Thermal performance of thermosyphons in series connected by thermal plugs

Experimental Thermal and Fluid Science 88 (2017) 409–422

Contents lists available at ScienceDirect

Experimental Thermal and Fluid Science journal homepage: www.elsevier.com/locate/etfs

Thermal performance of thermosyphons in series connected by thermal plugs C. Tecchio a, J.L.G. Oliveira a,⇑, K.V. Paiva a, M.B.H. Mantelli a, R. Gandolfi b, L.G.S. Ribeiro b a b

Mechanical Engineering Department, Lepten/Labtucal, Federal University of Santa Catarina, Florianópolis, SC 88040-900, Brazil Embraer S. A., São José dos Campos, SP 12227-901, Brazil

a r t i c l e

i n f o

Article history: Received 20 December 2016 Received in revised form 27 June 2017 Accepted 27 June 2017 Available online 28 June 2017 Keywords: Serial thermosyphons Thermal performance Dryout Filling ratio Aspect ratio

a b s t r a c t Thermal performance of a novel heat exchanger setup consisting of two thermosyphons connected in series was investigated. An intermediary heat transfer element (IHTE) transported heat from a fictitious component to be cooled to the evaporator of a second thermosyphon. Cylindrical and conical shaped couplings were used as the IHTE condenser geometry. The effects of IHTE filling ratio, inclination angle, aspect ratio and coupling geometries on the system thermal performance were evaluated. While increasing filling ratios promote increasing operation temperatures, decreasing filling ratios facilitate dryout. The critical input heat flux due to dryout decreases with decreasing thermosyphon inclination angle. Higher aspect ratio values imply higher operation temperatures. Both conical and cylindrical IHTE condensers were demonstrated as possible fitting elements for thermal couplings. Ó 2017 Elsevier Inc. All rights reserved.

1. Introduction Recent advances in electric-powered equipment have raised the demand for novel thermal control systems in compact, lightweight and high-performance devices. Moreover, the trend of miniaturization increases the need of developing reliable thermal management systems, capable of dissipating high heat transfer rates per area. Conventional cooling mechanisms by forced convection have been widely used to cool down electronic equipment. However, size, periodical maintenance, power consumption and low cooling efficiency are significant penalties when this approach is used. These drawbacks have motivated the introduction of novel cooling systems based on heat pipe technology. Thermosyphons are wickless gravity-assisted heat pipes vastly used in applications such as solar collectors, thermoelectric power generators and large equipment as industrial heat exchangers [1]. Due to the low overall thermal resistance, these devices allow heat transfer between a heat source and a heat sink even when they are subjected to low temperature differences. Moreover, they do not require power to work. A thermosyphon consists basically of a sealed evacuated container (tube, for instance) where of a suitable amount of working fluid is inserted. It is usually divided into three main parts: evaporator (where heat to be transferred is captured), adiabatic and condenser section (where the transferred heat is released) [2,3]. ⇑ Corresponding author. E-mail address: [email protected] (J.L.G. Oliveira). http://dx.doi.org/10.1016/j.expthermflusci.2017.06.021 0894-1777/Ó 2017 Elsevier Inc. All rights reserved.

Several parameters can affect the thermosyphon thermal performance, including: working fluid properties [4], inclination angle [5], geometry [6], vacuum level [2] and filling ratio [7] (FR ¼ V w =V ev ap , defined as the ratio between working fluid to evaporator volumes, V w and V ev ap , respectively [2]). A thermosyphon must be carefully designed, in order to avoid typical operating limits, such as dryout, flooding and boiling. These limits impose restrictions to the maximum heat transfer rate [1,2,8,9]. As stated by Mantelli [3] the two phase phenomena within thermosyphons are gravity driven, and so, they are expected transfer heat more efficiently when compared to other capillary driven devices, such as heat pipes. The use of two or more thermosyphons in series can be a feasible solution when, due to geometry limitations, the application of conventional thermosyphon or heat pipe technologies would lead to the design of very complex devices. Actually, the fabrication of complex appliances demands the welding of several pieces in different locations, which can favor air infiltration. On the other hand, smaller thermosyphons can be easier to fabricate, enabling proper evacuation, and so, resulting in devices with good thermal performances. However, care should be taken to guarantee minor thermal contact resistances in the connections between the thermosyphons. Moreover, in this kind of arrangement it is possible to have only one large evaporator which is able to collect heat from several independent sources. Fig. 1 illustrates this concept. The idea is making an analogy with electrical net, plugs and domestic appliances of a house, that the heat exchanger evaporator works as a

410

C. Tecchio et al. / Experimental Thermal and Fluid Science 88 (2017) 409–422

Nomenclature

v
C1

Dtps Dtto _ m

g

T cond T ev ap / Aev ap cp d dadin dev apout lad lev ap po psat qC1 qC2 qC3 qin

mean air velocity at the air duct, [m/s] interval of partial start-up, [s] transient operation interval, [s] mass flow rate, [kg/s] thermal efficiency condenser mean temperture, [°C] evaporator mean temperature, [°C] inclination angle from horizontal axis, [°] evaporator area, [m2] specific heat at constant pressure, [J/(kgK)] diameter, [m] adiabatic inner diameter, [m] evaporator outer diameter, [m] adiabatic lenght, [m] evaporator lenght, [m] operating pressure, [Pa] saturation pressure, [Pa] heat transfer rate at the condenser C1, [W] heat transfer rate at the condenser C2, [W] heat transfer rate at the condenser C3, [W] input power, [W]

‘‘heat sink net”, that ‘‘energizes” the several ‘‘heat sink thermal plugs”. Intermediary heat transfer elements (IHTEs) consisting of thermosyphons or heat pipes can work as the ‘‘thermal wires” transporting the heat (electrical current) from the source (domestic appliance) to the sink (electrical system net); see Fig. 1. As shown by Tecchio et al. [6], in applications such as avionics cooling the aircraft cabin-external air stream could work as the heat sink. One large thermosyphon can serve as heat exchanger between the aircraft external environment (condenser) and the fuselage internal ambient (evaporator with several thermal plugs installed). Heat from several independent sources could be captured by the evaporators of smaller IHTEs (thermosyphons or heat pipes) and released in the large thermosyphon evaporator through thermal plugs, which are installed in its wall. In this work, devices able to collect heat from several heat sources (which mimic electronic components to be cooled) and

q00in qout R T t T C1;in T C1;out T C2;in T C2;out T C3;in T C3;out To V ev ap Vw AR FR HES IHTE LHS RHS

input heat transfer flux, [W/m2] output power, [W] thermal resistance, [°C/W] temperature, [°C] time, [s] inlet temperature at condenser C1, [°C] outlet temperature at condenser C1, [°C] inlet temperature at condenser C2, [°C] outlet temperature at condenser C2, [°C] inlet temperature at condenser C3, [°C] outlet temperature at condenser C3, [°C] operation temperature, [°C] evaporator volume, [m3] working fluid volume, [m3] aspect ratio filling ratio heat exchanger system intermediary heat transfer element left hand side right hand side

to dissipate to a single heat sink are tested. In one of the experiments, two devices are associated in series: a thermosyphon condenser and a loop thermosyphon evaporator. These two thermosyphons are thermally coupled using thermal plugs with different geometries: cylindrical or conical; Fig. 1 illustrates these couplings. The thermosyphon which connects the heat source to evaporator of the loop thermosyphon is referred as intermediate heat transfer elements (IHTEs). The loop thermosyphon which receives heat from the IHTEs and releases it to the heat sink, is referred as heat exchanger system (HES). In another test one calorimeter, in which wall several thermal plugs are installed, was used to simulate a multiple plug loop thermosyphon evaporator. Tests are performed to evaluate the effects on IHTEs performance of: the thermosyphon design, the shape of the coupling plug between thermosyphons (conical or cylindrical), the filling ratio, the operating inclination angle and the input power.

Fig. 1. Intermediary heat transfer elements (IHTEs) connected in series with a common heat exchanger. Heat is transported from independent heat sources to the heat sink through conical and cylindrical couplings.

C. Tecchio et al. / Experimental Thermal and Fluid Science 88 (2017) 409–422

Therefore, the present work has three major objectives:  To propose a new solution for heat removal of devices located in hard to reach spots, consisting of the association of two or more thermosyphons or heat pipes in series;  To study the influence of the plug coupling shape (conical or cylindrical) between the condenser of the IHTEs and the HES evaporator or calorimeter.  To investigate the thermal performance of two thermosyphons connected in series. Special attention is given to the design parameters of the IHTE, which were tested until dryout limit was reached. The intermediary thermosyphons were tested under challenging operating conditions by employing low inclination angles (up to 5°). Layout restrictions may not allow the choice of inclination angles where efficient thermosyphon performance is expected; typically above 20° [10]. Filling ratios ranging from 0.6 to 1.2 were used. Geyser boiling occurrence was observed and its effect on thermal performance is analyzed.

2. Literature review Design features of thermosyphons were extensively investigated through experiments and numerical simulations in the last decades. The effects of filling ratio, inclination angle, working fluid properties and device geometry on thermal performance have been systematically analyzed [3]. In this section literature results concerning the subject of the present study are reviewed. The heat transfer performance of an inclined thermosyphon was experimentally investigated by Negishi and Sawada [11]. Water and ethanol were used as working fluids. They observed that, in steady state regime, the overall heat transfer coefficient increased with increasing the operating temperature. For water, they suggested for best device performance, inclination angles and filling ratios between 20–40° and 25–60%, respectively. Imura et al. [12] investigated the ideal filling ratio for thermosyphons vertically oriented, according to the magnitude of the critical heat flux and operating conditions. Best results were achieved with FR in the range 0.18–0.22. Imura [13], Larkin [14] and Harada et al. [15] suggested similar ranges for ideal filling ratios. Wang and Ma [16] developed a semi-empirical correlation to predict the optimum condensation heat transfer coefficient within inclined thermosyphons. The ideal inclination angle increased from 20 to 50° by increasing the filling ratio from 0.10 to 0.33. The ratio of heat transfer rates between inclined and vertical thermosyphons, q=q90 , was investigated by Payakaruk et al. [10], who observed that the ratio q=q90 was hardly affected by filling ratio modifications. On the other hand, the ratio q=q90 increased with decreasing latent heat of vaporization of the working fluid at inclination angles ranging from 20 to 70°. By using water as the working fluid, q=q90 achieved its maximum value with an inclination angle of 50°. Zuo and Gunnerson [17] stated that the optimum inclination angle may vary according to the operating conditions and device geometry. The critical heat flux for dryout occurrence in thermosyphons with small filling ratios was studied by Cohen and Bayley [18], Strel’tsov [19], Andros and Florschuetz [20]. Results of Cohen and Bayley model showed good agreement with the experimental results obtained by Shiraishi et al. [21] for filling ratios between 0.10 and 0.20. Faghri [2] and Park et al. [7] suggested that the critical heat flux is associated with the flooding limit, for filling ratios higher than 0.20. Noie et al. [5] performed experiments to evaluate the thermal performance of a water-filled thermosyphons, by modifying the

411

inclination angle (from 5° to 90°) and the filling ratio (from 0.15 to 0.30). The heat transfer coefficients at the condenser region increased with increasing filling ratio and reached the highest values with inclination angles in the range 30–45°. Shabgard et al. [22] proposed a numerical model to predict the transient operation of a vertical thermosyphons with various filling ratios, which optimum value was obtained considering that there is liquid film from the condenser to the evaporator end. They observed that, for this optimum filling ratio, the transient operation conditions were shortened. They suggested that a small amount of additional working fluid could prevent the breakdown of the liquid film. Agostini et al. [23] evaluated the thermal performance of a prototype based on heat pipe technologies, aiming the thermal control of electronics. The setup consisted of two-loop thermosyphons connected in series. A mechanical fitting yielded the heat transfer from the condenser of the first loop to the evaporator of the second loop. The prototype dissipated about 1 kW with a mean operating temperature of 110 °C. Oliveira et al. [24,25] and Tecchio et al. [6] have focused on developing a novel passive heat exchanger system for avionics cooling. The last work introduces the concept of intermediary heat transfer devices associated in series. However, a thorough investigation of optimum design parameters was not presented. The present work aims at revealing the effects of essential design parameters, such as filling ratio and inclination angle, on the thermal performance of the intermediary thermosyphons.

3. Experiments 3.1. Experimental setups Thermosyphons with cylindrical and conical shaped plug condensers were designed and tested as intermediary heat transfer elements (IHTEs) between a fictitious heat source and the evaporator of a loop thermosyphon heat exchanger system (HES). To test the IHTEs with cylindrical evaporator shape, a HES, already available in the laboratory was used; see Oliveira [24,25]. The use of a calorimeter instead of the HES evaporator for the conical plug testing is justified as follows: due to space constraints and several welding points air infiltration could occur in the HES evaporator. The vacuum procedure is not necessary by applying a calorimeter and, therefore, IHTE results would not be compromised by HES malfunctioning. In future work, the calorimeter (Fig. 3) will serve as the HES loop thermosyphon evaporator. Schematics of the cylindrical and conical appliances are presented in Figs. 2 and 3, where main dimensions are presented in millimeters. Red dots indicate thermocouple positions. The working principles of the tested devices are briefly described as follows. The input power to be removed from the heat source heats up the working fluid within the evaporator of the IHTEs, causing liquid to vapor phase change. The vapor flows from the IHTEs evaporator to the condenser, where condensation takes place, releasing latent heat of vaporization. The formed liquid returns to the evaporator by means of gravity. The same phase change phenomena drive the HES used to remove heat from the cylindrical shaped thermosyphon evaporator. As a result, the input power is transferred from a fictitious equipment to the heat sink through two thermasyphons in series. The HES depicted in Fig. 2 and used for testing IHTE with cylindrical condenser consists of a closed loop-thermosyphon composed by two condensers in parallel and one common evaporator. This loop thermosyphon was developed by Oliveira et al. [26] for airplane applications and was extensively investigated by Oliveira et al. [24,25]. The evaporator of this device

412

C. Tecchio et al. / Experimental Thermal and Fluid Science 88 (2017) 409–422

Fig. 2. Thermosyphon with cylindrical shaped condenser (in the illustration / ¼ 0). The main dimensions are shown in millimeters. Red dots denote temperature recording points. The angle / was varied from 0° to 5°. HES stands for a heat exchanger system (a loop-thermosyphon) connected to end heat sinks C1 and C2. The cylindrical inner condenser area is 0.0126 m2. (For interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article.)

Fig. 3. Thermosyphon with conical shaped condenser. The main dimensions are shown in millimeters. Red dots (A-D) denote temperature recording points. The angle / was kept in 5°. A calorimeter replaced the HES loop-thermosyphon. The conical inner condenser area is 0.0037 m2. (For interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article.)

consists of a partially-filled (with working fluid) cylindrical horizontal container; see Fig. 2. Two adiabatic lines connect the evaporator to the condensers and two dedicated lines redirect the condensate back to the evaporator basis (not shown here to simplify the illustration). An air duct and a calorimeter were employed as heat sinks for the HES condensers C1 and C2, respectively. A centrifugal fan was used to propel air inside the air duct with a mean
C1 . Inlet and outlet air duct temperatures are reprevelocity of of v sented by T C1;in and T C1;out , respectively. The air duct cross section area is approximately 0.0072 m2. The calorimeter comprises pipeli_ C2 nes where water flows internally with a mass flow rate of m removing heat from the condenser C2 after crossing a distance of

about 4 m. Inlet and outlet temperatures are denoted by T C2;in and T C2;out , respectively. The external area of condensers C1 and C2 are about 0.1 m2 each. One should note that the objective of this paper is to analyze the IHTE and so, only the HES evaporator (in thermal connection with the condenser of the IHTE) schematic suffices to analyze the present results. The IHTE shown in Fig. 2 consists of a cylindrical thermosyphon with an outer diameter of 12 mm and a length of 854 mm. The condenser section has a length of 400 mm and is in direct contact with the HES evaporator wall (details of the mechanical fitting can be observed in the cross section view H-H’ in Fig. 2). The mechanical coupling between the HES evaporator and the IHTE

413

C. Tecchio et al. / Experimental Thermal and Fluid Science 88 (2017) 409–422

condenser was designed to assure practical fitting and low contact thermal resistance. The thermal contact resistance observed in cylindrical contacts, such as the one shown in Fig. 2 can be quite high [6,27]. Therefore, other contact geometries should be considered. In this work, conical contacts are also studied. The IHTE depicted in Fig. 3 comprises a thermosyphon with a conical shaped condenser. The evaporator and adiabatic sections are made of the same tube and so they have constant cross sections, with outer diameter of 21 mm. Four conical plugs, designed to match the IHTE conical shaped condenser, were welded in the calorimeter wall. The heat transfer area of the calorimeter conical wall within the calorimeter was increased by means of fins. Cone and fin dimensions are presented in the expanded illustration in Fig. 3. The heat sources were simulated by cartridge resistances inserted within aluminum blocks, which are in close contact with the evaporator of the IHTEs. Thermal grease reduced the thermal contact resistance between these elements. To avoid heat losses to the environment, ceramic fiber blankets covered the external area of the prototypes. Distilled water was employed as the working fluid for the HES and for the IHTEs. Copper was chosen as the container material due to its compatibility with the working fluid (water), avoiding the formation of non-condensable gases. Before inserting the working fluid, the contaminants of the copper surfaces where removed and high vacuum procedures were performed to the containers. The experimental setups were instrumented as follows. Temperatures were recorded by means of K-type thermocouples, disposed as presented in Figs. 2 and 3. A National Instrument NI cDAQ-9178 was used as the data acquisition system. The input power to the embedded cartridge resistances, qin , was provided through a TDK Lambda power supply (GEN 300-5) and controlled virtually via software Labview 2013. A personal computer was employed for data storage and to control the experimental variables. The heat transfer conditions at the calorimeters shown in Figs. 2 and 3 were controlled by a Lauda Proline thermal bath. This equipment allows volumetric flow rates and temperatures ranging from 0 to 25 L/min and from 20 to 50 °C. Water was used as the coolant fluid at the calorimeters. Air velocity measurements were acquired at the air duct cross section by hot wire anemometry using a Kimo transmitter CTV 110, following the Log-Tchebycheff method [28].

input power ranging from 100 to 700 W in 150 W increments. Inclination angles, /, of 0°, 2° and 5° with respect to the horizontal axis and filling ratios of 0.8, 1 and 1.2 were employed in these cases. The aspect ratio (AR), defined here as

AR ¼

lad ; dadin

ð1Þ

was set to 25 for tests A1 to A5. In Eq. (1) lad and dadin stand for adiabatic section length and inner diameter, in that order. The HES operating conditions are as follows: A filling ratio (FR) of about 90% was used to ensure that the condenser area of the cylindrical IHTE would be flooded by the working fluid, preventing
C1 was adjusted to 3 m/s IHTE malfunctioning. At the air duct the v _ C2 with T C1;in of about 20 °C. At the calorimeter of condenser C2, m and T C2;in were set to 0.04 kg/s and 21 °C, respectively. The experi_ C2 ; T C1;in and T C2;in
C1 ; m mental variables of the HES heat sinks, v were kept constant during the tests A1 to A5. As shown in Tecchio et al. [6], changes in the heat sinks conditions of the HES hardly affect the IHTE’s thermal responses. The thermal performance of the thermosyphon with conical condenser plug (see Fig. 3) was evaluated in cases B1 to B6 with the input power ranging from 20 to 200 W in 20 W increments. The inclination angle was kept constant at 5° and two different aspect ratios were employed: 49 and 29. Three filling ratios (0.6, 0.8 and 1) were employed for each value of aspect ratio. In cases _ C3 , and the inlet temB1 to B6, the mean coolant mass flow rate, m perature at the calorimeter, T C3;in , were adjusted to 0.3 kg/s and 15 °C, respectively. Each input power level for every experimental case was kept constant during 20 min and then increased. Due to safety requirements, tests were interrupted when the evaporator temperature of the intermediary thermosyphons reached 100 °C. Note that the input power level has been decreased owing to the reduced condenser area. The cylindrical condenser area (cases A1-A5) is 0.0126 m2, whereas the conical condenser area (cases B1-B6) is 0.0037 m2; see the captions in Figs. 2 and 3. The condenser area was reduced by a factor 3.4. The decrease in the input power level to promote a limiting operation temperature of 100 °C was reduced by a similar factor. The pipe lines connecting the thermal bath to calorimeters C2 or C3 were also thermally insulated from the external environmental. Therefore, the temperature measurements taken at the inlet of the calorimeters, T C2;in ¼ 21 °C in cases A1 to A5 and T C3;in ¼ 15 °C in cases B1 to B6, correspond to the temperatures set at the thermal bath.

3.2. Experimental procedure 3.3. Data regression and uncertainty analysis A summary of the parameter combination for the experiments performed in this work is presented in Table 1. One can see that the thermosyphon with cylindrical shaped condenser (see Fig. 2) was evaluated under 5 different conditions (A1 to A5) with the

The overall thermal resistance, R, of the intermediary thermosyphons is calculated from the input power, qin , and temperature measurements as follows [2,29]:

Table 1 Summary of experiments. Setup

Case

Cylindrical plug (Fig. 2)

A1 A2 A3 A4 A5

100–700, 100–700, 100–700, 100–700, 100–700,

Conical plug (Fig. 3)

B1 B2 B3 B4 B5 B6

20–200, 20–200, 20–200, 20–200, 20–200, 20–200,

/ [°]

FR

AR

Condenser area [m2]

150 150 150 150 150

0 2 5 5 5

0.8 0.8 0.8 1 1.2

25 25 25 25 25

0.0126 0.0126 0.0126 0.0126 0.0126

20 20 20 20 20 20

5 5 5 5 5 5

0.6 0.8 1 0.6 0.8 1

49 49 49 29 29 29

0.0037 0.0037 0.0037 0.0037 0.0037 0.0037

Input power, Increment [W]

414

R¼

C. Tecchio et al. / Experimental Thermal and Fluid Science 88 (2017) 409–422

T ev ap  T cond ; qin

ð2Þ

where T ev ap and T cond stand for mean evaporator and condenser temperatures, in that order. For the IHTE with cylindrical plug, T ev ap is defined as the mean temperatures, red points A and B, and T cond , with the mean of points 1 and 2; see Fig. 2. The heat transfer rate to the air duct heat sink (C1), qC1 , is determined measuring the air mass flow rate and inlet and outlet temperatures as follows:

_ C1 cp ðT C1;out  T C1;in Þ; qC1 ¼ m

ð3Þ

where cp stands for the specific heat at constant pressure. The heat transfer rates to the calorimeters C2 and C3, qC2 and qC3 , are similar to qC1 . The system thermal efficiency, g, is assessed by [30]:

g¼

qout ; qin

ð4Þ

where, for the IHTE with cylindrical condenser, qout is the total heat rate removed from heat sinks C1 and C2, i.e. qout ¼ qC1 þ qC2 ; see Fig. 2. For the intermediary thermosyphon with conical plug qout is the heat rate removed from calorimeter C3, qout ¼ qC3 ; see Fig. 3. The operating pressures, po , for both HES and IHTE’s, are assessed by the saturated vapor pressure, po ¼ psat ðT o Þ, where T o represents the vapor temperature at the adiabatic lines [31,32]. For HES, the vapor temperature is assumed as the measurements at location 4 in Fig. 2. In this work, T o is also referred as the operation temperature and it is defined as the mean temperature value at locations B and C for the IHTE with cylindrical condenser in Fig. 2 and mean measurements at locations B, C and D for the IHTE with cylindrical plug in Fig. 3. The uncertainties of the experimental variables are summarized in Table 2. Errors of indirect measurements, such as R and g, are estimated with the aid of the error propagation method for a confidence interval of 95% [33,34]. In this case, the uncertainty u of an arbitrary variable G, is expressed by:

sffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi 2 Xn  @G uðGÞ ¼ uðx Þ i i¼1 @x i

ð5Þ

where G is represented by n input parameters, i.e. G ¼ f ðx1 ; x2 ; . . . ; xn Þ. The uncertainties of direct measurements are reported according to the instruments technical features. 4. Results 4.1. Thermosyphon with cylindrical condenser 4.1.1. Temperature distribution Temperatures of the IHTE and of the HES are shown for case A1 in the left hand side (LHS) axis of Fig. 4. The input power is pre-

Table 2 Summary of the experimental uncertainties. Parameter

Instrument or method

Uncertainty

Temperature

1.1 °C

Input power Air velocity Operating pressure, po

Omega K-type thermocouple TKD Lambda GEN 300–5 Transmitter CTV 110 po ¼ psat ðT o Þ

Overall thermal resistance, R

Eq. (2)

System thermal efficiency, g

Eq. (4)

3% 3% 12% maximum 22% maximum 25% maximum

sented in the right hand side (RHS) axis. Temperature recording locations are disposed as depicted in Fig. 2. In case A1, the inclination angle and the filling ratio were set to 0° and 0.8, respectively. One can see in Fig. 4(a) that the evaporator temperature (A) of the IHTE increases linearly with the input power. The IHTE works in transient regime. The regime operating condition is defined in this work in accordance with the change rate of the IHTE evaporator temperature (represented here as point A in Figs. 2 and 3): transient regime for values higher than 0.2 °C/min and quasi-steady state regime otherwise. The test was interrupted when the operation limiting temperature (100 °C) was achieved at time 26.5 min and with the input power set to 250 W. The temperature at the adiabatic section (point C) remained roughly constant at 31 °C for time P2.5 min. The large temperature difference between the thermosyphon adiabatic locations B and C, over 40 °C for time P22 min, reveals that the IHTE startup did not occur. Therefore, heat is not properly transferred from the evaporator to the IHTE condenser. As expected the thermosyphon performance is reduced when the inclination angle with the horizontal position is null (/ ¼ 0°). Heat is only transferred to the HES evaporator by heat conduction through the walls of the intermediary thermosyphon, while most of the input power is used to increase the internal energy of the IHTE casing. Temperature histories of HES are shown in Fig. 4(b) for case A1. The temperature in the right side of the HES evaporator (point 1, see Fig. 2) increases almost linearly for time P2.5 min. Similar behavior is observed for HES evaporator temperatures at locations 2 and 3 for time P12.5 min. The HES vapor line temperature (point 4) increases from 25 °C to approximately 26 °C for 7.5 6 time [min] 6 10 and then decreases smoothly. The thermal behavior within the adiabatic lines is possibly affected by the environment surrounding the HES vapor line (ambient air at 20 °C). The large temperature difference between the vapor line (point 4) and the HES evaporator (points 1, 2 or 3) exposes that vapor is hardly produced within the HES evaporator. Most of the energy released by the resistances heats up the IHTE element and only a fraction achieves the HES evaporator. This fraction is not sufficient to yield the HES start-up. In Fig. 4(a) and (b) oscillating thermal behavior can be noticed on the IHTE adiabatic locations B and C and on the HES evaporator (point 1) for time P15 min. These unexpected events might be related to geyser boiling instabilities; see Ref. [31] for example. Note that the fluctuations in temperature happen simultaneously for points B and C in the IHTE and for point 1 in the HES. The ITHE thermal performance was modified by increasing / from 0° to 2° or 5°. Start-up already occurred for / ¼ 2° and both HES and IHTE were able to work in quasi-steady state in cases A2 to A5. The IHTE temperature profile for cases A2 and A3 are similar and, for the sake of brevity, the presentation of only one case suffices here to the thermal analysis. The similitudes among these results will be shown and discussed further. Results for case A3 are presented in Fig. 5, where the temperature histories are shown in the LHS, and the input power, in the RHS. Results for the IHTE are shown in Fig. 5(a) and for the HES in Fig. 5(b). Four distinct patterns can be perceived in the IHTE thermal analysis of case A3: a transient regime, a transitional and unstable regime, a steady-state regime and a regime affected by dryout. Despite the change in the IHTE inclination angle, results for case A2 are similar and therefore are not presented. The IHTE thermosyphon works in transient regime for time 612.5 min; note the closely linear increase in evaporator temperature in Fig. 5(a) (point A). The abrupt increase in temperature at the adiabatic section (point C) at time 12.5 min indicates the thermosyphon start-up. The HES start-up only occurs at time 17.5 min; see the sudden temperature increase in the vapor line (point 4) in Fig. 5(b). This reduced interval of about 5 min is an

C. Tecchio et al. / Experimental Thermal and Fluid Science 88 (2017) 409–422

415

Fig. 4. Input power effect on the thermal behavior of the (a) intermediary thermosyphon with cylindrical condenser and (b) heat exchanger system for case A1 (/ ¼ 0 and FR = 0.8). Symbols represent temperature recordings and the solid line, the input power. Note the differences in Y-scale from Fig. 4(a) and (b).

indication of low contact thermal resistance between the ITHE cylindrical condenser and the HES evaporator. In the interval 12.5 6 time [min] 6 40 and with 100 6 qin ½W 6 250, a transitional and unstable regime occurs. Although start-up already happened, the change rate in IHTE temperatures in points A, B and C is significant but less pronounced than in the transient regime. The IHTE evaporator and adiabatic temperatures decrease unexpectedly at time 36.5 min; see temperature profiles of points A, B and C in Fig. 5(a). This event coincides with a decrease in the HES temperatures; see Fig. 5(b). The temperature decrease is followed by an oscillatory behavior, which is also noted for higher input powers; see the inset graph in the same figure. This sort of fluctuation is characteristic of geyser boiling occurrence; see Casarosa et al. [32], Khazaee et al. [35] and Lin et al. [31], and apparently occurs more intensely in the HES device; see Tecchio et al. [36]. The reduction in temperature at time 36.5 min is possibly associated to thermal resistance reduction at the HES evaporator induced by the geysering action. For 400 6 qin ½W 6 550, corresponding to the interval 40 6 time [min] 6 80, the IHTE works in quasi-steady regime. The IHTE thermosyphon works with a mean operation temperature, T o , below 60 °C. T o is defined as the mean temperature value at the adiabatic section (mean of measurements at locations B and C; Fig. 2). For input power equal to 700 W, the IHTE evaporator temperature (point A) increases almost linearly in a rate of about 7.6 °C/ min. At the same time temperatures in the adiabatic section (points B and C) slightly decrease. Actually, this effect shows that the IHTE thermal behavior is affected by dryout occurrence. The

liquid film thickness reduces at the evaporator end cap. As a result, a dry region is created; see Faghri [2]. Since evaporation hardly occurs in this regime, the input power is mainly converted in sensible heat and the IHTE evaporator temperature increases sharply. The IHTE thermal performance was modified by increasing FR from 0.8 (cases A2, A3) to 1.0 (A4) or 1.2 (A5). The IHTE thermal behavior is similar for cases A4 and A5, and therefore the presentation of only one case is sufficient for analysis. Despite the changes in the IHTE behavior, the HES thermal response did not show significant variations in cases A2-A5 as it will be shown further in this section. Fig. 6 shows the input power effect on the IHTE thermal performance for case A4 (/ ¼ 5 and FR = 1.0). Dry-out was avoided by increasing FR from 0.8 to 1.0 with qin up to 700 W. The start-up behavior was also modified; compare Figs. 5(a) and 6. In case A4, the temperature at the adiabatic section (point C) oscillates in the interval 11 6time [min] 6 25, indicating the occurrence of partial start-up before the full device operation. As stated by Mameli et al. [37], the full start-up is achieved when stability occurs in the two-phase flow motion: continuous vapor supply to the condenser as well as continuous liquid return to the thermosyphon evaporator. Stable operation only occurs for time >25 min. In this interval, temperature oscillations at point ‘‘C” are less pronounced. The IHTE works in quasi-steady regime for time P35 min (qin P250 W) with operating temperature of about 78 °C for qin ¼ 700 W. The partial startup period is defined as Dt ps and is approximately 14 min. The main thermal features in quasi-steady regime of cases A2A5 are summarized in Table 3. The IHTE operating pressure and the

416

C. Tecchio et al. / Experimental Thermal and Fluid Science 88 (2017) 409–422

Fig. 5. Input power effect on the thermal behavior of the (a) intermediary thermosyphon with cylindrical condenser and (b) heat exchanger system for case A3 (/ ¼ 5 and FR = 0.8). Symbols represent temperature recordings and the solid line, the input power.

Fig. 6. Input power effect on the thermal behavior of the intermediary thermosyphon with cylindrical condenser for case A4 (/ ¼ 5 and FR = 1.0). Symbols represent temperature recordings and the solid line, the input power.

input power at the full start-up, pio and qiin , are presented together with the maximum operating pressure and the input power without dry-out occurrence, piio and qiiin . Values for Dtps are provided for

the IHTE as well as HES operating pressures when pio and piio occur in the IHTE. The heat sink HES capacity is over 1 kW. HES dryout and partial start-up did not occur.

417

C. Tecchio et al. / Experimental Thermal and Fluid Science 88 (2017) 409–422

The operating pressure in the HES at full start-up is similar for cases A2-A5, about 4.3 kPa. piio increases with increasing IHTE filling ratio and ranges from 4.9 to 7.1 kPa. Since the HES filling ratio is kept constant and its heat sink capacity is significant, HES operating characteristics are hardly affected by the IHTE behavior. The IHTE operating pressures are similar for cases A2 and A3. In both cases Dt ps is null and the startup occurs for qin ¼ 100 W. By increasing the inclination angle from 2° to 5° the input power at dry-out occurrence was increased from 400 (qiiin ¼ 250) to 700 W (qiiin ¼ 550 W). In cases A4 and A5 the IHTE dry-out did not occur with the maximum input power, qiiin ¼ 700 W. Full start-up was reached with qin ¼ 250 W, whereas Dtps ranged from 10 to 14 min. The operating pressures pio and piio were affected by modifying the filling ratio from 1.0 to 1.2. piio ranged from 43.5 to 90 kPa. Table 3 also reveals that the IHTE operating pressure is increased with increasing FR from 0.8 to 1.0 or 1.2. piio reaches 90 kPa for a FR of 1.2. This is explained as follows: The required energy to yield liquid to vapor phase change increases with increasing FR; see qiin in Table 3. As a result, the period to achieve the full start-up increases with high FR values, while the evaporator temperatures also increase promoting high values of pio and piio . 4.1.2. Overall thermal resistance The thermosyphon thermal performance can be assessed with the aid of the overall thermal resistance, R, as shown in Eq. (2). Although this equation is only valid for quasi-steady regime, in order to show the dryout occurrence the overall thermal resistance

is shown in Fig. 7 along the whole IHTE operation for cases A1 to A3. Error bars for R are estimated following the procedure described in Section 3.3. In Fig. 7 one can see that, when the IHTE start-up does not occur or the thermosyphon works in unsteady regime, the overall thermal resistance increases almost linearly with the input power. That happened in case A1 for the whole operation range and in cases A2 and A3 before the start-up. Of course, the thermal resistance increases with increasing operation temperature since the thermosyphon does not transfer heat properly in this condition. Once the IHTE start-up occurs and in quasi-steady regime, the resistance R tends to reach a constant plateau, which decreases with increasing input power (see triangles and squares for cases A2 and A3, respectively, in Fig. 7). The discontinuities in R due to the increase in qin correspond to a brief unsteady response. In Fig. 7, this occurs for cases A2 and A3 after raising qin from 100 to 250 W which lasts about 5 min. A sudden increase in R happens if the input power is sufficiently high: this shows the onset of dryout. This is observed at time 48 min for case A2 and at time 80 min for case A3 with input powers of 400 W and 700 W, in that order. The critical input heat flux (q ¼ qin =Aev ap ) for dryout occurrence is estimated as 53 kW/m2 (/ ¼ 2 ) and as 92.8 kW/m2 (/ ¼ 5°). Here, Aev ap ¼ pdev apout lev ap , where dev apout and lev ap stand for evaporator outer diameter and length, respectively. Therefore, the critical heat flux of the intermediary thermosyphon increases with increasing inclination angle. Values of R below 0.05 °C/W occur with qin ¼ 400 or 550 W for inclination angles about 5°. However, these data show that thermosyphons with cylindrical

Table 3 Main thermal features for serial thermosyphons working in quasi-steady regime (cases A2-A5). Case (/; FR)

HES pio

A2 A3 A4 A5 i

o

(2 ;0.8) (5o ;0.8) (5o ;1.0) (5o ;1.2)

[kPa]

4.50 4.20 4.30 4.40

IHTE piio

[kPa]

Dt ps [min]

4.90 6.30 7.00 7.10

0.0 0.0 14.0 10.0

pio

[kPa]

7.40 7.60 25.30 55.00

qiin [W]

piio [kPa]

qiiin [W]

100 100 250 250

16.50 19.00 43.50 90.00

250 550 700 700

Experimental settings at the full start-up. ii Experimental settings for the maximum qin without dry-out occurrence.

Fig. 7. Effect of the inclination angle on the overall thermal resistance of the intermediary thermosyphon with cylindrical condenser for cases A1, A2 and A3.

418

C. Tecchio et al. / Experimental Thermal and Fluid Science 88 (2017) 409–422

shaped condensers can work properly as intermediary heat transfer elements (IHTE) even at low inclination angles (e.g. / from 2° to 5°). The filling ratio (FR) influence on the IHTE thermal performance is shown in Fig. 8 by plotting the IHTE overall thermal resistances in quasi-steady regime, defined in Eq. (2), as a function of the input power for cases A3, A4 and A5. FR varied from 0.8 in case A3 to 1 and 1.2 in cases A4 and A5, respectively, whereas the inclination angle was kept equal to 5° (see Table 1). In this figure, one can see that R decreases almost exponentially with the input power for the three FR cases studied. For a same input power, the overall thermal resistance increased by increasing the FR from 0.8 to 1.0 and 1.2, respectively. The filling ratio change also affects dryout occurrence. Note in Fig. 8 that the R value for qin ¼ 700 W is not plotted for case A3 (FR = 0.8), because dryout was reached. There-

fore, while increasing filling ratios promote increasing overall thermal resistances, decreasing filling ratios facilitate dryout. The optimum scenario among the set cases A1-A5 is the case A4 (/ ¼ 5o ; FR = 1.0). These settings promoted a stable and sufficient liquid refilling from the condenser to the evaporator, avoiding the occurrence of dry spots and keeping po ; T o and R in intermediary levels.

4.1.3. Thermal efficiency The system thermal efficiency (g) is evaluated with the aid of Eq. (4) and illustrated in Fig. 9 as a function of the input power for cases A2-A5. Results are only presented once the system (HES and IHTE, Fig. 2) reaches quasi-steady working conditions. With / ¼ 0o (case A1), the IHTE was not able to transfer heat properly

Fig. 8. Filling ratio effect on the IHTE overall thermal resistances as a function of the input power for cases A3, A4 and A5.

Fig. 9. Thermal efficiency for cases A2-A5.

C. Tecchio et al. / Experimental Thermal and Fluid Science 88 (2017) 409–422

and g is roughly zero. For the remaining cases g increases with increasing qin . This remark is explained as follows. The two-phase flow motion within the IHTE and HES becomes more intense with increasing qin . As a consequence the vapor flux to the HES condensers increases as well as the heat transferred to the heat sinks. The heat leakage magnitude also increases owing to higher operating temperatures despite the increase in g. The heat leakage to the environment by natural convection roughly varies from 30 to 55 W. For the same input power, g also increases with decreasing filling ratios. By reducing FR from 1.2 to 0.8, g increases up to 4%. With lower FR values, the heat leakage is reduced owing to the decrease in the IHTE operating temperature. In case A4, g is nearly 95% with qin =700 W.

4.2. Thermosyphon with conical condenser 4.2.1. Temperature distribution Thermosyphons with conical shaped condensers were also tested as intermediary heat transfer devices, but as already justified, in this case a calorimeter replaced the HES. The effects of filling ratio, input power and aspect ratio on the thermosyphon thermal performance are evaluated in cases B1 to B6; see Table 1. The inclination angle was set to 5° in these cases. Temperature histories for case B1 are presented in the LHS of Fig. 10. This case is representative of experiments with the same AR (cases B2 and B3). Measurements were performed at positions A, B, C and D as illustrated in Fig. 3. The input power provided to the IHTE evaporator is presented in the RHS. In Fig. 10, two distinct patterns can be observed in the thermal analysis of case B1: a transient regime and a steady-state regime. The interval of transient operation, Dt to , of about 63 min is indicated and represents the range where the change rate of the evaporator temperature (point A in Fig. 3) is higher than 0.2 °C/min. Note that the system start-up occurs at time 12.5 min with input power of 20 W, observed by the sudden raise in the adiabatic section temperature (point C). The evaporator temperature increases in a rate of about 0.65 °C/min during this interval. For time 6 63 min condensation takes place along the adiabatic section owing to low input power levels (qin 660 W) and to the reduced vapor flux towards the condenser; note that the temperature difference between points C and D in Fig. 10 becomes minor only for qin P 80 W.

419

For time 6 63 min, the vapor produced in the evaporator hardly achieves the condenser. As a consequence, the operation temperature of the IHTE rises continuously. Only for time P63 min and qin P 80 W the vapor transport to the condenser is effective and the thermosyphon works in quasi-steady regime. The temperature difference along the thermosyphon (between points A and D) becomes minor. An operation temperature (T o ) of about 80 °C is achieved with an input power of 200 W. Fig. 11 shows the temperature history for case B5. This case also illustrates the thermal behavior for cases with AR = 29, cases B4 and B6. The period of transient operation is reduced by decreasing the adiabatic section external area. Quasi-steady operation was reached with input power of 60 W in case B5. As a consequence operating temperatures and pressures were also reduced when AR = 29. Maximum temperature was about 70 °C. The main thermal characteristics of thermosyphons with conical condenser (cases B1 to B6) are summarized in Table 4. The interval of transient operation, Dtto , is presented together with the input power, the operation temperature and pressure in the transition from transient to quasi-steady operation regime, qiin ; T io and pio , respectively, besides the operation temperature and pressure when the input power is modified to 200 W, T iio and piio . Inspection of results in Table 4 reveals that T io and Dtto increase with increasing filling ratio for both aspect ratios. Decreasing filling ratios promote decreasing operation temperatures as expected: less heat is necessary to promote effective vapor transport towards the condenser. Of course, T io and Dtto increase with increasing aspect ratio: more energy at the evaporator is necessary to the vapor to surpass the additional pressure drop owing to a higher adiabatic section length. For cases with AR = 29 and FR = 0.6 or 0.8 (cases B4 and B5), the input power to promote the transition from transient to quasi-steady operation regime was 60 W, while qiin was 80 W for the remaining cases. For cases with FR = 0.6, when AR is reduced from 49 (case B1) to 29 (case B4) T iio is decreased in about 10 °C, whereas for FR of 0.8 and 1.0 T iio is decreased in about 24.5 °C. Both pio and piio decrease with decreasing AR or FR. Values of piio up to 96.7 kPa were registered for aspect ratios of 49, whereas for AR = 29 the operating pressures piio did not overcome 40 kPa. The optimum scenario among set cases B1-B6 is when AR = 29 and FR = 0.6. These settings allowed rapid start-up and reduced

Fig. 10. Input power effect on the thermal behavior of the intermediary thermosyphon with conical condenser plug for case B1 (FR = 0.6; AR = 49). Symbols represent temperature recordings and the solid line, the input power.

420

C. Tecchio et al. / Experimental Thermal and Fluid Science 88 (2017) 409–422

Fig. 11. Input power effect on the thermal behavior of the intermediary thermosyphon with conical condenser plug for case B5 (FR = 0.8; AR = 29). Symbols represent temperature recordings and the solid line, the input power.

Table 4 Main thermal features of thermosyphons with conical condensers (cases B1-B6).

i

Case (FR;AR)

Dt to [min]

qiin [W]

T io [°C]

pio [kPa]

T iio [°C]

piio [kPa]

B1 (0.6;49) B2 (0.8;49) B3 (1.0;49)

63.0 77.5 77.6

80 80 80

64.4 80.8 82.8

24.4 48.9 53.0

80.1 97.0 98.7

47.6 90.9 96.7

B4 (0.6;29) B5 (0.8;29) B6 (1.0;29)

52.7 53.8 62.1

60 60 80

54.9 56.3 66.0

15.7 16.7 26.1

70.2 70.6 75.9

31.4 32.0 40.0

Values in the transition from transient to quasi-steady operation regime. ii Values when qin reaches 200 W.

the vapor pressure drop along the adiabatic line, which promoted the lowest values of T iio and piio . The IHTE with conical condenser shows a higher operating temperature than the cylindrical one. Cases A3 and B5 assist this

remark. These cases present the same FR and inclination angles and similar AR values. In case A3, the IHTE with cylindrical shaped plug was able to transfer 250 W with T o of about 50 °C (see Fig. 8) whereas the IHTE with conical condenser transferred 200 W with

Fig. 12. Thermal efficiency for cases B1-B6.

C. Tecchio et al. / Experimental Thermal and Fluid Science 88 (2017) 409–422

T o of about 71 °C in case B5 (see Table 4). This is expected owing to the small condenser area provided by the conical contact surface (about 3.4 times smaller than the cylindrical condenser). As a result, the heat transfer is not effective unless the temperature difference (DT ¼ T o  T cond ) is high enough. The IHTE operating temperature decreases with increasing condenser area in a rate of about 0.24 °C/cm2. 4.2.2. Thermal efficiency Fig. 12 presents the thermal efficiency of the intermediary thermosyphon with conical condenser as a function of the input power for cases B1-B6. Results are only presented in quasi-steady working conditions. As explained in Section 4.1.3, g increases with increasing qin and with decreasing filling ratios. The efficiency also increases with decreasing aspect ratios. The heat leakage to the environment by natural convection is decreased by reducing the thermosyphon external area. In case B4 g is nearly 96% with qin =200 W. 5. Concluding remarks Fictitious heat sources, mimicking electrical devices, were cooled down by a heat sink via two thermosyphons associated in series. Thermosyphons with cylindrical and conical shaped condensers were experimentally evaluated as intermediary heat transfer elements. These arrangement assures practical mechanical assembly and a modularity concept for cooling of several independent heat sources. The effects of input power, filling ratio, inclination angle and aspect ratio of adiabatic section length to pipe diameter on the thermal performance of the intermediary thermosyphons were investigated. The thermosyphon with cylindrical condenser was evaluated with inclination angles in the range 0–5°, filling ratios in the range 0.8 to 1.2, input power up to 700 W and constant aspect ratio of 25. The thermosyphon with conical plug was tested with filling ratios ranging from 0.6 to 1.0, input power up to 200 W and aspect ratios of 49 and 29, with a constant inclination angle of 5°. Critical heat fluxes for dryout and the input power for thermosyphon start-up have been determined. The main concluding remarks of this study are:  Thermosyphons with cylindrical shaped condensers can work properly as intermediary heat transfer elements even at low inclination angles (/); e.g. 2–5°. In these conditions the critical input heat flux for dryout occurrence increased with increasing inclination angle, whereas the operation temperature increased with increasing filling ratio. The critical heat flux for dryout occurrence is estimated as 53 kW/m2 for / ¼ 2 and as 92.8 kW/m2 for / ¼ 5 with a filling ratio of 0.8. Decreasing filling ratios facilitate dryout.  By increasing the aspect ratios from 25 to 29 or 49 and by reducing the condenser area from 0.0126 to 0.0037 m2, the interval of transient operation and the working temperature in quasi-steady operation are increased for intermediary thermosyphons with conical shaped condenser. By decreasing the filling ratio, the intervals of transient operation and the working temperatures are decreased.

Acknowledgements We would like to express our gratitude to Embraer S.A. for providing essential information and resources to our research project. In particular, we are indebted to Kênia W. Milanez, Luiz Domingos,

421

Yago Braun, Clayton Müller and Leandro Setubal from the Labtucal laboratory.

References [1] D. Jafari, A. Franco, S. Filippeschi, P.D. Marco, Two-phase closed thermosyphons: a review of studies and solar applications, Renew Sust. Energy Rev. 53 (2016) 575–593. [2] A. Faghri, Heat Pipe Science and Technology, Taylor & Francis, Washington, DC, USA, 1995. [3] M.B.H. Mantelli, Thermosyphon Technology for Industrial Applications: Heat Pipes and Solid Sorption Transformations – Fundamentals and Practical Applications, CRC Press, Boca Raton, Florida, USA, 2013. [4] A. Amiri, R. Sadri, M. Shanbedi, G. Ahmadi, B. Chew, S. Kazi, M. Dahari, Performance dependence of thermosyphon on the functionalization approaches: an experimental study on thermo-physical properties of graphene nanoplatelet-based water nanofluids, Energy Convers. Manage. 92 (2015) 322–330. [5] S.H. Noie, M.R.S. Emani, M. Khoshnoodi, Effect of inclination angle anf filling ratio on thermal performance of a two-phase closed thermosyphon under normal operating conditions, Heat Transfer Eng. 28 (2007) 365–371. [6] C. Tecchio, K.V. Paiva, J.L.G. Oliveira, M.B.H. Mantelli, R. Gandolfi, L.G.S. Ribeiro, Passive cooling concept for onboard heat sources in aircrafts, Exp. Therm. Fluid Sci. 82 (2017) 402–413. [7] Y.J. Park, H.K. Kang, C.J. Kim, Heat transfer characteristics of a two-phase closed thermosyphon to the fill charge ratio, Int. J. Heat Mass Trans. 45 (2002) 4655– 4661. [8] M.S. El-Genk, H.H. Saber, Flooding limit in closed, two-phase flow thermosyphons, Int. J. Heat Mass Trans. 40 (1997) 2147–2164. [9] S. Chang, D. Lo, K. Chiang, C. Lin, Sub-atmospheric boiling heat transfer and thermal performance of two-phase loop thermosyphon, Exp. Therm. Fluid Sci. 39 (2012) 134–147. [10] T. Payakaruk, P. Terdtoon, S. Ritthidech, Correlations to predict heat transfer characteristics of an inclined closed two-phase thermosyphon at normal operating conditions, Appl. Therm. Eng. 20 (2000) 781–790. [11] K. Negishi, T. Sawada, Heat transfer performance of an inclined twophase closed thermosyphon, Int. J. Heat Mass Trans. 26 (1983) 1207– 1213. [12] H. Imura, K. Sasaguchi, H. Kozai, S. Numata, Critical heat flux in a closed twophase thermosyphon, Int. J. Heat Mass Trans. 26 (1983) 1181–1188. [13] H. Imura, Optimum fill quantity of liquid in two-phase closed thermosyphons (in japanese), Tech. Report Kumamoto Univ. 29 (1980) 9–18. [14] B.S. Larkin, An experimental study of the two-phase thermosyphon tube, T. Can. Soc. Mech. Eng. 14 (1971). [15] K. Harada, S. Inoue, J. Fujita, H. Suematsu, Y. Wakiyama, Heat transfer characteristics of large heat pipes (in japanese), Hitachi Zozen Tech. Rev. 41 (1980) 167–174. [16] J.C.Y. Wang, Y. Ma, Condensation heat transfer inside vertical and inclined thermosyphons, J. Heat Transf. 113 (1991) 777–780. [17] Z.J. Zuo, F.S. Gunnerson, Heat transfer analysis of an inclined two-phase closed thermosyphon, J. Heat Transf. 117 (1995) 1073–1075. [18] H. Cohen, F.J. Bayley, Heat transfer problems of liquid-cooled gas-turbine blades, Proc. Instn. Mech. Engrs. 169 (1955) 1063–1080. [19] A.I. Strel’tsov, Theoretical and experimental investigation of optimum filling for heat pipes, Heat Transfer-Soviet Res. 7 (1) (1975) 23–27. [20] F.E. Andros, L.M. Florschuetz, The two-phase closed thermosyphon: An experimental study with flow visualization, in: Proc. Symp. Workshop, Fort Lauderdale, Florida, 1976, pp. 547–550. [21] M. Shiraishi, M. Yoneya, A. Yabe, Visual study of operating limit in the twophase closed thermosyphon, in: Proc. 5th Int. Heat Pipe Conf., Tsukuba, Japan, 1984, pp. 11–17. [22] H. Shabgard, B. Xiao, A. Faghri, R. Gupta, W. Weissman, Thermal characteristics of a closed thermosyphon under various filling conditions, Int. J. Heat Mass Trans. 70 (2014) 91–102. [23] F. Agostini, M. Habert, F. Molitor, R. Flueckiger, L. Kaufmann, A. Bergamini, M. Rossi, S. Besana, Double-loop thermosyphon for electric components cooling, IEEE T. Comp. Pack. Man. 4 (2014) 223–231. [24] J.L.G. Oliveira, C. Tecchio, K.V. Paiva, M.B.H. Mantelli, R. Gandolfi, L.G.S. Ribeiro, Passive aircraft cooling systems for variable thermal conditions, Appl. Therm. Eng. 79 (2015) 88–97. [25] J.L.G. Oliveira, C. Tecchio, K.V. Paiva, M.B.H. Mantelli, R. Gandolfi, L.G.S. Ribeiro, In-flight testing of loop thermosyphons for aircraft cooling, Appl. Therm. Eng. 98 (2016) 144–156. [26] J.L.G. Oliveira, K.V. Paiva, M.B.H. Mantelli, R. Gandolfi, L.G.S. Ribeiro, Passive aircraft cooling systems and methods, United States application number or PCT International application number 14/322,133 filed on July 2, 2014. [27] T.R. Alvarenga, Interface para a transferência de calor entre trocador de calor passivo e termossifão em circuito para aplicações em aeronaves, Master’s thesis, Federal University of Santa Catarina, Florianópolis, 2016. [28] ASHRAE Handbook: Fundamentals, American Society of Heating, Refrigeration and Air-Conditioning Engineers, Atlanta, 2001.

422

C. Tecchio et al. / Experimental Thermal and Fluid Science 88 (2017) 409–422

[29] E.S.D.U, Heat pipes 81038 – Performance of two-phase closed thermosyphons, Engineering Sciences Data Unit, 2005. [30] S. Zeinali Heris, M. Fallahi, M. Shanbedi, A. Amiri, Heat transfer performance of two-phase closed thermosyphon with oxidized cnt/water nanofluids, Heat Mass Transfer 52 (2016) 85–93. [31] T.F. Lin, W.T. Lin, Y.L. Tsay, J.C. Wu, Experimental investigation of geyser boiling in an annular two-phase closed thermosyphon, Int. J. Heat Mass Trans. 38 (1995) 295–307. [32] C. Casarosa, E. Latrofa, A. Shelginski, The geyser effect in a two-phase thermosyphon, Int. J. Heat Mass Trans. 26 (1983) 933–941. [33] J.P. Holman, Experimental Methods for Engineers, McGraw-Hill Series in Mechanical Engineering, 1994.

[34] H.W. Coleman, W.G. Steele, Experimentation and Uncertainty Analysis for Engineers, Wiley-Interscience, New York, 1989. [35] I. Khazaee, R. Hosseini, S.H. Noie, Experimental investigation of effective parameters and correlation of geyser boiling in a two-phase closed thermosyphon, Appl. Therm. Eng. 30 (2010) 406–412. [36] C. Tecchio, J.L.G. Oliveira, K.V. Paiva, M.B.H. Mantelli, R. Galdolfi, L.G.S. Ribeiro, Geyser boiling phenomenon in two-phase closed loop-thermosyphons, Int. J. Heat Mass Trans. 111 (2017) 29–40. [37] M. Mameli, D. Mangini, G.F. Vanoli, L. Araneo, S. Filippeschi, M. Marengo, Multi-evaporator closed loop thermosyphon, in: 7th European-Japanese TwoPhase Flow Group Meeting, Zermatt, Switzerland, 2015.