Review

Phase change heat storage in an enclosure with vertical pipe

in the center
S. Lorente
a
, A. Bejan
b,
, J.L. Niu
c
a
Universit de Toulouse, UPS, INSA, LMDC (Laboratoire Matriaux et Durabilit des Constructions), 135, avenue de Rangueil, F-31 077 Toulouse Cedex 04, France
b
Duke University, Department of Mechanical Engineering and Materials Science, Durham, NC 27708-0300, USA
c
The Hong Kong Polytechnic University, Department of Building Services Engineering, Hong Kong, China
a r t i c l e i n f o
Article history:
Received 15 November 2013
Received in revised form 8 January 2014
Accepted 8 January 2014
Available online 1 February 2014
Keywords:
Melting
Energy storage
Phase change material
Scale analysis
Natural convection
a b s t r a c t
This paper documents the fundamentals of heat storage by melting a phase-change material inside a
cylindrical enclosure with a vertical heated pipe on its axis. The phenomenon is studied analytically
and numerically. The analysis consists of the scale analysis of the early stages of natural convection in
the liquid, and an analysis of the entire duration of the melting process. The numerical simulations cover
the entire process, and validate all the features predicted by theory.
2014 Elsevier Ltd. All rights reserved.
Contents
1. Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 329
2. Scale analysis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 330
3. Numerical model . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 331
4. Melting history. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 332
5. Concluding remarks . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 334
Acknowledgements . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 334
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 334
1. Introduction
The fundamentals of heat transfer in phase change materials
(PCM) have received considerable attention during the past two
decades. This is due to the fact that the heat transfer applications
of such materials are extremely broad: they range from thermal
energy storage and manufacturing, to metallurgy and the cooling
of electronics. For reviews on this body of work, see Refs. [19].
In recent years, we have seen a surge in interest in this domain
because of environmental concerns combined with fossil fuel
consumption and its consequences [1023].
The focus of the present paper is on thermal energy storage, and
more specically on latent heat thermal storage. One reason is that
air conditioning is responsible for a large fraction of the energy
consumption in buildings, which is why latent heat thermal stor-
age systems have been used as solution to mitigate the mismatch
between energy demand and supply. For example, when the en-
ergy supply is provided by solar collectors, the technique consists
of melting the PCM during day time, and releasing the stored en-
ergy (solidifying the PCM) during night time [24].
The main objective of this paper is to propose a theoretical anal-
ysis of the PCM melting process with natural convection in a cylin-
drical enclosure. The analysis is backed by full time, dependent
numerical simulations. Special attention is dedicated to predicting
the propagation of the phase-change front and the history and per-
formance of the phase-change process.
0017-9310/$ - see front matter 2014 Elsevier Ltd. All rights reserved.
http://dx.doi.org/10.1016/j.ijheatmasstransfer.2014.01.021

Corresponding author. Tel.: +1 919 660 5309; fax: +1 919 660 8963.
E-mail address: abejan@duke.edu (A. Bejan).
International Journal of Heat and Mass Transfer 72 (2014) 329335
Contents lists available at ScienceDirect
International Journal of Heat and Mass Transfer
j our nal homepage: www. el sevi er . com/ l ocat e/ i j hmt
2. Scale analysis
The phase change material is placed in a vertical cylindrical
enclosure of diameter D and height H. A heated pipe is placed on
the axis of the enclosure, Fig. 1. The outer surface of the enclosure
is adiabatic. The PCM is initially at a temperature T
0
lower than the
temperature of the pipe.
Modeling the PCM in liquid state, assume that the ow of the
liquid is incompressible. The ow and the heat transfer are gov-
erned by the laws of conservation of mass, momentum and energy
throughout in the enclosure,
rV 0 1
q
DV
Dt
rP lr
2
V F 2
qc
P
DT
Dt
kr
2
T 3
where V is the velocity vector of components (V
r
, V
h
, V
z
), and F is the
body force in the vertical direction. We assume that the liquid
PCM obeys the Boussinesq approximation, therefore q = q
0
[1 b(T T
0
)].
Immediately after the pipe temperature is set equal to T
w
, the
melt layer expands in the radial direction. Following the scale anal-
ysis developed in [1], from Eq. (3) we derive
DT
t
a
DT
d
2
4
and
t
d
2
a
5
where d is the radial thickness of the melted domain. Assuming that
convection develops, the scale analysis of the complete energy con-
servation equations shows the balance between inertia, convection
and conduction. As time passes, the inertia term decreases and the
balance between convection and conduction yields
Nomenclature
A cross section (m
2
)
C constant
c
P
heat capacity at constant pressure (kJ/kg K)
D diameter of the tank (m)
d diameter of the tube (m)
F volume force (N)
g gravity (m
2
/s)
H height of the tank (m)
h height of the boundary layers (m)
h
S
, h
F
specic enthalpy of solid, uid (J/kg)
k thermal conductivity (W/m K)
L
f
latent heat of fusion (J/kg)
m mass (kg)
P pressure (Pa)
Pr Prandtl number
Q heat transfer (J)
q
00
heat ux (W/m
2
)
Ra Rayleigh number
r radius (m)
T temperature (K)
t time (s)
U internal energy (J)
V (V
r
, V
h
, V
z
) velocity vector
z axial coordinate (m)
z
L
thickness of the melting front (m)
Greek letters
a thermal diffusivity (m
2
/s)
b thermal expansion coefcient (K
1
)
d radial thickness of the melted domain (m)
l dynamic viscosity (Pa s)
q density (kg/m
3
)
m kinematic viscosity (m
2
/s)
s temperature interval (K)
Subscripts
f nal
L latent
s sensible
w wall
0 reference
Symbol
nondimensional
Fig. 1. Cylindrical enclosure lled with phase change material (PCM).
330 S. Lorente et al. / International Journal of Heat and Mass Transfer 72 (2014) 329335
t
f

Hm
agbDT

1=2
6
where t
f
is the nal time of conduction, when convection becomes
dominant. Combining Eqs. (5) and (6) we obtain
d
f
HRa
1=4
H
7
where d
f
is the radial thickness of the melted domain at the end of
the conduction-dominated mode, and Ra
H
is the Rayleigh number
based on the height of the tube. Finally, replacing d
f
with the radius
of the PCM reservoir, D/2, we nd that in the long time limit when
the entire PCM is liquid
2H
D
Ra
1=4
H
8
is the criterion for the existence of distinct thermal boundary layers.
Convection exists in the enclosure when 2H=D < Ra
1=4
H
. Based on the
same argument, distinct velocity boundary layers are expected if
2H=D < Ra
1=4
H
Pr
1=2
Assume that the entire PCM is initially solid. The melting front
d(t) propagates radially from the tube wall into the PCM because
the tube wall temperature T
w
is above T
0
(Fig. 1). Assume that the
initial solid phase temperature is T
0
, which is the melting tempera-
ture of the PCM. We attach to the melting front d(t) a control vol-
ume that is very narrow in the radial direction and wide in the
angular direction. The control volume moves radially at the same
speed as the melting front. The law of conservation of energy states
that the heat transfer rate experienced by the control volume is bal-
anced by the change in the enthalpy ow rate through the volume
0 kA
@T
@r
qA
@d
@t

h
S
qA
@d
@t

h
F
9
where k is the thermal conductivity of the liquid. Because T
0
is the
melting temperature of the PCM, there is no heat transfer on the so-
lid side of the interface. Furthermore, A is the frontal area of the
control volume, h
S
and h
F
are the specic enthalpy of the solid
and liquid, respectively. Eq. (9) can be rewritten as
kA
@T
@r
AqL
f
@d
@t
10
where L
f
= h
F
h
S
is the latent heat of fusion. When the melt layer is
thin enough, the heat transfer across it is by conduction, the tem-
perature variation across it is linear, and Eq. (10) becomes
k
T
w
T
0
d
qL
f
@d
@t
11
Conduction is replaced by convection as time increases: boundary
layers of height h(t), Fig. 1, develop along the upper part of the tube,
and their thickness is
d
h
hRa
1=4
h
12
where Ra
h
is the Rayleigh number based on h. In the bottom part of
the reservoir, the melting front continues to be due to conduction,
whereas in the upper part it is due to conduction and convection.
In time, the convection zone expands downward.
When the Rayleigh number is high enough, the volume of liquid
PCM created in the upper part of the reservoir grows fast in the ra-
dial direction and creates a distinct layer of shallow liquid. Assume
that the ow is laminar, and the scale of the radius of the layer (r) is
greater than the scale of the vertical thickness z
L
, as shown in Fig. 1.
The prole of the intrusion layer is slender, and its ow is de-
scribed by the following boundary layer analysis. First, the conser-
vation of mass requires:
V
r
r

V
z
z
L
13
Near the tip of the layer the ow is two-dimensional, and the
momentum equation is a balance between buoyancy and friction,
gb
@T
@r
m
@
3
V
r
@z
3
14
The scale of @T/@r is DT/r, because the liquid temperature near the
center of the disc is T
w
and at the tip it is T
0
. The scale analysis ver-
sion of Eq. (14) is
gb
DT
r
m
V
r
z
3
L
15
The temperature gradient scale in the vertical direction is DT/z
L
, and
this drives the downward advance of the melting front z
L
(t), cf. Eq.
(11). The balance between vertical heat ux and latent heat ab-
sorbed at the melting front, kDT/z
L
qL
f
z/t, yields
z
L
t
kDT
qL
f

1=2
16
Repeating the preceding scale analysis, we obtain V
r
DT=r aDT=z
2
L
.
Combining this equation with mass conservation and Eq. (16), we
obtain the radial scale of the liquid layer,
r t
kDT
qL
f

5=4
gbDT
am

1=2
17
The ratio between the two scales of the horizontal layer is
z
L
r
Ra
1=5
r
18
which means that the vertical scale of the liquid layer evolves as
rRa
1=5
r
, where Ra
r
= gbDTr
3
/(ma).
3. Numerical model
Numerical simulations of the melting process were performed
based on Eqs. (1)(3), in the geometry represented in Fig. 1. Previ-
ous work [25,26] demonstrated that the isothermal tube is a very
good approximation of the conguration of pipes buried in a med-
ium with high thermal inertia. These preliminary studies consid-
ered the complete conjugate heat transfer process of uid
convection inside the pipes and transient conduction in the sur-
rounding solid. Because the residence time of the uid in the pipe
is much shorter than the time scale of the heat transfer between
the buried pipes and the surrounding medium, during most of
the heat transfer process the pipes are approximately isothermal,
at a temperature T
w
comparable with the inlet temperature of
the uid.
In the numerical simulations, the dimensions of the cylindrical
PCM were H = 1 m and D = 60 cm, while the tube diameter was
d = 2 cm. Parafn was chosen as PCM, and its characteristics are gi-
ven in Table 1. Parafn wax melts between T
0
= 303 K and
T
0
+ s = 304 K. The initial temperature of the PCM was T
0
= 303 K,
while the tube wall temperature T
w
= 323 K. The thermal conduc-
tivity and density are assumed to be the same in liquid and solid
states. The Rayleigh number is Ra
H
= 1.96 10
11
.
The phase change was modeled by means of the effective heat
capacity method by stating that between T
0
and T
0
+ s, the specic
heat of the PCM is
c

P
c
P

L
f
s
19
Below T
0
and above T
0
+ s, the specic heat has the value given in
Table 1. The modeling of the solid state was implemented by can-
celing the inertia terms in the momentum equation. For this reason,
we specied that the viscosity at temperatures below melting was
S. Lorente et al. / International Journal of Heat and Mass Transfer 72 (2014) 329335 331
four orders of magnitude higher than the one corresponding to li-
quid PCM [28].
The computations were performed by using a nite element
package [29]. In order to conrm that the solution is independent
of the size of the mesh, the solution was performed with a coarse
mesh and then was calculated again increasing the number of ele-
ments by a factor of 3 until the changes in the maximum temper-
ature became of order of 2%.
Fig. 2 shows the time evolution of the melting front. The con-
duction regime (Fig. 2(a)) is replaced by a boundary layer regime
which starts almost immediately (Fig. 2(b)). This agrees with the
analysis of Section 2. The upper region develops boundary layers
as shown in Fig. 2c. The height of the boundary layer increases in
time but its development is much more pronounced in the radial
direction, following the trend predicted in Eq. (7), Fig. 2(d). Once
the melting front has reached the wall of the container, it starts
moving downward as shown in Fig. 2(e)(h).
The conservation of energy requires
dU dQ 20
where U is the internal energy of the system and Q is the heat trans-
ferred to the entire volume of PCM. We have
Q
Z
T
T
0
mc
P
dT 21
The total mass of PCM consists of the mass of the solid (volume V
S
),
the mass of the liquid (volume V
L
), and the mass of PCM contained
in the melting front. The heat transferred during melting is given by
Q qL
f
V
L
22
therefore the total heat transferred to the PCM volume is
Q
Z
V
S
q
Z
T
T
0
mc
P
dT

dV
Z
V
L
q
Z
T
T
0
mc
P
dT

dV qL
f
V
L
23
When the temperature is lower than the melting temperature T
0
,
the heat is sensible and calculated for the mass of solid,
R
V
S
q
R
T
T
i0
mc
P
dT

dV. When T > T
0
, Eq. (23) accounts for the contribu-
tion of sensible heat and latent heat.
Fig. 3 shows the evolution of latent heat and sensible heat
stored in the PCM space, as a function of time. The latent heat con-
tribution is about ten times larger than the sensible heat contribu-
tion. Fig. 3 also shows the evolution of the melting fraction. In this
conguration, it would take more than 14 days to melt the entire
volume of PCM.
4. Melting history
The melting history (Fig. 3) and the time scale of the storage
process can be predicted. As shown in Fig. 1, we model the liquid
and solid conguration as a pool of liquid above a horizontal melt-
ing front. The height of the solid PCM is z(t), Fig. 4. In this congu-
ration there are two thermal resistances, between the pipe wall
and the liquid, and between the liquid and the melting front. Be-
cause the pipe surface lined by the natural convection boundary
layer is much smaller than the surface of the melting front, we as-
sume that the dominant resistance is between the pipe wall and
the liquid. This is equivalent to assuming that the liquid tempera-
ture is closely approximated by T
0
, which means that the temper-
ature difference across the boundary layer is close to DT = T
w
T
0
.
In order to predict the latent heat curve of Fig. 3, we proceed in
two steps. First, we estimate the instantaneous heat transfer rate
dQ/dt by focusing on the natural convection boundary layer,
dQ
dt
q
00
zpd 24
where q
00
is the z-averaged heat ux through the boundary layer.
When the boundary layer thickness is smaller than the pipe diam-
eter, the heat ux is given approximately by the solution for natural
convection on a plane vertical wall [1]
Table 1
Thermophysical properties for the parafn wax [27].
Density, q
(kg/m
3
)
Thermal
conductivity,
k (W/m K)
Heat capacity, c
P
(kJ/kg K)
Dynamic
viscosity, l (Pa s)
Latent heat of
fusion, L
f
(J/kg)
Melting
temperature,
T
0
(K)
Melting temperature
range, s (K)
Thermal expansion
coefcient, b (K
1
)
800 0.2 1.25 0.008 125,000 303 1 0.002
Fig. 2. The evolution of the melting front in a vertical plane through the centreline: (a) t = 500 s, (b) t = 10
3
s, (c) t = 10
4
s, (d) t = 5 10
4
s, (e) t = 10
5
s, (f) t = 5 10
5
s, (g)
t = 9 10
5
s, and (h) t = 1.2 10
6
s.
332 S. Lorente et al. / International Journal of Heat and Mass Transfer 72 (2014) 329335
Nu
z
C Ra
1=4
z
25
where C is approximately 0.2, and
Nu
z

q
00
DT

z
k
26
Combining Eqs. (24)(26), we obtain
dQ
dt
CpdkDT Ra
1=4
z
27
with the observation that Ra
z
= gbDTz
3
/am, where the height z is a
function of time, which must be determined.
Next, we focus on the horizontal melting front, the surface of
which is
p
4
D
2
d
2

, after assuming that the diameter of the melt
region around the pipe is comparable with d. The speed at which
the melting front descends is dz/dt, therefore
dQ
dt

dz
dt

p
4
D
2
d
2
qL
f
28
By eliminating dQ/dt between Eqs. (27) and (28) we arrive at
the differential equation for z(t). The dimensionless form of this
equation is
~
z
3=4

d
~
z
d
~
t
29
where
~
z
z
H
;
~
t
t
t
c
30
and the time scale is
t
c

HqL
f
D
2
d
2

4CdkDT Ra
1=4
H
31
Fig. 3. Sensible heat, latent heat, and melting front, as functions of time.
Fig. 4. Model with horizontal melting front, in the later stages of the process.
Fig. 5. The evolution of the horizontal melting front.
S. Lorente et al. / International Journal of Heat and Mass Transfer 72 (2014) 329335 333
Integrating Eq. (29) from
~
t 0 when ~z 1, we obtain the ~z
~
t
function
~
z 1
1
4
~
t

4
32
which has the asymptotes
~
z !1
~
t as
~
t !0 33
~
z !e
4
as e !0 34
where e 1
1
4
~
t 1. The curve ~z
~
t is shown in Fig. 5. In the same
gure we also plotted data extracted from the numerical experi-
ments. Good agreement between the theoretical curve provided
by Eq. (32) and the numerical results exists in the beginning. The
discrepancy increases when the horizontal melting front reaches
the bottom of the tank: this can be attributed to the one dimen-
sional character of the theoretical analysis.
Finally, Eq. (28) can be nondimensionalized as
d
e
Q ~z
3=4
d
~
t 35
where
e
Q
Q
p
4
D
2
d
2
HqL
f
36
Eliminating ~z
~
t between Eqs. (32) and (35), and integrating from
~
t 0 when
e
Q 0, we obtain
e
Q 1 1
1
4
~
t

4
37
The asymptotes of the
e
Q history are
e
Q !
~
t as
~
t !0 38
e
Q !1 e
4
as e !0 39
The
e
Q history and its rst asymptote are shown in Fig. 6 together
with the numerical results. Note the good agreement between the-
ory and numerical experiments.
5. Concluding remarks
In this paper we studied theoretically and numerically the pro-
cess of phase-change energy storage in a cylindrical enclosure
where the heating is provided by a vertical pipe placed coaxially
inside the enclosure. The theoretical part consisted of the scale
analysis of the early stages of natural convection in the liquid, fol-
lowed by a complete analysis of the later stages. The main theoret-
ical result is the predicted duration of the entire phase-change
process. This result is important in the design of phase-change
storage systems.
The numerical part consisted of simulations of the ow and
temperature elds in the liquid zone. The migration of the melting
front was followed in time. The numerical results obtained for the
history of energy storage and the movement of the melting front
agree well with the theoretical predictions.
On the practical side, this study shows that it is possible to rely
on theory to predict the evolution and performance of phase-
change storage processes. A cylindrical body of PCM with vertical
heated pipe in the center is perhaps the simplest possible congu-
ration, and yet a common element of many complete storage de-
signs, in which the fundamental characteristics of the melting
process are involved. Promising is the application of this dual ap-
proach (theory and numerical experiments) to the analysis and
optimization of more complicated phase-change heat storage
geometries such as the complex cavities in sensible-heat bodies
explored by Lorenzini et al. [30].
Acknowledgements
Profs. Lorente and Bejans work was supported by the National
Renewable Energy Laboratory. Prof. Nius work was supported by
Hong Kong Research Grant Council (RGC) General Research Fund
(RGC No. PolyU 5241/11E).
References
[1] A. Bejan, Convection Heat Transfer, fourth ed., Wiley, Hoboken, 2013.
[2] B. Zalba, J.M. Marin, L.F. Cabeza, H. Mehling, Review on thermal energy storage
with phase change: materials, heat transfer analysis and applications, Appl.
Therm. Eng. 23 (2003) 251283.
[3] M.M. Farid, A.M. Khudair, S.A.K. Razack, S. Al-Hallaj, A review on phase change
energy storage: materials and applications, Energy Convers. Manage. 45 (2004)
15971615.
[4] A. Shama, V.V. Tyagi, C.R. Chen, D. Buddhi, Review on thermal energy storage
with phase change materials and applications, Renewable Sustainable Energy
Rev. 13 (2009) 318345.
[5] M. Kenisarin, K. Mahkamov, Solar energy storage using phase change
materials, Renewable Sustainable Energy Rev. 11 (2007) 19131965.
[6] A. Shukla, D. Buddhi, R.L. Sawhney, Solar water heaters with phase change
material thermal energy storage medium: a review, Renewable Sustainable
Energy Rev. 13 (2009) 21192125.
[7] A. Dutil, D. Rousse, N. Ben Salah, S. Lassue, L. Zalzewski, A review on phase-
change materials: mathematical modeling and simulations, Renewable
Sustainable Energy Rev. 15 (2011) 112130.
[8] S.D. Sharma, K. Sagara, Latent heat storage materials and systems: a review,
Int. J. Green Energy 2 (2005) 156.
[9] M.F. Demirbas, Thermal energy storage and phase change materials: an
overview, Energy Sources Part B 1 (2006) 8595.
[10] I. Dincer, M.A. Rosen, Thermal Energy Storage Systems and Applications,
second ed., John Wiley and Sons Ltd., UK, 2002.
[11] P.D. Silva, L.C. Gonalves, L. Pires, Transient behavior of a latent-heat thermal
energy store: numerical and experimental studies, Appl. Energy 73 (2002) 83
98.
[12] A. Trp, K. Lenic, B. Frankovic, Analysis of the inuence of operating conditions
and geometric parameters on heat transfer in water-parafn shell-and-tube
latent thermal energy storage unit, Appl. Therm. Eng. 26 (2006) 18301839.
[13] P. Lamberg, R. Lehtiniemi, A.-M. Henell, Numerical and experimental
investigation of melting and freezing processes in phase change material
storage, Int. J. Therm. Sci. 43 (2004) 277287.
[14] A.H. Mosaffa, C.A. Infante Ferreira, M.A. Rosen, F. Talati, Thermal performance
optimization of free cooling systems using enhanced latent heat thermal
storage unit, Appl. Therm. Eng. 59 (2013) 473479.
[15] A. Dutil, D. Rousse, S. Lassue, L. Zalzewski, A. Joulin, J. Virgone, F. Kuznik, K.
Johannes, J.-P. Dumas, J.-P. Bdcarrats, A. Castell, L.F. Cabeza, Modeling phase
Fig. 6. Comparison between the theoretical and numerical results for the full life of
the melting process.
334 S. Lorente et al. / International Journal of Heat and Mass Transfer 72 (2014) 329335
change materials behavior in building applications: comments on material
characterization and model validation, Renewable Energy 61 (2014) 132135.
[16] Z.-X. Gong, S. Devahastin, A.S. Mujumdar, Enhanced heat transfer in free
convection-dominated melting in a rectangular cavity with an isothermal
vertical wall, Appl. Therm. Eng. 19 (1999) 12371251.
[17] K. Bilen, F. Takgil, K. Kaygusuz, Thermal energy storage behaviour of
CaCl
2
.6H
2
O during melting and solidication, Energy Sources Part A 37
(2009) 775787.
[18] A. Erek, Z. Ilken, M.A. Acar, Experimental and numerical investigation of
thermal energy storage with a nned tube, Int. J. Energy Res. 29 (2005) 283
301.
[19] K.A.R. Ismail, C.L.F. Alves, M.S. Modesto, Numerical and experimental study on
the solidication of PCM around a vertical axially nned isothermal cylinder,
Appl. Therm. Eng. 21 (2001) 5357.
[20] H.S. Choi, B.C. Shin, S.D. Kim, Heat transfer in a latent heat storage system,
Energy 14 (1989) 513524.
[21] A. Sari, K. Kaygusuz, Thermal energy storage system using stearic acid as a
phase change material, Sol. Energy 71 (2001) 365376.
[22] A. Sari, K. Kaygusuz, Thermal performance of palmitic acid as a phase change
energy storage material, Energy Convers. Manag. 43 (2002) 863876.
[23] K. Kaygusuz, A. Sari, Thermal energy storage system using a technical grade
parafn wax as latent heat energy storage material, Energy Sources 27 (2005)
15351546.
[24] A. Joulin, Z. Younsi, L. Zalewski, S. Lassue, D. R Rousse, J.-P. Cavrot,
Experimental and numerical investigation of a phase change material:
thermal-energy storage and release, Appl. Energy 88 (2011) 24542462.
[25] L. Combelles, S. Lorente, R. Anderson, A. Bejan, Tree-shaped uid ow and heat
storage in a conducting solid, J. Appl. Phys. 11 (2012) 014902, http://
dx.doi.org/10.1063/1.3671672.
[26] H. Kobayashi, S. Lorente, R. Anderson, A. Bejan, Serpentine thermal coupling
between a stream and a conducting body, J. Appl. Phys. 111 (2012) 044911,
http://dx.doi.org/10.1063/1.3689152.
[27] O. Bertrand, B. Binet, H. Combeau, S. Couturier, Y. Delannoy, D. Gobin, M.
Lacroix, P. LE Qur, M. Medale, J. Mencinger, H. Sadat, G. Vieira, Melting
driven by natural convection. A comparison exercise: rst results, Int. J. Therm.
Sci. 38 (1999) 526.
[28] W. Ogoh, D. Groulx, Effects of the number and distribution of ns on the
storage characteristics of a cylindrical latent heat energy storage system: a
numerical study, Heat Mass Transfer 48 (2012) 18251835.
[29] www.comsol.com.
[30] G. Lorenzini, C. Biserni, L.A. Isoldi, E.D. Dos Santos, L.A.O. Rocha, Constructal
design applied to the geometric optimization of Y-shaped cavities embedded
in a conducting medium, J. Electron. Packag. 133 (2011) 041008.
S. Lorente et al. / International Journal of Heat and Mass Transfer 72 (2014) 329335 335