Proceedings of the Canadian Society for Mechanical Engineering International Congress

Computational Fluid Dynamics Canada International Congress

CSME - CFD-SC2023
May 28-31, 2023, Sherbrooke, QC, Canada

Analysis of A Thermosyphon Heat Pipe with Modified Geometry

Ali Tarokh1*, Çetin Yavuz2, Aydın Dikici3

1
Department of Mechanical Engineering, Lakehead University, Thunder Bay, Canada
2
Department of Electrical and Energy, Tatvan Vocational School, Bitlis Eren University, Bitlis, Turkey
3
Department of Energy System Engineering, Fırat University, Elazığ, Turkey
* [email protected]

Abstract— In this study, an experimental and 3-dimensional is effectively used in engineering fields due to its visualization
numerical analysis of a thermosiphon heat pipe (THP) using possibilities and extremely efficient.
two-phase fluid cycling was performed to model the heat Kamburova et al. [8] carried out a numerical analysis in
transfer process by evaporation and condensation. R404A which they modeled the simultaneous evaporation and
refrigerant was chosen as the working fluid for the THP and condensation event in a thermosyphon in 2D. They used the
50% of the evaporator section was filled with the working fluid. VOF technique and distilled water as the working fluid in their
Numerical analysis was performed using ANSYS FLUENT work. As a result of their studies, they reported that they
program and VOF method. In the meantime, in order to model successfully realized the complex flow events occurring in the
the evaporation and condensation process successfully, a UDF thermosyphon. At the same time, they stated that it was in good
code written in C was used and introduced to the fluent agreement with the experimental results.
program. As a result of the study, it was seen that the Yang et al. [9] carried out an experimental and numerical
temperature values obtained from the experimental and analysis of the thermal properties of a heat pipe. Researchers
numerical analysis for the evaporator and condenser parts of the using UDF and VOF techniques in their studies stated that
THP were in good agreement. As a result of the numerical evaporation and condensation phenomena can be simulated with
analyzes carried out in 3D, the contours of the evaporation and these methods. At the same time, they stated that the axial
condensation process and the heat transfer process are presented velocity fluctuation in most areas of the heat pipe was very
visually. The analysis took a total of 200 seconds. small, and they reported that the velocity magnitudes changed
from 0 to the maximum value from the end of the evaporator and
Keywords-component; Thermosyphon heat pipe; Evaporation condenser sections, and the maximum value was maintained in
and condensation; ANSYS; VOF the adiabatic section.
Tarokh et al. [10] carried out a numerical analysis to improve
I. INTRODUCTION the heat transfer performance of a thermosyphon with a vortex-
Heat pipes and/or thermosiphons are heat transfer devices forming barrier placed in its body. The researchers stated that the
that perform the fluid cycle in two phases in a vacuumed closed positioning of the barrier in the adiabatic and condenser sections
tube. A thermosiphon consists of 3 main parts, known as the had a very good effect on the overall thermal resistance and
evaporator, where the heat is given to the working fluid, average wall temperature distribution.
adiabatic, and the condenser, where the condensation event is Fadhl et al. investigated a thermosyphon heat pipe in two
carried out. In the evaporator section, the working fluid is heated different studies. They used water, r134a and r404a as working
and evaporation is carried out. The resulting steam then reaches fluid. The researchers, who evaluated the thermosyphon heat
the adiabatic section and then the condenser section. Here, the pipe using different heat inputs, stated that the average
steam coming into contact with the surface of the condenser temperature values obtained from the numerical and
section below the saturation temperature condenses and returns experimental analyzes were in good agreement as a result of
to the evaporator section again. This cycle continues throughout their studies. At the same time, they have successfully
the operation of the THP [1]. THP's are used in many areas [2- visualized the evaporation and condensation phenomenon using
7]. As a result of the literature research, it has been seen that VOF and UDF techniques [11, 12].
many 2D numerical analyzes have been carried out for Kim et al. [13] conducted a study examining the effects of
thermosyphons. However, there is no comprehensive 3D mass transfer time relaxation parameters on a thermosyphon
numerical analysis that reaches steady state. Therefore, this heat pipe. Using the mass transfer relaxation parameter as 0.1 for
study is extremely important in this respect. As it is known, CFD evaporation and 4 different values for condensation, they stated
analysis, which is used in the solutions and visualization of that the most appropriate 𝛽𝑐 value was 0.1×( 𝜌𝐿 /𝜌𝑉 ). Using
complex flow events occurring in the interior of thermosiphons, water as the working fluid, researchers simulated evaporation
and condensation using VOF and UDF techniques in their

ID 370 – Heat transfer symposium 1

numerical analysis. Wang et al. [14] reported that the 𝛽𝑒 value Energy equation
should be approximately 10-1000 times greater than the 𝛽𝑐 𝜕
(𝜌𝑒) + ∇. [𝑢⃗ (𝜌𝑒 + 𝜌)] = ∇. (𝑘. ∇T) + 𝑆𝐸 (3)
value. 𝜕𝑡
The terms T, ρ, 𝑢 ⃗ , k, t, and SE in the energy equation are the
This is also an indication that the mass transfer time
mass-averaged temperature shared by the phases, density,
relaxation parameters can be selected in accordance with the
velocity, thermal conductivity, time, and source terms,
experimental results.
respectively. The term e given in the equation represents the
In this context, experimental and 3-dimensional numerical
specific internal energy and is expressed by the equation below.
analyzes of a Thermosiphon heat pipe with a finned structure
were carried out in the presented study. In the meantime, it is of 𝜌𝐿 𝛼𝐿 𝑒𝐿 +𝜌𝑉 𝛼𝑉𝑒𝑉
great importance to choose the appropriate mass transfer time 𝑒= (4)
𝜌𝐿 𝛼𝐿 +𝜌𝑉 𝛼𝑉
relaxation parameters for successful evaporation and
condensation analysis. The mentioned values were determined When Equation 4 is examined, it is seen that the terms 𝑒𝐿 and
as 0.1 for evaporation and 20 for condensation in this study. In 𝑒𝑉 exist. These stand for the specific internal energy of the liquid
addition, the numerical analyzes performed in this study were and vapor phases, respectively, and these terms defined by the
carried out using ANSYS FLUENT 2022R1 using high- caloric equation of state [19] are expressed in equations 5 and 6.
performance computing (HPC) systems from the Digital
Research Alliance of Canada. 𝑒𝐿 = 𝐶𝑝𝐿 (𝑇 − 𝑇𝑠𝑎𝑡 ) (5)
(𝑇
𝑒𝑉 = 𝐶𝑝𝑉 − 𝑇𝑠𝑎𝑡 ) (6)
II. MATERIALS AND METHOD
Also, the terms ρ and k given in the energy equation are
As mentioned before, complex flow events occurring inside volume-averaged parameters and are expressed as follows.
the thermosiphon heat pipes are carried out by numerical
analysis. However, the numerical analyzes in question are based 𝜌 = 𝛼𝑉 𝜌𝑉 + (1 − 𝛼𝑉 )𝜌𝐿 (7)
on the finite volume method and the analyzes performed with 𝑘 = 𝛼𝑉 𝑘𝑉 + (1 − 𝛼𝑉 )𝑘𝐿 (8)
two phases are much more difficult than those with single phase.
Because the interfaces between the phases are movable in the Momentum equation
analyzes carried out as two-phases. This situation brings with it 𝜕
(𝜌𝑢
⃗ ) + ∇. (𝜌𝑢
⃗𝑢⃗ 𝑇 ) = −∇𝑝 + ∇. [𝜇(∇𝑢 ⃗ 𝑇 )] + 𝜌𝑔 + 𝐹𝐶𝑆𝐹
⃗ + ∇𝑢 (9)
a more intense computational effort [15, 16, 17]. As a result of 𝜕𝑡
the literature research, it has been seen that the VOF method is The terms p, 𝑔, μ, and 𝐹𝐶𝑆𝐹 in the momentum equation are
the most suitable method in terms of its convenience in local pressure, gravity, viscosity, and volumetric surface tension
monitoring the interface, and this method has been used in many force, respectively. Here, the viscosity is volume-averaged and
studies in this field. With the VOF method, surface tracking of is expressed as follows. However, the volumetric surface tension
two or more fluids that do not mix with each other is carried out. force is expressed by equation 11 [20].
Therefore, the VOF method was used in the numerical analysis
in this study [15, 16, 17]. In this method, the volume fractions in 𝜇 = 𝛼𝑉 𝜇𝑉 + (1 − 𝛼𝑉 )𝜇𝐿 (10)
𝛼 𝜌 𝐶 ∇𝛼 +𝛼 𝜌 𝐶 ∇𝛼
both phases are added to the calculations and the sum of the 𝐹𝐶𝑆𝐹 = 2𝜎𝑙𝑣 ( 𝐿 𝐿 𝑉 𝑉 𝑉 𝑉 𝐿 𝐿 ) (11)
𝜌𝑉 +𝜌𝐿
volume fractions of both phases in any control volume is equal
The terms 𝜎𝑙𝑣 , 𝐶𝑉 , and 𝐶𝐿 given in Equation 11 are surface
to one. This situation is expressed in the equation below.
tension and surface curvetures for vapor and liquid, respectively.
𝛼l + 𝛼v = 1 (1) The expressions suggested by Lee and De schepper et al [21,
15] were used to calculate the mass and energy transfer in the
If αl=1, the cell is completely filled with liquid. numerical analysis of thermosiphon working in two phases.
If αv=1, the cell is completely filled with vapor. These expressions are given below:
For the mass transfer process;
If 0 < αl < 1, both liquid phase and vapor phase are present Evaporation: 𝑇𝑚𝑖𝑥 > 𝑇𝑠𝑎𝑡
in the cell [18].
𝑇𝑚𝑖𝑥 −𝑇𝑠𝑎𝑡
At the same time, since the analysis performed in this study 𝑆𝑀 = −𝛽𝑒 𝛼𝑙 𝜌𝑙 | | (12)
𝑇𝑠𝑎𝑡
was performed in two phases, the continuity, momentum, and 𝑇𝑚𝑖𝑥 −𝑇𝑠𝑎𝑡
𝑆𝑀 = 𝛽𝑒 𝛼𝑙 𝜌𝑙 | | (13)
energy equations given below are expressed as vapor and liquid 𝑇𝑠𝑎𝑡
phases. Condensation: 𝑇𝑚𝑖𝑥 < 𝑇𝑠𝑎𝑡
The equations used for numerical analysis of the 𝑇 −𝑇
𝑆𝑀 = 𝛽𝑐 𝛼𝑉 𝜌𝑉 | 𝑚𝑖𝑥 𝑠𝑎𝑡| (14)
thermosyphon heat pipe are given below. 𝑇𝑠𝑎𝑡
𝑇𝑚𝑖𝑥 −𝑇𝑠𝑎𝑡
𝑆𝑀 = −𝛽𝑐 𝛼𝑉 𝜌𝑉 | | (15)
𝑇𝑠𝑎𝑡
Continuity equation For the heat transfer process;
𝜕
(𝛼𝐿 𝜌𝐿 ) + ∇. (𝛼𝐿 𝜌𝐿 𝑢
⃗ ) = 𝑆𝑀 (2) Evaporation: 𝑇𝑚𝑖𝑥 > 𝑇𝑠𝑎𝑡
𝜕𝑡
𝑇 −𝑇
The terms 𝛼𝐿 , 𝜌𝐿 , 𝑢⃗ , t, and SM in this equation represent the 𝑆𝐸 = −𝛽𝑒 𝛼𝑙 𝜌𝑙 | 𝑚𝑖𝑥 𝑠𝑎𝑡| . 𝐿𝐻 (16)
𝑇𝑠𝑎𝑡
volume fraction for the liquid phase, density for the liquid phase, Condensation: 𝑇𝑚𝑖𝑥 < 𝑇𝑠𝑎𝑡
velocity, time, and the source term for mass, respectively. 𝑇 −𝑇
𝑆𝐸 = 𝛽𝑐 𝛼𝑉 𝜌𝑉 | 𝑚𝑖𝑥 𝑠𝑎𝑡| . 𝐿𝐻 (17)
𝑇𝑠𝑎𝑡
The terms Tmix, Tsat, LH, 𝛽𝑒 and 𝛽𝑐 in the equations given for
the mass and heat transfer process, the mixture temperature,

ID 370 – Heat transfer symposium 2

saturation temperature, latent heat, the mass transfer time
relaxation parameters for evaporation and condensation,
respectively. As mentioned before, these values should be
chosen in accordance with the experimental results in order to
perform the evaporation and condensation analyzes successfully
and to match the experimental and numerical results. In this
study, 𝛽𝑒 and 𝛽𝑐 parameters were determined as 0.1 and 20,
respectively. At the same time, the equations given above for
mass and heat transfer were written in UDF using the C
programming language and introduced to the fluent program.

A. Mesh and boundary conditions applied in numerical

Condenser
analysis
The thermosyphon heat pipe presented in this study has an
outer diameter of 25 mm and a total length of 700 mm. The
material from which THP is produced is aluminum. The lengths
of the evaporator, adiabatic and condenser sections are 120, 270
and 310 mm, respectively, and are shown in Fig. 1a. The
thermosiphon heat pipe consists of 3556874 nodes and 3430000
elements, and the mesh structure in question is indicated in Fig.
1b. The boundary condition for the inner walls of the
thermosiphon heat pipe is chosen as no-slip. The gravitational
acceleration is taken as -9.81 m/s2. A constant temperature
boundary condition has been applied for the evaporator and
condenser section, and these temperature values are 323.15 and
283.15 K, respectively. The zero-flux boundary condition was
chosen for the adiabatic section. The operating pressure value
was determined as 1084400 Pascal in accordance with the
saturation temperature taken as 293.15 K. The initialization
process was carried out at the saturation temperature after the
patch process was performed in such a way that 50% of the
evaporator section was the working fluid and the remaining
section was steam. The working fluid was used as R404A. The
analysis was carried out at 20 C saturation temperature and the Adiabatic
properties of the fluid were determined in accordance with this
temperature value and given in Table 1.

Evaporator

a- THP Sections b- Mesh

Figure 1. THP sections and mesh structure

Identify applicable sponsor/s here. (sponsors)

ID 370 – Heat transfer symposium 3
 Table 1. Properties of working fluid the first times when the analysis starts. As a result of heating the
Properties Primary phase Secondary phase
evaporator section, the temperature value of the working fluid
increased over time and evaporation started with the formation
ρ [kg/m ] 3
56.306 1067.3 of bubbles at the points where it reached the boiling point. These
μ [kg/m.s] 0.011849 0.00013748 bubbles that emerged later moved upwards in the pool and came
to the upper part of the pool. Here, the steam in the bubbles
k [W/m.K] 0.015178 0.065499 reached the upper part and was released and started to rise
𝜎𝑙𝑣 = 0.09805856 − 1.845 × 10−5 𝑇 − towards the condenser part with the effect of the pressure
σlv[11]
2.3 × 10−7 𝑇 2 (18) difference. Here, the vapor condensed in the form of droplets
Cp [J/kg.K] 1163.6 1502 came back to the evaporator section with the effect of gravity.
LH [kJ/kg] 374.74 228.75
As a result of this event, which took place during the numerical
analysis process, after a small decrease in the amount of liquid
The time step size used in the numerical analysis was chosen occurred, it remained stable after 200 seconds, and thus the
as 0.0005 and thus the courant number remained below 2 system reached a stable structure.
throughout the analysis. The SIMPLE algorithm has been
chosen for pressure-velocity coupling. First-order upwind
discretization, Geo-Reconstruct and Presto discretization were
chosen for momentum and energy equations, volume fraction
and pressure interpolation, respectively. Explicit formulation
was used in the numerical analysis performed within the scope
of the study. Residuals are determined as 10-4 for volume
fraction and velocity, and 10-6 for energy. The saturation
temperature and enthalpy value of the working fluid are
specified in the UDF code.
III. RESULTS AND DISCUSSION
The numerical analyzes carried out within the scope of this
study took a total of 200 seconds. The said numerical analysis
became steady state after 200 seconds. As a result of the constant
temperature boundary condition applied to the walls of the
evaporator section, the working fluid in this section is heated and
nucleate boiling occurs. The steam that emerges later reaches the
adiabatic and then the condenser section over time. The steam,
which comes into contact with the walls of the condenser section
below the saturation temperature, condenses here and reaches
the evaporator section again. This process was continued until
the system reached the steady state and at the end of 200 seconds,
when the system reached the steady state, the analysis was
completed. Images of the above-mentioned process are given
later in this section.
The average temperature values obtained from the
experimental and numerical results of the evaporator and
condenser section are given in Table 2.

Table 2. Average temperature values obtained from experiment and

numerical analysis and Relative error
Section TEXP,ave (K) TCFD,ave (K) Re(%)

Evaporator 320.97003 313.3941618 2,36

Condenser 297.4961 299.16323895 0.56

When the table above is examined, it is seen that the average

temperature values obtained from the experimental and
numerical analyzes are in perfect harmony with each other.
Especially when the relative error values are looked at, it can be
clearly seen that the condenser and evaporator sections fit very
well.
Fig. 2 shows the contours of the evaporation and
condensation event occurring in the THP. When these contours
Figure 2. Evaporation and condensation contours of THP in 0.1-0.2-5-15-
are examined, it can be seen that the pool boiling event occurs in 50-100-150 and 200 seconds, respectively
the form of nucleate boiling and the condensation takes place in

ID 370 – Heat transfer symposium 4

 Fig. 3 shows the temperature contours that occur inside the IV. CONCLUSIONS
THP. Between 0.1 and 15th seconds, the temperature increased In this study, an experimental and 3D numerical analysis of
in seconds and after 15th seconds, the average temperature value a thermosiphon was carried out. The average temperature values
of THP no longer changed and became uniform. When Fig. 3 is obtained from the experiments for the condenser and evaporator
examined, it can be easily seen that the temperature of the THP sections were compared with the values obtained from the CFD
no longer changes after 15th seconds. analysis. It has been seen that the results are in very good
agreement. especially the average temperature values of the
condenser section were found to be very close to the
experimentally obtained values. As a result of the CFD analysis
performed, the evaporation and condensation event was
successfully modeled and visualized.
ACKNOWLEDGMENT
The authors are deeply grateful to Mert GÜRTÜRK, Beşir
KOK, and Cihangir KALE for allowing the use of experimental
setups.
The authors thank Compute Canada for giving them access
to the computational resources.
This publication was produced from the PHD Thesis named
Realization of experimental and numerical analysis of a new
thermosiphon heat pipe that will realize the thermal energy
transfer from liquid to gas.
REFERENCES
[1] J. Legierski, B. Wie, G. De Mey, et al., Measurements and simulations of
transient characteristics of heat pipes, Microelectron. Reliab. 46 (1), 109–
115, 2006.
[2] Y.-C. Weng, H.-P. Cho, C.-C. Chang, S.-L. Chen, Heat pipe with PCM
for electronic cooling, Appl. Energy 88 (5), 1825–1833, 2011.
[3] D. Jafari, A. Franco, S. Filippeschi, P. Di Marco, Two-phase closed
thermosyphons: a review of studies and solar applications, Renew.
Sustain. Energy Rev. 53, 575–593, 2016.
[4] H. Shabgard, M.J. Allen, N. Sharifi, S.P. Benn, A. Faghri, T.L. Bergman,
Heat pipe heat exchangers and heat sinks: opportunities, challenges,
applications, analysis, and state of the art, Int. J. Heat Mass Transfer 89,
138–158, 2015.
[5] L.L. Vasiliev, S. Kakaç, Heat Pipes and Solid Sorption Transformations
Fundamentals and Practical Applications, Ed. CRC Press- Taylor &
Francis Group, 2013.
[6] T. He, C. Mei, J.P. Longtin, Thermosyphon-assisted cooling system for
refrigeration applications, Int. J. Refrig. 74, 165-176, 2017.
[7] A.A. Eidan, S.E. Najim, J.M. Jalil, Experimental and numerical
investigation of thermosyphone performance in HVAC system
applications, Heat Mass Transfer 52, 2879–2893, 2016.
[8] V. Kamburova, A. Ahmedov, I. Iliev, I. Beloev, I. Pavlović, Numerical
modelling of the operation of a two-phase thermosyphon, Thermal
Science 22, 1311-1321, 2018.
[9] K.M. Yang, Z. Cong & W.F. Wang. CFD simulation of flow and heat
transfer in a thermosyphon. Sustainable Development, 1038-1044, 2016.
[10] A. Tarokh, C. Bliss, A. Hemmati, Performance enhancement of a two-
phase closed thermosyphon with a vortex generator, Appl. Therm. Eng.,
182, 116092, 2021.
[11] B. Fadhl, L.C. Wrobel, H. Jouhara, Numerical modelling of the
Figure 3. Temperature contours of THP in 0.1-5-15-50-100-150 and 200
temperature distribution in a two-phase closed thermosyphon, Appl.
seconds, respectively
Therm. Eng. 60, 122–131, 2013.
[12] B. Fadhl, L.C. Wrobel, H. Jouhara, CFD modelling of a two-phase closed
thermosyphon charged with R134a and R404a, Appl. Therm. Eng. 78,
482–490, 2015.
[13] Y. Kim, J. Choi, S. Kim, Y. Zhang, Effects of mass transfer time relaxation
parameters on condensation in a thermosyphon, J. Mech. Sci. Technol. 29,
5497-5505, 2015.
[14] X. Wang, Y. Wang, H. Chen, Y. Zhu, A combined CFD/visualization
investigation of heat transfer behaviors during geyser boiling in two-phase
closed thermosyphon. Int. J. Heat Mass Transf. 121, 2020.

ID 370 – Heat transfer symposium 5

[15] S.C.K. De Schepper, G.J. Heynderickx, G.B. Marin, Modeling the [18] ANSYS FLUENT, Theory Guide (Release 13.0). Multiphase Flows,
evaporation of a hydrocarbon feedstock in the convection section of a ANSYS, Inc., November 2010, pp. 455-568 (chapter 17).
steam cracker, Computers & Chemical Engineering 33, 122-132, 2009. [19] H. Jouhara, A.J. Robinson, Experimental investigation of small diameter
[16] H.K. Versteeg, W. Malalasekera, An Introduction to Computational Fluid two-phase closed thermosyphons charged with water, FC-84, FC-77 and
Dynamics; the Finite Volume Method, 2nd ed., Prentice-Hall, Harlow, FC-3283, Appl. Therm. Eng. 30 (2–3), 201–211, 2010.
2007. [20] J.U. Brackbill, D.B. Kothe, C. Zemach, A continuum method for modeling
[17] J.D. Anderson, Computational Fluid Dynamics the Basics with surface tension, J. Comput. Phys. 100 (2), 335–354, 1992.
Applications, McGraw-Hill, New York, 1995. [21] W.H. Lee, Pressure iteration scheme for two-phase flow modeling, Multi-
phase Transp. Fundam. React. Safety, Appl. 1, 407–431, 1980.

ID 370 – Heat transfer symposium 6