Thermal Science and Engineering Progress 18 (2020) 100520

Contents lists available at ScienceDirect

Thermal Science and Engineering Progress

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

Transient simulation of finned heat sinks embedded with PCM for T

electronics cooling
Adeel Arshada, Mark Jabbala, Pouyan Talebizadeh Sardaria, Muhammad Anser Bashirb,
Hamza Farajic, Yuying Yana,d,
⁎

a
Fluids & Thermal Engineering (FLUTE) Research Group, Faculty of Engineering, University of Nottingham, Nottingham, NG7 2RD, UK
b
Department of Mechanical Engineering, Mirpur University of Science & Technology (MUST), Mirpur 10250 AJK, Pakistan
c
Physics Department, LPMMAT Laboratory, Faculty of Sciences Ain Chock, Hassan II University, Casablanca, Morocco
d
Research Centre for Fluids and Thermal Engineering, University of Nottingham Ningbo China, Ningbo 315100, China

ARTICLE INFO ABSTRACT

Keywords: This paper reports the two-dimensional (2D ) transient numerical simulation of a phase change material (PCM)
Passive cooling based finned heat sink to investigate the heat transfer performance for passive cooling of electronic devices. The
Finned heat sink finned heat sinks of 2 mm and 3 mm fin thickness are employed with a constant fin volume fraction of 9% , acting
Phase change material as thermal conductivity enhancer (TCE). The n-eicosane is employed as a PCM inside the heat sink to store the
Thermal conductivity enhancer
heat generated from the electronic device applied at the heat sink base. Transient numerical simulations are
N-eicosane
performed using finite-volume-method and conjugate heat transfer and melting/solidification phenomenon are
investigated by applying various power levels. The numerical results show that the employed PCM with low
temperature keeps the heat sink base temperature in lower limits and uniform melting is observed inside the
finned heat sink. With the increase of heating power level, the PCM melting time is decreased for fin thickness
heat sinks. By increasing the power level from 4 to 6 W, for the case of 3 mm fin thickness, the melting time
increases by 6.63%, 3.59% and 1.90% by 3 mm fin thickness heat sink, compared to the 2 mm fin thickness heat
sink. The developed equations of liquid fraction and modified Nusselt number are obtained as function of
modified Fourier number, Stefan number, and Rayleigh number which provide guidelines for generalizing the
thermal performance of PCM based finned heat sinks.

1. Introduction because of high heat storage density and isothermal process during
phase transition [6].
The miniaturization and advanced multi-functionality of electronic Several studies have been reported which introduced the finned
devices have undergone a great revolution in current era. Both features heat sinks as a TCE, filled with PCMs which have high specific and
result in the overheating of electronic components and lead to the un- latent heat of fusion with small change in volume. A numerical study
reliable and dramatic increase in operating temperature [1]. Around proposed by Nayak et al. [7] for TM of electronics using porous matrix
55% of failures in operating electronic devices are due to the increase in and finned (plate-type and rod-type) as a TCEs for PCM filled heat sinks.
device temperature [2]. Therefore, to overcome the overheating phe- TCEs of 5%, 10% and 15% volume fractions of TCE were selected in the
nomenon and efficient temperature distribution, a novel and efficient case of finned heat sinks. Authors marked two findings: firstly 10% had
cooling technology is needed to avoid any type of failures. Many re- the better thermal performance and secondly rod-type TCE was more
searchers are working on the electronic cooling system by using pie- significant in terms of convection of melt front of PCM to transfer heat,
zoelectric energy harvesting mechanism specifically in the field of which in turn lead to lower chip temperature. Shatikan et al. [8] con-
mechanical and aerospace engineering [3–5]. In recent years, the in- ducted the transient numerical simulation of a PCM-based heat sink at
vention of passive thermal management (TM) of electronics devices constant heat flux and performed a dimensional analysis. The results
using phase change materials (PCMs) and various thermal conductivity were presented with good agreement with melt fraction in terms of
enhancers (TCEs) have introduced a new direction in cooling technol- Nusselt, Fourier, modified Stefan and Rayleigh numbers. Further, Saha
ogies. PCMs are the best competitors for cooling of electronics devices and Dutta [9,10] proposed the heat transfer correlation of a plate-fin

⁎
Corresponding author at: Fluids & Thermal Engineering (FLUTE) Research Group, Faculty of Engineering, University of Nottingham, Nottingham, NG7 2RD, UK.
E-mail address: [email protected] (Y. Yan).

https://doi.org/10.1016/j.tsep.2020.100520

2451-9049/ Crown Copyright © 2020 Published by Elsevier Ltd. All rights reserved.
A. Arshad, et al. Thermal Science and Engineering Progress 18 (2020) 100520

Nomenclature Tm Melting temperature (K)

Tini Initial temperature (K)
Am Mushy zone Ts Solidus Temperature (K)
CFD Computational fluid dynamics Tl Liquidus temperature (K)
cp Specific heat (J/kg.K) Tref Reference temperature (K)
FVM Finite volume method t Time (sec)
Fo Fourier number tf Thickness of fin
Fo Modified Fourier number tp Thickness of PCM between two consecutive fins
g Gravity (m/s2 ) u Velocity components in x direction
H Total enthalpy (J/kg) VTCE Volume of TCEs
h Sensible enthalpy (J/kg) VHS Volume of heat sink
href Reference enthalpy (J/kg) v Velocity components in y direction
H Height of heat sink W Width of heat sink
hf Height of fin 2D Two dimensional
k Thermal conductivity (W/m.K) T Difference temperature (K)
LHSU Latent heat storage unit H Latent heat (J/kg)
L Latent heat of fusion (J/kg)
Nu Modified Nusselt number Greek letters
PCM Phase change material
P Pressure (Pa) Thermal expansion coefficient (1/K)
q Heat flux (W/m2) cp Thermal capacitance (J/m3 K)
Ra Rayleigh number Thermal diffusivity (m2 /s)
Ra Modified Rayleigh number µ Dynamic viscosity (Pa.s)
SIMPLE Semi-Implicit Pressure-Linked Equation Density (kg/m3 )
Ste Stefan number ref Reference density (kg/m3 )
Ste Modified Stefan number Porosity function
S Source term in momentum equation Volume fraction of TCEs
Sh Source term in energy equation Dimensionless temperature
TCE Thermal conductivity enhancer Dimensionless time
TM Thermal management fl Liquid fraction
T Temperature (K)

PCM based heat sink of various enclosures and heat fluxes. The authors proposed a numerical study to obtain the best fin size and location in a
suggested that a single correlation of Nu and Ra numbers is not enough rectangular enclosure to give the minimum melting time for a given fin-
to relate the performance of heat sink with melt fraction. Wang and to-PCM volume ratio to increase the thermal transport of the PCM. The
Yang [11,12] conducted a three dimensional numerical study for dif- authors used four different configurations and found that the fin-to-
ferent number of fins (0, 3 and 6 fins) at heating power level of 2, 3 and PCM volume ratio could be decreased by half and could increase the
4 W. The authors used the n-eicosane as a PCM and found good agree- total volume of the thermal energy storage. Further, it was concluded
ment with experimental results. Ye et al. [13] performed the simulation that the reduction in volume of fins at the top portion of the TES system
of a plate-fin heat storage unit and studied the thermal storage/release did not change the overall rate of PCM melting significantly. Ji et al.
characteristics of paraffin wax at different wall temperature and pre- [19] carried out a numerical investigation of a vertically heated rec-
sented the results in dimensionless variables. Pakrouh et al. [14] per- tangular enclosure to suppress the non-uniform heat transfer of PCM
formed the numerical optimization of number of fins, height, thickness melting with fins angled at 0°, + 15°, + 30°, 15° and 30° . It was found
and base thickness of heat sink using Taguchi method. The results found that 15° fins increased the PCM melting rate significantly and reduced
that the number of fins had the greatest influence, followed by fin the melting duration by 62.7% compared with the same length fins at
thickness and height at low critical temperature while base thickness 0°. Kazemi et al. [20] investigated the arrangement of longitudinal fins
had the least effect on the operating temperature. Kant et al. [15] from 60° to 120° to improve heat transfer by considering the triple-fin
studied the thermal performance of five different PCMs namely capric heat exchangers. The authors found that the best cases for double-fin
acid, lauric acid, myristic acid, palmitic acid and stearic acid filled in (reducing angle from 150° to 45°) and triple-fin (increasing angle from
aluminum containers. The results showed that maximum energy was 60° to 120°) resulted in 62% and 22.5% reduction in melting time, re-
stored by stearic acid and minimum for capric acid with the same spectively, compared with the simple heat exchanger. Kalbasi et al.
boundary conditions. Sahoo et al. [16] conducted the numerical si- [21] numerically investigated the optimum number of fins of PCM filled
mulation of orthotropic composite fins as the TCEs using n-eicosane as a heat sink and proposed relations for optimum fin number, fin spacing
PCM and reported that orthotropic fins showed better thermal perfor- and fin height for various input heat fluxes.
mance in reducing the base temperature compared to isotropic fins. Recently, Arshad and his co-authors [22–26] conducted parametric
Further, it was suggested that using orthotropic fins instead of isotropic experimental studies for square and circular profiles pin-fin heat sinks
fins minimized the thermal constraints, as well as weight, on the based on the number of fins, fin thickness and fin configuration filled
thermal electronic components. Vogel et al. [17] carried out experi- with six different types of PCM of varying melting temperatures and
mental and numerical studies of eutectic mixture of sodium nitrate and latent heat of fusion at constant and intermittent heating. The authors
potassium nitrate (KNO3 + NaNO3 ) as a PCM on a high temperature selected 9% volume fraction of TCEs and concluded that 2 mm and
flat-plate as TES system in a natural convection mode. A convective 3 mm fin thickness of square and circular fins, respectively, had the
enhancement factor was introduced and it was found that natural maximum thermal performance compared with no fin, 1 mm, 3 mm and
convection heat transfer enhancement was increased with higher 4 mm fin thickness PCM-filled heat sinks. Further, Ali et al. [27] con-
widths and smaller heights of TES system. Joshi and Rathod [18] ducted the experimental study using 2 mm fin thickness pin-fin heat

2
A. Arshad, et al. Thermal Science and Engineering Progress 18 (2020) 100520

sink to study the constant and intermittent operation conditions. The Table 1
results revealed that intermittent operating modes showed the max- Dimensions of heat sinks.
imum operation duration by increasing the operation cycles. Ashraf and Parameter Dimensions (mm)
his co-authors [28–30] used the various PCMs of different meting
temperature and latent-heat fusion and filled in circular and square pin- W 72
H 25
fin heat sink having inline and staggered fin distribution. Results re-
hf 20
vealed the best cooling performance of inline arrangement as compared
tf 2&3
to staggered for both circular and square pin-fins heat sinks.
This research activity presents the numerical simulation of finned
heat sinks using different fin thickness of 2 mm and 3 mm considering Table 2
constant fin volume fraction of 9% resulting in different fin numbers. Thermophysical properties of PCM and TCE.
Both heat sinks are filled with n-eicosane as the PCM and three different
Property n-Eicosane Aluminium
heating power levels of 4, 5 and 6 W are applied at the heat sink base to
explore the thermal enhancement performance of PCM based finned Tm (K) 309.5 –
heat sinks. This will eventually provide a better picture to select the Ts (K) 308.5 –
optimum configuration of PCM based finned heat sink for the solution Tl (K) 310.5 –
L (J/kg) 241,000 –
of electronic devices as passive cooling technology.
cp (J/kg.K) 2050 871
(kg/m3 ) 790 2719
2. Geometric and mathematical description k (W/m.K) 0.23 202.4
µ (Pa.s) 0.00355 –
2.1. Physics of the problem (1/K) 0.0001 –

In present study, the PCM based finned heat sinks of 2 mm and

walls and a heat flux is implemented at the bottom of the system. The
3 mm fin thickness, which was experimentally investigated by Arshad
dimensions the physical domain are mentioned in Table 1. The ther-
et al. [22,23] and Ashraf et al. [28,29], were numerically modelled to
mophysical properties of used PCM (n-eicosane) and aluminium are
explore the thermal performance of heat sink by varying input para-
listed in Table 2.
meters. The heat sinks were made of aluminium with constant volume
fraction of 9% which are acting as TCEs embedded with the PCM
2.2. Governing equations
namely n-eicosane and with fin height of 20 mm. The volume fraction of
TCEs denoted by is calculated using Eq. (1), which is actually the ratio
The governing equations are expressed based on the problem under
of volume occupied by the fins to the total empty volume of heat sink.
consideration in current study, as shown in Fig. 1. The enthalpy-por-
=
VTCE osity method is adopted to model the effect of phase transition in the
VHS (1) PCM based finned heat sink. The conjugate heat transfer has been
solved using continuity, momentum and energy equations. For the
In recent experimental research investigations, Baby and Balaji [31]
phase transition problem “melting and solidification” model has been
and recently Arshad and his co-authors [22,23,26] and Ashraf et al.
applied. The following assumptions are taken to model the PCM based
[28,29] reported that a finned heat sink of volume fraction of 9% pro-
finned heat sink in present numerical investigation [32–34]:
vided the best thermal performance for passive thermal management of

• Initially, the PCM and heat sink are at the melting temperature of
portable electronic devices. The system is designed based on the
average dimensions of portable hand-held electronic devices. The heat
PCM.
• The heat sink remains in solid-state and absorbs the sensible heat
sink were insulated from all sides to prevent any heat loss to the sur-
rounding during operation. The top surface of the heat sink was covered
only.
• The material of heat sink is homogeneous and isotropic.
with perspex sheet, to provide the insulation and to visualize the phy-

• Local thermal equilibrium exists between the fins and liquid PCM.
sical change in solid-liquid interface of PCM in melting and solidifica-

• Thermophysical properties of PCM and fins are remained constant

tion phases. The schematic of the physical domain studied in this paper
is illustrated in Fig. 1. In the system, 2D geometry of the both heat sinks
with temperature and phase.
is considered having fin thickness of 2 mm and 3 mm with adiabatic

Fig. 1. Schematic diagram of the physical domain used in current study.

3
A. Arshad, et al. Thermal Science and Engineering Progress 18 (2020) 100520

Fig. 3. Validation of present simulation with experimental results with no fin

heat sink.

v v v P 2v 2v
+u +v = +µ + + ref g T Tref + Sv
t x y y x2 y2
(4)
Energy conservation:

T T T 2T 2T
cp +u +v =k + + Sh
t x y x2 y2 (5)
where, P is the pressure, g is the gravitational acceleration, is the
density, µ is the dynamic viscosity, cp is the thermal capacitance, k is
the thermal conductivity, and t is the time, u and v are the velocity
components in x and y directions, respectively. Due to the direction of
gravitational acceleration in the negative y -direction, the Boussinesq
approximation is added as the sixth term by ref g (T Tref ) [35,36].
Fig. 2. (a)-Grid and (b) time-step independency test results. The value of = 0.001 K 1 has been chosen based on the analysis of the
detailed data reported by Humphries and Griggs [37]. The source term

• The liquid PCM is assumed to be laminar, unsteady, incompressible S in momentum equation, Eqs. (3) and (4) can be expressed as follows:
Newtonian fluid and subjected to the Boussinesq approximation. S= Am ( ) (6)
• The volume change of PCM is negligible during the melting and where is Am ( ) is the “porosity function” defined by Brent et al. [38]
solidification.
• The radiative heat transfer is negligible. which fulfil “the basic principle is gradually to reduce the velocities
from a finite value in liquid, to zero in full solid, over the computational
calls that are changing the phase”. Thus, nth fluid cells which are un-
The following governing conservation equations for mass, mo-
dergoing the mushy-zone phase change, Am ( ) is defined as follows to
mentum and energy can be written based the above assumptions: .
make the momentum equation to mimic Carman-Kozeny equations for
Mass conservation:
flow in porous media:
u v
+ =0 (1 )2
x y (2) Am ( ) = C
( 3 b) (7)
Momentum conservation: where b = 0.001 is merely a computational constant used to avoid di-
2u 2u vision by zero. The constant C is reflecting the morphology of the
u u u P
+u +v = +µ + + Su melting front which controls the degree of penetration of the convec-
t x y x x2 y2 (3)
tion flied into the mushy-zone phase change. The value of C = 105 is

Table 3
Grid independency analysis.
Grid systam Number of elements Melting time (s) Deviation (%) Average heat sink temperature (K) Deviation (%)

Grid A 18862 1945.00 0.00 319.64 0.00

Grid B 27126 1945.00 0.00 319.59 0.01
Grid C 42418 1950.00 0.26 319.50 0.03
Grid D 73970 1955.00 0.26 319.38 0.04
Grid E 167493 1980.00 1.28 318.80 0.18

4
A. Arshad, et al. Thermal Science and Engineering Progress 18 (2020) 100520

( H)
Sh =
t (12)

where

0 if T < Tm
H=
L if T > Tm (13)

The mathematical models used for the present simulation can be

found in Nayak et al. [7], Wang and Yang [11,12], Sahoo et al. [16] and
Shatikian et al. [8,39]. They simulated the heat transfer performance of
PCM in internal fins. A good agreement was found of numerical results
with experimental results reported by Fok et al. [40] and Hosseinizadeh
et al. [41]. Therefore, current method is used to simulate the pin-fin
heat sink filled with PCM a LHSU.

2.3. Initial and boundary conditions

The initial and boundary conditions are referenced to the experi-

mental study by Arshad et al. [23,22,26] and Ashraf et al. [28,29]. The
schematic of physical domain for the 2D case used in this study is shown
in Fig. 1. The outer boundary walls are defined as an adiabatic
boundary condition because the outer sides of heat sink are insulated
with thermal insulation material to prevent the heat loss towards the
surrounding in experimentation. Since, the phase change phenomenon
is occurring inside the PCM based finned heat sink in transient mode.
The initial and boundary conditions applied in the present case are as
follows to solve the governing equations:

1. Initial condition
t = 0, T = Tini = 298K , fl = 0

The following boundary conditions are applied for the solution of

governing equations:
2. No slip condition at the walls, i.e. u = v = 0
3. Heat flux supplied at the bottom,

T q
=
y x = 11 61mm
k
y=0

Fig. 4. Validation of the present results for (a) average temperature and (b)
liquid fraction of PCM. 4. Natural convective boundary condition with ambient at the top
wall, i.e.

taken in present study as recommended in several studies [11,12,39]. T

Furthermore, a mushy region lies between the liquid fraction 0 and 1. k =h T T
y
The can be defined as: y=H

0 if T < Ts
5. Adiabatic boundary condition at side wall and bottom,
H T Ts
if Ts < T < Tl
= = Tl Ts
L T
1 if T < Tl (8) =0 Along vertical walls
x x = 0, W

The total enthalpy of the PCM is the summation of sensible and

latent heat defined as:
T
=0 For bottom.
H=h+ H (9) y x = 0 11,61 72
y=0
where h is the sensible enthalpy which is defined as follows:
T
h = href + cp dT
Tref (10) 2.4. Numerical methodology

and H is the fractional latent heat of the PCM that may vary between 0
The commercial computational fluid dynamics (CFD) software
for solid and L (latent heat of fusion) for liquid which is expressed as:
ANSYS-FLUENT 19.1 was used to conduct the transient simulations in
H= L (11) present study. The conservation of mass, momentum and energy
equations described in Eqs. (2)–(5) are discretized by finite volume
The Sh is the energy source term in Eq. (3) represents the latent heat method (FVM) with double precision. The conjugate heat transfer has
storage due to melting and it is defined as: been solved using continuity, momentum and energy equations. For the

5
A. Arshad, et al. Thermal Science and Engineering Progress 18 (2020) 100520

Fig. 5. Variation of liquid fraction of 2 mm fin thickness at various time periods and input heat fluxes.

phase transition problem “melting and solidification” model has been avoid their effects on the numerical accuracy. Five different mesh size
applied. The solidification and melting model is based on the enthalpy- of elements sizes 18862, 27126, 42418, 73970 and 167493 were in-
porosity method. In this model, instead of explicitly tracking of melt vestigated. The result of liquid-fraction variations of each element size
fraction, the liquid volume of each cell represented as liquid fraction. is presented in Fig. 2a. The melting time and average heat sink base
For the pressure-velocity coupling, Semi-Implicit Pressure-Linked temperature for each gird size are presented in Table 3. The results
Equation (SIMPLE) algorithm has been used of Patanker [42]. For the reveals that the maximum deviation in melting time and average heat
pressure correction equation, Presto scheme is used while QUICK sink temperature are obtained of 1.28% and 0.18% , respectively. The
scheme is employed for the momentum and energy equations [43]. A three time-steps of t = 0.05, 0.1 and 0.2 s are examined and result is
second order upwind differencing scheme was employed for the dis- presented in Fig. 2b. The results illustrate no significance change is
cretization of convective terms in momentum and energy equations. observed in liquid fraction for different time-step sizes. The reason is
Gravity effects were also considered. The convergence criteria are set to that PCM upfront velocity and thermal front movement are low, re-
10 4 , 10 4 and 10 6 for continuity, momentum and energy equations, flecting a low Peclet number and Courant number situation. After
respectively. The grid and time independence tests were carried out to comparisons of results, the mesh with 42418 elements and time-step of

6
A. Arshad, et al. Thermal Science and Engineering Progress 18 (2020) 100520

Fig. 6. Comparison of isotherms contours at various time period of 2 mm and 3 mm fin thickness cases at power level of 5 W.

t = 0.1 s were selected in the current study for further simulations. To effect of heating power during the PCM phase transition inside the heat
compromise the accuracy and cost, the mesh size of 42418 elements is sink which averaged over the heat sink base.
selected.
qhf
Q =
k (Tm Tini ) (15)
2.5. Non-dimensional equations
The non-dimensional number utilized in transient heat conduction
The dimensionless analysis is performed in order to normalize the problems is Fourier number (Fo) which is defined as the ratio between
phase change heat transfer process and melting phenomenon of the n- the diffusive transport and the energy storage. Fo depicts the time for
eicosane embedded into the finned heat sinks. diffusion of heat across the PCM thickness and it can be calculated as
The dimensionless average heat sink base temperature ( ) is defined follows [32,45]:
to express the temperature distribution while melting inside the heat
sink [44]. t
Fo =
hf2 (16)
k (T Tini )
=
qhf (14) Stefan number (Ste) represents amount of superheating which is
experienced by the liquid-phase of the PCM. While melting, the effect of
The dimensionless heat flux (Q ) is defined to express the generalized convection is present only when the Ste number is greater than 1. Ste

7
A. Arshad, et al. Thermal Science and Engineering Progress 18 (2020) 100520

Fig. 7. Comparison of liquid fractions contours at various time period of 2 mm and 3 mm fin thickness cases at power level of 6 W.

number is defined as the ratio of the sensible heat to the latent heat, t cp q
= Fo × Ste = ×
which is defined as [32,45]: hf kL (20)

cp T To account the effect of free or natural convection heat transfer in

Ste =
L (17) molten PCM, Grashof (Gr) number or more comprehensively Rayleigh
(Ra) number is used which is associated with the buoyancy driven flow
where indicating the ratio of natural convection to conduction heat transfer.
qhf For a uniform heat flux, Ra based on the fin height as a characteristic
T= length is defined as [32,45]:
k (18)
2g cp qhf4
Therefore, Ste number is defined as: Ra =
k 4µ (21)
cp qhf
Ste =
(19) Heat transfer as a function of dimensionless number is described by
kL
modified Nusselt number (Nu ) which presents the time-dependent
The product of Ste and Fo numbers represents the dimensionless relationship between the instantaneous heat flux (q) and temperature
time for phase change without convection. difference ( T ), is defined as [32,39]:

8
A. Arshad, et al. Thermal Science and Engineering Progress 18 (2020) 100520

Fig. 11. The variation of in terms of Fo for different Q .

Fig. 8. Velocity vectors of PCM melting within the fins.

Fig. 12. The variation of in terms of for different Q .

Fig. 9. Temperature-time history of 2 mm and 3 mm fin thickness PCM filled

heat sinks at different heat fluxes.

Fig. 13. The variation of liquid fraction in terms of Fo for different Ste .

q (tp/2)
Fig. 10. Melt fraction curves of 2 mm and 3 mm fin thickness PCM filled heat Nu =
Tk (22)
sinks at different heat fluxes.
Where, T is the temperature difference between the base temperature and
melting temperature of the PCM, tp/2 is half-thickness of the PCM layer is
chosen as characteristics length proposed by Saha and Dutta [32] and

9
A. Arshad, et al. Thermal Science and Engineering Progress 18 (2020) 100520

Shatikian et al. [39]. The reason to choose the tp/2 as characteristics

length is that for tall enclosure, the aspect ratio of fin height and PCM
thickness (hf 2/ tp ) of the heat sink cavity is very large, which reduce the
effect of convection compared with conduction heat transfer in the PCM
domain results is a lower temperature difference between the two
consecutive fins. Furthermore, Saha and Dutta [32] reported the range
of aspect ratio of fin height and PCM thickness (1.939 < hf 2/ tp 5) for
tall enclosures which almost equal to 4.32 and 5.13 for 3 mm and 2 mm
fin thickness heat sinks. Therefore, the results of Nu number and liquid
fraction as a function of FoSteRa based on tp/2 as characteristic length
scale are expressed using modified Fo number, Ste number and Ra
number:
t
Fo =
(tp/2) 2 (23)

cp q (tp/2)
Ste =
kL (24)
2g cp q (tp/2) 4
Fig. 14. The variation of liquid fraction in terms of Fo Ste for different Ste . Ratp/2 =
k 4µ (25)
The mentioned dimensionless parameter are employed in various
previous studies to define the variation in heat sink temperature and
liquid fraction [8,13,16,32,44].

3. Model validation

3.1. Experimental validation for the heat sink

The current model is validated with experimental results of a no fin

heat sink filled with PCM having the dimension of similar dimensions
used in Ashraf et al. [28]. The RT-35HC was used as a PCM having
melting temperature of 35 °C at a input power level of 5 W. The results
of average temperature of heat sink was compared of both numerical
and experimental cases, as shown in Fig. 3. The results of both nu-
merical and experimental results showed a good agreement and re-
vealed that the present numerical model can be used for further si-
mulations.

3.2. Validation for the phase change process

Fig. 15. The variation of liquid fraction in terms of Fo Ste Ra for different Ste .
To validate and ensure the accuracy of the numerical approach, the
average temperature and liquid fraction of RT-58 PCM in a horizontal
triples tube LHS system with internal-external fins were determined and
compared with the experimental and numerical results presented by
Mat et al. [46]. For the studied geometry, the initial and wall tem-
peratures were considered 27°C and 120°C, respectively, where the
inner and outer diameters of the system are 25. 4 and 75 mm, respec-
tively. The copper fins have a length of 42 mm with 1 mm thickness
which are heated from the walls as shown. The results of mean tem-
perature and liquid fraction obtained from the present study are shown
in Fig. 4 and compared with Mat el al. [46], with excellent agreement
for both temperature and liquid fraction.

4. Results and discussion

4.1. Temperature and melt-fraction distribution

Fig. 5 presents the transient variation of liquid or melt fraction

contours of PCM for 2 mm fin thickness heat sink at various input
power level of 4–6 W for 10 min time intervals during melting. The
0.166 0.25
Fig. 16. Distribution of Nu as function of Fo Ste Ratp Ra hf and proposed variation in solid-liquid interface can also be depicted from the Fig. 5 as
correlation compared with the Shatikian et al. [39]. well. When the liquid fraction is 0, the PCM is fully in solid form and
when it is 1 the PCM is fully melted. The region between 0 and 1 is
known as the transition region where PCM exhibited both solid and
liquid phases called mushy zone. The melting of PCM increases with the

10
A. Arshad, et al. Thermal Science and Engineering Progress 18 (2020) 100520

increase of time. With the increase of input heat flux, the melting region 1605 s at 6 W for 2 mm and 3 mm fin thickness finned heat sink, re-
of PCM increases and therefore, a higher value of liquid fraction is spectively. Hence, the selection of PCM based heat sink as a passive
obtained at an identical time. In addition, the melting of PCM increases cooling media is based on the operating power of the electronic device.
with the increase of time at an identical input heat flux, as expected. At higher input power level, more PCM or a PCM with higher latent
The rate of PCM melting lower at low input heat flux as compared to heat of fusion is required, which can prolong the operating duration at
high heat flux as the time increases which is due to the rapid heat comfortable temperature level. Furthermore, it can be seen from Fig. 10
transfer into the PCM through fins, results in a higher rate of PCM that at each power level, the configuration of 3 mm fin thickness finned
melting. It can be seen from Fig. 5 that at 4 W, the PCM did not fully heat sink takes more time to complete the solid-liquid phase change
melt even after 40 min of heating. Contrarily, it is fully melted at 5 W process compared with 2 mm fin thickness finned heat sink when fully
and 6 W after 40 min. It is revealed that the higher the power level, filled with PCM. The higher time to complete the melting process leads
higher the heat transfer rate results in higher the liquid fraction rate or to lower the base temperature of the heat sink and thus 3 mm fin
melting rate of the PCM based finned heat sink. thickness finned heat sink has the best thermal performance to enhance
The rate of heat transfer in a PCM filled finned heat sink can been the passive cooling capability of electronic devices filled with PCMs.
from temperature contours shown in Fig. 6, which shows the distribu-
tion of temperature inside in the 2 mm and 3 mm fin thickness finned 4.3. Dimensionless temperature, melt-fraction and heat transfer analysis
heat sink for different time steps at a power level of 5 W during melting.
At a constant input power level, it can be seen clearly that the rate of Fig. 11 illustrates the distribution of dimensionless average heat
temperature rise is higher in 2 mm fin thickness heat sink as compared sink base temperature ( ) as a function Fourier number (Fo) at different
3 mm fin thickness heat sink in all times. For example, the temperature dimensionless heat fluxes (Q = 12, 15 and 18) for 2 mm and 3 mm fin
varies from 317.82 K to 319. 58 K for 2 mm fin thickness heat sink at thickness finned heat sink. For a constant Q = 12, the variation in of
45 min, whereas it varies from 315.7 to 317.5 K for the case of 3 mm fin finned heat sinks shows that uniform distribution of temperature inside
thickness heat sink. The higher temperature rise in 2 mm fin thickness the heat sink with the increase in Fo. Then, the rate of temperature
heat sink is due the more number of fins compared to the 3 mm fin enhancement reduces due to an almost constant temperature of the
thickness heat sink. Furthermore, the penetration of heat in PCM region PCM during the phase change and then increases sharply. At a constant
is more in case of 2 mm fin thickness heat sink causing a shorter melting Q , the temperature is almost similar for both 2 mm and 3 mm thickness
time which can be seen in Fig. 6 especially at time 30 min. of the fins. In addition, the is higher for a lower Q during the melting
A further comparison of liquid fraction contours for both fin process. However, PCM melting is terminated in a lower Fo for a higher
thickness heat sinks can be seen from Fig. 7. The finned heat sink of Q and then increases sharply after the melting process. The reason for
3 mm fin thickness has the lower rate of PCM melting due to the higher this can be explained from Eqs. (14) and (15). For constant properties of
heat transfer area, optimum number and distribution of fins. The lower PCM and a constant fin height, Fo number increases and decreases by
rate of PCM melting of 3 mm fin thickness heat sink enhances to absorb increasing Q . However, at constant Q , the varies with temperature
and transfer the heat more uniformly and effectively towards the am- difference ( T = T Tini ) and Fo. Moreover, at a constant Fo, a higher
bient, keeping the base temperature of heat sink at lower level thus is achieved for 2 mm fin thickness compared with 3 mm case at a
resulting in more efficient thermal cooling. Similar findings of 2 mm constant Q , which reflects the increase in the phase transition duration
and 3 mm fin thickness finned heat sink has been reported in Ref. [26]. . from solid to liquid results in improves the thermal performance of heat
The heating phenomenon inside the heat sink can be seen from the sink. As shown in Fig. 11, Fo alone is not sufficient to express the
velocity vectors shown in Fig. 8, which shows the formation of vortices generalize trend of temperature distribution because of the effect of
at the base and top of the PCM zone between the two consecutive fins. phase change process. Fig. 12 represents the distribution of as a
The direction of velocity vectors show that liquid PCM moves from function of dimensionless time ( ) at different Q . The similar trends of
bottom to top close to the walls and then converge at centre of the PCM by varying the Q , i.e. the at lower Q , reaches higher and it takes
zone while heating at the base. The formations of vortices are in two more time to complete the phase change process during melting. It can
main circulating patterns of the liquid PCM because of natural con- been seen from both finned heat sinks, the increases T increases with
vection and direction of gravity. the decrease of Q at the same due to T of energy storage increasing
constantly [13]. However, when the is larger than 1.25, the increases
4.2. Temperature and melt-fraction variation linearly. This is caused by only covers over the critical point
( = 1.25), as shown in Fig. 12.
The results are presented by comparing the temperature difference- Fig. 13 shows the variation of liquid fraction as a function of
time variations of 2 mm and 3 mm fin thickness finned heat sink at modified Fo number at different modified Ste numbers
different power levels in Fig. 9. Since the initial temperature is fixed (Ste = 0.12 0.21) for both finned heat sinks. It can be seen that the
which is equal to the ambient conditions, a temperature difference is 2 mm fin thickness finned heat sink has a higher liquid fraction com-
reported instead of temperature. The temperature difference pared to the 3 mm fin thickness finned heat sink at a constant Ste .
( T = T Tini ) represents the difference between the temperature of Furthermore, the liquid fraction varies with Ste similar to the , which
heated base of heat sink and initial temperature. As expected, the lower shows that Fo is further dependent on the other dimensionless para-
the power level, the higher the time needed to fully melt the n-eicosane meters to generalize the liquid fraction behaviour for transient phase
which happens when the PCM temperature rises above the melting change heating phenomenon. Higher liquid fraction is achieved by in-
point temperature. The maximum heat sink base temperature difference creasing the Ste at a constant Fo or a constant liquid fraction is
reaches at 21.68 °C and 18.31 °C at 4 W, 40.32 °C and 37.67 °C at 5 W, and achieved at a lower Fo for a higher Ste which is also reported in the
57.51 °C and 55.24 °C at 6 W with 2 mm and 3 mm fin thickness finned literature [47]. In addition, the Fo number is significantly reduced
heat sink, respectively. The melt fraction curves of both fin thickness from 0.23 to 0.12 with the increase of Ste from 0.12 to 0.21. The reason
finned heat sink at different power levels are shown in Fig. 9. At each behind such trends can be explained from Eq. (24). For the constant
power level, the melt fraction of different configurations finishes uni- PCM properties and PCM thickness layer, Ste number increases by
formly due to the natural convection heat transfer within the melt re- increasing Q . Therefore, at constant Fo , a higher liquid fraction is
gion and uniform distribution of the fins. In addition, the case of 3 mm obtained by increasing Q which takes less time to complete the melting
fin thickness takes a higher time to complete the melting phase com- process. From Fig. 13, it can been seen that 3 mm fin thickness finned
pared with the 2 mm fin thickness finned heat sink. The melting times heat sink has a higher Ste at constant Q , e.g. at Q = 12, Ste is 0.12 and
are 2565 s and 2730 s at 4 W, 1925 s and 1745 s at 5 W, and 1550 s and 0.14 for 2 mm and 3 mm fin thickness, respectively, which shows that

11
A. Arshad, et al. Thermal Science and Engineering Progress 18 (2020) 100520

3 mm fin thickness finned heat sink has a higher capacity to absorb the formed of vortices of the liquid PCM because of natural convection and
heat in form of latent heat. Moreover, It can be revealed that liquid direction of gravity. Further, the results indicated that finned heat sink
fraction is not just the function of Fo because the effect of phase change of 3 mm fin thickness has the better heat transfer performance to stable
are not included. Thus, liquid fraction is directly a function of Ste due the base temperature, enhances the better cooling capability and lower
to the fact that liquid fraction is changed directly with the inverse of the maximum operating temperature as compared to 3 mm fin thick-
latent heat of fusion based on the storage capacity of the PCM based ness finned heat sink. By increasing the power level from 4 to 6 W, for
finned heat sink. the case of 3 mm fin thickness, the melting time increases by
Therefore, for the generalization of melting phenomenon, the re- 6.63%, 3.59% and 1.90%, respectively, compared to the 2 mm fin
sults of liquid fraction are presented in Fig. 14 in terms of the product of thickness finned heat sink. The normalized new correlations for calcu-
Ste and Fo for different cases of Ste . It can be seen that the different lation of liquid fraction and Nu results reveal that liquid fraction and
lines of liquid fractions of are closer to each other. However, there is
0.155 0.01
Nu can be defined as a function of Fo Ste Ratp Ra hf and
still a deviation in completion of melting process for a higher Ste . As 0.166 0.25
Fo Ste Ratp Rahf , respectively, for both 2 mm and 3 mm fin
reported in literature [47], the independent melting process is suffi-
thickness finned heat sinks.
ciently generalized with product of Fo Ste . However, due to effect of
natural convection in melting front (especially for high Ste resulted
CRediT authorship contribution statement
from a high Q ), modified Rayleigh number is better to be included to
propose the effect of natural convection. Fig. 15 illustrate the variation
Adeel Arshad: Conceptualization, Methodology, Software,
of liquid fraction in terms of Fo Ste and Ra , namely Fo Ste Ratp Ra hf .
Validation, Formal analysis, Data curation, Writing - original draft.
For both fin thickness of finned heat sinks, the liquid fractions for all
Mark Jabbal: Resources, Writing - review & editing, Visualization,
Ste almost match each other, which shows a reasonable generalization
Supervision, Funding acquisition. Pouyan Talebizadeh Sardari:
of the liquid fraction in terms of the product of Fo Ste Ratp Ra hf . A curve
Methodology, Software, Validation, Writing - review & editing.
fitting analysis is performed to develop a correlation for the liquid
Muhammad Anser Bashir: Methodology, Software, Writing - review &
fraction as function of Fo Ste Ratp Ra hf for different fin thickness of
editing. Hamza Faraji: Writing - review & editing. Yuying Yan:
finned heat sinks:
Visualization, Supervision, Funding acquisition, Project administration.
fl = 50560x 4 + 5326.8x 3 124.41x 2 + 15.74x 0.0126 (R2 = 0.998)
(26) Declaration of Competing Interest
0.155 0.01
where, x = Fo Ste Ratp Rahf . Note that developed correlation is The authors declare that they have no known competing financial
reasonable for finned heat sink chosen in current study within the interests or personal relationships that could have appeared to influ-
parameter range: 0 < Fo < 0.26, 0.12 < Ste < 0.21, 1.54 × ence the work reported in this paper.
0.155 0.01
103 < Ratp < 4.60 × 103, 1.71 × 107 < Rahf < 2.56 × 107, 0 < fl <
0.155 0.01 Acknowledgement
1, 0 < Fo Ste Ratp Rahf < 0.063.. Finally, Fig. 16 illustrates the Nu
variation, defined in Eq. (22), with for both fin thickness at all Ste . A
This research is facilitated by the Faculty of Engineering, University
similar process is performed for Nu to find the suitable dimensionless
of Nottingham, UK research infrastructure. The first author (Adeel
parameter to generalize it. The results are presented as function of di-
0.166 0.25 Arshad) acknowledges the University of Nottingham for awarding him
mensionless numbers Fo Ste Ratp Rahf and a good agreement is
the Faculty of Engineering Research Excellence PhD Scholarship to pursue a
achieved for both cases of finned heat sinks. The results showed that
Ph.D. research program.
Nu is higher at the initial stage of melting process due to the less
thermal resistance. Afterwards, Nu decreases sharply until the melt
References
fraction is achieved about 0.3 and further decreases asymptotically
approaching a non-zero value, which represents the heat transfer in
[1] S.S. Murshed, C.N. de Castro, A critical review of traditional and emerging tech-
fully melted PCM under natural convection. The following relation of niques and fluids for electronics cooling, Renew. Sustain. Energy Rev. 78 (2017)
Nu can be defined to express the results presented in Fig. 16: 821–833, https://doi.org/10.1016/j.rser.2017.04.112.
[2] L. Yeh, Review of heat transfer technologies in electronic equipment, J. Electron.
b Packaging 117 (4) (1995) 333–339.
Nu = a + 0.166 0.25 [3] M. Eugeni, H. Elahi, F. Fune, L. Lampani, F. Mastroddi, G.P. Romano, P. Gaudenzi,
Fo Ste Ratp Ra hf +c (27) Numerical and experimental investigation of piezoelectric energy harvester based
on flag-flutter, Aerosp. Sci. Technol. 97 (2020) 105634, , https://doi.org/10.1016/
where constants a = 2, b = 4 , and c = 0.02 are obtained after the best j.ast.2019.105634.
[4] H. Elahi, M. Eugeni, P. Gaudenzi, Design and performance evaluation of a piezo-
fitting in current system. A good agreement was achieved with the
electric aeroelastic energy harvester based on the limit cycle oscillation phenom-
Shatikian et al. [39] proposed relation of Nu . enon, Acta Astronaut. 157 (2019) 233–240, https://doi.org/10.1016/j.actaastro.
2018.12.044.
[5] A. Arshad, M. Jabbal, Y. Yan, Synthetic jet actuators for heat transfer enhancement
5. Conclusions – a critical review, Int. J. Heat Mass Transf. 146 (2020) 118815, , https://doi.org/
10.1016/j.ijheatmasstransfer.2019.118815.
[6] A. Arshad, M. Jabbal, Y. Yan, Preparation and characteristics evaluation of mono
The transient 2D numerical simulation of PCM filled finned heat
and hybrid nano-enhanced phase change materials (NePCMs) for thermal man-
sinks are carried out to investigate the heat transfer and flow-field agement of microelectronics, Energy Convers. Manage. 205 (2020) 112444, ,
characteristics having the fin thickness of 2 mm and 3 mm, optimally https://doi.org/10.1016/j.enconman.2019.112444.
explored from the previous experimental investigations. Three different [7] K. Nayak, S. Saha, K. Srinivasan, P. Dutta, A numerical model for heat sinks with
phase change materials and thermal conductivity enhancers, Int. J. Heat Mass
input heating power level of 4, 5 and 6 W are provided at base of each Transf. 49 (11–12) (2006) 1833–1844, https://doi.org/10.1016/j.
heat sink. The numerical results show that finned heat sink filled with ijheatmasstransfer.2005.10.039.
PCM has the capability of uniform heat distribution inside the heat sink. [8] V. Shatikian, G. Ziskind, R. Letan, Numerical investigation of a PCM-based heat sink
with internal fins, Int. J. Heat Mass Transf. 48 (17) (2005) 3689–3706, https://doi.
Uniform melting of PCM is observed inside the fins and higher phase org/10.1016/j.ijheatmasstransfer.2004.10.042.
change duration is obtained for 3 mm fin thickness finned heat sink at [9] S.K. Saha, K. Srinivasan, P. Dutta, Studies on optimum distribution of fins in heat
each input heat flux. In addition, natural convection heating phenom- sinks filled with phase change materials, J. Heat Transfer 130 (3) (2008) 034505, ,
https://doi.org/10.1115/1.2804948.
enon is observed and velocity vectors reveal that melting of PCM occur [10] S.K. Saha, P. Dutta, Thermal management of electronics using PCM-based heat sink
from bottom to top of the finned heat sink. The circulating patterns are

12
A. Arshad, et al. Thermal Science and Engineering Progress 18 (2020) 100520

subjected to cyclic heat load, IEEE Trans. Components, Packaging Manuf Technol 2 A. Arshad, Thermal management of electronics: an experimental analysis of trian-
(3) (2012) 464–473, https://doi.org/10.1109/tcpmt.2011.2180021. gular, rectangular and circular pin-fin heat sinks for various PCMs, Int. J. Heat Mass
[11] Y.-H. Wang, Y.-T. Yang, Three-dimensional transient cooling simulations of a por- Transf. 123 (2018) 272–284, https://doi.org/10.1016/j.ijheatmasstransfer.2018.
table electronic device using PCM (phase change materials) in multi-fin heat sink, 02.044.
Energy 36 (8) (2011) 5214–5224, https://doi.org/10.1016/j.energy.2011.06.023. [30] H. Usman, H.M. Ali, A. Arshad, M.J. Ashraf, S. Khushnood, M.M. Janjua, S.N. Kazi,
[12] Y.-T. Yang, Y.-H. Wang, Numerical simulation of three-dimensional transient An experimental study of PCM based finned and un-finned heat sinks for passive
cooling application on a portable electronic device using phase change material, Int. cooling of electronics, Heat Mass Transf. 54 (12) (2018) 3587–3598, https://doi.
J. Therm. Sci. 51 (2012) 155–162, https://doi.org/10.1016/j.ijthermalsci.2011.08. org/10.1007/s00231-018-2389-0.
011. [31] R. Baby, C. Balaji, Experimental investigations on phase change material based
[13] W.-B. Ye, D.-S. Zhu, N. Wang, Numerical simulation on phase-change thermal finned heat sinks for electronic equipment cooling, Int. J. Heat Mass Transf. 55
storage/release in a plate-fin unit, Appl. Therm. Eng. 31 (17–18) (2011) (5–6) (2012) 1642–1649, https://doi.org/10.1016/j.ijheatmasstransfer.2011.11.
3871–3884, https://doi.org/10.1016/j.applthermaleng.2011.07.035. 020.
[14] R. Pakrouh, M. Hosseini, A. Ranjbar, R. Bahrampoury, A numerical method for [32] S. Saha, P. Dutta, Heat transfer correlations for PCM-based heat sinks with plate
PCM-based pin fin heat sinks optimization, Energy Convers. Manage. 103 (2015) fins, Appl. Therm. Eng. 30 (16) (2010) 2485–2491, https://doi.org/10.1016/j.
542–552, https://doi.org/10.1016/j.enconman.2015.07.003. applthermaleng.2010.06.021.
[15] K. Kant, A. Shukla, A. Sharma, Performance evaluation of fatty acids as phase [33] M.A. Bashir, A. Giovannelli, Design optimization of the phase change material in-
change material for thermal energy storage, J. Energy Storage 6 (2016) 153–162, tegrated solar receiver: a numerical parametric study, Appl. Therm. Eng. 160
https://doi.org/10.1016/j.est.2016.04.002. (2019) 114008, , https://doi.org/10.1016/j.applthermaleng.2019.114008.
[16] S.K. Sahoo, P. Rath, M.K. Das, Numerical study of phase change material based [34] A. Shahsavar, A. Shaham, P. Talebizadehsardari, Wavy channels triple-tube LHS
orthotropic heat sink for thermal management of electronics components, Int. J. unit with sinusoidal variable wavelength in charging/discharging mechanism, Int.
Heat Mass Transf. 103 (2016) 855–867, https://doi.org/10.1016/j. Commun. Heat Mass Transfer 107 (2019) 93–105, https://doi.org/10.1016/j.
ijheatmasstransfer.2016.07.063. icheatmasstransfer.2019.05.012.
[17] J. Vogel, J. Felbinger, M. Johnson, Natural convection in high temperature flat plate [35] J.M. Mahdi, E.C. Nsofor, Melting enhancement in triplex-tube latent heat energy
latent heat thermal energy storage systems, Appl. Energy 184 (2016) 184–196, storage system using nanoparticles-metal foam combination, Appl. Energy 191
https://doi.org/10.1016/j.apenergy.2016.10.001. (2017) 22–34, https://doi.org/10.1016/j.apenergy.2016.11.036.
[18] V. Joshi, M.K. Rathod, Constructal enhancement of thermal transport in latent heat [36] J.M. Mahdi, E.C. Nsofor, Melting enhancement in triplex-tube latent thermal energy
storage systems assisted with fins, Int. J. Therm. Sci. 145 (2019) 105984, , https:// storage system using nanoparticles-fins combination, Int. J. Heat Mass Transf. 109
doi.org/10.1016/j.ijthermalsci.2019.105984. (2017) 417–427, https://doi.org/10.1016/j.ijheatmasstransfer.2017.02.016.
[19] C. Ji, Z. Qin, Z. Low, S. Dubey, F.H. Choo, F. Duan, Non-uniform heat transfer [37] W.R. Humphries, E.I. Griggs, A design handbook for phase change thermal control
suppression to enhance PCM melting by angled fins, Appl. Therm. Eng. 129 (2018) and energy storage devices, NASA STI/recon technical report N 78.
269–279, https://doi.org/10.1016/j.applthermaleng.2017.10.030. [38] A.D. Brent, V.R. Voller, K.J. Reid, Enthalpy-porosity technique for modeling con-
[20] M. Kazemi, M. Hosseini, A. Ranjbar, R. Bahrampoury, Improvement of longitudinal vection-diffusion phase change: application to the melting of a pure metal, Numer.
fins configuration in latent heat storage systems, Renewable Energy 116 (2018) Heat Transfer 13 (3) (1988) 297–318, https://doi.org/10.1080/
447–457, https://doi.org/10.1016/j.renene.2017.10.006. 10407788808913615.
[21] R. Kalbasi, M. Afrand, J. Alsarraf, M.-D. Tran, Studies on optimum fins number in [39] V. Shatikian, G. Ziskind, R. Letan, Numerical investigation of a PCM-based heat sink
PCM-based heat sinks, Energy 171 (2019) 1088–1099, https://doi.org/10.1016/j. with internal fins: Constant heat flux, Int. J. Heat Mass Transf. 51 (5–6) (2008)
energy.2019.01.070. 1488–1493, https://doi.org/10.1016/j.ijheatmasstransfer.2007.11.036.
[22] A. Arshad, H.M. Ali, M. Ali, S. Manzoor, Thermal performance of phase change [40] S. Fok, W. Shen, F. Tan, Cooling of portable hand-held electronic devices using
material (PCM) based pin-finned heat sinks for electronics devices: Effect of pin phase change materials in finned heat sinks, Int. J. Therm. Sci. 49 (1) (2010)
thickness and PCM volume fraction, Appl. Therm. Eng. 112 (2017) 143–155, 109–117, https://doi.org/10.1016/j.ijthermalsci.2009.06.011.
https://doi.org/10.1016/j.applthermaleng.2016.10.090. [41] S. Hosseinizadeh, F. Tan, S. Moosania, Experimental and numerical studies on
[23] A. Arshad, H.M. Ali, W.-M. Yan, A.K. Hussein, M. Ahmadlouydarab, An experi- performance of PCM-based heat sink with different configurations of internal fins,
mental study of enhanced heat sinks for thermal management using n-eicosane as Appl. Therm. Eng. 31 (17–18) (2011) 3827–3838, https://doi.org/10.1016/j.
phase change material, Appl. Therm. Eng. 132 (2018) 52–66, https://doi.org/10. applthermaleng.2011.07.031.
1016/j.applthermaleng.2017.12.066. [42] S. Patankar, Numerical Heat Transfer and Fluid Flow, Hemisphere Publishing
[24] A. Arshad, H.M. Ali, S. Khushnood, M. Jabbal, Experimental investigation of PCM Corporation; McGraw-Hill Book Company, New York, 2018.
based round pin-fin heat sinks for thermal management of electronics: Effect of pin- [43] A. Shahsavar, A.A. Al-Rashed, S. Entezari, P.T. Sardari, Melting and solidification
fin diameter, Int. J. Heat Mass Transf. 117 (2018) 861–872, https://doi.org/10. characteristics of a double-pipe latent heat storage system with sinusoidal wavy
1016/j.ijheatmasstransfer.2017.10.008. channels embedded in a porous medium, Energy 171 (2019) 751–769, https://doi.
[25] H.M. Ali, A. Arshad, Experimental investigation of n-eicosane based circular pin-fin org/10.1016/j.energy.2019.01.045.
heat sinks for passive cooling of electronic devices, Int. J. Heat Mass Transf. 112 [44] S. Song, Q. Liao, W. Shen, Y. Ruan, J. Xu, Numerical study on laminar convective
(2017) 649–661, https://doi.org/10.1016/j.ijheatmasstransfer.2017.05.004. heat transfer enhancement of microencapsulated phase change material slurry
[26] H.M. Ali, A. Arshad, M. Jabbal, P. Verdin, Thermal management of electronics using liquid metal with low melting point as carrying fluid, Int. J. Heat Mass Transf.
devices with PCMs filled pin-fin heat sinks: a comparison, Int. J. Heat Mass Transf. 62 (2013) 286–294, https://doi.org/10.1016/j.ijheatmasstransfer.2013.03.013.
117 (2018) 1199–1204, https://doi.org/10.1016/j.ijheatmasstransfer.2017.10.065. [45] R. Srikanth, C. Balaji, Experimental investigation on the heat transfer performance
[27] H.M. Ali, A. Arshad, M.M. Janjua, W. Baig, U. Sajjad, Thermal performance of LHSU of a PCM based pin fin heat sink with discrete heating, Int. J. Therm. Sci. 111
for electronics under steady and transient operations modes, Int. J. Heat Mass (2017) 188–203, https://doi.org/10.1016/j.ijthermalsci.2016.08.018.
Transf. 127 (2018) 1223–1232, https://doi.org/10.1016/j.ijheatmasstransfer.2018. [46] S. Mat, A.A. Al-Abidi, K. Sopian, M. Sulaiman, A.T. Mohammad, Enhance heat
06.120. transfer for PCM melting in triplex tube with internal-external fins, Energy Convers.
[28] M.J. Ashraf, H.M. Ali, H. Usman, A. Arshad, Experimental passive electronics Manage. 74 (2013) 223–236, https://doi.org/10.1016/j.enconman.2013.05.003.
cooling: parametric investigation of pin-fin geometries and efficient phase change [47] Y. Xu, M.-J. Li, Z.-J. Zheng, X.-D. Xue, Melting performance enhancement of phase
materials, Int. J. Heat Mass Transf. 115 (2017) 251–263, https://doi.org/10.1016/j. change material by a limited amount of metal foam: Configurational optimization
ijheatmasstransfer.2017.07.114. and economic assessment, Appl. Energy 212 (2018) 868–880, https://doi.org/10.
[29] H.M. Ali, M.J. Ashraf, A. Giovannelli, M. Irfan, T.B. Irshad, H.M. Hamid, F. Hassan, 1016/j.apenergy.2017.12.082.

13