Cooling performance of flat plate heat pipes with different liquid filling ratios

International Journal of Heat and Mass Transfer 77 (2014) 874–882

Contents lists available at ScienceDirect

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

Cooling performance of flat plate heat pipes with different liquid filling ratios Jung-Shun Chen ⇑, Jung-Hua Chou Department of Engineering Science, National Cheng Kung University, Tainan 70101, Taiwan

a r t i c l e

i n f o

Article history: Received 11 September 2012 Received in revised form 21 May 2014 Accepted 10 June 2014

Keywords: Flat plate heat pipes Liquid filling ratio Maximum heat transport capability Thermal resistance Effective thermal conductivity Leakage

a b s t r a c t The effects of liquid filling ratios and leakage on the cooling performance of flat plate heat pipes (FPHPs) were examined experimentally in this study. With the size of 150 mm  50 mm  2.5 mm for all Al 6061 FPHPs filled with acetone (99.87% pure), the results showed that the one with the liquid filling ratio of 25% performed thermally the best. The corresponding maximum heat transport capability, minimum thermal resistance, and maximum effective thermal conductivity were about 47 W, 0.254 K/W, and 3150 W/m K, respectively. In contrast, improper vacuum and leakage would decrease the maximum thermal conductivities greatly to about 200–306 W/m K and 164 W/m K, respectively. The latter was similar to that of an aluminum block and performed the worst among all FPHPs. This situation should be avoided or carefully assessed as the maximum effective thermal conductivity decreased from the best by a factor of about 19.2. Ó 2014 Elsevier Ltd. All rights reserved.

1. Introduction Due to thinness in geometry, slim liquid crystal display TVs (LCDTVs) are becoming the main stream in TV markets because they occupy a small space and can be installed on a wall as wall decorations. For slim LCD-TVs to be achievable, a thin LCD module is essential. However, thickness reduction of present LCD modules always poses thermal issues, including thermal strain problems [1,2]. Therefore, heat pipe cooling modules are typically used in slim LCD-TVs to handle these issues for their smaller form factors. These heat pipe cooling modules have to be flat so that the thinness of LCD TVs could be accommodated. Conventionally, a flat heat pipe is made by flattening a round one. However, flattening has limitations because of material strength and wick structures. Hence, flat plate heat pipes (FPHPs) have been developed to resolve these limitations. The capillary structures of mesh, groove, and porous types are commonly used in conventional heat pipes. Their thermal performance is different. For horizontal installations, the thermal resistance in decreasing order is mesh, groove, and porous type. Taking a heat pipe of 150 mm in length (the length used in this study) as a specific example, the thermal resistances normalized by that of porous type are 1.36 and 1.68 for groove and mesh types, respectively. Moreover, gravity has a larger effect on the groove type heat pipes [3]. ⇑ Corresponding author. Tel.: +886 6 2757575 63342; fax: +886 6 2766549. E-mail address: [email protected] (J.-S. Chen). http://dx.doi.org/10.1016/j.ijheatmasstransfer.2014.06.029 0017-9310/Ó 2014 Elsevier Ltd. All rights reserved.

As conventional heat pipes, the capillary structures FPHPs are also mainly mesh, groove, and porous. Various techniques have been developed to fabricate them, such as diffusion wire bonding [4,5] for meshes, diamond cut [6] for microgrooves, etching for special grooves [7,8]. These methods generally can make smaller heat pipes and are suitable for cooling of electronic components. Mesh and groove capillary structures are popular for their relative simplicity in manufacturing and reasonable performance. In this regard, Lefevre et al. [9] showed that the maximal heat flux obtainable was 2 W/cm2 with a thermal resistance of 0.035 K/W for Tsat of 70 °C for their grooved Cu-methanol flat heat pipe. The results of Lefèvre et al. [10] indicated that the minimum thermal resistance was 3.7 times lower than that of the empty system for their grooved silicon–methanol flat heat pipe. Lips et al. [11] showed that a small vapor space thickness could reduce the thermal resistance the FPHP. Hence the filling ratio and vapor space thickness should be optimized. Lips et al. [12] showed the importance of liquid–vapor interface and the need for its accurate assessment. Moreover, Lips et al. [13] showed that dry out was not due to boiling but was a classical capillary limit. Lefèvre et al. [14] investigated a series of six different FPHPs made of either silicon or copper with different capillary structures and sizes. The results showed that 80 W of heat transfer could be achieved and new laws were in need for better understanding of the mechanisms involved. Lefèvre et al. [15] showed that the thermal performance FPHPs of the CuSn 325 square screen mesh and coarse screen mesh associated with rectangular grooves in methanol was not significantly

J.-S. Chen, J.-H. Chou / International Journal of Heat and Mass Transfer 77 (2014) 874–882

875

Nomenclature A Keff L Pi Qmax Rth Ta1 Ta2 Tc1

cross-sectional area of the FPHP, mm2 effective thermal conductivity, W/m K distance between Th1 point and Tc1 point, mm input heating power, W maximum heat transport capability, W thermal resistance, K/W temperature of the right adiabatic section, K temperature of the left adiabatic section, K representative temperature of the condenser section, K

different. Chien and Shih [16] studied the effect of mesh size, filling volume ratio, and inclination angle on the thermal resistance of their Cu-water FPHP. The results showed that a larger mesh size would yield a lower thermal resistance and the inclination angle had a larger effect on the condenser than on the evaporator. Hu and Jia [17] investigated the start-up performance of pulsating FPHPs. The results indicated that both heating power and filling ratio were important factors for the start-up of the pulsating FPHP. Wang et al. [18] presented the length effect of the evaporation and condensation sections on the thermal performance of their sintered porous FPHP. The results showed that the FPHP would dry out at a lower heating power with an increased condensation section length and would perform better when the condensation section length approached the evaporation section length. For FPHP applications in electronic cooling, Boukhanouf et al. [19] showed that the thermal spreading resistance of their Cu-water porous FPHP was about 40 times smaller than that of a solid copper block. This was considerably larger than that obtained by Lefèvre et al. [10]. For start-up and shut-down transient characteristics, Wang and Vafai [20,21] investigated the transient characteristics of their FPHP during start-up and shut-down operations. The results showed that the wick in the evaporator section created the largest thermal resistance; whereas the wick in the condenser section was second. Sonan et al. [22] predicted the transient performance of a FPHP by coupling the wall heat transfer calculation to the fluid flow in both vapor and liquid phases and demonstrated that the FPHP clearly worked as a very good thermal spreader. Harmand et al. [23] studied the transient thermal performance of the heat pipe used to cool electronic components in a starter–alternator and demonstrated that local hot points could be avoided even in complex and confined geometries. The above studies on FPHPs clearly indicates two key aspects of FPHPs. One is that they are good heat spreaders; the other is that their performance depends strongly on the geometry, capillary structure, and liquid filling ratios. The capillary structures reported for FPHPs in the literature mainly focused on sheets of meshes or grooves on one side. Hence, for practical considerations, this study examined the features of FPHPs with grooves on both sides made by Al extrusion because this technique could be readily adopted for mass production. In addition, the effect of leakage was also explored so that practical utilization of FPHPs could be properly addressed. All the Al FPHPs investigated had the same size and were filled with acetone. The main parameter investigated was the liquid filling ratio. The thermal performance was characterized by the thermal resistance, the effective thermal conductivity, and the maximum heat transport capability Qmax. Details are presented in the following sections. 2. FPHP fabrication The FPHP was fabricated as follows. Firstly, structural forming by extrusion of aluminum 6061 was performed to make the FPHP

Tcold Th1 Thot Tsat Vi Vo /i DT

cold temperature, K representative temperature of the evaporator section, K hot temperature, K saturation temperature, K liquid volume, cm3 channel space, cm3 liquid filling ratio for a single channel, % temperature different, K

with internal channels and associated capillary grooves. The extrusion process was tuned so that no cleaning was required for the FPHP. Secondly, the extruded FPHP was cut into a desired length and sealed at one end. Thirdly, acetone (chemical pure, 99.87% pure) was filled into the grooves to the required amount by a syringe with a volumetric scale. Fourthly, air in the FPHP was evacuated until the pressure was about 80 torrs. At that moment, the other end of the FPHP was sealed so that no extra outgassing was needed, thus completing the fabrication process. Lastly, the fabricated FPHP was set in a hot water tank with temperature of about 100 °C for reliability screening tests including leakage proof and quickness of heat transfer before characterizing the thermal performance. A typical fabricated FPHP is shown in Fig. 1(a) with a marked A–A cross sectional cut. The A–A cross-sectional cut is depicted in Fig. 1(b) which illustrates that the FPHP had fourteen channels across its width. Each channel had sixteen capillary grooves. The aspect ratio of each capillary groove was 2. The related sizes are summarized in Table 1. The liquid filling ratio (/i) is defined by Eq. (1):

/i 

Vi  100% Vo

ð1Þ

The channel volume Vo was fixed to 0.597 cm3 and /i was varied from 5% to 50% with an interval of 5%. That is, Vi was changed from 0.03 to 0.3 cm3 with an increment of 0.03 cm3 for each different /i test. Each FPHP had a specific /i. 3. Experimental approach A schematic diagram of the FPHP experimental set-up is shown in Fig. 2, (a) for the overall side view of the setup whereas (b) for the top view of the FPHP. The FPHP was placed horizontally on the heating and cooling modules. The FPHP was enclosed inside a temperature controlled space with a size of 1200 mm (length) by 500 mm (width) by 400 mm (height). The temperature inside the controlled space was automatically controlled at 40 ± 2°C by a cooling fan equipped with a temperature sensor. The contact thermal resistance between the FPHP and the heating module was reduced by both thermal grease and an insulation board under a pressure of 196 KN/m2 as shown in Fig. 2(a); similarly for the cooling module arrangement. The heating module was made of a Cu block (2 mm thick) which was heated by a resistance heater (0.25 mm thick) to provide input heating powers to the FPHP. The input heating power Pi was varied from 5 to 60 W with an increment of 5 W by adjusting the needed current and voltage to the heating element. The cooling module was made of an Al 6061 block which was cooled by a water cooling system. The water cooling system delivered cooling water (25 ± 1 °C) at a fixed flow rate of 1200 ± 20 cm3/min. The evaporation and condensation sections had the same size of 50 ± 1 mm (width) and 30 ± 1 mm (length). K-type thermocouples with an uncertainty of about ±1 °C were used to measure the FPHP wall

876

J.-S. Chen, J.-H. Chou / International Journal of Heat and Mass Transfer 77 (2014) 874–882

50mm

150mm

(a) An actual FPHP (top view)

(b) A-A cross-sectional view Fig. 1. Geometry of the FPHP.

Table 1 Sizes of the FPHP. FPHP size

Length 150 mm

Width 50 mm

Thickness 2.5 mm

Channel size

Width 3 mm

Height 1.7 mm

Number 14

Groove size

Depth 0.4 mm

Gap 0.2 mm

Pitch 0.2 mm

temperatures at eight positions as shown in Fig. 2(b). Temperature data were taken continuously by a PC data acquisition system throughout the experiment. The thermal resistance, effective thermal conductivity, hot and cold temperature difference, and maximum heat transport capability adopted for characterizing the performance of FPHPs are defined as follows. (1) Thermal resistance, Rth The thermal resistance is given by Eq. (2). The error was about ±0.009 K/W.

Rth ¼

T h1  T c1 Pi

ð2Þ

(2) Effective thermal conductivity, Keff The effective thermal conductivity Keff is defined by Eq. (3):

K eff ¼

L Rth  A

ð3Þ

Here L was fixed and equal to 100 mm. The error of Keff was about ±13.3 W/m K. (3) Temperature difference, DT The performance of FPHPs will deteriorate drastically when dry out occurs and will result in a significant temperature increase in the evaporation section. Thus, it is important to monitor the possibility of dry out. For this purpose, the temperature difference given by Eq. (4) was applied.

DT ¼ T hot  T cold

ð4Þ

The hot and cold temperatures Thot and Tcold at appropriate locations will be specified later. (4) Maximum heat transport capability, Qmax The maximum heat transport capability, Qmax, is the limit of the dissipated power which can be removed by a FPHP. It can be deduced from the variation of thermal resistance of the FPHP as shown in Fig. 3. The state of Qmax will be reached when the condensate cannot return to the evaporation section in time with an adequate quantity. This situation will result in an increase in thermal resistance due to reduced thermal performance. Hence, by sensing the characteristic change of the thermal resistance, Qmax can be obtained. The two dash lines in Fig. 3 are the respective regression curves through the thermal resistance data of the lower and higher input heating powers. These two curves will intersect with each other as shown in Fig. 3. The input heating power at the intersection point was taken as Qmax in this study and the error was about ±2.5 W.

877

J.-S. Chen, J.-H. Chou / International Journal of Heat and Mass Transfer 77 (2014) 874–882

C abinet s ize : 1200mm X 500mm X 400mm Ambient controlled by F an venlaon

T hermocouple wire

25mm 15mm

70mm

15mm

25mm

196K N/m 2 board s ize : 10mm 50mmX 30mmX 10mm

196K N/m 2 F P HP

board

15mm

PC 27.7mm

30mm

30mm P ump

board

C u plate (thicknes s : 2mm) electric heang Ins ulaon P laorm

Water C ooler

C ooler s ize: 50mmX 30mmX 15mm

Heater s ize: 50mmX 30mmX 0.25mm

C ooler Ins ulaon P laorm s ize: 50mmX 30mmX 15mm

Heater Ins ulaon P laorm s ize: 50mmX 30mmX 27.7mm

Air out

Air in, 25ഒ

P ower S upply

Heat exchanger

(a) Measurement setup (side view)

(b) FPHP geometry and thermocouple locations (top view) Fig. 2. Schematic diagrams of the experimental measurement setup.

4. Results and discussion 1.50

4.1. Effect of liquid filling ratio 1.25

Rth(K/W)

1.00

0.75

Qmax

0.50

0.25

0.00 25

30

35

40

45

50

55

60

Pi (W) Fig. 3. Determination of maximum heat transport capability Qmax.

65

For the effect of liquid filling ratio (/i), steady state conditions of the FPHP were examined. The differences in temperatures Th1, Th2, and Th3 were within 1 °C, and a similar situation also held for the temperatures Tc1, Tc2, and Tc3. Hence, the measured variations of Th1 of the evaporation section and Tc1 of the condensation section vs. input heating powers were used as representatives for thermal analysis. The trends of Th1 vs. input heating powers under various /i are shown in Fig. 4(a). Three features can be observed. Firstly, for both cases of /i = 5% and 10%, Th1 increased linearly with increasing heating powers, an indication of insufficient amount of filled liquid in the FPHP. The measurements were terminated at the heating power of 25 W for which Th1 was about 100 °C for these two cases. Secondly, at the same input heating powers, Th1 decreased as /i increased from 5% to 25%. This indicates that more filled acetone would result in more vaporized acetone to absorb more heat by phase change and to transport more dissipated heat. The condition of /i = 25% gave the best thermal performance as the temperature

J.-S. Chen, J.-H. Chou / International Journal of Heat and Mass Transfer 77 (2014) 874–882

450

φi

440

5% 10% 15% 20% 25% 30% 35% 40% 45% 50%

430 420 410 400

Th1 (K)

390 380 370 360 350 340 330 320 310 300 0

5

10

15

20

25

30

35

40

45

50

55

60

65

Pi (W) (a) Th1 temperature profiles versus input power (uncertainty about ±1 324

5% 10% 15% 20% 25% 30% 35% 40% 45% 50%

320 318 316

Temperature (K)

)

φi

322

314 312 310 308 306 304 302 300 298

Evaporator

Condenser

296 0

15

30

45

60

75

90

105

120

135

150

Position (mm) (b) Temperature distributions along the FPHPs with Pi = 5 W (uncertainty about ±1 400

)

φi

390

15% 20% 25% 30% 35% 40% 45% 50%

380 370

Temperature (K)

878

360 350 340 330 320 310

Evaporator

Condenser 300 0

15

30

45

60

75

90

105

120

135

150

Position (mm) (c) Temperature distributions along the FPHPs with Pi = 45 W (uncertainty about ±1 Fig. 4. Temperature distributions of FPHPs.

)

J.-S. Chen, J.-H. Chou / International Journal of Heat and Mass Transfer 77 (2014) 874–882

was the lowest for all the heating conditions. Thirdly, for /i in the range from 30% to 50%, the values of Th1 were greater than those of /i = 25%. In other words, the heat transfer capability of the FPHP with further increased /i in this latter range was not as good as that of /i = 25%, an indication of incomplete vaporization due to over filled liquid. That is, filling more liquid than needed for a particular heating condition would lead to an increased thermal resistance and resulted in a reduction in the heat transport capability. Whether the FPHP functioned properly as a phase change device or not could be revealed more clearly by the temperature distribution along its length. For simplicity, two cases of Pi being 5 and 45 W were taken as examples for demonstration. The corresponding measured temperature distributions are displayed in Fig. 4(b) and (c). For these two figures, the horizontal axis represents the distance from the condenser section to the evaporator section of the FPHP. The vertical axis represents the temperatures Tc1, Ta2, Ta1, and Th1, along the FPHP at the locations of 25, 40, 110, and 125 mm from the beginning side of the condenser section, respectively as shown in Fig. 2(b). For a properly functioned FPHP, Ta2, Ta1, and Th1 they should be close together. From the variations of Tc1, Ta2, Ta1, and Th1 shown in Fig. 4(b), it is clear that for all the liquid filling ratios from 5% to 60% at the low power of Pi = 5 W, none of the FPHPs performed as an effective phase change cooler. In instead, they acted

more like a thermal conductor and were acceptable because the evaporator temperatures were reasonable. In contrast, the results for Pi = 45 W as in Fig. 4(c) show that the FPHP with /i = 25% functioned completely as an efficient phase change heat spreader and performed the best. The FPHPs with /i = 20%, 30%, and 35% acted approximately as phase change cooling devices but with higher evaporator temperatures which might not be desirable. The differences between Ta1 and Th1 for /i = 5% and 10% of Fig. 4(b) for Pi = 5 W and /i = 15% of Fig. 4(c) for Pi = 45 W were 5.3, 4, and 41.6 °C, respectively. This temperature rise in the evaporation section was caused by insufficient filling of liquid which in turn resulted in inadequate condensate for heat removal. Thus, the vapor in the evaporation section would be heated continuously by the input heating power until thermal equilibrium being reached at a higher temperature. Hence, for Pi = 45 W, the temperature rise was considerably larger than those of Pi = 5 W. In contrast, with sufficient condensate in the evaporation section, the FPHP would behave as an efficient phase change cooler and vapor would not be over heated; thus resulted in a flat temperature distribution. Therefore, the right amount of liquid filling ratio was essential to the performance of FPHPs at higher input heating powers. To further quantify the thermal performance of the FPHPs with different /i, the related thermal resistance (Rth) and effective

4.0

Pi 5W 10W 15W 20W 25W 30W 35W 40W 45W 50W 55W 60W

3.5 3.0

Rth (K/W)

2.5 2.0 1.5 1.0 0.5 0.0 5%

10%

15%

20%

25%

30%

35%

40%

45%

50%

φi

(a)Thermal resistance profiles (uncertainty about ± 0.009 K/W) 3250

Pi

3000

5W 10W 15W 20W 25W 30W 35W 40W 45W 50W 55W 60W

2750 2500 2250

Keff (W/mK)

2000 1750 1500 1250 1000 750 500 250 0 5%

10%

15%

20%

879

25%

30%

35%

40%

45%

50%

φi

(b) Effective thermal conductivity profiles (uncertainty about ± 13.3 W/mK) Fig. 5. Rth and Keff profiles of FPHP.

880

J.-S. Chen, J.-H. Chou / International Journal of Heat and Mass Transfer 77 (2014) 874–882

thermal conductivity (Keff) were computed using Eqs. (2) and (3), together with the data displayed in Fig. 4. The results are illustrated in Fig. 5. From the variations of Rth shown in Fig. 5(a), two features can be noticed. One is that Rth reduced first and then increased as /i increased from 5% to 50% as expected for the reasons described in the first paragraph of this section. This trend was independent of input heating powers. The minimum Rth occurred around /i = 25% for Pi from 5 to 45 W, then shifted to larger /i for Pi larger than 45 W. The other is that as the input heating power increased, Rth decreased until Pi = 45 W; afterward, Rth increased as Pi increased further. That is, at Pi = 45 W all FPHPs with different /i had the smallest Rth. Hence, for the present FPHP condition, the maximum heat transfer was about 45 W with the minimum thermal resistance of about 0.254 K/W. The variation of Keff shown in Fig. 5(b) exhibited essential the same behavior of that of Rth with the best achievable Keff of 3150 W/m K under the condition of /i = 25% and Pi = 45 W.

335 330

Temperature (K)

325 320 315 310

Th1

305

Tc1 Ta1

300

Ta2 295 0

100

200

300

400

500

600

700

800

900

1000

Time (sec)

4.2. Thermal performance of the FPHP with liquid filling ratio of 25%

(a) Temperature profiles

Because the case of /i = 25% generally gave the best performance, its thermal features were further analyzed to determine the maximum heat transport capability Qmax. For this purpose, the variations Th1, Tc1, Ta1, and Ta2 vs. time are depicted in Fig. 6. It can be observed from Fig. 6(a) that these four temperatures increased rapidly to approach their respective steady state values. The corresponding temperature difference are shown in Fig. 6(b) for which DT1 = Th1  Tc1 and DT2 = Ta1  Ta2. The former was for identifying the condition of phase change and the latter was for the effectiveness of the adiabatic section of the FPHP. The results show that that DT1 increased first to 14.5 K at 100 s and then dropped noticeably afterward. This observable turning point corresponded to the phase change condition of the FPHP and revealed that the saturation temperature was about 318.35 K as deduced from Th1 shown in Fig. 6(a). Moreover, DT2 maintained relatively constant at 1 K, implying a good adiabatic condition of the adiabatic section as it should be. Using the method illustrated in Fig. 3 and the thermal resistance variation of /i = 25% for which the thermal performance was the best, the maximum heat transport capability Qmax deduced was about 47 W, slightly larger than 45 W illustrated in Fig. 5. For comparison under a comparable input heating power, it is interesting to note that the obtained minimum thermal resistance

20

ΔT1

18

ΔT2

16 14

10 8 6 4 2 0 0

100

200

300

400

500

600

700

800

900

1000

Time (sec) (b) Temperature difference profiles Fig. 6. Temperature and temperature difference of the FPHP (/i = 25%).

liquid ratio 5 % liquid ratio 10 % liquid ratio 25 % liquid ratio 25 % without vacuum one end broken Al 6061 block Cu block

460 450 440 430 420 410 400 390

Th1(K)

ΔT (K)

12

380 370 360 350 340 330 320 310 300 0

5

10

15

20

25

30

35

40

45

50

55

60

65

Pi (W) Fig. 7. Th1 distributions under imperfect situations.

J.-S. Chen, J.-H. Chou / International Journal of Heat and Mass Transfer 77 (2014) 874–882

881

5.5 5.0

liquid ratio 5% liquid ratio 10% liquid ratio 25% liquid ratio 25% without vacuum one end broken Al 6061 block Cu block

4.5 4.0

Rth (K/W)

3.5 3.0 2.5 2.0 1.5 1.0 0.5 0.0 0

5

10

15

20

25

30

35

40

45

50

55

60

65

Pi (W)

(a)Thermal resistance profiles 3250

liquid ratio 5% liquid ratio 10% liquid ratio 25% liquid ratio 25% without vacuum one end broken Al 6061 block Cu block

3000 2750 2500 2250

Keff (W/mK)

2000 1750 1500 1250 1000 750 500 250 0 0

5

10

15

20

25

30

35

40

45

50

55

60

65

Pi (W)

(b) Effective thermal conductivity profiles 425 400 375

liquid ratio 5% liquid ratio 10% liquid ratio 25% liquid ratio 25% without vacuum one end broken Al 6061 block Cu block

350

Keff (W/mK)

325 300 275 250 225 200 175 150 125 0

5

10

15

20

25

30

35

40

45

50

Pi (W)

(c) Partially enlarged portion of (b) Fig. 8. Rth and Keff profile curve for FPHP of various situations.

of the present was close to that of Lips et al. [11] in which Cu with n-pentane was used. But the corresponding liquid filling ratios were not the same, larger for the present study. This might be caused by the differences in the aspect ratio and gap of the groove

capillary structure. The aspect ratio and gap of the present study were twice and half of those of Lips et al., respectively. The differences also indicated the importance of proper liquid filling for heat pipes.

882

J.-S. Chen, J.-H. Chou / International Journal of Heat and Mass Transfer 77 (2014) 874–882

4.3. Imperfection effect In the fabrication process, two key steps which might deteriorate the performance of FPHPs were inadequate vacuum and improper sealing. Thus, their effects on the thermal performance were examined using two extreme situations; namely, no vacuum for the former, and broken one end to induce leakage for the latter. The variations of Th1 and Tc1 vs. input heating powers are shown in Fig. 7(a) and 7(b), respectively. In both figures, the results for /i = 5%, 10%, and 25% which were illustrated already in Fig. 4 are also included for comparison. The results show that Th1 was the largest for the case with one end broken, even higher than that of the case with an Al 6061 block heat sink. This was reasonable as the original space in the FPHP filled by liquid was replaced by air having a lower thermal conductivity. For the case of /i = 25% without vacuum, its Th1 was close to that of /i = 5% when Pi 5 15 W and approached that of /i = 10% at Pi = 25 W. But afterward it increased. That is, for lower input heating powers, the heat pipe behaved like an Al heat sink and phase change took place only at higher input heating powers. However, without vacuum, Th1 was always greater than that with a Cu block. This indicated that the FPHP with the Cu block heat sink performed better thermally than that with optimum liquid filling ratio of 25% without vacuum. It is also interesting to note that Th1 increased linearly as the input heating power increased for the FPHPs with /i = 5% and 10%, and with the Al and Cu block heat sinks. In contrast, Th1 varied piecewise linearly and nonlinearly for /i of 25% without and with vacuum, respectively. Namely, for FPHPs with either small liquid filling ratios or without vacuum, they acted more like metal heat sinks as shown by their linearly steady state temperature distributions. From Th1 given in Fig. 7 and measured Tc1, the associated Rth and Keff for imperfection situations were computed. The results are displayed in Fig. 8. Also for the purpose of comparison, Rth an Keff of aluminum and copper blocks are also included in the figure, in addition to those of /i = 5%, 10%, and 25% FPHPs. It can be observed that Rth for the case of /i = 25% without vacuum was about 10 times that with vacuum, and Keff was about 200–306 W/m K which was slightly larger than the thermal conductivity of aluminum as vaporization would occur only at higher input heating powers. For the leakaged FPHP with one end open, Rth was even larger than that without vacuum described above, and Keff was only about 164 W/m K, similar to that of aluminum 6061. In other words, leakage had a large adverse effect on the thermal performance of the heat pipe and should be carefully checked and corrected because Keff could be reduced by a factor of about 19.2 from the best value. If not, the leakage effect should be considered as the worst case and assessed with Keff of about 164 W/m K in practical designs. Fig. 8 also shows an interesting point at Pi = 5 W. At this input heating power, the Cu block, with Rth and Keff being about 2.02 K/ W and about 396 W/m K respectively, performed slightly better than the best FPHP of /i = 25% by a factor of about 1.08 in Keff. In contrast, when the FPHP behaves as an effective cooler, Keff was about 3150 W/m K, a factor of about 7.95 to that of Cu. Hence, proper functioning of heat pipes was essential to their successful applications. However, this factor of 7.95 was still much smaller than 40 obtained by the porous capillary structure of Boukhanouf et al. [19] and would deserve further attentions on the effect of mixed capillary structures. 5. Conclusions In this study, experiments were conducted to investigate the thermal performance of FPHPs with different liquid (acetone) filling ratios. Key findings were as follows:

(1) The optimal liquid filled ratio of the present FPHP was 25%, and the resulted maximum heat transport was about 47 W. The corresponding thermal resistance and effective thermal conductivity were 0.254 K/W and 3150 W/m K, respectively. Too much or insufficient filling of filled liquid would reduce the effective thermal Conductivity of the FPHP considerably. (2) The thermal conductivity of the FPHP without vacuumed was from about 200–306 W/m K; whereas that with leakage due to one end open was about 164 W/m K. Both situations greatly deteriorate the thermal performance of FPHPs and should be avoided. Conflict of interest None declared. References [1] S.S. Choi, H.C. Bae, W.J. Kim, J.C. Choi, B.C. Yang, E.J. Kang, T.S. Jang, Ultra-slim TV module technology, SID Symp. Dig. Tech. Papers 48 (3) (2009) 720–722. [2] X.P. Li, N.N. Gong, Y.J. Guan, G.M. Cheng, Thermal and stress analysis of rapid electric heating injection mold for a large LCD TV panel, Appl. Therm. Eng. 31 (17–18) (2011) 3989–3995. [3] B. Zohuri, Heat Pipe Design and Technology – A Practical Approach, CRC Press, Roca Raton, 2011, p. 207. [4] S. Launay, V. Sartre, M.B.H. Mantelli, K.V. de Paiva, M. Lallemand, Investigation of a wire plate micro heat pipe array, Int. J. Therm. Sci. 43 (5) (2004) 499–507. [5] Y.X. Wang, G.P. Peterson, Analysis of wire-bonded micro heat pipe arrays, J. Thermophys. Heat Transfer 16 (3) (2002) 346–355. [6] D. Plesch, W. Bier, D. Seidel, K. Schubert, Miniature heat pipes for heat removal from microelectronic circuits, Micromech. Sens. Actuators Syst. ASME-DSC 32 (1991) 303–314. [7] G.P. Peterson, A.B. Duncan, M.H. Weichhold, Experimental investigation of micro heat pipes fabricated in silicon wafers, J. Heat Transfer 115 (3) (1993) 751–756. [8] B. Badran, F.M. Gerner, P. Ramadas, T. Henderson, K.W. Baker, Experimental results for low-temperature silicon micromachined micro heat pipe arrays using water and methanol as working fluids, Exp. Heat Transfer 10 (4) (1997) 253–272. [9] F. Lefevre, R. Rulliere, G. Pandraud, M. Lallemand, Prediction of the temperature field in FPHP with micro-grooves – experimental validation, Int. J. Heat Mass Transfer 51 (15–16) (2008) 4083–4094. [10] F. Lefevre, R. Rullière, S. Lips, J. Bonjour, Confocal microscopy for capillary film measurements in a flat plate heat pipe, J. Heat Transfer 132 (3) (2010) 1–6. [11] S. Lips, F. Lefèvre, J. Bonjour, Combined effects of the filling ratio and the vapour space thickness on the performance of a flat plate heat pipe, Int. J. Heat Mass Transfer 53 (4) (2010) 694–702. [12] S. Lips, J. Bonjour, F. Lefèvre, Investigation of evaporation and condensation processes specific to grooved flat heat pipes, Front. Heat Pipes 1 (2) (2010) 1–8. [13] S. Lips, F. Lefevre, J. Bonjour, Physical mechanisms involved in grooved flat heat pipes: experimental and numerical analyses, Int. J. Therm. Sci. 50 (7) (2011) 1243–1252. [14] F. Lefevre, S. Lips, R. Rulliere, J.B. Conrardy, M. Raynaud, J. Bonjour, Flat plate heat pipes: from observations to the modeling of the capillary structure, Front. Heat Pipes 3 (1) (2012) 1–9. [15] F. Lefevre, J.B. Conrardy, M. Raynaud, J. Bonjour, Experimental investigations of FPHP with screen meshes or grooves covered with screen meshes as capillary structure, Appl. Therm. Eng. 37 (2012) 95–102. [16] L.H. Chien, Y.C. Shih, An experimental study of mesh type flat heat pips, J. Mech. 27 (2) (2011) 167–176. [17] C.F. Hu, L. Jia, Experimental study on the startup performance of flat plate pulsating heat pipe, J. Therm. Sci. 20 (2) (2011) 150–154. [18] S. Wang, J. Chen, Y. Hu, W. Zhang, Effect of evaporation section and condensation section length on thermal performance of flat plate heat pipe, Appl. Therm. Eng. 31 (14–15) (2011) 2367–2373. [19] R. Boukhanouf, A. Haddad, M.T. North, C. Buffone, Experimental investigation of a flat plate heat pipe performance using IR thermal imaging camera, Appl. Therm. Eng. 26 (17–18) (2006) 2148–2156. [20] Y. Wang, K. Vafai, An experimental investigation of the transient characteristics on a flat-plate heat pipe during startup and shutdown operations, J. Heat Transfer – Trans. ASME 122 (3) (2000) 525–535. [21] Y. Wang, K. Vafai, An experimental investigation of the thermal performance of an asymmetrical flat plate heat pipe, Int. J. Heat Mass Transfer 43 (19) (2000) 2657–2668. [22] R. Sonan, S. Harmand, J. Pelle, D. Leger, M. Fakes, Transient thermal and hydrodynamic model of flat heat pipe for the cooling of electronics components, Int. J. Heat Mass Transfer 51 (25–26) (2008) 6006–6017. [23] S. Harmand, R. Sonan, M. Fakes, H. Hassan, Transient cooling of electronic components by flat heat pipes, Appl. Therm. Eng. 31 (11–12) (2011) 1877– 1885.