonic power

Fig. 11(a) and (b) show variations of average heat transfer coefficient (have) in saturated boiling region with heat flux at four ultrasonic power level (12.5 W, 25 W, 37.5 W, 50 W) for each transducer and three ultrasonic actuations (NU, IU, IOU) for G = 121.1 kg/m2·s, f = 23 kHz, θ = 25°. Under the experimental conditions, the change of have with heat flux in the presence of ultrasound at each power level presents a slowly increase to the maximum (3508.61 W/m2·K for IU, 4117.88 W/m2·K for IOU) at around 15 kW/m2 and power level 50 W, then followed by a slightly reduction, and finally approach the have at around 21 kW/m2 without ultrasound. The trend is similar to the results drawn in Fig. 10. Differently, the average heat transfer coefficient increases with the ultrasonic power rising, but reduces with the ultrasonic frequency rising. It could be because the higher ultrasonic power prevents the bubble merger to decrease the length of confined bubble due to the stronger action of secondary Bjerknes forces so that augment the flow boiling heat transfer under experimental conditions. Moreover, it can be seen from Fig. 11(a)-(b) that the IOU (each transducer 25 W, total ultrasound power 50 W) has a slightly better heat transfer efficiency than the IU 50 W.

Fig. 11(c) and (d) depict variations of heat transfer enhancement ratios (Uh) in saturated boiling region at four ultrasonic power (12.5 W, 25 W, 37.5 W, 50 W) and two ultrasonic actuations (IU, IOU). The diagram describes that the enhancement efficiency by ultrasound is proportional to ultrasonic power. The enhancement ratios for the power 12.5 W by IU vary from 0.01 to 0.13, while the ratios by IOU range from 0.04 to 0.45. However, when the power is 50 W, the heat transfer enhancement ratios by IU and IOU are in the range of 0.05 ~ 1.45, 0.09 ~ 0.82 respectively.

Ultrasonic radiation angle

Fig. 12(a) and (b) present variations of average heat transfer coefficient (have) in saturated boiling region with heat flux at three radiation angles (0°, 25°, 45°) and three ultrasonic actuations (NU, IU, IOU) for G = 121.1 kg/m2·s, f = 23 kHz. Under the experimental conditions, the variation of have with heat flux in the presence of ultrasound at each radiation angle is a slow increase to the maximum (3880.26 W/m2·K for IU, 4447.51 W/m2·K for IOU) at around 15 kW/m2 and radiation angle 45° following a slightly reduction, and then close to the have at approximate 21 kW/m2 without ultrasound. It is worthy to note that in the presence of IOU, the rises in the radiation angle are able to increase the critical heat flux at which ultrasound has little influence on the heat transfer.

Fig. 12

Variations of average heat transfer coefficient (have) and heat transfer enhancement ratio (Uh) in saturated boiling region with heat flux at different radiation angle and ultrasonic actuations.

Fig. 12(c) and (d) depict variations of heat transfer enhancement ratios (Uh) in saturated boiling region at three ultrasonic radiation angle (0°, 25°, 45°) and two ultrasonic actuations (IU, IOU). The diagram describes that the enhancement efficiency by ultrasound is increasing with the rise of ultrasonic radiation angle. The enhancement ratios for the radiation angle 0° by IU vary from 0.02 to 0.35, while the ratios by IOU range from 0.03 to 0.65. However, when the radiation angle is 45°, the heat transfer enhancement ratios by IU and IOU are in the range of 0.06 ~ 0.62, 0.13 ~ 0.88 respectively. That may be interpreted by the increasing in the ultrasonic pressure amplitude induced by the increment of the radiation angle (seen in the Fig. 2(d)) under this experimental condition [17], which promotes the reinforcing effects of ultrasound on heat transfer.

Heat transfer characteristics under ultrasonic field

Inlet subcooling

Fig. 13(a) and (b) show that variations of average heat transfer coefficient (have) and heat transfer enhancement ratio (Uh) in saturated boiling region with inlet subcooling at three ultrasonic actuations (NU, IU, IOU) for G = 151.4 kg/m2·s, Pus = 50 W, f = 23 kHz, θ = 25°, qeff = 15.4 kW/m2. It can be found from Fig. 13(a) that have without ultrasonic field almost presents a slightly rise with the inlet subcooling increasing, which indicates that inlet subcooling has slight effects on flow boiling heat transfer. In the presence of both ultrasound actuations, the variable trend of have is similar to that without ultrasound except obvious increments in have when the ΔTsub < 6 °C. The increments from the 4 to 12 °C for NU, IU, IOU are 384.42, 711.22, 812.00 W/m2·K, respectively (Table 2). Moreover, enhancement efficiency by IOU is obviously stronger than that by IU. For the Fig. 13(b), we can see that changes of heat transfer enhancement ratio at two ultrasonic actuations show a similar trend of a steady state. That implies that the modes of ultrasonic actuation have no significant influence on the intensified effects of inlet subcooling after the ΔTsub greater than 8 °C, because the higher ΔTsub is, the shorter the length of elongating confined bubble is, which leads to the increases of have. The heat transfer enhancement ratio by IOU is in the range of 0.49 ~ 0.62, while that by IU varies from 0.23 to 0.35. It can be concluded from the diagram that the enhancement ratio by IOU at each subcooling is almost twice that by IU.

Fig. 13

Variations of average heat transfer coefficient (have) and heat transfer enhancement ratios (Uh) in saturated boiling region with inlet subcooling at NU, IU, IOU.

Table 2

The average heat transfer coefficient (have) at different inlet subcooling for NU, IU, IOU.

△Tsub (°C)	have (W/m2 K)
	NU	IU	IOU
4	2677.46	3304.99	4013.11
6	2799.70	3524.37	4542.61
8	2853.64	3865.65	4625.42
10	2951.38	3961.90	4755.23
12	3061.88	4016.21	4825.11

Outlet vapor quality

Fig. 14(a) describes variations of average heat transfer coefficient (have) in saturated boiling region with outlet vapor quality at three ultrasonic actuations (NU, IU, IOU) for G = 151.4 kg/m2·s, 181.6 kg/m2·s, Pus = 50 W, f = 23 kHz, θ = 25°. The diagrams show that when the outlet vapor quality is<0.2, ultrasound has obvious influence on flow boiling heat transfer at same outlet vapor quality. Without ultrasonic field, the have increases stably with the rise in . However, in the presence of ultrasound, the have slightly rises firstly with the increasing and subsequently decreases sharply after the is larger than 0.17, and finally approaches the value without ultrasound. That implies that the effects of ultrasound on heat transfer intensification weaken gradually with the rising at high vapor quality (). Fig. 14(b) presents variations of the heat transfer enhancement ratio (Uh) in saturated boiling region with outlet vapor quality for corresponding conditions. The enhancement ratios at two ultrasonic excitations decreases with the increment of . Especially for the , there is a sharp fall in Uh. It can be seen that Uh for IOU is in the range of 0.04 ~ 0.82, while Uh for IU varies from 0.02 to 0.47. When , Uh with and without ultrasound is nearly the same under the experimental conditions, which indicates that the ultrasound has little influence on the flow boiling heat transfer.

Fig. 14

Variations of average heat transfer coefficient (have) and heat transfer enhancement ratio (Uh) in saturated boiling region with outlet vapor quality at NU, IU, IOU.

Conclusions

In this study, a novel rectangular minichannel heat exchanger with ultrasonic transducers inside the inlet and outlet plenum, which can generate three types of ultrasonic actuation (NU, IU, IOU), is developed to explore experimentally the impacts of ultrasonic parameters (e.g., frequency, power, ultrasonic radiation angle) and ultrasound excitations on R141b flow boiling heat transfer enhancement at different heat fluxes, mass velocities. Also, Flow visualization is performed to investigate the motion behaviors of bubbles with and without ultrasound. The important findings are obtained from this work as follows:

Flow visualization analyses show that in the subcooling boiling section, ultrasound can promote rapid bubble motion, bubble detachment from heating wall surface and thereby new bubble generation; some bubbles whose diameter smaller than the minichannel width move with up and down vibration perpendicular to the fin between two minichannel fins along the flow direction. Also, in the saturated boiling section, ultrasound can decrease the length of confined bubble. Moreover, the types of ultrasound actuation (IU, IOU) have distinct influence on the speed of bubble motion. In the subcooling region, the speed of the bubble motion increases to 2.3 times and 1.2 times by IU and IOU respectively, while in the saturated region, the speed rises to 2.4 times and 1.5 times by IU and IOU respectively.

Ultrasound have remarkably intensified effects on flow boiling heat transfer in minichannel. The maximum augmentation ratios of have by IU and IOU are 1.62 and 1.88 respectively. However, when the heat flux is approximately greater than 15 kW/m2 under experimental conditions, the intensification effects in saturated boiling section will be weakened with the heat flux increase.

Under the experimental conditions (ultrasonic frequency 20 kHz ~ 40 kHz, ultrasonic power level 0 ~ 50 W, ultrasonic radiation angle 0 ~ 45°), ultrasonic frequency has adverse effects on flow boiling heat transfer augmentation in minichannel, while ultrasonic power level and the radiation angle have positive influence on have in saturated boiling region. The maximum heat transfer enhancement ratio of have, namely 1.88, can be obtained at the 23 kHz, 50 W and θ = 45° under IOU.

In the presence of ultrasound field (23 kHz, 50 W, 25°), vapor quality has significant influence on the heat transfer enhancement efficiency. Under present experimental conditions, the intensified effects of ultrasound on flow boiling heat transfer is prominent at low vapor (); the effects of ultrasonic actuation on heat transfer intensification weaken gradually with the increase in the vapor quality when the . Whereas, inlet subcooling has no significant influence on the intensified effects of ultrasound.

CRediT authorship contribution statement

Fan Yu: Conceptualization, Visualization, Methodology, Formal analysis, Writing - original draft, Writing - review & editing. Xiaoping Luo: Funding acquisition, Project administration, Supervision, Writing - review & editing. Bolin He: Software, Resources, Writing - review & editing. Jian Xiao: Visualization, Resources. Wen Wang: Visualization. Jinxin Zhang: Writing - original draft.

Declaration of Competing Interest

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.