The study on the difference of the start-up and heat-transfer performance of the pulsating heat pipe with water−acetone mixtures

International Journal of Heat and Mass Transfer 77 (2014) 834–842

Contents lists available at ScienceDirect

International Journal of Heat and Mass Transfer journal homepage: www.elsevier.com/locate/ijhmt

The study on the difference of the start-up and heat-transfer performance of the pulsating heat pipe with wateracetone mixtures Yue Zhu, Xiaoyu Cui ⇑, Hua Han, Shende Sun University of Shanghai for Science and Technology, 516 Jungong Road, Yangpu District, Shanghai, PR China

a r t i c l e

i n f o

Article history: Received 17 December 2013 Received in revised form 28 April 2014 Accepted 25 May 2014

Keywords: Pulsating heat pipe Mixture Start-up Heat-transfer performance

a b s t r a c t An experimental study was conducted to investigate the start-up and heat-transfer performance of a closed-loop pulsating heat pipe with wateracetone mixtures (at mixing ratios of 13:1, 4:1, 1:1, 1:4 and 1:13) and pure water and acetone under various filing ratios (3570%) and heat inputs (10100 W). The closed-loop pulsating heat pipe was vertically placed and bottom-heated (i.e., heating wires were wrapped on the evaporation section) with inner and outer diameters of 2.0 and 4.0 mm, respectively. It was observed that (1) compared with pure water, the pulsating heat pipe with water acetone mixtures of mixing ratios of 13:1, 1:1, 1:4 and 1:13 possessed improved start-up performances, which could be initiated under a heat input of 10 W and filling ratios of 35% and 45%. (2) Under low filling ratios (i.e., 35% and 45%), the pulsating heat pipe with wateracetone mixtures (i.e., at mixing ratios of 4:1, 1:1, 1:4 and 1:13) presented improved performance against the onset of dry-out conditions compared with PHPs using pure water and acetone. Under a heat input of 50 W, the thermal resistances of the PHP with water–acetone mixtures (i.e., at mixing ratios of 4:1, 1:1, 1:4 and 1:13) decreased from 33.6% to 68.9% compared with pure working fluids. The addition of a fraction of pure water into pure acetone (e.g., the 13:1 wateracetone mixture) was found to be effective against dry-out. Conversely, adding a fraction of pure acetone into pure water (e.g., the 1:13 wateracetone mixture) did not prevent the onset of dry-out. (3) For high filling ratios (i.e., 62% and 70%) for which dry-out conditions are rarely encountered, the heat-transfer performances of the pulsating heat pipe with wateracetone mixtures (at mixing ratios of 13:1, 4:1, 1:1, 1:4 and 1:13) were not as efficient as that of the pulsating heat pipe with pure fluids. In contrast with the minima of mixtures under certain heat inputs, the maximum thermal resistances of pure water and acetone decreased by 45.8% and 38.7%, respectively. Ó 2014 Elsevier Ltd. All rights reserved.

1. Introduction A novel type of heat pipe patented by Akachi [1], the pulsating heat pipe (PHP), has been introduced as a promising solution for small spaces with high heat flux due to the excellent heat-transfer performance, rapid thermal response and compact structure [2,3]. In terms of structure, the PHP can be categorized as two types, an open or a closed-loop pulsating heat pipe. Because the working fluid can flow continuously, the latter typically possesses greater heat-transfer performance compared with that of the former. Due to the small inner diameter of the PHP, the working fluids naturally distribute to form liquid–vapor slugs when the PHP is charged with fluid after evacuation. Due to the non-uniform distribution of liquid–vapor slugs, the pressure difference between parallel pipes gradually increases as external heating or cooling ⇑ Corresponding author. Tel.: +86 13166199495. E-mail address: [email protected] (X. Cui). http://dx.doi.org/10.1016/j.ijheatmasstransfer.2014.05.042 0017-9310/Ó 2014 Elsevier Ltd. All rights reserved.

sources are applied, which promotes the working fluid to flow inside the pipes. Once a PHP is initialized, the working fluids absorb heat at the hot end and flow to the cool end to release the heat. The operation and heat-transfer characteristics of the PHP are relatively complex and are determined by the type of working fluid, the filling ratio (FR), the heat input, the inclination and the number of turns [4]. With respect to the working fluid, pure fluids have been studied in several works [5,6]. However, an experimental study of the PHP with mixtures has yet to be conducted. In a few thermodynamic studies, the influence of the fluid mixture was investigated [7–12]. Long and Zhang [7] used cryogenic heat pipes with N2Ar mixtures (at a mole ratio of 50:50), pure N2 and Ar solutions. The study found that N2Ar mixtures broadened the range of functional temperature from 64 to 150 K. In other studies 8, a heat pipe with a 0.05 M 2-propanol (diluted in water) mixture was studied at different inclinations. Compared with pure water, the critical heat flux of the heat pipe with

835

Y. Zhu et al. / International Journal of Heat and Mass Transfer 77 (2014) 834–842

Nomenclature K I Q R T U

coverage factor current (A) heat input (W) thermal resistance (°C W1) temperature (°C) voltage (V)

Subscript c e i sat max min

0.05 M 2-propanol increased by 52%. Nuntaphan et al. [9] experimentally investigated the heat-transfer performance of heat pipes with methanolwater and TEG–water mixtures. It concluded that a heat pipe with a methanolwater mixture had a better heattransfer performance compared with a pipe with only pure water. Due to the evident contribution of the heat-transfer performance of TEG, the heat transferred in the heat pipe with TEGwater varied with the concentration of TEG. The literature 10 indicated that in contrast with pure water (0.41 cc or 1.9 cc), a heat pipe with wateralcohol mixtures (0.41 cc or 1.9 cc) presented better heattransfer performance. Jouhara et al. [11] indicated the advantage of utilizing an ethanolwater mixture (95.63% ethanol and 4.37% distilled water by weight) in the two phase closed thermosyphon (TPCT) at a horizontal orientation. Compared with pure water, the TPCT with the ethanol mixture operated at lower temperatures with dampened temperature fluctuations at the evaporator. At a lower heat transfer rate and a vertical orientation, the thermal resistance of the TPCT was also lower than that of pure water. Wang et al. [12] experimentally investigated the influence of CuO (1.0 wt.%) nanofluids on the heat-transfer performance of an inclined miniature mesh heat pipe. It noted that the total thermal resistance of the heat pipe reduced by approximately half to when water-based nanofluids were substituted for pure water as the working fluid. With respect to the PHP, Chu et al. [13,14] proposed an azeotropic binary working fluid in the PHP application. Shi et al. [15] observed that the thermal resistance of the PHP with a water– methanol mixture was lower than that of the PHP using pure methanol. Burban et al. [16] carried out open loop pulsating heat pipe experiments with a water–n-pentane mixture, using 2:3 and 1:3 mixing ratios, under an FR of 60%. It mentioned that the heat-transfer performance of the PHP with the mixture was better than that of the PHP with pure water or n-pentane. However, the range of the FR and ratios of mixtures studied in these studies were relatively limited. A PHP study focusing on mixtures with a wider FR range and diverse mixing ratios has not yet been performed. Materials with low boiling points (i.e., generating bubbles at lower temperature) and high (dP/dT)sat values (i.e., producing stronger pressure impulses) have advantages in rapid initializations of PHPs. Working fluids with high specific and latent heats can absorb more heat at the hot end and efficiently transfer heat

condensation section evaporation section index of thermocouple saturation maximum minimum

to the cool end. As seen in Table 1, because pure acetone has a lower boiling point and a higher (dP/dT)sat value, it can help activate the PHP at a relatively lower heat input. The high specific and latent heat properties of pure water provide favorable thermodynamic characteristics for absorbing and efficiently carrying energy. Therefore, it can be deduced that these two materials are thermodynamically complementary. Due to the properties of zeotropic mixtures, the wateracetone mixture exists shifting concentrations at vapor–liquid equilibrium, unlike pure fluids. In this study, an experimental investigation of heat-transfer performances of a PHP using pure water, pure acetone and wateracetone mixtures of various mixing ratios (i.e., 13:1, 4:1, 1:1, 1:4 and 13:1) under a wide FR range (i.e., 3570%) and varying heat inputs (i.e., 10–100 W), will be carried out. Through analyzing and comparing the temperature oscillations and thermal resistances, start-ups and heat-transfer performances at low (i.e., 35% and 45%) and high FR (i.e., 62% and 70%) for PHPs running mixtures and pure fluids will be discussed. The results obtained in this study may provide insight for improving the heat-transfer performance of pulsating heat pipes and initialize new investigations on using mixtures as working fluids.

2. Experimental apparatus and data processing 2.1. Experimental apparatus As shown in Fig. 1, the experimental setup was composed of a PHP, the heating and cooling system, the charging and evacuating system and the data acquisition system. The PHP was axially oriented and was fabricated using pure copper with inner and outer diameters measuring 2 and 4 mm, respectively. The pipes were bent into several turns to form a closed loop structure with 10 parallel pipes, such that neighboring pipes were 20 mm apart as measured from the pipe centers. The PHP was sequentially divided into following sections according to functionality: a condensation section (applied via forced-air cooling), an adiabatic section (the transition section between the condensation and evaporation section) and an evaporation section (applied via a heat load), whose lengths were 80, 20 and 80 mm, respectively. The evaporation section was wrapped with a 0.2 mm diameter NiCr heating wire. The evaporation and adiabatic sections were installed in an insulation

Table 1 Physical properties of working fluids at standard atmospheric pressure [17]. Type of working fluid

Acetone Water

Boiling point TS

Density ql

°C

kg (m ) (20 °C)

Specific heat Cpl kJ (kg °C)1 (20 °C)

56.2 100.0

792 998

2.35 4.18

3 1

Conductivity kl

0.170 0.599

W (m °C) (20 °C)

1

Latent heat Hfg kJ kg1

(dp/dT)sat  103

523 2257

Pa °C (20 °C)

Dynamic viscosity g  106 Pa s (20 °C)

Surface tension r  103 N m1 (20 °C)

1.11 0.14

0.32 1.01

23.7 72.8

1

836

Y. Zhu et al. / International Journal of Heat and Mass Transfer 77 (2014) 834–842

Cooling system Axial flow fan (1.5 m/s)

PHP

Agilent 34970A

112 T2

TND-500VA

Data acquistion

Heating system

Fig. 1. Schematic of experimental setup.

chamber (i.e., an organic glass chamber) fabricated as a dual-layer structure with an aluminum foil-layer fixed on the internal wall to prevent dissipative heat loss (the overall heat loss was approximately 2.8%). The condensation section was placed into an air duct with a rectangular cross-section, and a fan at the bottom of the duct (providing an air flow velocity of 1.5 m/s). An Agilent 34970A with 34901A multi-channel module was used to record data at a 60 channels per second scanning velocity. In this study, 20 channels were activated; thus, these channels were scanned up to 3 times per second. According to the Nyquist–Shannon sampling theorem, the frequency for collecting temperature (3 Hz) is greater than twice the frequency of a temperature oscillation (0.333 Hz) [18]. Fig. 2 shows the placement of the thermocouples. In this study, twenty thermocouples (T-type, ±0.1 °C after calibration) were installed. Nineteen were attached on the outer pipe wall to monitor the temperatures of the PHP, and one thermocouple was fixed on the outer wall of the chamber to detect heat loss from the insulation chamber. Among the nineteen thermocouples, six were fixed on the turns for the condensation section, numbered 1–6, five were fixed on turns of the evaporation section, numbered 7–11, and the remaining thermocouples were fixed at the center of the pipe lengths, numbered 12–19. 2.2. Data processing The thermal resistance has been used as a crucial indicator of the heat-transfer performance of the PHP [3,19]. Here, the thermal resistance can be defined as the ratio between the average temperature difference of the evaporation section (Te, 6 measuring points) and condensation section (Tc, 5 measuring points) and the heat input ðQ_ Þ:

Tc ¼

6 1X Ti 6 i¼1

ð1Þ

Te ¼

11 1X Ti 5 i¼7

ð2Þ

R¼

Te  Tc Q_

ð3Þ

where Tc and Te are the averages of the temperatures of the corresponding points for the condensation section and evaporator, respectively, and Q_ is the heat load supplied by heating wires. To ensure the temperatures were at steady state, data were collected for the final 10 min of the experiment. The oscillating working fluids in the PHP required approximately 10 min to reach a steady state. Based on the following equations,

Q_ ¼ U  I; R¼

Te  Tc Q_

ð4Þ ð5Þ

standard uncertainties can be expressed:

sffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi  2  2 dU dI þ ; U I vffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi ! u 2  2 dR u dT e dT c dQ_ : ¼t þ þ R Te  Tc Te  Tc Q_ dQ_ ¼ Q_

ð6Þ

ð7Þ

A calibrated T-type thermocouple (±0.1 °C) was employed to measure the temperature and an Agilent 34970A (±0.0256 °C) unit was used to record data. The voltage and the current of the power

Y. Zhu et al. / International Journal of Heat and Mass Transfer 77 (2014) 834–842

U max ¼

837

dR  K ¼ 2:41%  2  4:82%: R

By checking the uncertainty for each heat input and working fluid, this is the maximum uncertainty for a small heating power and a low temperature difference. 3. Characteristics of the phase change of mixtures

(a) Size of the PHP and arrangement of thermocouples

(b) Cross-section of the PHP

In this study, the phase diagrams were plotted by Aspen plus (version 11.1), and the pressure and temperature were, respectively fixed at 1 atm and 25 °C (Fig. 3). Fig. 3(a) shows the phase diagram between the temperature and the mole fraction of the wateracetone mixture, where Tsat Component water and Tsat Component acetone are the saturation temperatures of water and acetone, respectively. The ‘b’ represents the mixture concentration before the phase change; during the phase change, mixture concentrations in the vapor phase (b2) and the liquid phase (b1) varies with the temperature. It can be found that there exists a concentration shift between the vapor and liquid phase for the component in the wateracetone mixtures, which is different from that of the pure fluid (i.e., at the endpoints of the curves). The bubble line represents the boiling point temperatures for different mixture ratios; the dew line represents the condensation point temperatures for different mixture ratios. All boiling and dew points of various mixtures are between saturation temperatures of pure water and acetone. Fig. 3(b) shows the phase diagram between the pressure versus the mole fraction of the wateracetone mixture, where Psat Component water and Psat Component acetone represent the saturation pressures of component water and acetone, respectively. Similar to Fig. 3(a), all saturation pressures of water–acetone mixtures are between that of pure water and acetone. Unlike pure water and acetone, the wateracetone mixture permitted a wider range of possible concentrations shifts between the vapor and liquid phase and wider ranges of saturation temperatures and pressures. It is probable that these features of wateracetone mixtures can be exploited to explore new heattransfer characteristics in a PHP. 4. Thermal behaviors of the PHP with pure fluids and mixtures 4.1. The Start-up performance of the PHP with pure fluids and mixtures

(c) Three dimensional diagram of the PHP Fig. 2. The schematic diagram of the PHP.

supply were both accurate to 0.5. At 10 W, the ranges of the voltage and current were 0–75 V and 0–1 A, respectively. Under these conditions, the expected uncertainty can be calculated: U = 31.07 V, I = 0.324 A, (Te  Tc)min = 12.6 °C (for pure water, FR = 62%). Then, the above equations can be resolved:

sffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi  2  2 0:5%  75 0:5%  1 þ ¼ 1:96%; 31:07 0:324 sffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi  2 dR 0:1 þ 0:0256 ¼ 2 þ ð1:96%Þ2  2:41%: R 12:6 dQ_ ¼ Q_

Given a coverage factor (K) of 2, the expected maximum uncertainty (Umax) is

Fig. 4 depicts temperature variations of the hot and cool ends of the PHP using pure water and the wateracetone mixture (i.e., a mixing ratio = 13:1), under an FR of 45% and a heating input of 10 W. The start-up performance evaluates the thermal response and heat-transfer performance under a relatively low heat load of the PHP, which has been studied in previous works [3,4]. For the PHP with a wateracetone mixture (mixing ratio = 13:1, shown in Fig. 4(a)), the temperatures of the hot and cool end initially increased with time after the heating power was applied to the PHP. While a few oscillations were observed, the hot and cool ends generally increased in temperature until 375 s (line A), indicating that the vaporliquid slugs in the pipes maintained a stagnant status with an occasional local oscillation. At 375 s, the heat could not be effectively transferred from the hot to cool end. Thus, the temperature difference of the two ends remained relatively large for the remainder of the run. Thereafter, abrupt temperature changes appeared simultaneously at both ends (e.g., 23.8–27.8 °C over 15 s for the cool end; 40.4–34.8 °C over 15 s for the hot end), indicating that the working fluid started to circulate in the PHP, whereby a large amount of heat

838

Y. Zhu et al. / International Journal of Heat and Mass Transfer 77 (2014) 834–842

110

b the original concentration b1 the concentration in the liquid phase b2 the concentration in the vapor phase

P = 1 atm

Temperature/

100

Tsat Component water

90 80

Dew line Bubble line

70 60 b1

b

Tsat Component actone

b2

50 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 13:14:1 1:1 1:4 1:13

1

Mole fraction (component actone)

a Temperature−Mole fraction (Component acetone) 35 Psat Component actone

30

Pressure/kPa

25

Bubble line

20 15 10

Dew line 5 Psat Component water

0 13:1 4:1 1:1

1:4

1:13

Mole fraction (Component actone)

b Pressure−Mole fraction (Component acetone) Fig. 3. Phase diagrams of wateracetone mixture.

absorbed at the hot end by a vapor–liquid slug was forced to the cool end to release the heat by another working fluid slug. Following the simultaneous temperature increases, the temperature curves of the two ends began to continuously oscillate at various amplitudes and frequencies, indicating that the PHP was initialized. In the PHP with pure water, shown in Fig. 4(b), there were temperature changes for the two ends, but they were comparatively slight (e.g., 32.6–37.4 °C over 16.18 s for the cool end; 40.51–40.08 °C over 35.5 s for the hot end). Instead of generating consistent oscillations as seen in the PHP with the water–acetone working fluid, the temperatures of the two ends continued to increase after the initial abrupt temperature changes. Although the vapor–liquid slugs were recirculating, the working fluid returned to a stagnant state because the generated impetus was not sufficient to overcome the resistance of the vapor–liquid slugs. Table 2 provides a summary of the start-up performance of the PHP. In contrast with the PHP with pure water, the PHP with the wateracetone mixture has a better start-up performance (except for the 4:1 water–acetone mixture) due to:

(a) The boiling points of water–acetone mixtures were lower than that of pure water, as shown in Fig. 3(b). These mixtures required less heat to reach their boiling points. (b) Due to the relatively low specific and latent heats of pure acetone (2.35 kJ (kg °C)1 and 523 kJ kg1 at 1 atm, respectively), compared with pure water, wateracetone mixtures have correspondingly lower specific and latent heats. Hence, for the same heat capacity, more energy was carried in water–acetone mixtures, causing stronger pressure impulses to push working fluids. 4.2. The heat-transfer performance of the PHP with low FR (i.e., 35% and 45%) Generally, under a low FR (i.e., 35% and 45%), dry-out likely occurs in the evaporation section at higher heat inputs (i.e., P50 W), leading to the deterioration of heat-transfer rates. Thus, the ability to prevent dry-out conditions determines the heattransfer performance of the PHP under low FR. In this study, experiments were compulsorily terminated when the temperature of the evaporation section exceeded 130 °C.

839

Y. Zhu et al. / International Journal of Heat and Mass Transfer 77 (2014) 834–842

45 Line A

40

Abrupt temperature changes

T/

35 Pre start-up (375 second)

30

Start-up (consistent oscillations)

25 20

T3

T8

15 0

200

400

600

800

1000

1200

t/s (a) 13:1 water−acetone mixture (FR = 45%) 55 Line B

50

Pre temperature changes (410 second)

45

Slight temperature changes

Failed to start up (no consistent oscillation)

T/

40 35 30 25 20 15 0

200

400

600

800

1000

1200

1400

1600

t/s (b) Pure water (FR = 45%) Fig. 4. Temperature versus time of the PHP using a water–acetone mixture and pure water.

Table 2 Start-up status for the PHP with various working fluids (i.e., FR = 35%, 45%, 62%).

Water Acetone 13:1 (Wateracetone) 4:1 (Wateracetone) 1:1 (Wateracetone) 1:4 (Wateracetone) 1:13 (Wateracetone)

35% Start-up (10 W)

45% Start-up (10 W)

 U U  U U U

 U U  U U 

Acetone Water/acetone 13:1

2

Water/acetone 4:1 Water/acetone 1:1

Note: The start-up status evaluates for the appearance of abrupt temperature changes followed by consistent oscillations.

R/(°C W−1)

Type

Water

2.5

Water/acetone 1:4

1.5

Water/acetone 1:13

1 0.5 0 0

20

40

60

80

100

Q/ W

Figs. 5 and 6 display the variations of thermal resistances of the PHP with wateracetone mixtures (at mixing ratios of 13:1, 4:1, 1:1, 1:4 and 1:13) and pure water and acetone under 35% and 45% FR. At a 35% FR, initially, the thermal resistances of wateracetone mixtures and pure fluids decreased with increasing heat input (1020 W) until the onset of dry-out. When the heat input exceeded 20 W, the thermal resistance of pure acetone started to increase from 0.87 (20 W) to 1.77 °C W1 (50 W). Similarly, the thermal resistances of pure water and the wateracetone mixture (mixing ratio = 13:1) also increased, indicating significant dry-out conditions in the pipes. On the onset of dry-out, the temperature difference between the hot and cool end widens, causing an increase in the thermal resistance. However, thermal resistances of most of the water–acetone mixtures (i.e., mixing ratios of 4:1, 1:1, 1:4, and 1:13) decreased from 33.6% to 68.9% as compared to pure fluids at the heat input of 50 W, indicating that there was only slight to no dry-out in the PHP. Furthermore, the PHP with a 1:13 wateracetone mixture could still safely operate and

Fig. 5. Thermal resistance versus heat input of the PHP for mixtures and pure fluids (FR = 35%).

maintain a low thermal resistance at a 65 W heat input (R = 0.81 °C W1). Similarly, for 45% FRs, the PHPs with pure fluids (i.e., pure water or pure acetone) and a 13:1 wateracetone mixture exhibited severe dry-out conditions. PHPs with other water–acetone mixture ratios (i.e., mixing ratios of 1:1, 1:4 and 1:13) maintained functions with relatively low thermal resistance under a 65 W heat input (R = 0.6, 0.5 and 0.55 °C W1, respectively). From the analysis of PHP performances at FRs of 35% and 45%, it can be concluded that working fluids comprising wateracetone mixtures at high acetone fractions (i.e., mixing ratios of 4:1, 1:1, 1:4 and 1:13) prevented the onset of dry-out when compared with pure water, pure acetone and low acetone fractions of water– acetone mixtures.

840

Y. Zhu et al. / International Journal of Heat and Mass Transfer 77 (2014) 834–842

Water Acetone Water/acetone 13:1 Water/acetone 4:1 Water/acetone 1:1 Water/acetone 1:4 Water/acetone 1:13

R/(°C W−1)

2 1.5

62% Test 1

2.5

62% Test 2

2 R/(°C W−1)

2.5

1

62% Test 3

1.5 1 0.5

0.5 0

0

40

60

80

40

100

Q/ W Fig. 6. Thermal resistance versus heat input of the PHP for mixtures and pure fluids (FR = 45%).

The improved ability against the onset of dry-out conditions of the water–acetone mixtures may be attributed to the following properties: (a) During the phase change process in water–acetone mixtures, acetone abundantly entered the vapor phase while water largely remained in the liquid phase (e.g., the mole fraction of acetone in the 13:1 water–acetone mixture reached 0.57 at 78.8 °C). In the PHP, due to the relatively high saturation pressure and (dP/ dT)sat value, vapor slugs were enriched with acetone vapor and functioned as the impetus to drive the working fluid flow, while water remained in the liquid phase to prevent dry-out in the evaporation section. (b) For the PHP with water–acetone mixtures because the component acetone possessed relatively high saturation pressure, the pressure of vapor slugs for water–acetone mixtures was higher compared to pure water, increasing the saturation temperature of the component water. Therefore, the component water could still maintained in the liquid phase at higher heat input (e.g., P50 W) to wet the evaporation section. The addition of a relatively low fraction of pure water into pure acetone (i.e., the 1:13 wateracetone mixture) improved the prevention of dry-out conditions in the PHP, as confirmed by the decreasing thermal resistances in the low FR (35% and 45%) and heat input of 50 W experiments (62.9% and 71.1% decreases, respectively, compared with pure acetone). For the 1:13 water– acetone mixture, the concentration of acetone in the vapor slugs was relatively high (e.g., the mole fraction of acetone in the vapor phase reached 0.92 at 57.26 °C). The saturation temperature for water was higher, enabling water to remain in the liquid phase at higher heat inputs (e.g., P50 W). Conversely, for low concentrations of acetone (e.g., the 13:1 wateracetone mixture) because the concentration of acetone was relatively low in the vapor phase (e.g., the mole fraction of acetone in the vapor phase reached 0.57 at 78.8 °C), the pressure of the vapor slugs was similar to that of the PHP with pure water, resulting in only a slight increase in the saturation of water. Thus the PHP performance of a 13:1 water– acetone mixture was similar to that of pure water, as observed in the onset of dry-out conditions. 4.3. The heat-transfer performance of the PHP with high FR (i.e., 62% and 70%) Fig. 7 shows the thermal resistance of the PHP with pure methanol at a 62% FR. Although the three tests were carried out at the same FR, heat input and working fluid, the curves were relatively different at the low heating input, illustrating the weak repeatability of the PHP, under high FRs (i.e., 62% and 70%) and low heat input (i.e., 620 W). This is likely attributed to the increased amount and random distribution of the working fluid. Hence,

60

80

100

Fig. 7. Thermal resistance versus heat input of the PHP for pure methanol (FR = 62%).

following discussions focused on PHP performance at higher heat input (i.e., P35 W). Figs. 8 and 9 show the variations in thermal resistances of PHPs with wateracetone mixtures (i.e., mixing ratios of 13:1, 4:1, 1:1, 1:4 and 1:13) and pure water and acetone under the FRs of 62% and 70%. The following observations can be made: (1) For PHPs using pure water and acetone, the thermal resistance of pure water and acetone were initially similar until the heat input exceeded 50 W. From 50 to 100 W, thermal

Water Acetone Water/acetone 13:1 Water/acetone 4:1 Water/acetone 1:1 Water/acetone 1:4 Water/acetone 1:13

1 0.8 R/(°C W−1)

20

20

Q/ W

0.6 0.4 0.2 0 35

55

75

95

Q/ W Fig. 8. Thermal resistance versus heat input of the PHP for mixtures and pure fluids (FR = 62%, heat input = 35100 W).

Water Acetone Water/acetone 13:1 Water/acetone 4:1 Water/acetone 1:1 Water/acetone 1:4 Water/acetone 1:13

1 0.8 R/(°C W−1)

0

0

0.6 0.4 0.2 0 35

55

75

95

Q/ W Fig. 9. Thermal resistance versus heat input of the PHP for mixtures and pure fluids (FR = 70%, heat input = 35100 W).

Y. Zhu et al. / International Journal of Heat and Mass Transfer 77 (2014) 834–842

resistances of pure acetone and water increased from 0.01 to 0.11 (62%) and 0.08 (70%) °C W1, respectively. The heattransfer performance of the PHP with pure water was better than that of the PHP with pure acetone. The improved heattransfer performance of the PHP with pure water could be attributed to the increasing ability to efficiently transfer energy (due to its higher specific and latent heats). At high FRs (i.e., 62% and 70%), because the onset chance of dryout is negligible, factors determining the heat-transfer performance of a PHP are primarily influenced by the ability to transfer the energy of working fluids and the flow velocity inside the PHP. According to the physical properties listed in Table 1, due to its low dynamic viscosity and high (dP/dT)sat value, pure acetone can flow rapidly in the PHP. Due to its relatively high specific and latent heats, pure water can efficiently transfer energy. For heat inputs from 35 to 50 W, the flow velocity and the capability to carry energy both exerted comparable influence on the heat-transfer performance. However, for higher heat input (i.e., P50 W), the PHP with pure water had higher heat-transfer performance, illustrating that, due to an inertial effect coupled with a decreasing viscosity effect at higher heat input, both water and acetone can achieve high flow velocities and thus, the increased ability to transfer energy and exert more significant impacts on the heat-transfer performance. (2) The thermal resistances of wateracetone mixtures (i.e., at mixing ratios of 13:1, 4:1, 1:1, 1:4 and 1:13) varied for heat inputs from 35 to 50 W. However, at higher heat input (i.e., P50 W), the thermal resistances of the mixtures appeared more uniform. Based on the earlier discussion in (1), the PHP with water–acetone mixtures, whose ability to carry energy varied with different mixing ratios, should also exhibit varied heat-transfer performances at higher heat input. However, the experimental results showed relatively similar PHP performances with the various water–acetone mixtures at higher heat input (i.e., P50 W). This indicates that the ability to transfer energy may play a relatively minor role in the heat-transfer performance of the PHP with water– acetone mixtures. (3) By comparing wateracetone mixtures (i.e., at mixing ratios of 13:1, 4:1, 1:1, 1:4 and 1:13) and pure fluids (i.e., pure water and acetone), at a heat input range from 35 to 50 W, the thermal resistances of the PHP with wateracetone mixtures were higher than that of pure fluids. In contrast with the minima of water–acetone mixture for certain heat inputs, the maximum thermal resistances of pure water and acetone decreased by 38.7% and 35.6%, respectively. Although the thermal resistance of pure acetone gradually approached that of the wateracetone mixtures at higher heating values (i.e., P50 W), the thermal resistance of pure water remained relatively low, with the thermal resistance decreasing 37–45.8% compared with the minima of mixtures. In general, at high FRs (i.e., 62% and 70%), heat-transfer performances of a PHP with wateracetone mixtures were weaker than that of the PHP with pure fluids. In summary, it can be deduced that the flow velocity of the working fluid fail to reach a sufficiently high rate at higher heat input, thereby decreasing the ability to transfer energy leading to a decrease in the heat-transfer performance. The low flow velocity is likely attributed to the following factors:  For the wateracetone mixture, during the process of phase change, the water fraction would increase with an increase of the heat input (e.g., the mole fraction of water for a 13:1 water–acetone mixture increased from 0.13 to 0.88 when the temperature increased from 57.2 to 67.3 °C). Due to the increasing fraction of water in the liquid slugs, the increasingly diluted

841

liquid acetone fraction would rarely contact the liquid–vapor interface, thereby decreasing the amount of acetone transferring to the vapor phase. Thus, the impetus to drive the flow decreased, leading to a relatively low flow velocity.  Compared with pure fluids, which can readily achieve a pressure equilibrium inside the PHP, water–acetone mixtures behave differently due to their zeotropic properties. Hence, less impetus was generated by the resultant variable pressure differences inside the PHP with the water–acetone mixtures, resulting in relatively low flow velocities. At high FRs, the PHP with pure water possessed improved heattransfer performance compared with that of the PHP with pure acetone. For water–acetone mixtures with various mixing ratios (i.e., 13:1, 4:1, 1:1, 1:4 and 1:13), the heat-transfer performances of the PHPs were similar at the higher heat input. By comparing the heat-transfer performances of the pure fluids and water–acetone mixtures, the PHP with the former as working fluids exhibited higher heat-transfer performance. 5. Conclusions In this study, an experimental investigation of a closed PHP with inner and outer diameters of 2.0 and 4.0 mm, respectively, using water–acetone mixtures (i.e., at mixing ratios of 13:1, 4:1, 1:1, 1:4 and 1:13) and pure water and acetone was carried out. Through a detailed analysis of temperature and thermal resistance variations of the mixtures and pure fluids, differences in the startup and heat-transfer performances between water–acetone mixtures and pure fluids were elucidated: (1) For start-up performances, PHPs with water–acetone mixtures (i.e., at mixing ratios of 13:1, 4:1, 1:1 and 1:4) were better than PHPs with pure water at FRs of 35%, 45% and 62%. At a heat input of 10 W, temperature variations of water–acetone mixtures (i.e., at mixing ratios of 13:1, 1:1, 1:4 and 1:13) were observed after an initial temperature increase at both the hot and the cool ends of the PHP. Consistent temperature oscillations were observed at both ends. For the PHP with pure water, there was no abrupt temperature variation and only a slight temperature change followed by inconsistent oscillations. (2) For low FRs (i.e., 35% and 45%), apart from low acetone fraction water–acetone mixtures (i.e., a 13:1 wateracetone ratio), water–acetone mixtures exhibited better performance against dry-out in the PHP compared with pure water and acetone working fluids. The PHP with water–acetone mixtures (i.e., at mixing ratios of 4:1, 1:1, 1:4 and 1:13) could still function well with comparatively low thermal resistances which decreased from 33.6% to 68.9% compared with pure working fluids at a heat input of 50 W. The addition of a fraction of water to pure acetone (i.e., at a 1:13 wateracetone ratio) was found to improve the performance against dry-out in the PHP, as confirmed by the decreased thermal resistances of the PHP of 62.9% and 71.1% compared with that of pure acetone for the FRs of 35% and 45% and heat input of 50 W, respectively. However, the addition of a fraction of acetone to pure water (i.e., at a 13:1 water–acetone ratio) did not improve the start-up performance. (3) With respect to high FRs (i.e., 62% and 70%), the heat-transfer performance of the PHP with pure fluid was superior to that of the PHP with water–acetone mixtures (at mixing ratios of 13:1, 4:1, 1:1, 1:4 and 1:13). For the range of heat input from 35 W to 50 W, thermal resistances of pure water and acetone were lower than those of the water–acetone

842

Y. Zhu et al. / International Journal of Heat and Mass Transfer 77 (2014) 834–842

mixtures. In contrast with the minima of water–acetone mixtures for certain heat inputs, the maximum thermal resistances of pure water and acetone lowered by 38.7% and 35.6%, respectively. Although the thermal resistance of pure acetone increased and approached that of the water– acetone mixture at higher heat input (i.e., >50 W), the PHP with pure water still maintained a low thermal resistance, which decreased from 37% to 45.8% compared with the minima of mixtures. Conflict of Interest None declared. Acknowledgement This study was supported by the National Natural Science Foundation of China under Grant No. 51076104. References [1] H. Akachi, Structure of a heat pipe. US Patent No. 4921041. [2] J. Qu, H.Y. Wu, P. Cheng, X. Wang, Non-linear analyses of temperature oscillations in a closed-loop pulsating heat pipe, Int. J. Heat Mass Transfer 52 (15–16) (2009) 3481–3489. [3] J. Qu, H.Y. Wu, P. Cheng, Start-up, heat transfer and flow characteristics of silicon-based micro pulsating heat pipes, Int. J. Heat Mass Transfer 55 (21–22) (2012) 6109–6120. [4] X.D. Liu, Y.P. Chen, M.H. Shi, Dynamic performance analysis on start-up of closed-loop pulsating heat pipes (CLPHP), Int. J. Therm. Sci. 65 (2013) (2013) 224–233. [5] M. Mameli, M. Marengo, S. Khandekar, Local heat transfer measurement and thermo-fluid characterization of a pulsating heat pipe, Int. J. Therm. Sci. 75 (2014) (2014) 140–152.

[6] M. Mameli, M. Marengo, S. Zinna, Numerical model of a multi-turn closed loop pulsating heat pipe: effects of the local pressure losses due to meanderings, Int. J. Heat Mass Transfer 55 (4) (2012) 1036–1047. [7] Z.Q. Long, P. Zhang, Heat transfer characteristics of thermosyphon with N2–Ar binary mixture working fluid, Int. J. Heat Mass Transfer 63 (2013) (2013) 204– 215. [8] K.M. Armijo, V.P. Carey, An analytical and experimental study of heat pipe performance with a working fluid exhibiting strong concentration Marangoni effects, Int. J. Heat Mass Transfer 64 (2013) (2013) 70–78. [9] A. Nuntaphan, J. Tiansuwan, T. Kiatsiriroat, Enhancement of heat transfer in thermosyphon air preheater at high temperature with binary working fluid: a case study of TEGwater, Appl. Therm. Eng. 22 (3) (2002) 251–266. [10] S. Raffaele, D.F. Nicola, F. Rainondo, A. Yoshiyuki, Heat pipes with binary mixtures and inverse Marangoni effects for microgravity application, Acta Astronaut. 61 (1–6) (2007) 16–26. [11] H. Jouhara, Z. Ajji, Y. Koudsi, H. Ezzuddin, N. Mousa, Experimental investigation of an inclined-condensation section wickless heat pipe charged with water and an ethanolwater azeotropic mixture, Energy 61 (2013) (2013) 139–147. [12] P.W. Wang, X.H. Chen, Z.H. Liu, Y.P. Liu, Application of nanofluid in an inclined mesh wicked heat pipes, Thermochim. Acta 539 (2012) (2012) 100–108. [13] H.R. Chu, G.Z. Xie, L. Liu, Discussion on thermodynamic properties of binary mixture as working fluid for pulsating heat pipe, Refrig. Air-Cond. 10 (6) (2012) 21–25 (in Chinese). [14] H.R. Chu, G.Z. Xie, L. Liu, Discussion on thermodynamic properties of binary mixture adaptation of pulsating heat pipe, Refrig. Air-Cond. 25 (3) (2011) 216– 219 (in Chinese). [15] W.X. Shi, W.Y. Li, L.S. Pan, X.F. Tan, H.T. Yun, Study on heat transfer properties of aqueous ethanol pulsating heat pipe, J. Mech. Eng. 47 (24) (2011) 117–121 (in Chinese). [16] G. Burban, V. Ayel, A. Alexandre, P. Lagonotte, Y. Bertin, C. Romestant, Experimental investigation of a pulsating heat pipe for hybrid vehicle applications, Appl. Therm. Eng. 50 (1) (2013) 94–103. [17] L. Zhu, J. Liu, B. Wang, Z.H. Wang, Principles of Chemistry Engineering, Petroleum Industry Press, 2007, pp. 520–521 (in Chinese). [18] J. Qu, Study on Operational Mechanism and Enhanced Heat Transfer of Micro/ Miniature Pulsating Heat Pipes and Micro-grooved Heat Pipes, Ph.D. Thesis, Shanghai Jiao Tong University, China, 2010. [19] K.H. Chien, Y.T. Lin, Y.R. Chen, K.S. Yang, C.C. Wang, A novel design of pulsating heat pipe with fewer turns applicable to all orientations, Int. J. Heat Mass Transfer. 55 (21–22) (2012) 5722–5728.