Download as pdf or txt
Download as pdf or txt
You are on page 1of 20

Physica A 509 (2018) 316–335

Contents lists available at ScienceDirect

Physica A
journal homepage: www.elsevier.com/locate/physa

Melting process in porous media around two hot cylinders:


Numerical study using the lattice Boltzmann method

Mahmoud Jourabian a , A. Ali Rabienataj Darzi b , Davood Toghraie c , ,
Omid ali Akbari d
a
Doctoral School in Earth Science and Fluid Mechanics (ESFM), University of Trieste (UNITS), Trieste, Italy
b
Faculty of Mechanical Engineering, University of Mazandaran, Babolsar, Iran
c
Department of Mechanical Engineering, Khomeinishahr Branch, Islamic Azad University, Khomeinishahr, 84175-119, Iran
d
Young Researchers and Elite Club, Khomeinishahr Branch, Islamic Azad University, Khomeinishahr, Iran

highlights

• Numerical study of melting process in porous media using LBM is investigated.


• Influences of porosity and thermal conductivity ratio between PCM and porous structure are investigated.
• Enhancing the thermal conductivity ratio of the porous matrix results in increase of melting rate.

article info a b s t r a c t

Article history: Instantaneous melting of ice is accelerated inside a horizontal rectangular cavity with
Received 31 January 2018 two vertically arranged cylinders using metallic porous matrix made of Ni-Steel alloys.
Received in revised form 15 April 2018 The enthalpy-based lattice Boltzmann method (LBM) with a double distribution function
Available online 20 June 2018
(DDF) is employed at the representative elementary volume (REV) scale. Single-phase
Keywords: model is used and there is a local thermal equilibrium between porous structure and
Porous media ice. Inserting metallic porous material into the base PCM results in the enhancement of
Cylinder the heat conduction and weakening of the natural convection flow. Concentric pattern of
Convection the solid–liquid interface persists in porous samples comparing to the pure PCM melting.
PCM Reducing porosity causes decrease of the full melting time and thermal storage capacity of
LBM the system. Thermal conductivity ratio has to be enlarged in porous samples with higher
porosity from energy saving viewpoint.
© 2018 Elsevier B.V. All rights reserved.

1. Introduction

In recent years, the heat transfer enhancement in mechanical and electronic instruments have attracted the researchers
and industrialists such as energy systems, fluidized bed microfluidic and nanofluidic [1–8], Phase change material and other
applications [9–12]. There are many works in literature that focused on melting and solidification of phase change materials
(PCMs) inside spherical containers [13,14], shell-and-tube heat storage unit [15], enclosure with discrete protruding heat
sources [16,17], concentric and eccentric horizontal annulus [18], single-tube [19] and multi-tube heat exchangers [20].
Solid–liquid phase change processes around different arrays of cylinders appear in many applications such as melting of
soil, removal of heat from engines or thermal storage systems. The low thermal conductivity of PCMs is main disadvantage

∗ Corresponding author.
E-mail address: Toghraee@iaukhsh.ac.ir (D. Toghraie).

https://doi.org/10.1016/j.physa.2018.06.011
0378-4371/© 2018 Elsevier B.V. All rights reserved.
M. Jourabian et al. / Physica A 509 (2018) 316–335 317

Nomenclature
c Streaming speed [m s−1 ]
ci Discrete lattice velocity i
cp Heat capacity [J kg−1 K−1 ]
cs Speed of sound [m s−1 ]
Da Darcy number
En Enthalpy [J]
Ens Enthalpy of the solid phase [J]
Enl Enthalpy of the liquid phase [J]
fl Liquid fraction
Fo Fourier number
Fε Geometric function
g Acceleration due to gravity [m s−2 ]
k Thermal conductivity [W m−1 K−1 ]
K Permeability [m2 ]
l Length scale [m]
Lf Latent heat [J kg−1 ]
P Pressure [kg m−1 s−2 ]
Pr Prandtl number
R Source term of temperature field
Ra Rayleigh number
Ste Stefan number
t Time [s]
T0 The initial temperature of PCM [K]
T1 The temperature of heated surfaces [K]
Tm Melting temperature [K]

u Velocity [m s−1 ]
Greek symbols
α Thermal diffusivity [m2 s−1 ]
β Thermal expansion coefficient [K−1 ]
∆t Lattice time step size
∆x Lattice cell size
ε Porosity
θ Dimensionless time (=Ste×Fo)
υ Kinematic viscosity [m2 s−1 ]
ρ Density [kg m−3 ]
σ Thermal capacity ratio
τ Lattice relaxation time
Ψ Scalar function
Ω Viscosity ratio
Superscripts
* dimensionless symbol
Subscripts
i Direction
m Solid matrix
f Liquid phase of PCM
s Solid phase of PCM
e Effective
318 M. Jourabian et al. / Physica A 509 (2018) 316–335

from energy saving point of view. Some methods have been proposed to improve the heat transfer in base PCM such as
using nanofluids [21–39], metallic porous matrices and so on. Impregnating high thermal conductivity metallic foams and
porous matrices with base PCM is a promising method to improve thermal performance of latent heat thermal energy storage
systems (LHTESS). The metallic porous matrix material has a lightweight, high surface area for transferring heat and high
thermal conductivity.
From an experimental viewpoint, Wu and Zhao [40] investigated NaNO3 melting inside a vertical cylindrical container
with metallic foams. In the heating solid NaNO3 process, the heat transfer rate was enhanced using metal foams made of
copper and steel alloys by 210%. In the process of heating fluid PCM, natural convection flow was weakened by porous
structures. Heat transfer performance of composites was no longer better than that of pure NaNO3 . Li et al. [41] numerically
and experimentally examined the paraffin melting embedded with the copper porous foams in a plexiglass cavity heated
from the left side.
Natural convection flow was prevailing in the pure PCM. The improvement of heat conduction overcompensated
the suppression of natural convection of melted PCM in the foam-PCM composites. The uniformity of the temperature
distribution inside the foam-PCM composite was augmented by abating porosity. The solid/liquid movement rate was bigger
for composites with low porosity. Chen et al. [42] numerically and experimentally studied at the pore scale melting of Paraffin
wax with aluminum foams inside a cubic enclosure as the thermal storage container.
Heat conduction mainly dominated in the metallic matrix. Effect of the metal foam structure on the heat transfer
is significant during the melting of PCM. Convection flow had little influence. Existence of metal foam engendered
augmentation of the flow resistance in the domain. Trelles and Dufly [43] applied the matrix-based enthalpy formulation in
a 3D finite volume method (FVM) to investigate the thermoelectric cooling in a porous latent heat energy storage apparatus.
Smaller porosities (meaning more aluminum) abated the thermal resistance of the device and diminished its effective energy
storage capacity. A larger porosity (meaning more PCM) did not show desired temperature of water. Mesalhy et al. [44]
conducted a FVM-based numerical modeling in a non-staggered grid to solve governing equations of melting in the porous
media inside cylindrical annulus. Using a solid matrix with a higher porosity and thermal conductivity enhanced the thermal
performance of the system. Huber et al. [45] derived a double distribution function (DDF)-lattice Boltzmann formulation
based on the D2Q5-D2Q9 lattice models to investigate melting of a pure substance in a synthetic porous cavity heated
from below. Gao et al. [46] developed a thermal lattice Boltzmann method (TLBM) at the representative elementary volume
(REV) scale to explore solid–liquid phase change processes coupled with natural convection in the metal foam under the
non-equilibrium thermal condition. While lowering the porosity resulted in the enhancement of the melting rate, the foam
pore size affected the process insignificantly. The Brinkman–Forchheimer extended Darcy model together with the enthalpy
method in the finite volume method (FVM) were utilized by Tian and Zhao [47] to research influences of copper metal foams
on melting of RT58 in a square cavity. Intensification of the conduction heat transfer exceeded the weakening of natural
convection flow in porous samples. Gao and Chen [48] proposed a LB-based model for the non-linear convection–diffusion
equation to examine melting process with the natural convection in the porous media at the REV scale. By increasing the
Rayleigh number, Darcy number and porosity, the influence of the natural convection on the process became stronger.
Jourabian et al. [49] employed an enthalpy-based LBM to simulate ice melting inside a horizontal annulus filled with the
Al2 O3 porous matrix. Liu et al. [50] employed the enthalpy-porosity approach at the pore scale to scrutinize the melting of
the paraffin-RT58 in a shell-and-tube latent heat thermal storage system (LHTSS) filled with a copper porous matrix. As the
pore size was decreased, the process was decelerated and the effect of the natural convection flow was disappeared. Liu
and He [51] developed an enthalpy-based multiple-relaxation-time (MRT)-LB model for simulating instantaneous melting
in a porous cavity at the REV scale. The conduction was prevailing at the beginning of process while in later times the phase
change interface was gradually deformed. Tao et al. [52] implemented an iterative enthalpy-based formulation in the DDF-LB
model to research melting of the copper foam/paraffin (PCM) composite inside a cavity. Enhancing the pore size in the low
porosity or reducing the pore size in the high porosity improved significantly uniformity of the heat transfer. As could be
seen, inserting metallic porous matrix into base PCM may have two effects. From an energy saving point of view, it promotes
thermal conductivity of the base PCM and heat transfer performance of the system while it substantially suppresses the
natural convection flow. Keep in mind that the natural convection flow may expedite the solid–liquid phase change process
in pure PCM.
The lattice Boltzmann method (LBM) has evolved as a complementary approach to solve the governing equations
of the melting process in a porous medium. Because of its particulate nature, the LBM has some benefits over the
conventional computational fluid dynamics (CFD) methods for instance handling the complicated boundary condition with
the high speed of running. Applications of LBM could be found in [53,54]. Comparisons between LBM and finite volume
method for the natural convection flow inside cavities and enclosures could be seen [55]. A quick survey in the literature
ascertains that enthalpy-based methods were employed mostly because they eradicate the numerical instability and sharp
discontinues by introducing a mushy zone. There is no need for tracking the solid–liquid interface throughout the entire
process.
There are generally two scales that have been adopted in the LB simulations of flow fields in porous media. They are pore
scale and representative elementary volume (REV) scale. The pore-scale is the simplest tool to apply the LBM in porous media.
It cannot be used for big domains. On the contrary, the REV scale is represented as a minimum element in which the scale
characteristics of porous medium can hold and is much larger than the pore scale. No one in the past investigated melting
in a water-saturated porous medium around two vertically arranged cylinders in a horizontal rectangular cavity. Sasaguchi
M. Jourabian et al. / Physica A 509 (2018) 316–335 319

et al. [56] and Sugawara and Beer [57] studied the phase change process of pure PCM around two and four cylinders in a
rectangular cavity, respectively.
The purpose of this work is to expedite melting of ice using the steel-Ni porous matrix in a horizontal rectangular cavity
with two vertically arranged hot cylinders. The enthalpy-based LBM with the DDF model is used at the REV scale. There
is local thermal equilibrium between porous medium and PCM. The Prandtl number, Stefan number, Rayleigh number
and Darcy number are equal to 6.2, 1, 105 and 10−3 , respectively. Governing equations of the solid–liquid phase change
process in the porous medium at the REV scale are provided in Section 2. In Section 3, the DDF model to solve velocity and
temperature equations and boundary conditions are described. In Section 4, results of pure PCM melting and PCM melting
in porous medium in cavities are compared against existing data. In Section 5, mesh requirement and computer system on
which simulations are performed are stated. In Section 6, the case under investigation in this study and initial conditions
are mentioned. In Section 7, influences of porosity 0.6< ε <0.8 and thermal conductivity ratio 15<Kr<35 on instantaneous
temperature contours, streamlines and liquid fractions are illustrated to find the best case from an energy saving perspective.

2. Governing equations

The flow is laminar and incompressible in the liquid phase. The incompressibility condition is checked through defining
a characteristic velocity in the natural convection regime from a LB formulation point of view. Slight density difference is
considered using the Boussinesq approximation. There is no chemical reaction, viscous dissipation or radiation heat transfer.
The PCM and porous medium are isotropic, homogeneous and immovable. The PCM and porous matrix in the composite are
in the local thermal equilibrium. The thermo-physical properties of the PCM and porous medium are constant during the
process. Density of liquid phase is equal to that of solid phase.
Based on these assumptions, the Navier–Stokes equations for the natural convection flow in a porous medium at the REV
scale are [58–60]:
∇.⃗u = 0
∂ u⃗ (
( )
) u⃗ 1 (1)
+ u⃗ .∇ = − ∇ (ε p) + νe ∇ 2 u⃗ + F⃗
∂t ε ρ
νe ∇ 2 u⃗ is the Brinkman term. It accounts for the presence of a solid boundary. F⃗ is the total body force. It is evaluated from
[50–60],
εν ε Fε ⏐ ⏐ ε3 d2p
F⃗ = − ⃗ − √ ⏐u⃗ ⏐ u⃗ + εG
u ⃗, K = (2)
K K 150 (1 − ε)2
K is the permeability of the porous medium and dp is the solid particle diameter.
Eq. (1) simplifies to the standard Navier–Stokes (NS) equation for free flows if porosity ε is unity.
Additionally, owing to the presence of the porous medium, first and second terms are defined empirically. They are
accounted for the linear (Darcy) and nonlinear (Forchheimer) drag forces, respectively.
In Eq. (2), Fε is the geometric function and the last term is the external force due to the buoyancy:
⃗ = −⃗g β (T − Tm )
G (3)
Note that T represents the temperature of the liquid phase.
In agreement with the experimental data obtained by Ergun [61], the geometric function is [62]:
1.75
Fε = √ (4)
ε 150ε
The volume-averaged energy equation for the natural convection melting in porous media is [59,60],
∂T Lf ∂ f l
σ + u⃗ .∇ T = ∇ (αe ∇ T ) − ε (5)
∂t cpf ∂ t
Note that σ is the thermal capacity ratio. It is specified as [48]:
ρ cp
( )
σ = ε + (1 − ε) ( )m (6)
ρ cp f
The effective thermal conductivity and thermal diffusivity are calculated by [48]:
ke
ke = ε kf + (1 − ε) km , kf = ks , αe = ( (7)
ρ cp
)
f

In the enthalpy-based methods, the local enthalpy is divided into two components, the sensible and latent heat. It is
estimated from [59,60],
En = cp T + Lf fl (8)
320 M. Jourabian et al. / Physica A 509 (2018) 316–335

It is an iterative method. The liquid fraction is subsequently updated [59,60],


En < Ens = cp Tm

⎪ 0

⎨ En − En
s
fl = Ens ≤ En ≤ Ens + Lf (9)
En − Ens
⎩ l


1 En > Ens + Lf

By establishing some non-dimensional quantities, dimensionless forms of governing equations are derived [59,60],
⃗l
u νeT − T0 t αf g β (T1 − T0 ) l3
⃗=
U , T∗ = , Ω=
, Fo = t ∗ = 2 , Ra =
αf T1 − T0 νf l νf αf
(10)
cpf (T1 − T0 ) K νf pl2 X Y
Ste = , Da = 2 , Pr = , P = ,x= ,y=
Lf l αf ρf ν f α f l l
l, which is the length scale in this study, is equivalent to the height of the cavity.
The properties defined by the subscript f are of the base PCM. νf and αf are the kinematic viscosity and thermal diffusivity
of the base PCM, respectively. Dimensionless forms are [59,60],

∇.U⃗ = 0 (11)

( )
1 ∂U

( ⏐ ⏐)
1 1 ( ) 1 1 Fε g⃗
+ ⃗ .∇ U⃗ = −∇ P +
U ∇ 2 U⃗ − + √ ⏐U⃗ ⏐ U + RaT ∗ ⏐ ⏐
⏐ ⏐
(12)
Pr ε fl ∂ t ∗ (ε fl )2 ε fl Da Pr Da ⏐g⃗ ⏐

∂T ∗ 1 ∂ fl
σ + U⃗ .∇ T ∗ = ∇ αe ∇ T ∗ − ε
( )
(13)
∂t ∗ Ste ∂ t ∗

3. LBM

3.1. LB equation for the velocity field

To simulate isothermal incompressible flow in porous media at the REV scale, Guo and Zhao [34] presented a LB equation
for a porous medium with porosity ε . It was emphasized that in order to collect correct equations for density and velocity
fields, force terms such as the Darcy’s term (linear drag force), Forchheimer’s expansion (nonlinear drag force) and external
forces should be incorporated properly. To accomplish, the porosity and a general force term should be considered in the
equilibrium distribution function (EDF) and the evolution equation, respectively. Normally, in the LBM the velocity field is
calculated by quantities fi (x,t) showing the particle density distribution (PDF).
The evolution of fi (x,t) is determined by [59,60],
eq
fi (x, t ) − fi (x, t )
fi (x + c⃗i ∆ t , t + ∆ t) − fi (x , t) = − + ∆t c⃗i Fi (14)
  
 τ

  
 force term
Streaming
Collision

The EDF is calculated from [24],


[ ]
eq 3 9 3
fi = ωi ρ 1 + 2 c⃗i . u⃗ + (c⃗ . u
4 i
⃗ )2 − ⃗ . u⃗
u (15)
c 2ε c 2ε c 2

For simplicity and without losing generalization, we presume the D2Q9 model. It is revealed in Fig. 1. Weights and local
particle velocities c⃗i are,
4


⎪ i=0
9



1

ωi = i = 1, 2, 3, 4 (16)

⎪ 9
⎩1


i = 5, 6, 7, 8

36

⎨(0, 0) i = 0

c⃗i = √(cos[(i − 1)π/2], sin[(i − 1)π/2])c i = 1, 2, 3, 4 (17)


2(cos[(2(i − 5) + 1)π/4], sin[(2(i − 5) + 1)π/4])c i = 5, 6, 7, 8

M. Jourabian et al. / Physica A 509 (2018) 316–335 321

Fig. 1. 2D nine-velocity models.

Fig. 2. Curved wall geometry.

The propagation speed c is equal to unity. The density and velocity are [48],

ρ (x , t) = fi (x , t) (18)
i

∑ F⃗
ρ u⃗ (x , t) = c⃗i fi (x , t) + (19)
2
i

The velocity of melted PCM in porous media is computed using an auxiliary velocity [48],

V⃗
⃗=
u √ (20)
c0 + c02 + c1 |V |

∑ c⃗i fi εG

V⃗ = + (21)
ρ 2
i
322 M. Jourabian et al. / Physica A 509 (2018) 316–335

Fig. 3. Bounce-back boundary conditions. (a) ∆=1/2, the perfect bounce-back without interpolation. (b) ∆<1/2, the bounce-back with interpolations before
the collision with the wall located at xw. (c) ∆ >1/2, the bounce-back with interpolations after the collision with the wall.

c0 and c1 are estimated by [48],


1( νε ) ε Fε
c0 = 1+ , c1 = √ (22)
2 2K 2 K
Force term for the fluid flow in the porous medium is [48],

⎡ ( ) ⎤
(
1
)
3c⃗i .F⃗ ⃗ F⃗ : c i c⃗i
9 u ⃗ .F⃗
3u
Fi = ωi ρ 1− ⎣ + − ⎦ (23)
2τ c2 εc 4 εc 2

The kinematic viscosity in the LBM is determined as a function of relaxation time:


c2
νe = (τ − 0.5)cs2 ∆ t , cs2 = (24)
3
It is obvious that the value of the relaxation time must be bigger than 0.5 to make the viscosity positive.
√ The incompressibility condition in the natural convection regime can be ensured if the characteristic velocity
g⃗ × β × ∆T × l is small in comparison with the fluid speed of sound.

3.2. LB equation for the temperature field

Guo and Zhao [63] developed a LB formulation for the natural convection flow in a porous medium at the REV scale.
By establishing a new distribution function, calculation of the temperature field is achieved. It was stated that restrictions
like slight variability of the temperature, unstable analysis or fixed Prandtl number would be removed in the DDF model.
Treatment of the temperature is simplified if the compression work induced by the pressure and viscous heat dissipation
are absent.
M. Jourabian et al. / Physica A 509 (2018) 316–335 323

Fig. 4. Comparison of streamline and isotherm (a). present study and (b) Guo and Zhao [63].

Fig. 5. Comparison of dimensionless temperature at the mid-height of the cavity predicted by this study with that of Guo and Zhao [39] for ε = 0.4, Pr=1.0,
Da=10−4 , Ra=107 .

Shi and Guo [64] recommended a LB formulation to solve the advection–diffusion equation which included nonlinear
convective and diffusive terms. According to Shi and Guo [64], this nonlinear equation with the term R(x,t) as the source
324 M. Jourabian et al. / Physica A 509 (2018) 316–335

Fig. 6. Comparison of phase field, temperature field and flow field for pure paraffin wax melting against [42].

Fig. 7. Comparison of the instantaneous average melting front position for pure paraffin wax melting against [42].
M. Jourabian et al. / Physica A 509 (2018) 316–335 325

Fig. 8. Configuration of the early LB work [48] for convection melting in the porous media.

Fig. 9. Comparison of the average Nusselt number versus the dimensionless time between the present study and that of Gao and Chen [48].

term is:
∂t ψ + ∇.B (ψ) = ∇. [α∇ D (ψ)] + R (x, t ) (25)
ψ is the temperature considered as the scalar field. B(ψ ) and D(ψ ) are functions of ψ . Accordingly, the temperature
distribution function (TDF) is,
1 eq
gi (x + c⃗i ∆t , t + ∆t) = gi (x , t) − (gi (x , t) − gi (x , t)) + Ri ∆t (26)
τT
τT is a relaxation time for the temperature. This EDF is computed as [64]
c2ψ I c2I
[ ]
eq 3 9
gi = ωi ψ + 2 c⃗i . B + 4 (E − ): (c⃗i c⃗i − ) (27)
c 2c 3 3
326 M. Jourabian et al. / Physica A 509 (2018) 316–335

Fig. 10. Irregular thermal storage considered here.

I is called the unit tensor. E(ψ ) is the second order moment of the EDF,

E(ψ ) = E0 (ψ ) + cs2 D(ψ )I (28)


E0 (ψ ) is a tensor function of the scalar field. More details are in [64].
ψ and B(ψ ) are,
8 8 8
∑ ∑ eq
∑ eq
ψ= gi = gi , B (ψ) = c⃗i gi (29)
i=0 i=0 i=0

The source term in the scalar equation is evaluated based on Shi and Guo [64]:
c⃗i .B′ (φ) ∂B τT − 21 Lf ∂ f l
( )
Ri = ωi R 1+λ , B′ (ψ) = , λ= , R = −ε (30)
cs2 ∂ψ τT cpf ∂ t

Finally, the source term takes the following form:


( )[
τT − 1
c⃗i .⃗ fl (t + ∆t) − fl (t )
]
1 u
Ri = −ωi ε 1+ 2
(31)
σ Ste τT σ cs2 ∆t
In accordance with Eq. (27), the equilibrium temperature distribution function (ETDF) becomes
[ ]
eq 3 9 3
gi = ωi T 1+ c⃗ . u
2 i
⃗+ (c⃗ . u
4 i
⃗ )2 − ⃗ . u⃗
u (32)
σc 2σ 2 c 2σ 2 c 2
M. Jourabian et al. / Physica A 509 (2018) 316–335 327

Fig. 11. Streamlines and temperature contours for the pure PCM melting, Pr=6.2, Ra=105 and Ste=1.

The temperature and thermal diffusivity are,


8
∑ σ c2
T = gi , α= (2τT − 1) ∆t (33)
6
i=0

3.3. Boundary condition in LBM

Bounce-back boundary scheme proposed Ladd [65] is used to obtain no-slip velocity conditions. On the left wall of a
cavity, leaving distribution functions at i=1, 5 and 8 directions are unknown, which can be determined from,
f1 = f3
f5 = f7 (34)
f8 = f6

The walls of enclosure are adiabatic. No heat conduction is in the normal direction. For instance, for right wall, the
unknown distribution function, g3 is,

g(3, n, j) = g(3, n − 1, j) (35)

Here n represents nodes on the right wall while n − 1 represents nodes adjacent to the right wall.
Fig. 2 illustrates a section of the curved wall geometry differentiating a solid region from a fluid where the black small
circles on the boundary are xw , the open circles characterize boundary nodes in the fluid region xf and the gray solid circles
designate those in the solid region xb .
328 M. Jourabian et al. / Physica A 509 (2018) 316–335

Fig. 12. Streamlines and temperature contours for ε = 0.8, Pr=6.2, Ra=105 , Ste=1, Da=10−3 and (a) Kr=15, (b) Kr=25, (c) Kr=35.

In the boundary condition, to apply the streaming steps to the fluid nodes xf both PDFs at the position xb must be specified.
The intersected link in the fluid region is 0≤∆ ≤1:
 
 xf − xw 
∆=   (36)
xf − xb 
The half-way bounce back defined as the no-slip boundary condition has the value of 0.5 to the boundary wall as shown
in Fig. 3a.
M. Jourabian et al. / Physica A 509 (2018) 316–335 329

Fig. 13. Instantaneous variation of average melting front position for (a) ε = 0.8, (b) ε = 0.7, (c) ε = 0.6.

Because of the curved boundaries, delta values in the interval of (0, 1) are now achievable. Fig. 3b portrays the bounce
back behavior of a surface with a delta value smaller than 0.5 and Fig. 3c represents the bounce back behavior of a wall with
delta bigger than 0.5. In all cases, the reflected distribution function at xf is indefinite. Since the fluid particles in the LBM are
always assumed to move one cell length per time step (∆x=∆t=1), the fluid particles would rest at an intermediate node xi .
To compute the reflected distribution function in the node xf , an interpolation procedure needs to be performed.
For treating the velocity field in a curved boundary, the method is based on [66] while for handling the temperature field
the method is related to the extrapolation scheme of the second-order accuracy clarified in [67].

3.3.1. Velocity in curved boundary condition


To estimate the PDF in the solid region fi (xb , t ) based upon the boundary nodes in fluid region, the bounce-back boundary
conditions are incorporated with the interpolations plus a one-half grid spacing correction at the boundaries.
After that, the Chapman–Enskog expansion for the post-collision distribution function is manifested as:
3
fi xb + c⃗i ∆t , t + ∆t = (1 − χ) fi xf + c⃗i ∆t , t + ∆t + χ fi∗ (xb , t ) + 2ωi c⃗i . u
( ) ( )
⃗w (37)
c2
eq
) 3
fi∗ (xb , t ) = fi xf , t + ωi ρ xf , tc⃗i . u ⃗ bf − u⃗ f , c⃗i ≡ −⃗ci
( ) ( ( )
(38)
c2
(2∆ − 1) 1
= u⃗ ff = u⃗ xff , t , χ = ,⇐ 0 ≤ ∆ <
( )
⃗ bf
u (39)
(τ − 2) 2
1 3 (2∆ − 1) 1
⃗ bf
u = (2∆ − 3) u⃗ f + ⃗w , χ =
u ,⇐ ≤ ∆ < 1 (40)
2∆ 2∆ τ − 1/2 2
⃗ w is the velocity of the solid wall. u⃗ bf is the imaginary velocity defined for the interpolation procedure.
where u
330 M. Jourabian et al. / Physica A 509 (2018) 316–335

Table 1
Validation of average Nusselt number for natural convection flow in porous media.
Present Guo and Zhao [63] Succi et al. [68] Chen and Doolen [69]
Ra = 105 1.065 1.066 1.064 1.067
Ra = 106 2.57 2.603 2.58 2.55
Ra = 107 7.66 7.788 7.677 7.81

3.3.2. Temperature in curved boundary condition


Each TDF has two separate components: equilibrium and non-equilibrium
neq eq
gi (xb , t ) = g (xb , t ) + gi (xb , t ) (41)
i

By replacing Eq. (41) in Eq. (26) and in the absence of the source term, we get,
( )
eq 1 neq
gi xb + c⃗i ∆t , t + ∆t = g (xb , t ) + 1 − g (xb , t )
( )
(42)
i τT i

Both functions on the right hand side (RHS) of Eq. (42) have to be specified. The equilibrium and non-equilibrium parts
read as,
( )
eq 3 9 3
g
i
(xb , t ) = ωiT Tb∗ 1 + 2 c⃗i . u⃗ b + 4 (c⃗i . u⃗ b )2 − 2 u⃗ b . u⃗ b (43)
c 2c 2c

Tb∗ = Tb1 ⇐ ∆ ≥ 0.75 (44)

Tb∗ = Tb1 + (1 − ∆) Tb2 ⇐ ∆ < 0.75 (45)

Tw + (∆ − 1) Tf
Tb1 = (46)

2Tw + (∆ − 1) Tff
Tb2 = (47)
1+∆
Here, Tf and Tff denote the fluid temperature in nodes xf and xff , respectively.
The next task is to determine the second function. As a second-order accurate approximation, this function can be
computed as:
neq neq neq
(xb , t ) = ∆ gi xf , t + (1 − ∆) gi xff , t
( ) ( )
gi (48)
From the Chapman–Enskog analysis, it could be said that:
neq
gi (x, t ) = gi1 (x, t ) δ x (49)
eq
Readers note that gi0 (x, t ) has the same order as (x, t ) and gi
neq neq
gi1 (xw , t ) − gi 1
xf , t = O (δ x) , gi (xw , t ) − gi xf , t = O δ x2
( ) ( ) ( )
(50)
It can be proven that
neq neq
(xw , t ) − gi xff , t = O δ x2
( ) ( )
gi (51)

4. Validations

First, we simulate the natural convection flow in a plain porous cavity. The cavity is heated from the left wall and it is
cooled from the right wall, while the top and bottom walls are adiabatic. Variables Pr=1.0, ε =0.4, Da=10−4 and Rayleigh
numbers ranging from 105 to 107 are selected.
Resembling to the work of Guo and Zhao [63], the mesh size is a 2D (128×128) lattice and the relaxation times are
τT =τ =0.513. As revealed in Table 1, we analyze the average Nusselt number with previous studies [63,68,69].
As can be found from Fig. 4, we establish a schematic comparison between streamlines and isotherms in this study with
those of Guo and Zhao [63] for Ra=107 . Dimensionless temperature profiles at half height of the enclosure are obtained from
this study and Guo and Zhao [63]. Fig. 5 indicates that results are in good agreement.
Convection controlled melting of pure paraffin wax in a rectangular cavity, which is heated from the left wall T1 >T m and
is cooled from the right wall T0 , is taken for validation. Top and bottom walls are adiabatic and the subcooling is disregarded
M. Jourabian et al. / Physica A 509 (2018) 316–335 331

T0 =T m . The mesh size is 100×65. It is obvious from Fig. 6 that the phase field, temperature field and flow field at θ =0.2
agree well with those of Chen et al. [42] for Ra=105 , Pr =5 and Ste = 1. It is interesting to state that at θ =0.2, the shape of the
solid–liquid interface is affected mainly by the natural convection flow and there is no effect of the conduction heat transfer.
Also, we compare the instantaneous average melting front position from present study and Chen et al. [42].
It is shown in Fig. 7. As would be seen, the LBM code reproduces satisfactorily past experimental and numerical data on
pure PCM melting. In this study, the variable Sav is computed from [45]:
∫ L
Sav (t ) = xm dy ∝ L2 Ra1/4 θ (52)
0

It must be mentioned that xm is the distance between the melting front and left wall. It is valid if the melting front is
deformed in the natural convection-driven melting.
The third validation is accomplished using LB simulation of the melting in a square cavity filled with a porous matrix.
Results are compared against study of Gao and Chen [48]. The schematic of the geometry is seen in Fig. 8.
Here, T0 is the temperature of the cold wall in the right side. It is equal to the dimensionless melting temperature.
We choose these variables, ε =0.4, Ω =1, ks =kf =km , Tm =T 0 , Da=10−4 , Pr=5.0, Ste=1.0 and Rayleigh numbers of 2×106 ,
4×106 and 8×106 .
As is shown in Fig. 9, the average Nusselt number on the hot surface agrees acceptably with results of Gao and Chen [48].

5. Mesh independency

Regarding the size of the mesh in this study, multitudinous grid sizes are checked to guarantee independency of melt
fractions. Subsequently, an arrangement of 100 × 200 grids is found suitable in this study. All simulations are performed on
a computer with Dual cores CPU and 4 G RAM. When the PCM is fully melted, simulation terminates.

6. Configurations under investigation

The irregular thermal storage unit investigated here is sketched in Fig. 10. The space between two vertically arranged hot
cylinders and rectangular cavity is filled with water as a PCM. It is embedded in the Ni-steel porous matrix.
The thermophysical properties of porous matrix and water [70–72] are listed in Table 2. The Prandtl number, Stefan
number, Rayleigh number and Darcy number are 6.2, 1, 105 and 10−3 , respectively. Porosity is in this range 0.6< ε <0.8.
The variable 15<Kr <35 is the ratio between thermal conductivity of the porous medium to that of pure PCM. The vertical
and horizontal walls are adiabatic. Here, T0 is the initial temperature of the solid PCM. Keep in mind that the dimensionless
time θ is equal to the Stefan number multiplied by Fo. It is indicated in Eq. (10).
At θ = 0, the temperature of cylinders is raised to T1 which is higher than the initial temperature of the PCM. Solid
subcooling abates the melting rate. Consequently, to get higher efficiency in heat transfer enhancement techniques the
subcooling case is disregarded, T0 =T m .
Curved boundary condition for the velocity and temperature fields in LB framework is in [73–75]. No magnetic field
[76–82] is present in the flow field. For Rayleigh number of 105 , no turbulence is observed in the flow field, and hence, the
process remains symmetric. Half of the domain is shown.

7. Results

In Fig. 11, instantaneous temperature contours (right) and streamlines (left) are shown for pure PCM melting.
At initial stages, the conduction heat transfer dominates in the liquid phase. It is between the solid phase and surface of hot
cylinder. Isotherms have a concentric pattern. At time θ =9.99×10−3 , as the process advances more, the natural convection
flow becomes dominant above the cylinders and a pair of recirculating vortices are appeared around each cylinder.
It should be pointed out that the melted PCM in the vicinity of the hot cylinder surface departs symmetrically toward the
top. It is cooled and is shifted downward repeatedly adjacent to the solid–liquid interface.
It is worth mentioning that the melting fronts are deformed. They are deviated from the early concentric shape. The
centers of these vortices are shifted upwardly as well. It could be said that the melting rate is bigger in upper parts of
configuration due to the natural convection flow. Simultaneously, the PCM is melted below cylinders by the conduction
heat transfer.
At θ = 2.99 × 10−2 , the progression of upper phase change front is hindered by the upper adiabatic wall.
The liquid film, which is formed around the lower cylinder, still continues to be enlarged. It is obvious that the ice above
each cylinder is almost melted by the natural convection flow up to this moment. At time θ =5.99×10−2 , two liquid phases
are joined together. At time θ =7.99×10−2 most of ice is melted. And, further energy input from cylinders just overheats
above liquid phase. This overheating is inevitable because there is solid PCM which is not melted at the bottom. As can be
understood at θ =0.13, the melting front at the bottom is parallel to the bottom wall. It proceeds sluggishly due to increasing
thermal resistance of the liquid film. It must be emphasized that two vortexes persist.
Bigger vortex is in the middle and smaller one is at the top of the configuration.
332 M. Jourabian et al. / Physica A 509 (2018) 316–335

Table 2
Thermo-physical properties of PCM and porous matrix.
%Ni-steel PCM
ρ [kg m−3 ] 8100 997.1
µ [Pa s] – 8.9 × 10−4
cp [J kg−1 K−1 ] 460 4179
k [W m−1 K−1 ] 9–21 0.6
β [K−1 ] (10–16) × 10−6 2.1 × 10−4

Table 3
Decrease in the full melting time of porous samples compared to the pure PCM.
ε = 0.8 ε = 0.7 ε = 0.6
Kr = 15 −68% −91% −94%
Kr = 25 −92% −95% −97%
Kr = 35 −95% −97% −98%

Table 4
Increment in the average melt front in porous samples compared to the pure PCM.
ε = 0.8 ε = 0.7 ε = 0.6
Kr = 15 +13% +47% +82%
Kr = 25 +51% +92% +152%
Kr = 35 +84% +143% +282%

By comparing dimensionless times which are shown, it is recognizable that full melting time of pure PCM is prolonged
in this configuration due to sluggish melting of the solid PCM below the lower cylinder. It could be seen in [56].
Accordingly, from an energy-saving point of view, it is highly recommended to increase the heat transfer rate in this
region particularly. However, shifting the bottom cylinder toward the bottom wall may eradicate the problem but it is not
the purpose of this study. Keep in mind that the total melting time in this simulation is equal to θ =0.24.
The effects of the thermal conductivity ratio of the porous medium/PCM composites on temporal temperature contours
and streamlines are depicted in Fig. 12. The porosity and Darcy number are 0.8 and 10−3 , respectively.
As the thermal conductivity of the porous material augments, the melting rate enhances and the phase change interface
develops faster in comparison with the pure PCM melting. As stated before, the insertion of metallic porous materials into
the PCM generates significant intensification of the conduction heat transfer. There is persistent evolution of the concentric
pattern of the phase change front. It is obvious that there is no effect of the natural convection flow in the melt zone.
Direction of vortices is not upwardly in contrast to the pure PCM melting. They are approximately on the horizontal
middle line of configuration The more the thermal conductivity of the porous medium increases, the higher the melting
rate is and the smaller the full melting time of the porous medium/PCM composites becomes. Lastly, there is a large flow
resistance in the metallic porous structure. In spite of suppression of the natural convection flow, the thermal performance
of the porous medium/PCM composites is significantly higher than that of pure PCM.
Huber et al. [45] stated that one variable of interest frequently investigated in this engineering point of view is the
average melting front position. It describes the evolution of the system. Time-dependent variation of the average melting
front positions in porous medium/PCM composites is plotted in Fig. 13.
Different porosities and thermal conductivity ratios are considered. Full melting time of the pure PCM is not shown. It
results in a better evaluation between liquid fractions of composite samples. The slope of all graphs indicates the melting
rate.
At the beginning of the process, the melt fractions are equal approximately and their sharp steepness demonstrates the
high melting rate where the conduction is the main mode of the heat transfer between the hot surface and the solid phase.
As time elapses more, the effect of the insertion of the porous material becomes more evident so that samples with the
higher thermal conductivity ratio and lower porosity have bigger melting rates. It leads ultimately to the decrease of the full
melting time. From a theoretical point of view, this finding can be valid by Eq. (7).
In Table 3, influences of porosity and thermal conductivity ratio on full melting times of porous medium/PCM composites
are illustrated. By increasing the thermal conductivity ratio and decreasing the porosity of porous medium/PCM composites,
the full melting time diminishes.
In porous medium/PCM composites, PCM melts homogeneously and rapidly around each cylinder. There is no thermal
resistance of the liquid film at the bottom of the configuration.
Though reducing porosity causes decrease of PCM content and thermal storage capacity of the system. Samples with
porosity of 0.8 (20 percent decrease of PCM) are more effective than samples with porosity of 0.6 (40 percent decrease of
PCM). Also increasing thermal conductivity ratio Kr is more effective in samples with porosity of 0.8.
Average melting front positions in porous medium/PCM composites are compared to that in the pure PCM in Table 4 at
θ =0.001. Inserting metallic porous material into the base PCM results in noticeable enlargement of the melt zone.
M. Jourabian et al. / Physica A 509 (2018) 316–335 333

So, to get higher efficiency in this thermal storage unit, it is suggested to use porous medium/PCM composite with porosity
of 0.8 and Kr=35.

8. Conclusion

For pure PCM melting, the solid PCM at the bottom of the configuration melts slowly due to the increasing thermal
resistance of the liquid film. Full melting time of the pure PCM is prolonged.
Enhancing the thermal conductivity ratio or reducing the porosity of the porous matrix results in the increase of melting
rate. In porous samples, PCM is melted uniformly and rapidly around each cylinder and no thermal resistance of the liquid
film is allowed at the bottom.
The embedment of Ni-steel porous matrix in ice engenders the domination of the conduction heat transfer and significant
suppression of the natural convection flow in the melt zone. Reducing porosity engenders reduction of the thermal storage
capacity. To achieve higher heat transfer efficiency in this thermal storage unit, it is shown to improve the thermal
conductivity ratio in samples with a larger porosity.

References

[1] M. Heydari, D. Toghraie, O.A. Akbari, The effect of semi-attached and offset mid-truncated ribs and water/TiO2 nanofluid on flow and heat transfer
properties in a triangular microchannel, Therm. Sci. Eng. Prog. 2 (2017) 140–150.
[2] M.R. Shamsi, O.A. Akbari, A. Marzban, D. Toghraie, R. Mashayekhi, Increasing heat transfer of non-Newtonian nanofluid in rectangular microchannel
with triangular ribs, Physica E 93 (2017) 167–178.
[3] A.A.A. Arani, O.A. Akbari, M.R. Safaei, A. Marzban, A.A.A.A. Alrashed, G. Reza Ahmadi, T.K. Nguyen, Heat transfer improvement of water/single-wall
carbon nanotubes (SWCNT) nanofluid in a novel design of a truncated double layered microchannel heat sink, Int. J. Heat Mass Transfer 113 (2017)
780–795.
[4] O. Rezaei, O.A. Akbari, A. Marzban, D. Toghraie, F. Pourfattah, R. Mashayekhi, The numerical investigation of heat transfer and pressure drop of
turbulent flow in a triangular microchannel, Physica E 93 (2017) 179–189.
[5] Ramin Sarlak, Shahrouz Yousefzadeh, Omid Ali Akbari, Davood Toghraie, Sajad Sarlak, Fattah assadi, The investigation of simultaneous heat transfer
of water/Al 2 O 3 nanofluid in a close enclosure by applying homogeneous magnetic field, Int. J. Mech. Sci. 133 (2017) 674–688.
[6] Ramin Mashayekhi, Erfan Khodabandeh, Mehdi Bahiraei, Leyli Bahrami, Davood Toghraiee, Omid Ali Akbari, Application of a novel conical strip insert
to improve the efficacy of water–Ag nanofluid for utilization in thermal systems: A two-phase simulation, Energy Convers. Manage. 151 (2017)
573–586.
[7] Davood Toghraie, Mohammad Mehdi Davood Abdollah, Farzad Pourfattah, Omid Ali Akbari, Behrooz Ruhani, Numerical investigation of flow and heat
transfer characteristics in smooth, sinusoidal and zigzag-shaped microchannel with and without nanofluid, J. Therm. Anal. Calorim 131 (2) (2018)
1757–1766. http://dx.doi.org/10.1007/s10973-017-6624-6.
[8] Mohammad Parsaiemehr, Farzad Pourfattah, Omid Ali Akbari, Davood Toghraie, Ghanbarali Sheikhzadeh, Turbulent flow and heat transfer of
water/Al2O3 nanofluid inside a rectangular ribbed channel, Physica E 96 (2018) 73–84.
[9] Gh.R. Ahmadi, D. Toghraie, O.A. Akbari, Efficiency improvement of a steam power plant through solar repowering, Int. J. Exergy 22 (2) (2017) 158–182.
[10] O.A. Akbari, A. Marzban, Gh.R. Ahmadi, Evaluation of supply boiler repowering of an existing natural gas-fired steam power plant, Appl. Therm. Eng.
124 (2017) 897–910.
[11] Gh.R. Ahmadi, D. Toghraie, A. Azimian, O.A. Akbari, Evaluation of synchronous execution of full repowering and solar assisting in a 200 MW steam
power plant, a case study, Appl. Therm. Eng. 112 (2017) 111–123.
[12] Gh.R. Ahmadi, D. Toghraie, O.A. Akbari, Solar parallel feed water heating repowering of a steam power plant: A case study in Iran, Renew. Sustain.
Energy Rev. 77 (2017) 474–485.
[13] S.F. Hosseinizadeh, A.A. Rabienataj Darzi, F.L. Tan, Numerical investigations of unconstrained melting of nano-enhanced phase change material
(NEPCM) inside a spherical container, Int. J. Therm. Sci. 51 (2012) 77–83.
[14] S.F. Hosseinizadeh, A.A. Rabienataj Darzi, F.L. Tan, J.M. Khodadadi, Unconstrained melting inside a sphere, Int. J. Therm. Sci. 63 (2013) 55–64.
[15] H. Ait Adine, H. El Qarnia, Numerical analysis of the thermal behavior of a shell-and-tube heat storage unit using phase change materials, Appl. Math.
Model. 33 (2009) 2132–2144.
[16] M. Faraji, H. El Qarnia, Numerical study of melting in an enclosure with discrete protruding heat sources, Appl. Math. Model. 34 (2010) 1258–1275.
[17] H. El Qarnia, A. Draoui, E.K. Lakhal, Computation of melting with natural convection inside a rectangular enclosure heated by discrete protruding heat
sources, Appl. Math. Model. 37 (2013) 3968–3981.
[18] A.A. Rabienataj Darzi, M. Farhadi, K. Sedighi, Numerical study of melting inside concentric and eccentric horizontal annulus, Appl. Math. Model. 36
(2012) 4080–4086.
[19] A.A. Rabienataj Darzi, M. Farhadi, M. Jourabian, Melting and solidification of pcm enhanced by radial conductive fins and nanoparticles in cylindrical
annulus, Energ. Convers. Manage. 118 (2016) 253–263.
[20] A.A. Al-Abidi, S. Mat, K. Sopian, M.Y. Sulaiman, A.T. Mohammad, Experimental study of melting and solidification of pcm in a triplex tube heat
exchanger with fins, Energy Build. 68 (2014) 33–41.
[21] A. Karimipour, M.H. Esfe, M.R. Safaei, D. Toghraie Semiromi, S. Jafari, S.N. Kazi, Mixed convection of copper–water nanofluid in a shallow inclined lid
driven cavity using the lattice Boltzmann method, Physica A 402 (2014) 150–168.
[22] H. Alipour, A. Karimipour, M.R. Safaei, D. Toghraie Semiromi, O.A. Akbari, Influence of T-semi attached rib on turbulent flow and heat transfer
parameters of a silver-water nanofluid with different volume fractions in a three-dimensional trapezoidal microchannel, Physica E 88 (2017) 60–76.
[23] O.A. Akbari, M. Goodarzi, M.R. Safaei, M. Zarringhalam, G.R. Ahmadi Sheikh Shabani, M. Dahari, A modified two-phase mixture model of nanofluid
flow and heat transfer in 3-d curved microtube, Adv. Powd. Technol. 27 (2016) 2175–2185.
[24] O.A. Akbari, H. Hassanzadeh Afrouzi, A. Marzban, D. Toghraie, H. Malekzade, A. Arabpour, Investigation of volume fraction of nanoparticles effect and
aspect ratio of the twisted tape in the tube, J. Therm. Anal. Calorim 129 (3) (2017) 1911–1922. http://dx.doi.org/10.1007/s10973-017-6372-7.
[25] A. Karimipour, A. Hossein Nezhad, A. D’Orazio, M.H. Esfe, M.R. Safaei, E. Shirani, Simulation of copper–water nanofluid in a microchannel in slip flow
regime using the lattice Boltzmann method, Eur. J. Mech. B Fluids 49 (Part A) (2015) 89–99.
[26] A. Behnampour, O.A. Akbari, M.R. Safaei, M. Ghavami, A. Marzban, Gh.R. Ahmadi Sheikh Shabani, M. Zarringhalam, R. Mashayekhi, Analysis of heat
transfer and nanofluid fluid flow in microchannels with trapezoidal, rectangular and triangular shaped ribs, Physica E 91 (2017) 15–31.
[27] A. Karimipour, New correlation for Nusselt number of nanofluid with Ag / Al2 O3 / Cu nanoparticles in a microchannel considering slip velocity and
temperature jump by using lattice Boltzmann method, Int. J. Therm. Sci. 91 (2015) 146–156.
334 M. Jourabian et al. / Physica A 509 (2018) 316–335

[28] Ahmad Reza Rahmati, Omid Ali Akbari, Ali Marzban, Davood Toghraie, Reza Karimi, Farzad Pourfattah, Simultaneous investigations the effects of
non-Newtonian nanofluid flow in different volume fractions of solid nanoparticles with slip and no-slip boundary conditions, Therm. Sci. Eng. Progr.
(2017) http://dx.doi.org/10.1016/j.tsep.2017.12.006.
[29] O.A. Akbari, D. Toghraie, A. Karimipour, A. Marzban, G.R. Ahmadi, The effect of velocity and dimension of solid nanoparticles on heat transfer in
non-Newtonian nanofluid, Physica E 86 (2017) 68–75.
[30] M.H. Esfe, W.M. Yan, M. Akbari, A. Karimipour, M. Hassani, Experimental study on thermal conductivity of DWCNT-ZnO/water-EG nanofluids, Int.
Commun. Heat Mass Transfer 68 (2015) 248–251.
[31] M. Esfandiary, B. Mehmandoust, A. Karimipour, H.A. Pakravan, Natural convection of Al2 O3 –water nanofluid in an inclined enclosure with the effects
of slip velocity mechanisms: Brownian motion and thermophoresis phenomenon, Int. J. Therm. Sci. 105 (2016) 137–158.
[32] O.A. Akbari, A. Karimipour, D. Toghraie Semiromi, M.R. Safaei, H. Alipour, M. Goodarzi, M. Dahari, Investigation of Rib’s height effect on heat transfer
and flow parameters of laminar water-Al2O3 nanofluid in a two dimensional rib-microchannel, Appl. Math. Comput. 290 (2016) 135–153.
[33] O.A. Akbari, D. Toghraie, A. Karimipour, Impact of ribs on flow parameters and laminar heat transfer of water–aluminum oxide nanofluid with different
nanoparticle volume fractions in a three-dimensional rectangular microchannel, Adv. Mech. Eng. 7 (11) (2015) 1–11.
[34] M.R. Safaei, O. Mahian, F. Garoosi, K. Hooman, A. Karimipour, S.N. Kazi, S. Gharehkhani, Investigation of micro and nano-sized particle erosion in a
90o pipe bend using a two-phase discrete phase model, Sci. World J. 2014 (2014) 740578.
[35] A. Karimipour, H. Alipour, O.A. Akbari, D. Toghraie Semiromi, M. Hemat Esfe, Studying the effect of indentation on flow parameters and slow heat
transfer of water-silver nanofluid with vrying volume fraction in a rectangular two-dimensional microchannel, Ind. J. Sci. Tech. 8 (15) (2015) 51707.
[36] Farzad Pourfattah, Mahdi Motamedian, Ghanbarali Sheikhzadeh, Davood Toghraie, Omid Ali Akbari, The numerical investigation of angle of attack of
inclined rectangular rib on the turbulent heat transfer of water-Al 2 O 3 nanofluid in a tube, Int. J. Mech. Sci. 000–132 (2017) 1–11.
[37] Q. Gravndyan, O.A. Akbari, D. Toghraie, A. Marzban, R. Mashayekhi, R. Karimi, F. Pourfattah, The effect of aspect ratios of rib on the heat transfer and
laminar water/TiO2 nanofluid flow in a two-dimensional rectangular microchannel, J. Mol. Liq. 236 (2017) 254–265.
[38] M.H. Esfe, A.A.A. Arani, A. Karimiopour, S.S.M. Esforjani, Numerical simulation of natural convection around an obstacle placed in an enclosure filled
with different types of nanofluids, Heat Transfer Res. 45 (3) (2014) 279–292.
[39] O.A. Akbari, D. Toghraie, A. Karimipour, Numerical simulation of heat transfer and turbulent flow of water nanofluids copper oxide in rectangular
microchannel with semi attached rib, Adv. Mech. Eng. 8 (4) (2016) 1–25.
[40] Z.G. Wu, C.Y. Zhao, Experimental investigations of porous materials in high temperature thermal energy storage systems, Sol. Energy 85 (2011)
1371–1380.
[41] W.Q. Li, Z.G. Qu, Y.L. He, W.Q. Tao, Experimental and numerical studies on melting phase change heat transfer in open-cell metallic foams filled with
paraffin, Appl. Therm. Eng. 37 (2012) 1–9.
[42] Z. Chen, D. Gao, J. Shi, Experimental and numerical study on melting of phase change materials in metal foams at pore scale, Int. J. Heat Mass Transfer
72 (2014) 646–655.
[43] J.P. Trelles, J.J. Dufly, Numerical simulation of porous latent heat thermal energy storage for thermoelectric cooling, Appl. Therm. Eng. 23 (2003)
1647–1664.
[44] O. Mesalhy, K. Lafdi, A. Elgafy, K. Bowman, Numerical study for enhancing the thermal conductivity of phase change material (PCM) storage using
high thermal conductivity porous matrix, Energ. Convers. Manage. 46 (2005) 847–867.
[45] C. Huber, A. Parmigiani, B. Chopard, M. Manga, O. Bachmann, Lattice Boltzmann model for melting with natural convection, Int. J. Heat Fluid Flow 29
(2008) 1469–1480.
[46] D.Y. Gao, Z.Q. Chen, M.H. Shi, Z.S. Wu, Study on the melting process of phase change materials in metal foams using lattice Boltzmann method, Sci.
China Technol. Sc. 53 (2010) 3079–3087.
[47] Y. Tian, C.Y. Zhao, A numerical investigation of heat transfer in phase change materials (PCMs) embedded in porous metals, Energy 36 (2011) 5539–
5546.
[48] D. Gao, Z. Chen, Lattice boltzmann simulation of natural convection dominated melting in a rectangular cavity filled with porous media, Int. J. Therm.
Sci. 50 (2011) 493–501.
[49] M. Jourabian, M. Farhadi, A.A. Rabienataj Darzi, Lattice Boltzmann investigation for enhancing the thermal conductivity of ice using Al2 O3 porous
matrix, Int. J. Comput. Fluid D. 26 (2012) 451–462.
[50] Z. Liu, Y. Yao, H. Wu, Numerical modeling for solid–liquid phase change phenomena in porous media: Shell-and-tube type latent heat thermal energy
storage, Appl. Energy 112 (2013) 1222–1232.
[51] Q. Liu, Y.L. He, Double multiple-relaxation-time lattice Boltzmann model for solid–liquid phase change with natural convection in porous media,
Physica A 438 (2015) 94–106.
[52] Y.B. Tao, Y. You, Y.L. He, Lattice Boltzmann simulation on phase change heat transfer in metal foams/paraffin composite phase change material, Appl.
Therm. Eng. 93 (2016) 476–485.
[53] A. Karimipour, A. Hossein Nezhad, A. D’Orazio, E. Shirani, Investigation of the gravity effects on the mixed convection heat transfer in a microchannel
using lattice Boltzmann method, Int. J. Therm. Sci. 54 (2012) 142–152.
[54] A. Karimipour, A.H. Nezhad, A. D’Orazio, E. Shirani, The effects of inclination angle and Prandtl number on the mixed convection in the inclined lid
driven cavity using Lattice Boltzmann Method, J. Theoret. Appl. Mech. 51 (2) (2013) 447–462.
[55] M. Goodarzi, M.R. Safaei, A. Karimipour, K. Hooman, M. Dahari, S.N. Kazi, E. Sadeghinezhad, Comparison of the finite volume and lattice Boltzmann
methods for solving natural convection heat transfer problems inside cavities and enclosures, Abstr. Appl. Anal. 1–15 (2014) 762184.
[56] K. Sasaguchi, K. Kusano, R. Viskanta, A numerical analysis of solid–liquid phase change heat transfer around a single and two horizontal, vertically
spaced cylinders in a rectangular cavity, Int. J. Heat Mass Transfer 40 (1997) 1343–1354.
[57] M. Sugawara, H. Beer, Numerical analysis for freezing/melting around vertically arranged four cylinders, Heat Mass Transfer 45 (2009) 1223–1231.
[58] Z.L. Guo, T.S. Zhao, Lattice Boltzmann model for incompressible flows through porous media, Phys. Rev. E 66 (2002) 036304.
[59] M. Jourabian, M. Farhadi, A.A. Rabienataj Darzi, Constrained ice melting around one cylinder in horizontal cavity accelerated using three heat transfer
enhancement techniques, Int. J. Therm. Sci. 125 (2018) 231–247.
[60] M. Jourabian, M. Farhadi, A.A. Rabienataj Darzi, Accelerated melting of PCM in a multitube annulus type thermal storage unit using lattice Boltzmann
simulation, Heat Transfer - Asian Res. 46 (8) (2017) 1499–1525.
[61] S. Ergun, Flow through packed columns, Chem. Eng. Prog. 48 (1952) 89–94.
[62] K. Vafai, Convective flow and heat transfer in variable porosity media, J. Fluid Mech. 147 (1984) 233–259.
[63] Z.L. Guo, T.S. Zhao, A lattice Boltzmann model for convection heat transfer in porous media, Numer. Heat Transfer Part B 47 (2005) 157–177.
[64] B.C. Shi, Z.L. Guo, Lattice Boltzmann model for nonlinear convection–diffusion equations, Phys. Rev. E. 79 (2009) 016701.
[65] A.J.C. Ladd, Numerical simulations of particulate suspensions via a discretized Boltzmann equation part I: theoretical foundation, J. Fluid Mech. 271
(1994) 285–309.
[66] Z.L. Guo, C. Zheng, B.C. Shi, An extrapolation method for boundary conditions in lattice Boltzmann method, Phys. Fluid 14 (2002) 2007–2010.
[67] R. Mei, D. Yu, W. Shyy, Force evaluation in the lattice Boltzmann method involving curved geometry, Phys. Rev. E 65 (2002) 1–14.
[68] S. Succi, E. Foti, F. Higuera, Three-dimensional flows in complex geometries with the lattice Boltzmann method, Europhys. Lett. 10 (1989) 433–438.
[69] S. Chen, G.D. Doolen, Lattice Boltzmann method for fluid flows, Annu. Rev. Fluid Mech. 30 (1998) 329–364.
M. Jourabian et al. / Physica A 509 (2018) 316–335 335

[70] Mohammad Hussein Bahmani, Ghanbarali Sheikhzadeh, Majid Zarringhalam, Omid Ali Akbari, Abdullah A.A.A. Alrashed, Gholamreza Ahmadi Sheikh
Shabani, Marjan Goodarzi, Investigation of turbulent heat transfer and nanofluid flow in a double pipe heat exchanger, Adv. Powder Technol. 29 (2)
(2018) 273–282. http://dx.doi.org/10.1016/j.apt.2017.11.013.
[71] Rahim Hosseinnezhad, Omid Ali Akbari, Hamid Hassanzadeh Afrouzi, Mohit Biglarian, Ali Koveiti, Davood Toghraie, The numerical study of heat
transfer of turbulent nanofluid flow in a tubular heat exchanger with twin twisted-tapes inserts, J. Therm. Anal. Calorim 132 (1) (2018) 741–759.
http://dx.doi.org/10.1007/s10973-017-6900-5.
[72] Ali Heydari, Omid Ali Akbari, Mohammad Reza Safaei, Masoud Derakhshani, Abdullah A.A.A. Alrashed, Ramin Mashayekhi, Gholamreza Ahmadi Sheikh
Shabani, Majid Zarringhalam, Truong Khang Nguyen, The effect of attack angle of triangular ribs on heat transfer of nanofluids in a microchannel, J.
Therm. Anal. Calorim 131 (3) (2018) 2893–2912. http://dx.doi.org/10.1007/s10973-017-6746-x.
[73] M. Jourabian, M. Farhadi, K. Sedighi, A.A.R. Darzi, Y. Vazifeshenas, Melting of NEPCM within a cylindrical tube: numerical study using the lattice
Boltzmann method, Numer. Heat Transfer Part A 61 (2012) 929–948.
[74] M. Jourabian, M. Farhadi, A.A. Rabienataj Darzi, Outward melting of ice enhanced by Cu nanoparticles inside cylindrical horizontal annulus: lattice
Boltzmann approach, Appl. Math. Model. 37 (2013) 8813–8825.
[75] M. Jourabian, M. Farhadi, Melting of nanoparticles-enhanced phase change material (NEPCM) in vertical semicircle enclosure: numerical study, J.
Mech. Sci. Technol. 29 (2015) 3819–3830.
[76] M. Afrand, S. Rostami, M. Akbari, S. Wongwises, M.H. Esfe, A. Karimipour, Effect of induced electric field on magneto-natural convection in a vertical
cylindrical annulus filled with liquid potassium, Int. J. Heat Mass Transfer 90 (2015) 418–426.
[77] M. Afrand, D. Toghraie, A. Karimipour, S. Wongwises, A numerical study of natural convection in a vertical annulus filled with gallium in the presence
of magnetic field, J. Magn. Magn. Mater. 430 (2017) 22–28.
[78] M. Afrand, Using a magnetic field to reduce natural convection in a vertical cylindrical annulus, Int. J. Therm Sci. 118 (2017) 12–23.
[79] M. Afrand, S. Farahat, A.H. Nezhad, G.A. Sheikhzadeh, F. Sarhaddi, Numerical simulation of electrically conducting fluid flow and free convective heat
transfer in an annulus on applying a magnetic field, Heat Transfer Res. 45 (2014) 749–766.
[80] M. Afrand, S. Farahat, A.H. Nezhad, G.A. Sheikhzadeh, F. Sarhaddi, 3-D numerical investigation of natural convection in a tilted cylindrical annulus
containing molten potassium and controlling it using various magnetic fields, Int. J. Appl. Electromagnetics Mech. 46 (2014) 809–821.
[81] M. Afrand, S. Farahat, A.H. Nezhad, G.A. Sheikhzadeh, F. Sarhaddi, S. Wongwises, Multi-objective optimization of natural convection in a cylindrical
annulus mold under magnetic field using particle swarm algorithm, Int. Commun. Heat Mass Transfer 60 (2015) 13–20.
[82] M. Mahmoodi, M.H. Esfe, M. Akbari, A. Karimipour, M. Afrand, Magneto-natural convection in square cavities with a source–sink pair on different
walls, Int. J. Appl. Electromagnetics Mech. 47 (2015) 21–32.

You might also like