International Communications in Heat and Mass Transfer 125 (2021) 105289

Contents lists available at ScienceDirect

International Communications in Heat and Mass Transfer

journal homepage: www.elsevier.com/locate/ichmt

Investigation of flat plate type pulsating heat pipes via flow

visualization-assisted experiments: Effect of cross sectional ratio
Burak Markal a, *, Ayse Candere b, Mete Avci a, Orhan Aydin b
a
Department of Mechanical Engineering, Recep Tayyip Erdogan University, 53100 Rize, Turkey
b
Department of Mechanical Engineering, Karadeniz Technical University, 61080 Trabzon, Turkey

A R T I C L E I N F O A B S T R A C T

Keywords: In the present study, combined influence of filling ratio, geometry, inclination angle and heating power on heat
Pulsating heat pipe transfer characteristics of the pulsating heat pipes (PHPs) are investigated accompanying analysis of complex
Thermal behavior thermophysical mechanism. Three different flat-plate type closed loop PHPs covering uniform or non-uniform
Visualization
channel-pairs are investigated at different angles (0◦ − 90◦ ) and filling ratios (50% and 80%). It is concluded
Experimental investigation
that phase change phenomena and bubble dynamics play key roles on thermal characteristics. In addition to
unbalanced capillary force, evaporation momentum force is an important factor to improve operational per­
formance of non-uniform PHPs in the lack of gravity support. Non-uniform design decreases the influence of
gravity on results. At horizontal position for medium filling ratio (FR = 50%), uniform PHP (PHP1) collapses.
Compared to angled conditions, maximum heat input of PHP1 drops from 65 W to 45 W (nearly 30% decrease).
At horizontal orientation, non-uniform PHPs maintain operation, and perform better performance than PHP1.
For a relatively high filling ratio (FR = 80%), effect of inclination angle is more obvious, especially for uniform
PHP. In the existence of liquid-excess (at FR = 80%), a new type operation mode is defined as column type bulk
condensation flow.

gaps regarding the relative/combined influence of many parameters

1. Introduction (due to complicated flow structure), and (2) operational limitations
especially regarding the lack of gravity. Pulsating heat pipes operate as a
Effective heat removal from surfaces is one of the main requirements result of pressure fluctuations caused by generating and collapsing vapor
for the advanced technological devices being exposed to high thermal bubbles, and pressure forces caused by bursting vapor bubbles in the
loads. This requirement increases importance of enhanced thermal evaporator section. The operational performance depending on internal
management methods such as different types of heat pipes (wicked/ perturbations can be significantly enhanced by designing non-uniform
conventional, thermosiphon, pulsating and rotational type heat pipes, cross sections due to occurrence of different unbalanced forces. In
etc.) [1–5], vortex tubes [6], impinging jets [7], and flow boiling in addition, as stated in the comprehensive review study of Nazari et al.
small channels [8]. Among the mentioned techniques, pulsating heat [14], attempts regarding working fluid are another important aspect for
pipes provide both the advantages of passive operation and phase- higher performance. In the following, some efforts on PHPs are
change-flow; and also compared to the other kinds of passive two- summarized:
phase systems such as conventional counterparts (wicked heat pipes), Khandekar and Groll [15] underlined the complexity of the flow
they also offer advantages like easy construction, low cost and nature in the PHPs, and focused on operational behavior of a single loop
application-flexibility. Due to these advantages, popularity of PHPs in capillary PHP. Heat transferring sections (evaporator and condenser)
various fields such as space applications [9], solar based systems (col­ were made of copper, while the other parts constituted from glass for
lectors, water heaters) [10,11] and, cooling of electronic devices [12] visualization. Similar to those for the multi-turn ones, the two-phase
has been increasing. It should be stated that a comprehensive list of instabilities having dynamic character were observed, and the flow
applications regarding PHPs was well addressed by a review study of patterns were remarked as the dominant factor for the overall perfor­
Nazari et al. [13] In addition to these critical usages in practices, mance. Xu et al. [16] analyzed flow structures in a closed loop PHP
research interests are also motivated by two following reasons: (1) the (CLPHP) via flow visualization. They observed that distribution of flow

* Corresponding author.
E-mail address: [email protected] (B. Markal).

https://doi.org/10.1016/j.icheatmasstransfer.2021.105289

Available online 24 April 2021

0735-1933/© 2021 Elsevier Ltd. All rights reserved.
B. Markal et al. International Communications in Heat and Mass Transfer 125 (2021) 105289

tests were only performed for the filling ratio of 50% at vertical orien­
Nomenclature tation. The asymmetric ratio showed a heat flux-dependent effect. At
high heat fluxes, increasing asymmetric ratio had an adverse effect on
Abia bubble interface area [m2] the performance, while it had a positive influence at low and medium
Act total heat transfer area of the relevant channel [m2] heat fluxes. Also, it was stated that decreasing aspect ratio (total width
cp specific heat [J kg− 1 ◦ C− 1] over total height of a channel pair) lead to a reduction in flow resistance.
hfv latent heat of evaporation [J kg− 1] Sun et al. [24] investigated the role of evacuation pressure on PHP
ṁ mass flowrate [kg s− 1] performance. It was concluded that higher evacuation pressure meant
Qi input heating power [W] better overall performance; however, different fluids showed different
Qc heating power rejected to heat sink [W] operational behavior under various pressures. Nazari et al. [25] exper­
Rth thermal resistance [◦ C W− 1] imentally examined a two-turn capillary type PHP filled partially by
T temperature [◦ C] graphene oxide nanofluid. Nearly a 42% decrease in thermal resistance
was obtained by using nanofluid. However, high concentration of par­
Greek symbols ticles caused worse heat transfer performance than pure base fluid.
ρ density [kg m− 3] Dehshali et al. [26] focused on the characteristics of rotational CLPHPs.
Subscripts They underlined that increasing rotational speed generally decreases
c condenser thermal resistance, and thus improves heat transfer. The study showed
e evaporator possible applications of PHPs in rotating devices. Sedighi et al. [27]
hsi heat sink inlet focused on geometrical differences to improve the performance of the
hso heat sink outlet PHPs. They applied an extra branch to the evaporator zone of the single
v vapor turn PHP to boost the circulation. The filling ratio and the angles ranged
from 40% to 70% and 0 to 90◦ , respectively. Use of this branch enhanced
Abbreviations mixing of cold and hot fluids, and one directional flow of the fluid was
FP-CLPHP flat plate closed loop pulsating heat pipe supported. As a result, they observed a significant performance
FR filling ratio enhancement via the modified PHP. After experimental observations,
IA inclination angle Yoon and Kim [28] detected a relation between vapor distribution and
PHP pulsating heat pipe heating power. They stated that distribution of the vapor phase played a
primary role on the thermal performance of a PHP, and they proposed a
model presenting the abovementioned relation. Jo et al. [29] studied
thermal mechanisms dominating operation of a micro PHP filled by
patterns changed from channel to channel. The basis of oscillating ethanol. The orientation and the filling ratio were set to 90◦ and 50%,
mechanism was attributed to the bubble formation, deformation and respectively. Regarding an ongoing debate, they concluded that the
merging processes. Qu and Ma [17] dealt with the relations between the contribution of two phase heat transfer was much more than its single
startup characteristics and the influential parameters through a theo­ phase counterpart. In the study conducted by Li et al. [30], water-
retical analysis. It was disclosed that forming cavities or in other words ethylene glycol (as a base fluid) was fed by graphene nano-platelets
increasing the roughness of the internal surfaces would induce early for obtaining working fluid. When the concentration value of gra­
startup. Also, the working fluid type and the bubble shape were defined phene was increased, thermal resistance decreased; this was attributed
as the other effective factors. In horizontal or top-heating modes, failure to the improvement of Brownian motion and effect of thermophoresis
of PHPs having few numbers of turns was examined by Chien et al. [18], phenomenon. Wan et al. [31] focused on the combined effects of check
and the asymmetric channel configuration was proposed as a solution valve usage, gravitational force and uneven heat load on capillary PHPs.
for the first time. It was stated that providing uneven capillary force They concluded that any binary combination of these three influential
between the parallel channels could provide the operation of the PHP factors increased the thermal performance; however, simultaneous
under no gravity support. Goshayeshi et al. [19] focused on magnetic usage all of these mentioned techniques could lead to undesired thermal
field effect on thermal characteristics of PHPs. A type of nanofluid performance. In a recent study, Markal and Varol [32] performed a
named as Kerosene/Fe2O3 was used in five-turn capillary-tube-type comprehensive study to experimentally investigate a non-uniform
PHP. Under magnetic field effect with nanofluid, they obtained CLPHP filled by ternary mixtures. It was concluded that at vertical
improvement nearly by 16% in thermal performance. Kwon and Kim orientation, the CLPHP charged by ternary mixtures (water:metanol:
[20] studied with a micro flat-plate closed loop pulsating heat pipe (FP- pentane = 1:2:3 and 1:3:2) showed obviously better thermal perfor­
CLPHP) having five turns (fluids were ethanol or FC-72). They investi­ mance compared to the case with pure fluids. Four forces playing
gated effect of width ratio between parallel channels for a constant dominant role in the performance of non-uniform PHPs were defined:
filling ratio of 50%. The capillary pressure was formed as a result of (1) surface tension force, (2) shear force, (3) momentum force and (4)
different capillary hydraulic diameter of the neighboring channels. This gravitational force.
kind of geometrical arrangement could provide orientation independent After summary of the relevant literature, it can be concluded that
performance. As a continuation of their previous study [19], under geometrical non-uniformity of the channels of a PHP is very influential
magnetic field, Goshayeshi et al. [21] experimentally studied effect of tool regarding operational performance, and there are also some con­
tilt angle on a CLPHP filled by ferrofluid. Interestingly, they declared tradictory results regarding the sole or combined effects of geometrical
that improvement in performance with increasing angle was over at 75◦ ; and operational parameters (because of the complicated nature of the
and the performance deteriorated for higher angles. In other study, flow field). On the other hand, there are only a few studies focused on
Goshayeshi et al. [22] experimentally investigated effects of particle size asymmetrical channel cross section; and to the best of the authors’
and type on the performance of a nano-fluid (Fe2O3/Kerosene) charged knowledge, there is no study investigated effect of cross sectional ratio
PHP under magnetic field. They stated that addition of Fe2O3 nano­ for high filling ratios and/or there is no study investigated the effect of
particles enhances thermal characteristics of the PHP. They investigated filling ratio for different cross sectional ratios. However, it is known that
six nanoparticles, and 20 nm γ-Fe2O3 nanoparticles present the best high filling ratio is an important operational parameter due to delay of
results regarding heat transfer. Jang et al. [23] investigated both the the dryout; and thus, there are well-known studies only focused on high
effects of cross section and aspect ratio of the channel pairs on the filling ratios such 70% [31] and 80% [33].Therefore, the need for new
thermal characteristics of a PHP. The PHP was filled with FC-72, and the studies or new experimental database is clear. With this research

2
B. Markal et al. International Communications in Heat and Mass Transfer 125 (2021) 105289

Table 1 analysis, the top section of the PHPs is closed via a tempered glass. Also,
Details of the experimental conditions. to prevent leakage of the fluid from the PHP, an extra channel is
Condition/Case Range or Remarks engraved outside the flow channels; and a viton O-ring is placed there.
Over the channels, a transparent cover plate (tempered glass) was
Heat input (HI) 5 W to Maximum allowable value (up to Te = 110 ◦ C; step:5 W)
Filling ratio (FR) 50% and 80% placed, and the transparent plate was strongly compressed to the copper
Inclination angle 0◦ and 90◦ PHP via a brass frame (a special geometry). Circumferentially placed 32
(IA) M2 bolts were used to carry out this process. The details of the relevant
Working Fluid Ethanol geometry were presented in a previous article of the corresponding
PHP1 Uniform PHP (channel height 2 mm and width 2 mm
throughout the PHP)
author of the present paper [32]. Ethanol is used as the working fluid,
PHP2 Non-uniform PHP (channel height is constant as 2 mm; and in a and it is charged into the PHP through the charging channel represented
turn, one of the parallel channels has 2 mm width and the other in Fig. 2a. The charging channel is also connected to a vacuum pump; in
has 1.5 mm width (details are presented in Fig. 2a) this way, before every experiment, a vacuuming process is applied via a
PHP3 Non-uniform PHP (channel height is constant as 2 mm; and in a
vacuum pump. For the vacuuming process (nearly 20 min), we used a
turn one of the parallel channels has 2 mm width and the other
has 1 mm width (details are presented in Fig. 2a) manometer and a vacuum pump, and we read the value of − 1.012 bar
(as gauge pressure) via the manometer. This also corresponds to 125 Pa
(or 0.94 Torr) as absolute value (the atmospheric pressure is taken as
interest, the objectives of the present study are: (1) to investigate com­ 1.01325 bar).
plex thermo-physical mechanisms in FP-CLPHPs, (2) to examine the The PHP is placed onto two heat transfer subsections. One of them is
effect of cross sectional ratio on thermal performance for different filling the heating plate and the other one is the heat sink. The details regarding
ratios, and (3) to analyze combined influence of filling ratio, geometry test section components and their placement (3D partial section view of
(non-uniformity of channel pairs), inclination angle and heating power test section) were clearly demonstrated in Refs. [32, 34]. The heating
on heat transfer characteristics through the flow patterns. Flow visual­ plate has nearly rectangular form with overall dimensions of 20 mm ×
ization is used as an effective tool for analysis of the flow nature. As the 40 mm and 9 mm (length x width x height). Inside the heating plate, two
originality, to the best of the authors’ knowledge, in this study, the effect cartridge heaters are placed, and these heaters are connected to a power
of cross sectional ratio of parallel channels in a FP-CLPHP filled by supply. Top surface of the heating plate is just beneath the evaporator
ethanol is experimentally investigated for different filling ratios at mini section, and the desired heat load is applied through this section. On the
scale for the first time. Details regarding experimental conditions are other hand, the heat sink includes two subsections: (1) the 2 mm thick
presented in Table 1. copper plate placed just beneath the condenser section of the PHP and
(2) the polycarbonate plate including flow passages of cooling water.
2. The test apparatus and details related to data reduction This polycarbonate plate is screwed to the copper section of the heat
sink. The total thickness of the heat sink is 12 mm. The cooling water
2.1. Presentation of the experimental setup flows in the passages inside the heat sink to absorb the heat from the
condenser section of the PHP. In the cooling water circulation line,
The main devices located in the experimental setup are presented in different devices are used (see Fig. 1). Temperature of the cooling liquid
Fig. 1. Three different, flat plate type closed loop pulsating heat pipes is set via a temperature bath. In the present conditions, the temperature
(FP-CLPHPs) are used. The material of the heat pipes with eight turns is of the cooling liquid is 20 ◦ C. The cooling water is drawn from the
copper. The difference between the geometries stems from difference in temperature bath by means of a micro-pump. On the other hand, the
the channel width ratio of the parallel channels. Details are presented in flowrate is adjusted via a flow controller. The flowrate of cooling water
Fig. 2. As seen in the relevant figure (Fig. 2a), the PHPs are named as is set to 20 ml min− 1. It should be noted that the platform areas of the
PHP1, PHP2 and PHP3, and the values of the channel widths are written. heating plate and the heat sink are 20 mm × 40 mm and 40 mm × 40
Information presented regarding widths shows the non-uniformity of the mm, respectively (length x width).
PHPs. The cross section of the channels has rectangular shape. The One of the other important sections is the visualization side of the
channel heights are constant (2 mm), and thus, the width ratio means experimental setup. For visualization, a high speed camera is used. The
the cross sectional ratio as well. For visualization and flow pattern camera is connected to a computer; and an illuminator is also used to

Fig. 1. General schematic of experimental setup.

3
B. Markal et al. International Communications in Heat and Mass Transfer 125 (2021) 105289

Fig. 2. Details for heat pipe geometries (a) and locations of thermocouples (b).

make the images clearer. The camera is fixed to the PHP as a face to face 2.2. Experimental procedure
way, and the system is placed on a moving mechanism to be oriented at
the desired angle (see Fig. 1). In all of the zones (evaporator, adiabatic The procedure followed during the experiments can be mainly
and condenser), three T-type thermocouples (wire size of nearly 250 defined as follows:
μm) are used to measure the temperature values. The locations of the
thermocouples are schematically shown in Fig. 2b. Heights of the • Inside of the PHP is vacuumed using the vacuum pump
thermocouple-grooves are constant as 1 mm in each region. In this way, • The working fluid (ethanol) is degassed via boiling process. This
the thickness between the bottom surface of the flow passages and the process is applied to diminish the adverse effects of any available
thermocouples is 1 mm. In the evaporator and condenser regions, non-condensable gases, which is also experienced by Takawale et al.
thermocouples are placed radially, and thus, the width of the thermo­ [35].
couple grooves is 3 mm for these regions. In the adiabatic region, there • The cooling water loop is activated.
are three axial groves with a 1 mm width. Also, just before and after the • The PHP filled is with ethanol at the desired filling ratio. The filling
heat sink, two T-type thermocouples are also used to control the inlet ratio means ratio of working fluid volume over PHP total internal
temperature of the cooling water and to calculate the heat rejected. All volume at laboratory conditions.
of the temperature readings are taken via a data logger. Details • The angle of the PHP is set to the desired value.
regarding the modes/codes of the devices used in the present study were • The power supply is activated.
clearly addressed in Refs. [5, 32].

4
B. Markal et al. International Communications in Heat and Mass Transfer 125 (2021) 105289

Table 2
Input heating powers and corresponding heat fluxes.
Input heating power, Qi [W] 5 10 15 20 25 30 35 40 45 50 55 60 65
− 2
Heat flux [W cm ] 0.63 1.25 1.88 2.50 3.13 3.75 4.38 5.00 5.63 6.25 6.88 7.50 8.13

• Initial heating power is set to 5 W, and it increases with 5 W steps. In images are taken at a frame rate of 1000 (1000 fps), and 60 temperature
each value, when the steady-state conditions occur, temperature readings are performed for each station totally at nearly 37 s. It should
measurement and the flow images are taken. be stated that platform area of the evaporator section, or in other words,
• When the average temperature of the evaporator section becomes the heating area of all the PHPs (PHP1, PHP2 and PHP3) is same (20
approximately 110 ◦ C, the system is stopped for system security. mm × 40 mm; clearly seen in Fig. 2). To make clearer the results and to
Then, for another test condition (PHP type, filling ratio or inclination present more general results for further studies, the values of input
angle), the above steps are repeated. heating powers and corresponding heat fluxes are presented in Table 2. .

2.3. Data reduction 3.1. Influence of heating power, channel cross sectional ratio and
inclination angle
In the present paper, it is mainly focused on the combined effects of
the parameters, and the dominance of geometrical factors on the Performance evaluation of pulsating heat pipes should be conducted
thermo-hydraulic coupling. In this context, a comprehensive evaluation depending on many criteria. In this regard, in the present study, a
covering flow patterns and the trends of the thermal performance in­ comparative and simultaneous investigation of thermal resistance,
dicators (evaporator temperature, thermal resistance, startup perfor­ maximum heat input, startup performance, evaporator temperature and
mance, stable operation) is conducted. Undoubtedly, as also underlined flow structures (transitions/patterns) are presented.
by Xu et al. [36], the total thermal resistance is among the most
important parameters denoting heat transport performance. The ther­ 3.1.1. Influence of heating power at various inclination angles
mal resistance is presented as follows: Variation of the thermal resistance and the corresponding evaporator
and condenser temperatures with the heating power are presented in
(Te − Tc )
Rth = (1) Fig. 3 for different angles at the filling ratio of 50%. The relevant graphs
Q include all the PHPs. In the graphs, as stated in Fig. 2, PHP1, PHP2 and
In the above equation, T denotes the temperature; and, the subscripts PHP3 signify the heat pipe geometries having 2–2 mm, 2–1.5 mm and
of c and e represent condenser and evaporator sections, respectively. 2–1 mm parallel channel widths, respectively. All the channels have
They are all average values of the temperature readings in the relevant same height of 2 mm. As clearly seen from Fig. 3, in most cases, as a
regions. On the other hand, Q is the mean value of the input heating general evaluation, when the applied heat input increases the thermal
power (Qi) and rejected heating power (Qc), as shown below: resistance of all the geometries firstly decreases and then performs a
horizontal trend. This trend is obvious especially at all the positive-angle
Qi + Qc
Q= (2) (> 0◦ ) orientations (see Fig. 3c, e and g). As stated in the literature [39],
2
at saturated conditions, the variation of pressure with the temperature,
Here, it should be stated that Eq. (2) is a well-accepted approach in (dP/dT)sat, is a quite important thermo-physical property of the working
the literature [18,32,33,37]. As it is clearly stated by Spinato et al. [37], fluid. This property is addressed as the source of rapid fluid motion and
this approach/equation can be used to take the heat-loss effect into greater fluctuations by Bastakoti et al. [40]. In fact, this explains the
consideration in calculations. In this regard, the value calculated via Eq. underlying reason of heating-based enhancement in heat transfer per­
(2) is substracted from input heating power (Qi) to obtain heat loss. In formance. Increasing temperature leads to an increase in the pressure at
the present study, depending on the experimental range, minimum, which the liquid slug evaporates, and thus the pressure force applied by
maximum and average heat loss values, respectively, correspond to vapor plug in the evaporator region increases [31]. The pressure dif­
4.9%, 21.3% and 13% of input heating power. Heat input is provided via ference between the evaporator and condenser regions increases, the
the power supply, and the rejected one is calculated via the temperature vapor can push the liquid slug in a more easy way, the circulation rate
values read at the inlet (Thsi) and outlet (Thso) of the heat sink. increases, and as a result the thermal resistance decreases. However, it
should be underlined that the flow in pulsating heat pipes is two phase
Qc = ṁcp (Thso − Thsi ) (3)
flow having instantaneous changes in terms of both the direction and
Here, the mass flow rate of the cooling water is represented with ṁ, phase type. Therefore, flow pattern distributions and relevant heat
and the specific heat of the water is represented by cp. transfer mechanisms should also be discussed in order to explain the
The present study is an experimental one, and thus, to obtain the results caused by heat input based changes. In this regard, at the con­
uncertainty of the total thermal resistance, the well-known method dition of PHP1, FR = 50% and IA = 60◦ , some flow images taken at
defined in Ref. [38] is applied. The minimum and maximum values of different heating powers (15, 35 and 60 W) are presented in Fig. 4. When
the relevant uncertainty (for total thermal resistance) are 3.7% and the heating power increases gradually, significant variations occur in
5.5%, respectively. Also, it should be noted that the uncertainty in the flow pattern distributions. At 15 W (Fig. 4a), there are only a few
(calibrated) temperature readings is ±0.1 ◦ C. separate vapor slugs, most of the evaporator section is filled by liquid
phase. In some channels, long bubbles representing thin film evapora­
3. Results and discussion tion and/or condensation begin to form. This flow pattern distribution,
in fact, shows early stages of pulsating/oscillating motion. This image is
The present study focuses on complicated physics caused by strong compatible with the corresponding data in the resistance curve (see
relations between thermal and hydrodynamic phenomena, and the in­ Fig. 3e) in which a sudden drop is observed at the heating power of 15
fluence of operational and geometrical parameters on the thermal per­ W. This change in the resistance curve represents the startup of the heat
formance. In this context, the effect of cross sectional variation (may be pipe, as stated in the literature [41]. With increasing the heating power,
stated as asymmetry ratio) is investigated for different filling ratios (50% as seen in Fig. 4b, number of liquid slug and vapor bubbles increase, and
and 80%) and orientations (0◦ − 90◦ ). The images representing the flow they spread throughout all the PHP. This condition corresponds to a
structures are actively used in the scrutinizing process. In the study, flow quite low resistance value (see Fig. 3e). Therefore, it can be concluded

5
B. Markal et al. International Communications in Heat and Mass Transfer 125 (2021) 105289

Fig. 3. Variations in thermal resistance and evaporator and condenser temperatures of all the PHPs at FR = 50%.

6
B. Markal et al. International Communications in Heat and Mass Transfer 125 (2021) 105289

Fig. 4. Flow images at the conditions: PHP1, IA = 60◦ , FR = 50%, 15 W (a), 35 W (b), 60 W (c).

7
B. Markal et al. International Communications in Heat and Mass Transfer 125 (2021) 105289

Fig. 5. Flow images at the conditions: IA = 90◦ , FR = 50%, Qi = 50 W, PHP1 (a), PHP2 (b) and PHP3 (c).

8
B. Markal et al. International Communications in Heat and Mass Transfer 125 (2021) 105289

Fig. 6. Flow images at the conditions: IA = 0◦ , FR = 50%, Qi = 35 W, PHP1 (a) and PHP3 (b).

that phase change phenomena and relevant flow pattern types play a key type array is more obvious. PHP performs proper operation in the
role on thermal characteristics. Herein, there are two important reasons relevant conditions. In the heat pipe literature, relative contribution rate
related to heat-increment-based performance enhancement: (1) of sensible or latent heat transfers over the total heat transfer is still a
increasing of the flow circulation rate and (2) active role of the conflicting topic. However, as clearly seen from Figs. 4b and c, at me­
convective boiling and condensation mechanism. Especially the second dium to high heat inputs, most of the flow passages are covered by
one is related to flow pattern type. Increasing number of the vapor elongated vapor bubbles or smaller separate vapor plugs. Bubbles or
bubbles means longer thin-film contacting the internal walls of the PHP. vapor plugs are surrounded by thin liquid film. Thus, heat is absorbed
Also, fluctuations in the flow field increase. Effective heat carrying via phase change in the form of bubble generation and (mostly) thin film
components reproduce. The liquid slugs become smaller, thus the mass evaporation. Similarly, condensation process occurs through thin liquid
pushed by vapor bubbles decreases. Also, as seen from Fig. 3f, temper­ film meaning higher heat transfer coefficient. Therefore, as convenient
ature of the evaporator region also increases with heat input increase. to the results presented by Jo et al. [29]; from the present results, it can
Therefore, the pressure force performed by growing bubbles strengths, also be concluded that phase change heat transfer is more effective.
and the circulation improves. When the heat further increases (Fig. 4c,
60 W), the flow phenomena occurring in the medium heating powers 3.1.2. Channel cross sectional ratio at various inclination angles
(35 W) become more active. In this context, nucleation improves and the From Fig. 3, it is clearly seen that there is a considerable difference
bubble population increases. The role both of the nucleation and the between the positive-angle (IA = 30◦ , 60◦ and 90◦ ) and horizontal (IA =
convective boiling in the total heat transfer processes enhances. Espe­ 0◦ ) cases. In the positive-angle cases, thermal resistance decreases with
cially, here, as underlined by Xu et al. [42], importance of bubble increasing heating power for all PHPs. For the lower inclination angles
population on heat transfer performance of the PHPs should be under­ (IA = 30◦ ), the difference between the thermal resistances is a bit more
lined. The content of the liquid collected at the bottom level of the U- clear. However, the relevant difference obviously decreases in higher
turns in the evaporator section decreases. Due to the overall temperature inclination cases (IA = 60◦ and 90◦ ). The probable reason is related to
increase throughout the whole channels, much of the fluid content can the relative contribution of the gravitational force depending on the
survive as separate vapor bubbles and smaller liquid slugs. The train- angular variation. Generally, a sudden drop occurs in the thermal

9
B. Markal et al. International Communications in Heat and Mass Transfer 125 (2021) 105289

resistance data at the heating power of 15 W; which is an indicator of the increases. Thermal resistance curve of PHP1 shows nearly horizontal
startup of the PHPs. This case is valid for all the PHPs at IA = 30◦ and trend. This trend means that PHP1 cannot start operation in the hori­
60◦ . In the case of 90◦ , PHP1 and PHP2 can start up operation at the zontal position. In other words, the uniform-channel heat pipe (PHP1)
heating powers of 10 W, while the PHP3 again shows an active perfor­ collapses. The maximum heat input of PHP1 drops from 65 W to 45 W,
mance at 15 W. From Fig. 3, for inclination angles of 30◦ , 60◦ and 90◦ , it which means nearly 30% decrement. In the horizontal orientation, PHP2
is seen that thermal resistance of PHP2 is the lowest one in general and PHP3 maintain the operation, and perform better thermal perfor­
manner. However, the relative magnitude of the thermal resistance is mance than PHP1. To visually show the situation, two different images
not the only criteria or indicator to make a judgement about which heat taken for PHP1 and PHP 3 are presented in Fig. 6a and b, respectively. In
pipe presents the better performance. In the thermal resistance curves Fig. 6, the operational conditions corresponds to FR = 50%, IA = 0◦ and
(Fig. 3c, e and g), different from the relative magnitude of the thermal heating power of 35 W. The images should be simultaneously analyzed
resistance values, the maximum heat load reached by the heat pipes via the data presented in Fig. 3a. As seen in Fig. 6a, for PHP1, there is no
shows a different trend. As an example, for IA = 90◦ , PHP1, PHP2 and nucleation phenomenon in the evaporator; and/or there are no separate
PHP3 reach the maximum allowable temperature at 65 W, 50 W and 55 vapor plugs or liquid slugs throughout the flow passages in distributed
W, respectively. Also, from the corresponding evaporator temperature way in different locations. On the other hand, in Fig. 6b, for PHP3, there
curves (Fig. 3d, f and h), it can be understood that evaporator temper­ are lots of separate vapor plugs and liquid slugs; and they distribute
atures of the PHP1 are significantly lower than those for the other PHPs. throughout the flow passages. The reason of the collapse for PHP1 is
On the other hand, with decreasing inclination angle, for example for IA explained with the lack of gravitational force due to the horizontal po­
= 30◦ (see Fig. 3d), PHP1, PHP2 and PHP3 reach the maximum allow­ sition and the insufficient perturbations due to the less turn number
able temperature at 60 W, 50 W and 45 W, respectively. Thus, it can be [37]. It should be noted that PHP1 has a uniform channel structure. The
concluded that for the conditions of FR = 50% and IA = 30◦ − 90◦ , PHP1 lack of internal pressure perturbations and gravity support can be
shows the best thermal performance, while PHP2 and PHP3 present compensated via unbalanced capillary force. However, there is an upper
comparable results. At the vertical position (90◦ ), PHP3 is better than limit of this support. Kwon and Kim [20] proposed equilibrium for this
PHP2; however, with decreasing angle, PHP2 shows better thermal limit over the relative magnitudes of viscous pressure drop and capillary
performance than PHP3. Especially, as stated above, compared to PHP2 pressure difference. In addition to the definition of Kwon and Kim [20],
and/or PHP3, in spite of higher thermal resistance values of PHP1 in it is proposed that the evaporation momentum force which is handled in
some experimental conditions, evaporator temperature of PHP1 is lower flow boiling literature [43] may be considered in the evaluation of re­
and maximum heat input value of PHP1 is higher. This is actually related sults. This force is presented as below:
to definition of thermal resistance, see Eq. (1). In fact, in the relevant ( )2
conditions, in addition to evaporator temperature, condenser tempera­ Abia Qi
Femf = (4)
ture of PHP1 is also lower than those for PHP2 or PHP3 (see Figs. 3b, d, f ρv Act hfv
2

and g); however, difference between evaporator and condenser tem­

where ρv is vapor density, Act is the total heat transfer area of the rele­
peratures can be higher than that of PHP2 or PHP3. Therefore, in some
vant channel, Abia is the bubble interface area and hfv is the latent heat of
cases, lower evaporator temperatures are obtained in spite of higher
evaporation, respectively. According to the Eq. (4), decreasing channel
thermal resistance values. It should be underlined that, as stated above,
heat transfer area or channel dimension leads to an increase in the
the relative magnitude of the thermal resistance is not the only criterion
evaporation momentum force. In the non-uniform PHPs (PHP2 and
or indicator to reach a decision about which heat pipe presents better
PHP3), one of the channels is narrower, and thus, the cross sectional
performance. In this regard, to understand the underlying reason of the
area of the relevant channel is smaller than the neighboring channel.
large difference between the uniform and non-uniform channels, the
Therefore, the force denoted with Eq. (4) has different strengths in
advantage of flow visualization technique should be taken. Fig. 5 in­
neighboring channels (parallel channels in a channel pair). This in­
cludes the flow images of the whole PHPs at the conditions of IA = 90◦ ,
creases internal pressure perturbations or unbalanced pressure distri­
FR = 50% and Qi = 50 W. As seen from the instantaneous images in
bution between channels, which supports operational performance of
Figs. 5a, b and c, the flow patterns and their distribution are nearly
the PHP, and thus, provides advantage against the condition having lack
similar. However, via the long term observations during the visualiza­
of gravity support. In brief, the pressure perturbations increase, unbal­
tion test, it is seen that the condensation process in the condensation
anced pressure forces occur in the evaporator section, and the evapo­
zone is more effective in PHP1 (see Fig. 5a) compared to the non-
ration momentum force helps the fluid flow in addition to the
uniform PHPs. The condensation rate and the bubble activity in the
unbalanced capillary force in lack of gravity support. It should be noted
condenser are much better for PHP1. Therefore, the heat carried from
that decreasing channel diameter increases evaporation momentum
the evaporator region can be effectively rejected from the condenser
force; however, this force may also cause retardant effect on the flow for
region. Also, flow vibrations caused by vapor bubbles enhance the
the return of fluid in the condenser section by generating flow resis­
convective heat transfer. Effective rejection of heat from the condenser
tance. Therefore, the cross sectional ratio of the parallel channels should
region prevents the early increase of evaporator temperature and, thus,
be optimized. As it is seen from Fig. 3, except the vertical positions
extends the operational range (maximum heating power). Furthermore,
representing the gravity effect, PHP2 (2 mm × 1.5 mm) shows relatively
for PHP2 and PHP3, the flow components (vapor-liquid plugs) slow
better performance than the PHP3 (2 mm × 1 mm). On the other hand,
down especially in the turns of the condensation region. Thus, from the
in the positive-angle positions (IA = 30, 60 and 90◦ ), the gravitational
visual observations and an overall evaluation of all the measurements
force which is one of the most effective driving forces, takes an impor­
and geometrical differences, it is thought that there are four influential
tant role on the flow mechanism. However, the role of gravity on the
factors/phenomena: (1) the role of effective forces, (2) the total liquid
thermal performance can be obviously decreased by using non-uniform
content (3) the effective heat transfer area and (4) heat conduction in the
channel configuration.
heat pipe material.
The results cannot be only explained or discussed via the relative
The general reason underlying the non-uniform channel-designs is to
differences of the effective forces defined above. The liquid content of
generate an unbalanced capillary force to boost the flow motion in PHPs.
the PHPs is another important parameter affecting the results. At the
This is especially important for horizontal positions represent the lack of
same filling ratio, PHP1 includes the highest content of working fluid.
gravity support. The present results also prove the importance of non-
PHP1 contains nearly 12.5% and 22% higher working fluid contents
uniform designs. When the PHPs are placed in the horizontal position
compared to PHP2 and PHP3, respectively. Therefore, more fluid can
(Figs. 3a and b), completely different outcomes are attained. For IA =
easily absorb heat in the evaporator region. If the flow motion occurs,
0◦ (see Fig. 3a), the difference between the thermal resistance curves

10
B. Markal et al. International Communications in Heat and Mass Transfer 125 (2021) 105289

Fig. 7. Variations in thermal resistance and evaporator temperature of all the PHPs at FR = 80%.

11
B. Markal et al. International Communications in Heat and Mass Transfer 125 (2021) 105289

Fig. 8. Effects of filling ratio and inclination angle on heat transfer characteristics for all PHPs.

the fluid can transport much more thermal energy to the condenser re­ of the PHPs are same. Therefore, PHP1 has the thinnest fins but the
gion; and thus, more heat can be rejected from the condenser region to largest heat transfer area. Increasing the heat transfer area enhances
the heat sink. Therefore, one of the probable reasons to obtain better thermal performance. On the other hand, it should be noted that the
thermal performance for PHP1 (lower evaporator temperature and smallest heat transfer area belongs to PHP3. This kind of geometrical
higher maximum heat input value) in the angled positions is the higher difference also engenders another influential factor. This factor is the
fluid content. heat conduction in the heat pipe material. Copper is used as the heat
Another important factor occurring due to the geometry differences pipe material, which is one of the well-known heat spreader materials
is the difference between the heat transfer areas. Here, heat transfer area having quite higher thermal conductivity (nearly 395 W m− 1 ◦ C− 1).
means the wetted area by the fluid, in other words, the internal surface Especially flat plate pulsating heat pipes are mono-block structures and
area contacted by fluid. The turn number, and thus the channel numbers convenient to axial conduction and parallel channel interaction in the

12
B. Markal et al. International Communications in Heat and Mass Transfer 125 (2021) 105289

Fig. 9. Results obtained for the empty PHPs.

radial direction. In this context, fin thickness difference probably has an

influence on the thermal behavior.
Detailed evaluation regarding the filling ratio of 50% is performed in
the previous paragraphs. The results belong to the filling ratio of 80%
are presented in Fig. 7. When the filling ratio increases to 80%, signif­
icant differences occur, and thus, the reaction of the channels
completely changes. The first clear change occurs at the horizontal po­
sition. In IA = 0◦ , the curves of resistance versus heating power of all the
PHPs shows a nearly horizontal trend. The curve of PHP1 is nearly pure
horizontal, which denotes that any motion does not occur in PHP1, and/
or PHP1 completely collapses. In higher heating powers, in the resis­
tance curve of PHP2 and PHP3, a small decrease is seen. This means that
Fig. 10. Flow image at the conditions: PHP1, FR = 80%, IA = 90◦ and Qi =
some vapor bubbles occur in the evaporator region, and evaporation
50 W.
takes place. For FR = 80%, the dependence of PHPs on the inclination
angle is more obvious, especially for uniform channel PHP. For PHP1, in
On the other hand, in the vertical position, the thermal resistance values
addition to the relative magnitude of the resistance values, the
of PHP1 decrease with increasing heating power for both the filling
maximum heating powers are 45 W, 50 W, 55 W and 65 W for 0◦ , 30◦ ,
ratios of 50% and 80%. However, both the thermal resistances and the
60◦ and 90◦ , respectively. The influence of gravity is more obvious on
evaporator temperatures of PHP1 are lower in the filling ratio of 50%.
the thermal performance of PHP1 compared to other PHPs. It can be
Therefore, the performance of PHP1 significantly decreases with
concluded that non-uniform design decreases the influence of gravity on
increasing filling ratio. The underlying reasons can be easily explained
the results. The details related to the effect of filling ratio are clearly
via the flow visualization results. For PHP1, in the filling ratio of 80%,
presented in the next section on the basis of the critical inclination an­
there are many bubbles generating in the evaporator section. However,
gles (0◦ and 90◦ ).
in general, these bubbles are in the form of small vapor plugs swimming
in the liquid medium. Generally, each vapor plug existing in any column
3.2. Influence of filling ratio at various inclination angles rises up in the relevant column, and condensates just at the beginning of
the condenser section. There is not a clear circulatory flow motion. The
The filling ratio is one of the most important parameters, which condenser section is largely covered by the liquid phase. The mass of the
significantly influences the thermal characteristics of heat pipes. How­ liquid columns in front of the advancing bubbles are greater. Pushing of
ever, its effect cannot be evaluated independently from other influential these heavier liquid columns is not easy for vapor plugs. The underlying
parameters. Thus, in Fig. 8, combined effects covering the filling ratio, basic physical reason of all these phenomena is the excess of liquid
PHP types and extreme inclination angles are shown. Also, to better content. Excess of the liquid content restricts the circulation and pul­
evaluate or compare the results obtained for the charged PHPs, the ones sating/oscillating motion and, thus, the flow patterns and transitions
of empty PHPs are presented in Fig. 9. As seen, in the empty case, the change. Effective thin film evaporation and condensation significantly
thermal resistance show horizontal or slightly increasing trend; and the decrease. The flow phenomena defined in the previous section for the
relevant values are higher from Rth = 1. Also, system reaches the filling ratio of 50% disappear for the filling ratio of 80%. For a com­
maximum heat flux at 35 W or 40 W. After stating the characteristics of parison, Fig. 10 is presented for the filling ratio of 80% as the corre­
empty cases, the discussion of the charged PHPs is presented as in the sponding conditions of Fig. 5a. When Fig. 5a and Fig. 10 are compared,
following: From Fig. 8a, for the inclination angle of 0◦ , it is seen that the the abovementioned differences can be clearly seen. Therefore, it can be
data of PHP1 shows an approximately horizontal trend independently concluded that effective absorption of heat in the evaporator region, and
from the filling ratio. This result means the collapse of PHP1 at hori­ the thin film evaporation mechanism phenomena are not solitarily
zontal orientation; which is also validated through the visualization adequate; in addition, the fluid should have the flexibility in the flow
experiments. Therefore, it is concluded that uniform channel PHPs are passages to move easily.
not suitable for horizontal positions, or in other words, they cannot For non-uniform PHPs (PHP2 and PHP3), at the filling ratio of 50%,
maintain a continuous operation being independent of the orientation. the dependence on the orientation is significantly less compared to

13
B. Markal et al. International Communications in Heat and Mass Transfer 125 (2021) 105289

Fig. 11. Flow images at the conditions: FR = 80%, IA = 90◦ and Qi = 45 W: PHP2 (a) and PHP3 (b).

PHP1. Especially for PHP2, this result is much clearer. The closer trends filling ratio variation (from FR = 50% to 80%) is seen. In the existence of
and magnitudes of the data obtained for PHP2 are seen in Figs. 8c and d. higher amount of working fluid, PHP3 cannot show the similar opera­
The basic reason behind this behavior is the cross sectional variation tional characteristics of PHP1 and PHP2. A flow image of PHP3 is pre­
between the neighboring channels. This geometric structure can tolerate sented in Fig. 11b for the conditions of FR = 80%, IA = 90◦ and Qi = 45
the lack of gravity effect by providing unbalanced evaporation mo­ W. In the relevant conditions, only some channels are active, and the
mentum force and unbalanced capillary forces. The details of the rele­ active ones are generally the larger channels. The number of vapor
vant forces are discussed in the previous section. However, filling ratio bubbles is obviously less compared to other PHPs. In addition to the
plays a key role even on the non-uniform PHPs. When the filling ratio above-mentioned mechanisms, there should be an additional factor
increases from 50% to 80%, both the thermal resistances and the based on the geometrical variation to explain the phenomenon. The
evaporator temperatures significantly increase (see Figs. 8c to f). The underlying reason is related to the increasing viscous pressure drop
reason is related to higher amount of liquid covering most of the internal increasing with channel cross sectional difference, as also stated by
volume; which restricts the circulation. The flow image for PHP2 at FR Kwon and Kim [20].
= 80%, IA = 90◦ and Qi = 45 W is presented in Fig. 11a. In some
channels existing near the middle section of the heat pipe, there is 4. Conclusions
partial vapor plug (and thus liquid slug) transitions among a few
channels. However, this does not reflect the general phenomena. The This study is related to FP-CLPHPs, and the influence of cross
general behavior is in the form of column based flow motion, in which sectional variation between neighboring channels is investigated for
tiny bubbles generated in the evaporator region transform to vapor slug different filling ratio and orientations under different heat loads. Flow
and rise in their own column. Then, due to the liquid excess, they visualization is effectively used. Important outcomes are summarized
generally condense at the condenser inlet or even before the condenser below:
inlet. The above-mentioned type operation mode in the existence of
excess liquid content can be called as column type bulk condensation flow, • Influence of gravity is more obvious on thermal performance of PHP1
which is introduced to the literature for the first time. In this mode, the compared to other PHPs. Non-uniform design significantly decreases
vapor bubbles condense as a whole and are added to the liquid medium influence of gravity on the results. At the horizontal orientation for
existing just beneath or above them. medium filling ratio (FR = 50% and IA = 0◦ ), thermal resistance
Lastly, depending on an increase in the filling ratio, the following can curve of PHP1 shows nearly horizontal trend, which means that
be noted for PHP3. As it is seen from Fig. 7, in the filling ratio of 80%, PHP1 collapses. Compared to angled conditions, maximum heat
PHP3 shows the worst thermal performance. Also, from Fig. 8e, a sig­ input of PHP1 drops from 65 W to 45 W, which means nearly 30%
nificant change in the thermal resistance data of PHP3 depending on the

14
B. Markal et al. International Communications in Heat and Mass Transfer 125 (2021) 105289

decrement. In the horizontal orientation, PHP2 and PHP3 maintain [12] F. Xie, X. Li, P. Qian, Z. Huang, M. Liu, Effects of geometry and multisource heat
input on flow and heat transfer in single closed-loop pulsating heat pipe, Appl.
the operation, and perform better thermal performance than PHP1.
Therm. Eng. 168 (2020) 114856.
• In addition to unbalanced capillary force, evaporation momentum [13] M.A. Nazari, M.H. Ahmadi, R. Ghasempour, M.B. Shafii, O. Mahian, S. Kalogirou,
force is also an important factor to boost the operational perfor­ S. Wongwises, A review on pulsating heat pipes: from solar to cryogenic
mance of non-uniform PHPs in the lack of gravity support. It should applications, Appl. Energy 222 (2018) 475–484.
[14] M.A. Nazari, M.H. Ahmadi, R. Ghasempour, M.B. Shafii, How to improve the
be noted that decreasing channel diameter increases evaporation thermal performance of pulsating heat pipes: a review on working fluid, Renew.
momentum force. However, this force may also cause retardant ef­ Sust. Energ. Rev. 91 (2018) 630–638.
fect on the flow for the return of fluid in the condenser section by [15] S. Khandekar, M. Groll, An insight into thermo-hydrodynamic coupling in closed
loop pulsating heat pipes, Int. J. Therm. Sci. 43 (2004) 13–20.
generating flow resistance. Therefore, the cross sectional ratio of the [16] J.L. Xu, Y.X. Li, T.N. Wong, High speed flow visualization of a closed loop pulsating
parallel channels should be optimized. For this reason, in most cases heat pipe, Int. J. Heat Mass Transf. 48 (2005) 3338–3351.
PHP2 presents better thermal performance than PHP3. [17] W. Qu, H.B. Ma, Theoretical analysis of startup of a pulsating heat pipe, Int. J. Heat
Mass Transf. 50 (2007) 2309–2316.
• Performance of uniform PHP significantly decreases with increasing [18] K.H. Chien, Y.T. Lin, Y.R. Chen, K.S. Yang, C.C. Wang, A novel design of pulsating
filling ratio. Underlying physical reason is the excess of liquid heat pipe with fewer turns applicable to all orientations, Int. J. Heat Mass Transf.
content. 55 (2012) 5722–5728.
[19] H.R. Goshayeshi, M. Goodarzi, M. Dahari, Effect of magnetic field on the heat
• In the filling ratio of 80%, for PHP2 and PHP1, general behavior is in transfer rate of kerosene/Fe2O3 nanofluid in a copper oscillating heat pipe, Exp.
the form of column based flow motion, in which tiny bubbles Thermal Fluid Sci. 68 (2015) 663–668.
generated in the evaporator region transform to vapor slug and rise [20] G.H. Kwon, S.J. Kim, Experimental investigation on the thermal performance of a
micro pulsating heat pipe with a dual-diameter channel, Int. J. Heat Mass Transf.
in their own column. Then, due to the liquid excess, they generally
89 (2015) 817–828.
condense at the condenser inlet or even before the condenser inlet. [21] H.R. Goshayeshi, M. Goodarzi, M.R. Safaei, M. Dahari, Experimental study on the
The above-mentioned type operation mode in the existence of excess effect of inclination angle on heat transfer enhancement of a ferrofluid in a closed
liquid content is named as column type bulk condensation flow. loop oscillating heat pipe under magnetic field, Exp. Thermal Fluid Sci. 74 (2016)
265–270.
• The most important advantage of non-uniform PHPs is that they can [22] H.R. Goshayeshi, M.R. Safaei, M. Goodarzi, M. Dahari, Particle size and type effects
properly operate at horizontal position (lack of gravity) for suitable on heat transfer enhancement of Ferro-nanofluids in a pulsating heat pipe, Powder
filling ratios. Regarding the non-uniform PHPs, to effectively trans­ Technol. 301 (2016) 1218–1226.
[23] D.S. Jang, J.S. Lee, J.H. Ahn, D. Kim, Y. Kim, Flow patterns and heat transfer
port thermal energy from evaporator to condenser region, new ge­ characteristics of flat plate pulsating heat pipes with various asymmetric and
ometries may be designed or more effective working fluids can be aspect ratios of the channels, Appl. Therm. Eng. 114 (2017) 211–220.
used. In this regard, thermal transport capabilities of non-uniform [24] C.H. Sun, C.Y. Tseng, K.S. Yang, S.K. Wu, C.C. Wang, Investigation of the
evacuation pressure on the performance of pulsating heat pipe, Int. Commun. Heat
PHPs may improve. Mass Transf. 85 (2017) 23–28.
[25] M.A. Nazari, R. Ghasempour, M.H. Ahmadi, G. Heydarian, M.B. Shafii,
Experimental investigation of graphene oxide nanofluid on heat transfer
Declaration of Competing Interest enhancement of pulsating heat pipe, Int. Commun. Heat Mass Transf. 91 (2018)
90–94.
The authors declared that there is no conflict of interest. [26] M.E. Dehshali, M.A. Nazari, M.B. Shafii, Thermal performance of rotating closed-
loop pulsating heat pipes: experimental investigation and semi-empirical
correlation, Int. J. Therm. Sci. 123 (2018) 14–26.
Acknowledgement [27] E. Sedighi, A. Amarloo, M.B. Shafii, Experimental investigation of the thermal
characteristics of single-turn pulsating heat pipes with an extra branch, Int. J.
Therm. Sci. 134 (2018) 258–268.
The present study was supported by The Scientific and Technological [28] A. Yoon, S.J. Kim, Understanding of the thermo-hydrodynamic coupling in a micro
Research Council of Turkey (TUBITAK) with the project number of pulsating heat pipe, Int. J. Heat Mass Transf. 127 (2018) 1004–1013.
[29] J. Jo, J. Kim, S.J. Kim, Experimental investigations of heat transfer mechanisms of
217M341. a pulsating heat pipe, Energy Convers. Manag. 181 (2019) 331–341.
[30] Z. Li, M.M. Sarafraz, A. Mazinani, H. Moria, I. Tlili, T.A. Alkanhal, M. Goodarzi, M.
References R. Safaei, Operation analysis, response and performance evaluation of a pulsating
heat pipe for low temperature heat recovery, Energy Convers. Manag. 222 (2020)
113230.
[1] M.A. Nazari, M.H. Ahmadi, M.S. Adeh, M.B. Shafii, M. Goodarzi, A review on
[31] Z. Wan, X. Wang, C. Feng, Heat transfer performances of the capillary loop
application of nanofluid in various types of heat pipes, J. Cent. South Univ. 26
pulsating heat pipes with spring-loaded check valve, Appl. Therm. Eng. 167 (2020)
(2019) 1021–1041.
114803.
[2] M.M. Sarafraz, I. Tlili, Z. Tian, M. Bakouri, M.R. Safaei, Smart optimization of a
[32] B. Markal, R. Varol, Experimental investigation and force analysis of flat-plate type
thermosyphon heat pipe for an evacuated tube solar collector using response
pulsating heat pipes having ternary mixtures, Int. Commun. Heat Mass Transf. 121
surface methodology (RSM), Physica A 534 (2019) 122146.
(2021) 105084.
[3] M.M. Sarafraz, Z. Tian, I. Tlili, S. Kazi, M. Goodarzi, Thermal evaluation of a heat
[33] K.S. Yang, Y.C. Cheng, M.C. Liu, J.C. Shyu, Micro pulsating heat pipes with
pipe working with n-pentane-acetone and n-pentane-methanol binary mixtures,
alternate microchannel widths, Appl. Therm. Eng. 83 (2015) 131–138.
J. Therm. Anal. Calorim. 139 (2020) 2435–2445.
[34] B. Markal, K. Aksoy, The combined effects of filling ratio and inclination angle on
[4] K. Hooman, M.R. Safaei, H. Togun, M. Dahari, Thermohydraulics of the liquid films
thermal performance of a closed loop pulsating heat pipe, Heat Mass Transf. 39
in rotating heat pipes, Int. J. Numer. Method H. 30 (2020) 2861–2866.
(2020).
[5] B. Markal, R. Varol, Thermal investigation and flow pattern analysis of a closed-
[35] A. Takawale, S. Abraham, A. Sielaff, P.S. Mahapatra, A. Pattamatta, P. Stephan,
loop pulsating heat pipe with binary mixtures, J. Braz. Soc. Mech. Sci. Eng. 42
A comparative study of flow regimes and thermal performance between flat plate
(2020) 549.
pulsating heat pipe and capillary tube pulsating heat pipe, Appl. Therm. Eng. 149
[6] B. Markal, O. Aydin, M. Avci, Exergy analysis of a counter-flow Ranque-Hilsch
(2019) 613–624.
vortex tube having different helical vortex generators, Int. J. Exergy 10 (2012)
[36] D. Xu, T. Chen, Y. Xuan, Thermo-hydrodynamics analysis of vapor–liquid two-
228–238.
phase flow in the flat-plate pulsating heat pipe, Int. Commun. Heat Mass 39 (2012)
[7] B. Markal, The effect of total flowrate on the cooling performance of swirling
504–508.
coaxial impinging jets, Heat Mass Transf. 55 (2019) 3275–3288.
[37] G. Spinato, N. Borhani, J.R. Thome, Operational regimes in a closed loop pulsating
[8] B. Markal, A. Candan, O. Aydin, M. Avci, Experimental investigation of flow
heat pipe, Int. J. Therm. Sci. 102 (2016) 78–88.
boiling in single minichannels with low mass velocities, Int. Commun. Heat Mass
[38] S.J. Kline, F.A. McClintock, Describing uncertainties in single-sample experiments,
Transf. 98 (2018) 22–30.
Mech. Eng. 75 (1953) 3–8.
[9] D. Mangini, M. Mameli, A. Georgoulas, L. Araneo, S. Filippeschi, M. Marengo,
[39] Y. Zhang, A. Faghri, Advances and unsolved issues in pulsating heat pipes, Heat
A pulsating heat pipe for space applications: ground and microgravity experiments,
Transfer Eng. 29 (2008) 20–44.
Int. J. Therm. Sci. 95 (2015) 53–63.
[40] D. Bastakoti, H. Zhang, W. Cai, F. Li, An experimental investigation of thermal
[10] R. Heydarian, M.B. Shafıi, A.R. Shirin-Abadi, R. Ghasempour, M.A. Nazari,
performance of pulsating heat pipe with alcohols and surfactant solutions, Int. J.
Experimental investigation of paraffin nano-encapsulated phase change material
Heat Mass Transf. 117 (2018) 1032–1040.
on heat transfer enhancement of pulsating heat pipe, J. Therm. Anal. Calorim. 137
(2019) 1603–1613.
[11] M.M. Sarafraz, I. Tlili, M.A. Baseer, M.R. Safaei, Potential of solar collectors for
clean thermal energy production in smart cities using Nanofluids: experimental
assessment and Effciency improvement, Appl. Sci. 9 (2019) 1877.

15
B. Markal et al. International Communications in Heat and Mass Transfer 125 (2021) 105289

[41] V.G. Pastukhov, Yu.F. Maydanik, Development of a pulsating heat pipe with a [43] T. Alam, W. Li, F. Yang, W. Chang, J. Li, Z. Wang, J. Khan, C. Li, Force analysis and
directional circulation of a working fluid, Appl. Therm. Eng. 109 (2016) 155–161. bubble dynamics during flow boiling in silicon nanowire microchannels, Int. J.
[42] Y. Xu, Y. Xue, H. Qi, W. Cai, Experimental study on heat transfer performance of Heat Mass Transf. 101 (2016) 915–926.
pulsating heat pipes with hybrid working fluids, Int. J. Heat Mass Transf. 157
(2020) 119727.

16