10th IHPS, Taipei, Taiwan, Nov. 6-9, 2011 Characteristics of an Open-Loop Pulsating Heat Pipe and Flow Visualization Using a Transparent Tube Koji Fumoto a, Masahiro Kawaji b and Tsuyoshi Kawanami c a Hirosaki University, 3 Bunkyo-cho, Hirosaki, 0368561, Japan b City College of New York, New York, NY 10032, USA c Kobe University, 1-1 Rokkodai-cho, Nada-ku, Kobe, 6578501, Japan Tel : +81-172-39-3676, Fax: +81-172-39-3676, E-mail:kfumoto@cc.hirosaki-u.ac.jp ABSTRACT This paper presents preliminary experimental results obtained by using a transparent tube to perform a visualization experiment for a pulsating heat pipe (PHP). The tube has internal and external diameters of 2.0 mm and 4.0 mm, respectively. The PHPs were made from Teflon tubes in the form of single serpentine channels. The number of channels in the PHP was fixed at 12. The working fluids employed were water, ethanol, and self-rewetting fluids. The experimental results indicate that performance depends on the thermophysical properties of working fluids and the initial movement of the PHP operation. Moreover, in the start-up period, no nucleation was observed, and the vapor plugs at the evaporator section were formed mainly by the breakup of liquid slugs. At a steady operation state, large amplitudes of self-sustained oscillations were observed. Keywords: Pulsating heat pipe, Open loop, Flow visualization, Transparent tube 1. INTRODUCTION Pulsating (or oscillating) heat pipes (PHPs) are one of the latest developments in the area of heat pipe technology (Akachi, H. et al., 1995). The PHP is a relatively new type of a heat transfer device, which may be classified into a special category of heat pipes. The working fluid in a PHP oscillates in its axial direction unlike a conventional heat pipe, where the working fluid inside the heat pipe circulates continuously by capillary forces between the heat source and the heat sink in the form of countercurrent flows. The PHP has the advantage of not needing a wick structure to transport the liquid. There is also no pump; therefore, the PHP is passive and needs no power for operation other than the rejected heat. Recently, some studies (Mishkinis, D. et al., 2006) have been conducted on enhanced heat pipes using special working fluids, such as nanofluids or low-boiling fluids. Riehl and Santos (2011) presented an investigation on nanofluids applied to LHPs. According to them, when using a nanofluid in a PHP, its thermal performance could be improved. Moreover, certain studies (Fumoto, K. et al., 2010) have presented self-rewetting fluids as working fluids for PHP. Studies have also been conducted on flow visualization for PHP. Fluid flow inside the heat pipe is very complex because of phase changes and interactions between the liquid slugs and the vapor plugs. For example, Tong et al. (2001) performed flow visualization for a closed-loop PHP. It was observed that during the start-up process, the working fluid oscillates with a large amplitude. Wook et al. (1999) conducted flow visualization for a PHP made from a brass plate and an acrylic plate. The above literature survey shows that most of the previous works focused on the thermal performance evaluation or simple flow visualization. Therefore, there are limited studies on the oscillating behavior in which a thermal performance enhancement of the PHP is performed. Based on the above situation, the main purpose of this study is to observe oscillating behavior using various working fluids in the open-loop PHP (OLPHP) that is made of a transparent Teflon tube. We have also clarified the heat transport characteristics using a self-rewetting fluid as the working fluid. 2. EXPERIMENTAL SETUP AND PROCEDURE 2.1. Experimental setup of PHP Figure 1 illustrates the experimental setup, which consists of an OLPHP, an electrical power input source, a temperature data logger, and a PC. As shown in Figure 2, the PHP assembly was made of a transparent Teflon tube having an external diameter of 4.0 mm and a wall thickness of 1.0 mm. Thus, the internal diameter of the capillary tube is 2.0 mm. The bending radius of - 190 -
the U-bend in the evaporator section and the inverted U-bend in the condensation section is 8.0 mm. These form 24 snake-shaped PHP structures. The total height of the PHP is 160 mm, and its total width is 120 mm. The height of the evaporator section and the condensation section are 70 mm each. To measure the temperatures of the heat pipes, six K-type thermocouples with a wire diameter of 0.1 mm were attached to the heat pipe surface at three locations in the evaporator section (T 1 T 3 ) and three locations in the condenser section (T 4 T 6 ). Additionally, the heater temperature (T h ) and air temperature (T air ) were measured. All the thermocouple signals were sampled and recorded by a PC-based data acquisition system at a frequency of 3 Hz. A three-way valve was attached to the heat pipe at the top to allow for vacuuming and filling of the heat pipe with a working fluid using a syringe. The PHP was oriented vertically with an aluminum block in which the cartridge heater was attached to the evaporator section at the bottom, and a cooling block was attached to the condenser section. Both the heater and the cooler blocks were attached to the aluminum fin with a thin layer of thermal grease. A transparent Teflon tube was sandwiched between the aluminum fins. Between the evaporator and the condenser sections was an adiabatic section of 20 mm. The entire heat pipe excluding the cooling fin was insulated with 10 mm insulated brick and polystyrene layers. A fan driven by a DC 12 V motor was placed at the back of the condenser section at different distances to provide cooling air flow (22ºC 25ºC) at a velocity of 0 5.2 m/s. The power supplied to the cartridge heater block was varied from 0 to 120 W. Three types of working fluids, i.e., methanol, deionized water, and a self-rewetting fluid were used in the present study. The self-rewetting fluid was prepared by adding 1-pentanol to deionized water to prepare an aqueous solution of specified mass concentrations. 2.2. Experimental procedure For charging a pulsating heat pipe to a specified fill ratio, the heat pipe was first vacuumed to a pressure of 92 kpa using a gauge, and by turning the three-way valve, a specified amount of the working fluid was injected using the syringe. The mass of the heat pipe was weighed using an electronic balance before and after the filling operation to confirm the fill ratio. The power supplied to the cartridge heater block was adjusted after a steady-state temperature condition was reached over the entire surface of the heat pipe. When either a dry-out condition or the evaporator section temperature of 110ºC was reached, switching off the heater terminated the experiment. Figure 1 Experimental apparatus Figure 2 Heat pipe Experiments were conducted for the evaporator temperature range of 25ºC 110ºC, a filling ratio of 0 50 vol%, and concentrations of 1-pentanol in the working fluid of 0.5 wt%. The measurement uncertainties in the present work are as follows. Temperature readings: ±2.5% Filling ratio: ±0.4% Self-rewetting fluid concentration: ±0.02% Heater power: ±1.0% Thermal resistance: ±3.7% 3. RESULTS AND DISCUSSION 3.1. Self-rewetting fluid - 191 -
The thermophysical properties of the working fluids are listed in Table 1. Pentanol (C5H12O) is an alcohol, and the surface tension of aqueous solutions of alcohols is known to show unusual variations with temperature. Figure 3 shows the variations in surface tensions of pentanol, water, ethanol, and aqueous pentanol (Vochten, R. et al, 1973; JSME Data book 1991; and Jasper, J. J., 1972) with temperature. The surface tensions of water, ethanol, and pure pentanol decrease monotonically with an increase in temperature; in contrast, the surface tension of aqueous pentanol decreases gradually with an increase in temperature, reaching a minimum at approximately 60ºC and increasing gradually at higher temperatures. As shown in Figure 4, the temperature coefficient of surface tension for the aqueous pentanol solution is positive and above 60ºC; because of this, the boiling behavior of the solution may be altered, and surface tension-driven convection may occur. Hence, this working fluid is expected to have a positive effect on the heat transport performance. σ [mn/m] Surface tension 80 70 60 50 40 30 20 Ethanol (9) (17) 10 10 20 30 40 50 60 70 80 90 Figure 3 Surface tension of water, ethanol, pentanol, and its aqueous solution (from reference 7, 8, and 9) Ordinary liquids Self-rewetting fluids Temperature Pure Water (17) Pentanol aq. sol. (7) (15) T [ ] Pentanol (8) (16) Low temperature liquid Flow direction Liquid (9) Figure 4 Schematic diagram of surface tension-driven convection Vapor High temperature liquid Heating wall Table 1 Properties of working fluids Molar mass [g/mol] Density [g/cm³] Boiling point [ ] Latent heat [MJ/kg] Solubility in water [g/ml] 3.2. Overview of capillary tube The inner PHP tube hydraulic diameters were very important parameters for oscillation. According to Akachi et al. (1996), to ensure that the fluid forms stable vapor and liquid slugs randomly inside the PHP, the chosen fluid should have properties satisfying the following criteria for the upper limit of the PHP diameter (d): σ d 2 ( ρ L ρ V )g 1 2 (1) where σ is the surface tension and g is the gravitational force. ρ L and ρ V are the mass densities of the fluid and vapor, respectively. The amount of liquid phase in the PHP depends on the fill ratio. It was observed that vapor-liquid slug train units were randomly distributed in the PHP. Figure 4 shows the coexistence of liquid slugs and vapor plugs of different lengths in their initial condition in the PHP. They were placed with vertical orientation before heat was applied. This was done because the force of surface Working fluid (at 20, 1atm) Water Ethanol Pentanol 18.02 46.7 88.2 0.998 0.7835 0.814 100 78.6 137.9 2.26 0.854 0.59 3.3/100 Figure 4 Slug-train unit in capillary tube (vertical orientation) tension was much higher than the gravitational force. 3.3. Flow pattern at evaporator region - 192 -
Figure 5(a) (c) shows a series of photographs when heat at 80 W is applied to the PHP. The fill ratio is 50 vol%. The parameters in these figures are the working fluids, i.e., (a) water, (b) ethanol, and (c) a self-rewetting fluid (C = 0.5 wt%). The white arrows indicate the directions of slug movement with respect to the time indicated. Figure 5(a) proves that the liquid slug settles into the short slug. The length of PHP liquid slugs is very short; it is less than around 0.3 mm. This phenomenon occurs because of the high surface tension of the working fluid. For ethanol (Figure 5(b)), the separation behavior of the liquid slug could not be observed. However, it was confirmed that minute droplets adhered to the internal capillary wall. The droplets on the internal wall are cleaned out with the transfer of the liquid slug (indicated with white arrows). The liquid slug becomes smaller with the transfer and cannot be finally maintained in the slug (225 ms). This is because of the low surface tension of ethanol. For the pentanol aqueous solution (Figure 5(c)) having a surface tension between that of water and ethanol, the liquid slug shows the characteristics of both ethanol and water. It was proven that a large number of bubbles were generated in the fluid slug because of the self-rewetting phenomena. According to Nishiguchi and Shoji (2009), when using the self-rewetting aqueous solution, a mode of boiling similar to micro-bubble emission boiling (MEB) is observed. Moreover, according to Wijk et al. (1956), the changes in the boiling and slug flow characteristics of the working fluid were observed after adding the self-rewetting fluid. The nucleation, growth, and departure of bubbles are affected by the behavior of the liquid on the heated surface. 0ms 50ms 100ms (a) Water 0ms 100ms 150ms (b) Ethanol 225ms 0ms 50ms 150ms 200ms (c) Pentanol aq. sol. (C=0.5wt%) α=50 Vol%, Q=80W Figure 5 Characteristics of liquid slug flow 3.4. Temperature distribution Figure 6 shows the temperature distribution obtained for several working fluids such as water, ethanol, and aqueous pentanol (0.5 wt%) when the fill ratio is 50 vol%. This figure also shows the temperature distribution of the voided PHP (0 vol%). The y- and x-axes show the temperature and location of the PHP, respectively. In general, the surface temperature of the PHP has a temperature gradient. However, in this experiment, a large temperature gradient was not observed for the PHP. This is because the thermal conductivity of the Teflon tube is quite low. In this figure, for the pentanol aqueous solution, the temperature difference between the heating area and the cooling area is smaller than that of other working fluids. In addition, it is understood that the temperature difference of water is greater than that of ethanol, probably because of the higher evaporative latent heat of the former. 150ms - 193 -
T [ ] 120 80 40 Heater 0 0 40 80 120 160 x [mm] Figure 6 Tube temperature distribution 4. CONCLUSIONS A series of experiments was performed to investigate the performance improvement resulting from the use of a self-rewetting fluid in the pulsating heat pipe. By adding pentanol to water to produce a self-rewetting working fluid, the effects of flow behavior on the working fluid and other parameters were investigated. The following conclusions can be drawn from the current experimental data: 1. The self-rewetting fluid improved the boiling and bubble-generation behavior in the evaporator section. Therefore, liquid slug oscillation in the pulsating heat pipe increases as compared to the water- or ethanol-filled PHP. 2. The temperature difference of the pentanol aqueous solution is smaller than that of ethanol and water. 3. It is possible to obtain frequency oscillation by a solution concentration using a self-rewetting fluid solution of 0.5 wt% or less as the working fluid. NOMENCLATURE Q=80W, α=50vol% : Ethanol : Water : Pentanol ( : 0Vol%) Cooler C : Mass concentration, wt% d : diameter, mm Q : Heater power, W T : Temperature, T : Temperature Difference, x : Position, mm α : Fill ratio, Vol% (=V liq /V tot ) σ : Surface tension, mn/m ρ : Density, kg/m 3 Subscripts : 1~6 : Thermocouple location air : Air h : Heater liq : Liquid L : Fluid tot : Total V : Vapor ACKNOWLEDGEMENT The authors would like to acknowledge the financial support given by a research grant from the Ministry of Education, Science, Sports and Culture, Grant-in-Aid for Scientific Research (C), 20560205, 2008. REFERENCES [1] Akachi, H., Motoya, S., and Maezawa, S., Thermal Performance of Capillary Tunnel Type Flat Plate Heat Pipe, Proc. of 7th Int. Heat Pipe Conference, Albuquerque, New Mexico, 1995. [2] Mishkinis, D., Corbierre, M. K., Wang, G., and Nikanpour, D., Nanofluids As Working Media for Loop Heat Pipes, Proc. of 36th Int. Conference on Environmental Systems (ICES), Norfolk, VA USA, 2006. [3] Riehl, R. R. and Nadjara, S., Water-Copper Nanofluid Application in an Open Loop Pulsating Heat Pipe, App. Thermal Eng., Available online, 2011. [4] Fumoto, K., Kawaji, M., and Kawanami, T., Study on a Pulsating Heat Pipe with Self-Rewetting Fluid, J. Electronic Packaging, vol. 132, No. 031005, pp. 1 4, 2010. [5] Tong, B. Y., Wong, T. N., and Ooi, K. T., Closed-Loop Pulsation Heat Pipe, App. Thermal Eng., vol. 21, pp. 1845 1862, 2001. [6] Wook, H. L., Hyun, H. L., Jeung, S. J., H. K., and Jong, S. K., Flow visualization of oscillating capillary tube heat pipe, 11th International Heat Pipe Conference, vol. 2, Japan, pp. 131 136, 1999. [7] Vochten, R. and Peter, G., Study of the Heat of Reversible Adsorption at the Air-Solution Interface, J. Colloind ano Interface Sci., 42, pp.320 327, 1973. - 194 -
[8] JSME Data Book: Thermophysical Properties of Fluid 4th Edition, 1991.(in Japanese) [9] Jasper, J. J., The Surface Tension of Pure Liquid Compounds, J. Physical and Chemical Reference Data 1, pp. 841 1009, 1972. [10] Akachi, H., Polásvek, F., and Svtulc, P., Pulsating Heat Pipes, Proc. of 5th Int. Heat Pipe Symp., pp. 208-217, 1996. [11] Nishiguchi, S. and Shoji, M., A Study of Critical Heat Flux of Butanol Aqueous Solution, Proc. ASME 2009 7th International Conf. on Nanochannels, Microchannels, and Minichannels, South Korea, pp. 43 48, 2009. [12] Wijk, W. R., Vos, A. S., and Stralen, S. J. D., Heat Transfer to Boiling Binary Liquid Mixtures, Chemical Engineering Science, vol. 5, pp. 68 80,1956. - 195 -