International Journal of Heat and Mass Transfer 122 (2018) 1283–1297

Contents lists available at ScienceDirect

International Journal of Heat and Mass Transfer

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

Numerical simulation of double diffusive mixed convective nanofluid

flow and entropy generation in a square porous enclosure
S. Hussain a,b,⇑, K. Mehmood a, M. Sagheer a, M. Yamin a
a
Department of Mathematics, Capital University of Science & Technology (CUST), Islamabad, Pakistan
b
Institut für Angewandte Mathematik (LS III), Technische Universität, Dortmund, Germany

a r t i c l e i n f o a b s t r a c t

Article history: In the present work, the numerical analysis of a double diffusive mixed convective alumina-water nano-
Received 3 December 2017 fluid flow in a square porous lid driven cavity is investigated to determine the influence of different phys-
Received in revised form 15 February 2018 ical parameters on the heat transfer and entropy generation. The upper wall of the enclosure is moving to
Accepted 19 February 2018
the right while all the other walls are at rest. The flow is generated due to the motion of the top wall and
the buoyancy forces that are produced due to the difference in temperature. A monolithic Galerkin finite
element approach together with geometric multigrid technique has been adopted to solve governing
Keywords:
equations for various governing parameters. Analysis has been shown in the form of streamlines, iso-
Double diffusion
Nanofluid
therms and isoconcentration, tables and plots. The influence of various physical parameters on the flow,
Entropy in specific ranges such as the Richardson number (0:01 6 Ri 6 5), Darcy number (0:00001 6 Da 6 0:01),
Galerkin finite element method porosity parameter (0:2 6  6 0:8), Lewis number (0:1 6 Le 6 7), buoyancy ratio parameter
Porous cavity (
2 6 Br 6 2), heat generation/absorption parameter (
0:4 6 q 6 0:4), chemical reaction parameter
Chemical reaction (0 6 Kr 6 0:04) as well as the nanoparticles volume fraction (0 6 / 6 0:04) are investigated and findings
are very closely comparable to the previous analysis for the special cases in the literature.
Ó 2018 Elsevier Ltd. All rights reserved.

1. Introduction ble diffusion with free convection in a fluid-saturated porous cav-

ity. Mahapatra et al. [14] studied the effects of buoyancy ratio on
The combined temperature and concentration gradients phe- double-diffusive natural convection in a lid-driven cavity. Xu
nomenon in a porous medium is termed as double-diffusive convec- et al. [15] investigated the double diffusive natural convection
tion. Double diffusion study has gained an extensive attention for and oscillation characteristics in an enclosure filled with porous
many years based on its geophysical as well as the industrial appli- medium. Goyeau et al. [16] presented the free convection in porous
cations. Some applications of double diffusive convection such as media along with double diffusive phenomenon. The thermosolu-
biology, geosciences, astrophysics, chemical reaction are mentioned tal convection in an enclosure inserted with two porous layers
in [1,2], and its significance is also observed by Mansour et al. [3], was investigated by Bennacer et al. [17]. They found the rate of
Joly et al. [4], Platten [5], Patha et al. [6] and Bahloul et al. [7]. The heat and mass transfer as a weak function of Darcy number. Fur-
analysis of the double diffusive convection has been reported thermore, Mamou et al. [18] conducted the double diffusive con-
numerically in an enclosure by Hyun and Lee [8] and Lee and Hyun vection numerically in a porous enclosure, where at the vertical
[9]. Numerical results of their study were compared favorably with sides of the enclosure heat and mass fluxes are imposed. In recent
previous results obtained through experiments. The study of double past, Basak et al. [19] analyzed the impact of thermal boundary
diffusion in a vertical enclosure was reported by Mamou et al. [10]. conditions on free convection problem inside a cavity filled with
In literature, double diffusion with various aspects has been porous medium. Forchheimer utilized a model by introducing a
studied in the porous media. Shermet et al. [11] have considered non-linear inertial term [20]. This model successfully solves the
the nanofluid filled porous cavity to study the double diffusive problems with higher porosity values, as well.
mixed convection. Lin [12] performed transient convective heat The convective flows in the lid driven cavity carries significant
transfer in a porous medium. Nithiarasu et al. [13] studied the dou- use in various industrial applications including crystal growth,
solar collectors, food processing, oven drying and electronic cards
⇑ Corresponding author at: Department of Mathematics, Capital University of cooling, etc. [21]. Nithyadevi et al. [22] considered the effects of
Science & Technology (CUST), Islamabad, Pakistan. inclination angle and non-uniform heating on mixed convection
E-mail address: [email protected] (S. Hussain). of a nanofluid filled porous enclosure with active mid-horizontal

https://doi.org/10.1016/j.ijheatmasstransfer.2018.02.082
0017-9310/Ó 2018 Elsevier Ltd. All rights reserved.
1284 S. Hussain et al. / International Journal of Heat and Mass Transfer 122 (2018) 1283–1297

Nomenclature

Be Bejan number SDC dimensionless entropy generation due to solutal con-

Br buoyancy ratio parameter centration
c dimensional solute concentration (kg m
3 ) ST;avg dimensionless total average entropy generation
C dimensionless solute concentration T temperature (K)
D mass diffusivity (m2 s
1 ) U; V dimensionless velocity components
Da Darcy number, K/L2 x; y dimensional space coordinates (m)

1
Cp specific heat (J kg K
1 ) U0 lid velocity
k thermal conductivity (W m
1 K
1 ) X; Y dimensionless space coordinates
Gr Grashof number, bgDTL3 /m2f
ds diameter of nanoparticle (nm) Greek symbols
g gravitational acceleration (m s
2 ) a thermal diffusivity (m2 s
1 )
Kr dimensionless chemical reaction parameter  porosity of the porous medium
L length of cavity (m) b thermal expansion coefficient (K
1 )
K permeability of porous media (m2 ) / volume fraction of the nanoparticles
KE kinetic energy (J) h dimensionless temperature
km effective thermal conductivity of porous medium (W m kinematic viscosity (m2 s
1 )
m
1 K
1 ) q density (kg m
3 )
Nuavg average Nusselt number l dynamic viscosity (kg m
1 s
1 )
Nu local Nusselt number
p pressure (N m
2 ) Subscripts
P dimensionless pressure avg average
Pr Prandtl number, mf =af
c cold
q dimensionless heat generation/absorption parameter f fluid
Re Reynolds number, U 0 L/mf h hot
Ri Richardson number, Gr/Re2 nf nanofluid
SHT dimensionless entropy generation due to heat transfer
s nanoparticles
SFF dimensionless entropy generation due to fluid friction

moving. Rashad et al. [23] examined mixed convection of localized It was noticed that the heat transfer reduces due to the existence of
heat source/sink in a nanofluid-filled lid-driven square cavity with different obstacles. Influence of MHD on heat transfer and entropy
partial slip. Biswal et al. [24] discussed the analysis of heatline generation in an enclosure saturated with nanofluid was discussed
based visualization for thermal management during mixed convec- by Mehrez et al. [45]. It was observed that the average Nusselt
tion of hot/cold fluids within entrapped triangular cavities. Muth- number and entropy generation enhance due to an increase in
tamilselvan et al. [25] studied the mixed convection numerically to nanoparticles volume fraction.
analyze the impact of magnetic field on the flow in a lid driven cav- In this work, we shall perform the numerical simulation to
ity. Öztop et al. [26] examined the heat transfer numerically by investigate the double-diffusive mixed convection and entropy
conjugate mixed convection in a lid driven cavity considered with generation in alumina-water nanofluid filled lid-driven square por-
thick bottom wall. Sharif [27] considered the mixed convection ous cavity considering the effects of internal heat generation/
inside a shallow tilted cavity with hot and cooled moving lids on absorption and chemical reaction. According to a careful literature
its top and bottom respectively. Effect of nanofluid on the mixed survey, such type of study with this configuration and effects has
convection flow inside a lid driven cavity partially heated from bot- not been considered and investigated yet.
tom is observed by Mansour et al. [28]. The study of mixed convec-
tion in a cubic double lid driven enclosure is investigated by 2. Problem formulation
Ouertatani et al. [29]. Enhancement of the heat transfer by the
mixed convection in a lid driven wavy surface cavity using Taguchi 2.1. The problem configuration
approach was considered by Mamourian et al. [30].
Entropy generation suppresses the thermodynamic efficiency of The problem consists of a porous cavity saturated with nano-
a system. It indicates the location of a system in which more fluid (see Fig. 1). The top wall of the cavity is moving to the right
energy dissipation occurs. Bejan [31] has investigated the funda- with velocity U 0 while all the remaining walls are at rest. The left
mental principles to mitigate the entropy generation. Since vertical wall is maintained at hot temperature T h and higher mass
entropy is one reason out of many for the wastage of energy in heat concentration ch whereas the right vertical wall is kept at cold tem-
transfer process, therefore sometimes it becomes necessary to perature T c and low mass concentration cc . Furthermore, the top
measure entropy generation in a very accurate way. More study and the bottom walls are assumed to be adiabatic. The fluid motion
on entropy generation can be consulted from [32–42]. Mahmoudi is generated in the cavity because of the movement of the top wall
et al. [43] analyzed the nanofluid flow in a cavity along with heat and buoyancy forces. The porous medium is assumed to be isotro-
generation/absorption and entropy effects and it was observed that pic and homogenous. It is assumed that slipping effect between
heat generation/absorption parameter has no effect on entropy for any two phases is negligible. Viscous dissipation in the energy

10 6 q 6 5. It was also shown that addition of nanoparticles has equation is neglected [46]. Moreover, the fluid is considered as
reduced the entropy generation. Moreover, nanoparticles have an incompressible and Newtonian. The constant thermo-physical
significant effect for large values of Hartmann number. Selime- properties of base fluid and solid particles may be consulted from
fendigil and Öztop [44] have investigated the impact of heat [47]. However, the density in the buoyancy forces is calculated by
generation and magnetic field in an enclosure filled with nanofluid. the Boussinesq model.
 S. Hussain et al. / International Journal of Heat and Mass Transfer 122 (2018) 1283–1297 1285

The nondimensional governing equations are given as follows:

@U @V
þ ¼ 0; ð6Þ
@X @Y

 !
1 @U @U @P 1 lnf @2U @2U lnf
U þ V ¼
þ
þ
2 @X @Y @X Re qnf mf @X 2
@Y 2 qnf mf

qffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
1 1:75
 U
pffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi 3 U 2 þ V 2 U; ð7Þ
ReDa 150Da2

 !
1 @V @V @P 1 lnf @2V @2V
U þ V ¼
þ þ
2 @X @Y @Y Re qnf mf @X 2 @Y 2
!
qf qs bs
þ 1
/þ / Riðh þ BrC Þ
qnf qf bf

qffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
lnf 1 1:75

V
pffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi 3 U 2 þ V 2 V;
qnf mf ReDa 150Da2
ð8Þ
!
@h @h 1 anf km @2h @2h 1
U þV ¼ þ þ qh; ð9Þ
@X @Y RePr af keff @X 2 @Y 2 RePr
Fig. 1. Schematic diagram of the physical model.
!
2.2. The governing equations @C @C  @2C @2C
U þV ¼ þ
KrC: ð10Þ
@X @Y RePrLe @X 2 @Y 2
The equations which govern the physical problem are given by
The particular boundary conditions of the problem are given by
@u @ v
þ ¼ 0; ð1Þ
@x @y
 Top wall:

 ! U ¼ 1; V ¼ 0; @h
¼ 0; @C
¼0
qnf @u @p lnf lnf @Y @Y
@u @2u @2u
u þ v ¼
þ þ
u  Left wall:
2 @x @y @x  @x2 @y2 K U ¼ 0; V ¼ 0; h ¼ 1; C¼1
1:75qnf pffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi  Bottom wall:

pffiffiffiffiffiffiffiffiffiffiffiffi 3 u2 þ v 2 u; ð2Þ U ¼ 0; V ¼ 0; @h
¼ 0; @C
¼0
150K 2 @Y @Y
 Right wall:

 !
U ¼ 0; V ¼ 0; h ¼ 0; C¼0
qnf @v @v @p lnf @2v @2v
u þv ¼
þ þ þ ðqbT Þnf gðT
T c Þ
2 @x @y @y  @x2 @y2
2.4. The nanofluid properties
lnf
þ ðqbc Þnf gðc
cc Þ
v
K Following are the effective nanofluid properties [48,49]:
1:75qnf pffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi

pffiffiffiffiffiffiffiffiffiffiffiffi 3 u þ v2 v;
2 ð3Þ
150K 2 qnf ¼ ð1
/Þqf þ /qs ; ð11Þ
! ðqC p Þnf ¼ ð1
/ÞðqC p Þf þ /ðqC p Þs ; ð12Þ
@T @T @2T @2T Q0
u þv ¼ anf þ þ ðT
T c Þ; ð4Þ anf ¼
knf
; ð13Þ
@x @y @x2 @y2 ðqC p Þm ðqC p Þnf
! ðqbÞnf ¼ ð1
/ÞðqbÞf þ /ðqbÞs : ð14Þ
@c @c @2c @2c
u þv ¼ De þ
k1 ðc
ch Þ: ð5Þ Thermophysical properties of the base fluid and the Alumina
@x @y @x2 @y2
nanoparticles can be found in [47,50,51]. There has been noticed
a significant impact of Brownian motion on nanofluid’s thermal
2.3. The dimensionless governing equations conductivity. In this respect, the following model was suggested
by Koo and Kleinstreuer [52]:
Following transformations are used to convert the system
(1)–(5) into nondimensional form keff ¼ kstatic þ kBrownian ; ð15Þ
"   #
x u y T
Tc 3 ks =kf
1 /
X¼ ; U¼ ; Y ¼ ; h¼ ; kstatic ¼ kf 1 þ     ; ð16Þ
L U0 L Th
Tc ks =kf þ 2
ks =kf
1 /
v p K
V ¼ ; P¼ ; Da ¼ 2 Following relations are used to simulate the effective properties
U0 qnf U 0
2
L related to porous medium [53,54,47]:
U0 L gbT ðT h
T c ÞL a mf k1 L
Re ¼ ; Ri ¼ ; Le ¼ ; Pr ¼ ; De ¼ D; Kr ¼ ðqC p Þm ¼ ð1
ÞðqC p Þp þ ðqC p Þnf ; ð17Þ
mf U 20 D af U0
km ¼ ð1
Þkp þ knf : ð18Þ
bc ðch
cc Þ Q 0 L2 c
cc
Br ¼ ; q¼ ; C¼ :
bT ðT h
T c Þ ðqC p Þm anf c h
cc Here the glass fiber is used to simulate the porous medium
[55,56].
1286 S. Hussain et al. / International Journal of Heat and Mass Transfer 122 (2018) 1283–1297

2.5. Calculation of the Nusselt and Sherwood numbers weak formulation of the governing equations, the velocity, temper-
ature, concentration and pressure components are discretized with
The local Nusselt and Sherwood numbers at the left vertical standard Q 2 and Pdisc elements, respectively (see [59] for details).
1
wall are given by [57] We consider sequences of grids which are generated by uniform


km @h refinement from a coarsest mesh with only one cell. Starting from
Nu ¼
; ð19Þ the coarsest grid defined as grid level ‘ ¼ 1, we generate the grid of
kf @X X¼0

 level ‘ þ 1 by dividing each quadrilateral cell of grid level ‘ into
@C
Sh ¼
: ð20Þ four new quadrilaterals connecting the midpoints of opposite
@X X¼0 edges. Thus, the computational domain is discretized into a num-
The average Nusselt and Sherwood numbers at the left vertical ber of quadrilateral cells Xk ; k 2 N at the given space level. On each
wall can be expressed by quadrilateral, the standard biquadratic element Q 2 is defined by
Z introducing the four additional mid-side node points, together
1
Nuavg ¼ Nu dY; ð21Þ with a ninth node at the centre. To construct the local shape func-
0 tions on an arbitrary physical element Xk , we make use of a refer-
Z 1 ence coordinate system. A local coordinate system ðn; gÞ is
Shavg ¼ Sh dY: ð22Þ ^ k ¼ ½
1; 12 be the reference element located at
0 introduced. Let X
the center of this coordinate system ðn; gÞ. A one-to-one bilinear
mapping between the physical and the reference element is
2.6. Second law analysis ^ k ! Xk . Once the basis functions have been
referred to F k : X
defined on the reference element in terms of reference coordinates,
In view of local thermodynamic equilibrium of linear transport
the inverse mapping F
1 : Xk ! X^ k can be employed to get back to
theory [58], the dimensional form of local entropy generation for a k

two dimensional heat and fluid flow may be expressed as: the physical space. In this case, there are nine shape functions, four
"

2 # associated with the vertices, four with the edge mid-points and
2
k @T @T Th þ Tc one internal (or bubble) function (see [59,60] for further details).
sHT ¼ þ ; T0 ¼ ð23Þ
T 20 @x @y 2 On the other hand, the discontinuous P 1 finite element space on
each quadrilateral Xk consists of piecewise linear polynomials
l   which are discontinuous across inter-element boundaries. Three
sFF ¼ u2 þ v 2 shape function are defined locally at the center of each element.
KT 0
"

2
2 # These shape functions correspond to the function value and both
l @v @u @ v
2
@u
þ 2 þ2 þ þ ; ð24Þ of its partial derivatives. There are two possible choices for the
T0 @x @y @y @x
Pdisc
1 finite element: the unmapped approach (where the finite ele-
"
ment space is spanned by the constant 1 and the global coordinates
2
2 #
RDe @c @c x and y) or the mapped approach (where the finite element space is
sDC ¼ þ
C0 @x @y defined locally by means of the constant 1 and the local coordi-





 nates n and g) (see [59,60] for further details).
RDe @c @T @c @T
þ þ : ð25Þ Hence, the Q 2 =Pdisc element consists of 9 local degrees of free-
C0 @x @x @y @y 1
dom (DOFs) for each velocity, temperature, concentration compo-
The dimensionless forms of the local entropy generation are nent and 3 DOFs for a piecewise linear discontinuous pressure
given by Eqs. (26)–(28) approximation in each element (see [60] for details). This leads
"

 # to a total of 39 degrees of freedoms in each element consisting of
2 2
knf @h @h velocity, temperature, concentration and pressure variables. The
SHT ¼ þ ; ð26Þ
kf @X @Y Picard iteration is employed to deal with the fully discretized non-
"

2
2 #
u   lnf @U
2
@V @U @V linear systems at the given space level. Finally, the associated lin-
SFF ¼ 1 U 2 þ V 2 þ u1 2 þ2 þ þ ; ð27Þ
Da lf @X @Y @Y @X earized subproblems are solved using monolithic geometric
"

 #




 multigrid solver with a smoother based on blocking of all cell
2 2
@C @C @C @h @C @h unknowns (see [60] for details) and the solution is updated with
SDC ¼ u2 þ þ u3 þ : ð28Þ
@X @Y @X @X @Y @Y the obtained correction in each nonlinear step. The convergence
criteria for the nonlinear iteration is set as follows:
Here, the irreversibility ratios u1 ; u2 and u3 can be expressed by

Eqs. (29)–(31) Cnþ1
Cn

6


2 6 10 : ð32Þ
lT 0U0 Cnþ1
u1 ¼ ; ð29Þ
k Th
Tc

2 Here, C represents the general solution component. Further, it was
RD T c
c
u2 ¼ e 0 h c ; ð30Þ also ensured that the nonlinear iteration stops if the L2 -norm of the
c0 k T h
T c

 nonlinear residual drops down below 10
6 .
RDe ch
cc
u3 ¼ : ð31Þ
k Th
Tc Table 1
Results comparison of Nuavg to the previous literature for Gr ¼ 100; Pr ¼ 0:71.

Re 1000 400 100

3. The numerical procedure
Present work 6.58 4.07 2.03
Sheremet and Pop [61] 6.70 4.09 2.05
3.1. The discretization scheme Saha et al. [62] 6.28 3.97 2.01
Zheng et al. [63] 6.61 4.14 –
The finite element method (FEM) based on the Galerkin version Sharif [64] 6.55 4.05 –
Iwatsu et al. [65] 6.33 3.84 1.94
is utilized to solve the governing equations. After establishing the
 S. Hussain et al. / International Journal of Heat and Mass Transfer 122 (2018) 1283–1297 1287

Table 2 3.2. Validation of the code and grid independence study

Grid independence study for water-alumina.

‘ NEL DOFs Nuavg Shavg Table 1 shows the code validation for the average Nu and the
4 64 1348 5.421914 5.077505
results are in good comparison with that of [61–65]. A comparison
5 256 5124 5.956191 5.670900 of velocity with the previous study has also been published [66].
6 1024 19,972 6.154290 5.846686 Grid independent study is displayed in Table 2 for
7 4096 48,852 6.365928 6.047921 Re ¼ 100; Ri ¼ 1; Da ¼ 0:01; Br ¼ 1; Kr ¼ 0; q ¼ 0; / ¼ 0:04; Pr ¼ 6:2
8 16,384 313,348 6.510575 6.188277
together with total number of elements (NEL) and total degrees of
9 65,536 1,249,284 6.594470 6.270507
10 262,144 4,988,932 6.601463 6.286729 freedom (DOFs). The present code was tested for grid indepen-
dence by calculating the u and v velocities in the middle of the

Fig. 2. Velocities in the middle of the cavity for different uniform grids.

Table 3
Variation of different quantities with respect to q and Ri.

Ri Quantity q ¼
0:4 q ¼
0:2 q¼0 q ¼ 0:2 q ¼ 0:4

0:01 Nuavg 5.307138 5.262815 5.217928 5.172465 5.12641
Shavg 4.935111 4.935123 4.935134 4.935146 4.935157
havg 0.414202 0.417076 0.419995 0.422961 0.425975
SHT;avg 5.419055 5.417222 5.415865 5.415005 5.414667
SFF;avg 0.320764 0.320764 0.320764 0.320764 0.320764
SDC;avg 0.09811 0.098077 0.098052 0.098037 0.09803

0:1 Nuavg 5.472346 5.428816 5.384734 5.340089 5.294866

Shavg 5.098508 5.098568 5.098629 5.098691 5.098752
havg 0.415503 0.418392 0.421326 0.424307 0.427336
SHT;avg 5.584165 5.583195 5.582721 5.582769 5.583362
SFF;avg 0.32092 0.320922 0.320923 0.320925 0.320926
SDC;avg 0.101099 0.101082 0.101073 0.101074 0.101085

1 Nuavg 6.678053 6.636501 6.59447 6.551949 6.508929

Shavg 6.270814 6.270666 6.270507 6.270338 6.270157
havg 0.429107 0.431891 0.434714 0.437577 0.44048
SHT;avg 6.78428 6.787386 6.791064 6.795337 6.800231
SFF;avg 0.326415 0.326433 0.326452 0.326471 0.32649
SDC;avg 0.122827 0.122883 0.12295 0.123027 0.123116

5 Nuavg 9.127596 9.085781 9.043571 9.00096 8.957941

Shavg 8.589708 8.589212 8.588681 8.588113 8.587509
havg 0.450341 0.452846 0.45538 0.457943 0.460537
SHT;avg 9.225198 9.231904 9.239207 9.247128 9.255687
SFF;avg 0.371364 0.371401 0.371438 0.371476 0.371515
SDC;avg 0.167019 0.16714 0.167272 0.167416 0.167571
1288 S. Hussain et al. / International Journal of Heat and Mass Transfer 122 (2018) 1283–1297

cavity and portrayed by Fig. 2. Initially, mesh containing only a sin- The variation of the Nuavg ; Shavg ; havg ; SHT;avg ; SFF;avg and SDC;avg with
gle element at level ‘ ¼ 1 is taken then it is refined gradually up to respect to q and Ri has been elaborated by Table 3. It is observed
finer levels. Since there exists very small difference in average Nus- that the Nuavg declines with an increase in q and amplifies with
selt and Sherwood number values at both the levels 9 and 10 an augmentation in Ri. Both the Sherwood number and the average
therefore the former is utilized for computational purpose. temperature enhance with a rise in q and Ri. An increase is seen in
SHT;avg with an enhancement in Ri whereas a decrease has been
4. Results and discussion noticed in it with an increase in q. The same is the case with
SDC;avg . The SFF;avg rises slightly with an amplification in q while it
A double diffusive mixed convective square porous lid driven enhances steeply with an increase in Ri.
cavity saturated with alumina-water nanofluid has been consid- The impact of  on the streamlines, isotherms and isoconcentra-
ered. It is numerically examined the effect of different physical tions contours is portrayed by Fig. 3. It is expected that since the
parameters on the entropy generation and the heat transfer. The porosity is directly associated with the permeability of the porous
standard values of different parameters are such as Re ¼ 100, medium, a significant change in the porosity leads to an altering of
/ ¼ 0:04, Ri ¼ 1, Pr ¼ 6:2, Da ¼ 0:01,  ¼ 1, Le ¼ 1, Br ¼ 1, q ¼ 0 the flow pattern and heat transfer in the cavity. For small values of
and Kr ¼ 0 unless these are mentioned, otherwise. the , a circular rotating cell in the centre of the cavity is seen. With

Fig. 3. Streamlines, isotherms and isoconcentration contours for different .

 S. Hussain et al. / International Journal of Heat and Mass Transfer 122 (2018) 1283–1297 1289

Fig. 4. Streamlines, isotherms and isoconcentration contours for different Kr.

an increase in the porosity parameter, this cell adopts a rectangular core region of the cavity. This indicates an enhancement in the
shape covering almost the whole cavity indicating the stronger convection process due to large void portion in the porous med-
movement of the nanofluid. Most of the isotherms seem to be par- ium. As long as the isoconcentrations are concerned, a small plume
allel to the vertical hot wall for small  that shows the dominance pointing rightwards in the above left portion of the cavity is
of conduction. Increasing the porosity parameter up to  ¼ 0:8 observed. With an augmentation in the , this plume becomes
results these contours to be parallel to the horizontal walls in the significantly large, moving to the right wall passing through the
1290 S. Hussain et al. / International Journal of Heat and Mass Transfer 122 (2018) 1283–1297

Fig. 5. Streamlines, isotherms and isoconcentration contours for different Le.

centre of the cavity. This shows a considerable increase in the mass processes. Some of them include evaporation from the water body
transfer for the large porous matrix. surface, transfer of energy in a cooling wet tower and the desert
The influence of Kr on the isotherms, streamlines and isocon- cooler flow. The chemical reaction parameter has no clear effect
centrations contours is shown in Fig. 4. The effect of chemical reac- on the streamlines and the isotherms. The isoconcentrations are
tion is not neglected in this work due to its applications in many significantly affected with a mild increase in the Kr. The mass
 S. Hussain et al. / International Journal of Heat and Mass Transfer 122 (2018) 1283–1297 1291

Fig. 6. Streamlines, isotherms and isoconcentration contours for different Br.

transfer is reduced considerably with an amplification in Kr that is site behavior is observed that is the sign of significant decline in
also evident from the concentrations contours. In the absence of the mass transfer.
the chemical reaction, a small number of contours are restricted The effect of Le on the streamlines, isotherms and isoconcentra-
to the left concentrated wall, i.e., most of the lines are shifted to tions contours is depicted by Fig. 5. As the Lewis number repre-
the right wall of the cavity. With an enhancement in Kr, an oppo- sents the measure of the thermal diffusivity to the mass
1292 S. Hussain et al. / International Journal of Heat and Mass Transfer 122 (2018) 1283–1297

Fig. 7. Variation of Nuavg ; Shavg ; havg ; SHT;avg ; SFF;avg ; SDC;avg ; ST;avg ; Be and kinetic energy as a function of nanoparticle volume fraction.

diffusivity of a fluid, so the larger value of Le represents a relatively 0:0519 and 0:0546 at Br ¼
2 and Br ¼ 2, respectively. In convec-
low mass diffusivity value. The streamlines are almost similar for tion dominant regime, the isotherms and isoconcentration show
all values of Le, i.e., the strength of fluid flow is slightly decreased significant distributions with an increase in the buoyancy parame-
with an increase in Le. The thickness of the solutal boundary layer ter Br. These become parallel to the horizontal walls at the centre
near the heated vertical wall becomes thinner than the thermal of the cavity.
boundary layer with an augmentation in Le. This means that the The effect of nanoparticle volume fraction on
thermal resistance is higher than the solutal resistance and there- Nuavg ; Shavg ; havg ; SHT;avg ; SFF;avg ; SDC;avg ; ST;avg ; Be and kinetic energy is
fore the mass transfer rate is higher than the heat transfer rate. illustrated by Fig. 7. For pure fluid ð/ ¼ 0Þ, the average Nusselt
The influence of Br on the streamlines, isotherms and isocon- number is low that gradually increases with an increase in /. A
centrations contours is portrayed by Fig. 6. It is noticed that oppos- maximum heat transfer occurs for the free convection flow regime
ing buoyancy force (Br ¼
2) causes to decline the flow of fluid. As ðRi ¼ 5Þ. The average Sherwood number is maximum for the case
Br increases, the obtained buoyancy force increases which results of pure fluid and free convective flow regime that decreases
in the form of the movement of fluid in the whole cavity. This slightly with an enhancement in /. The average temperature
can also be observed from the stream function value that are decreases with an increase in /. The entropy generation due to
 S. Hussain et al. / International Journal of Heat and Mass Transfer 122 (2018) 1283–1297 1293

Fig. 8. Variation of Nuavg ; Shavg ; havg ; SHT;avg ; SFF;avg ; SDC;avg ; ST;avg ; Be and kinetic energy as a function of lewis number.

the heat transfer and fluid friction increases with an augmentation to all of the considered sources declines with an augmentation
in / whereas SDC;avg decreases with an enhancement in /. It is wor- in Le that is more amplified for the case of free convection
thy noticed that SHT;avg and ST;avg are almost the same which shows ðRi ¼ 5Þ. The same is the case for the Bejan number and kinetic
that a large part of the entropy generation is constituted due to the energy.
heat transfer. The Bejan number and the kinetic energy diminish The impact of the Darcy number on Nuavg ; Shavg ; havg ; SHT;avg ;
with a rise in /. SFF;avg ; SDC;avg ; ST;avg ; Be and kinetic energy is displayed by Fig. 9. For
The influence of the Lewis number on Nuavg ; Shavg ; havg ; SHT;avg ; the smallest Darcy number ðDa ¼ 0:0001Þ, the average Nusselt
SFF;avg ; SDC;avg ; ST;avg ; Be and kinetic energy is depicted in Fig. 8. It number and the average Sherwood number are minimum. An
is seen that the average Nusselt number is maximum for the increase in Da causes both the numbers to increase significantly.
lowest Lewis number ðLe ¼ 0:1Þ which gradually decreases with For the forced convection regime, i.e., Ri ¼ 0:01; 0:1, a small aug-
a rise in Le. An opposite behavior is shown by the average Sher- mentation in the heat and mass transfer takes place whereas for
wood number. The average temperature is maximum for Le ¼ 0:1 the mixed and the free convection regimes, a considerable increase
that gradually reduces and becomes minimum for the largest in the heat and the mass transfer occurs. For the smallest value of
Lewis number, i.e., Le ¼ 7. The average entropy generation due the Darcy number, the average temperature in the cavity is almost
1294 S. Hussain et al. / International Journal of Heat and Mass Transfer 122 (2018) 1283–1297

Fig. 9. Variation of Nuavg ; Shavg ; havg ; SHT;avg ; SFF;avg ; SDC;avg ; ST;avg ; Be and kinetic energy as a function of Darcy number.

same for all the considered three convective flow regimes. An in the porosity parameter whereas an opposite effect of  on
increase in Da makes the average temperature to enhance and it SFF;avg and kinetic energy is observed.
is maximum for the largest Da ðDa ¼ 0:01Þ and is more pronounced The influence of the buoyancy ratio parameter Br on the
for the case of free convection. The same is the case for Nuavg ; Shavg ; havg ; SHT;avg ; SFF;avg ; SDC;avg ; ST;avg ; Be and kinetic energy is
SHT;avg ; SDC;avg and the ST;avg . An opposite effect of Da on SFF;avg is portrayed in Fig. 11. It is observed that Br ¼
1 is the critical value.
observed, i.e., SFF;avg is maximum for the smaller Da and minimum Beyond this value the behavior of a quantity suddenly changes. For
for the larger Da. The Bejan number and the kinetic energy enhance the forced convective regimes, all the above mentioned physical
in a steep way for the increasing Da. quantities increase mildly with an increase in Br whereas these
The effect of the porosity parameter  on Nuavg ; Shavg ; enhance quickly after the critical value of Br for the cases of mixed
havg ; SHT;avg ; SFF;avg ; SDC;avg ; ST;avg ; Be and kinetic energy is shown in convection and free convection.
Fig. 10. A small difference in the values of the average Nusselt
number with respect to Ri is noticed for the smallest , i.e., 5. Conclusions
 ¼ 0:2. Increasing the  causes to amplify the heat transfer rate
that is more pronounced for the case of free convection. An oppo- In this work, the influence of the internal heat generation/
site behavior for Shavg is observed for . The average temperature is absorption and chemical reaction on the double diffusive mixed
minimum for  ¼ 0:4. For the small value of ; SHT;avg ; SDC;avg ; ST;avg convective nanofluid saturated square porous cavity are numeri-
and Be have minimum values. These increase steeply with a rise cally investigated. The governing equations resulting from physical
 S. Hussain et al. / International Journal of Heat and Mass Transfer 122 (2018) 1283–1297 1295

Fig. 10. Variation of Nuavg ; Shavg ; havg ; SHT;avg ; SFF;avg ; SDC;avg ; ST;avg ; Be and kinetic energy as a function of .

model are discretized by the Galerkin weighted residual finite ele-  The average entropy generation due to fluid friction SFF;avg rises
ment procedure. Some remarkable points of this work may be con- with an amplification in q; Ri and / while it reduces with an
cluded as follows: increase in the Kr; Da;  and Le.
 An increase is observed in SDC;avg with an enhancement in Ri; Da
 The average Nusselt number increases with an amplification in and  whereas a decrease has been noticed in it with an increase
Ri; /; Da and  whereas it decreases with a rise in q; Kr and Le. in q; Kr; / and Le.
 The average Sherwood number augments with an increase in  An increase in the kinetic energy is observed with an augmen-
Ri; Da; q; Kr and Le while it declines with an increase in / and . tation in Ri; Kr and Da whereas a decline in it is seen with a
 The kinetic energy, Nuavg ; Shavg ; havg ; SHT;avg ; SFF;avg ; SDC;avg and Be growth in /;  and Le.
decrease up to a critical value of the buoyancy ratio parameter,  The Bejan number increases with a rise in Ri; Da and  while it
i.e., Br ¼
1 and beyond that value, all of these quantities reduces with an increase in Kr; / and Le.
enhance.
 The average temperature increases with a growth in Ri; q and Kr
whereas it decreases with / and Le.
 An increase is seen in SHT;avg with an enhancement in Ri; /; Da Conflict of interest
and  whereas a decrease has been noticed in it with an increase
in q; Kr and Le. The authors declared that there is no conflict of interest.
1296 S. Hussain et al. / International Journal of Heat and Mass Transfer 122 (2018) 1283–1297

Fig. 11. Variation of on Nuavg ; Shavg ; havg ; SHT;avg ; SFF;avg ; SDC;avg ; ST;avg ; Be and kinetic energy as a function of Br.

Acknowledgments References

Calculations have been carried out on the LiDOng cluster at TU [1] C. Beghein, F. Haghighat, Numerical study of double-diffusive natural
convection in a square cavity, Int. J. Heat Mass Transfer 35 (1992) 833–846.
Dortmund. The support by the LiDOng team at the ITMC at TU [2] P. Cheng, Heat transfer in geothermal system, Adv. Heat Transfer 14 (1979) 1–
Dortmund is gratefully acknowledged. We would like to thank 105.
the LiDOng cluster team for their help and support. We also used [3] A. Mansour, A. Amahmid, M. Hasnaoui, M. Bourich, Multiplicity of solutions
induced by thermosolutal convection in a square porous cavity heated from
FeatFlow (http://www.featflow.de) solver package and would like below and submitted to horizontal concentration gradient in the presence of
to acknowledge the support by the FeatFlow team. Sorret effect, Numer. Heat Transfer 49 (2006) 69–94.
[4] F. Jolly, P. Vasseur, G. Labrosse, Sorret-driven thermosolutal convection in a
vertical enclosure, Int. Commun. Heat Mass Transfer 27 (2000) 755–764.
Appendix A. Supplementary material [5] J.K. Platten, The Sorret effect: a review of recent experimental results, J. Appl.
Mech. 73 (2006) 5–15.
[6] M.K. Patha, P.V.S.N. Murthy, G.P.R. Sekhar, Sorret and Dufour effect in a non-
Supplementary data associated with this article can be found, in Darcy porous medium, J. Heat Transfer 128 (2006) 605–610.
the online version, at https://doi.org/10.1016/j.ijheatmasstransfer. [7] A. Bahloul, N. Boutana, P. Vasseur, Double diffusive and Sorret-induced convection
in a shallow horizontal porous layer, J. Fluid Mech. 491 (2003) 325–352.
2018.02.082.
 S. Hussain et al. / International Journal of Heat and Mass Transfer 122 (2018) 1283–1297 1297

[8] J.M. Hyun, J.W. Lee, Double-diffusive convection in a rectangle with [37] A. Mchirgui, N. Hidouri, M. Magherbi, A.B. Brahim, Entropy generation in
cooperating horizontal gradients of temperature and concentration double-diffusive convection in a square porous cavity using Darcy Brinkman
gradients, Int. J. Heat Mass Transfer 33 (1990) 1605–1617. formulation, Transp. Porous Media 93 (2012) 223–240.
[9] J.W. Lee, J.M. Hyun, Double-diffusive convection in a rectangle with opposing [38] R.K. Nayak, S. Bhattacharyya, I. Pop, Numerical study on mixed convection and
horizontal and concentration gradients, Int. J. Heat Mass Transfer 33 (1990) entropy generation of Cu-water nanofluid in a differentially heated skewed
1619–1632. enclosure, Int. J. Heat Mass Transfer 85 (2015) 620–634.
[10] M. Mamou, P. Vasseur, E. Bilgen, Analytical and numerical study of double [39] F. Selimefendigil, H.F. Öztop, A.J. Chamkha, MHD mixed convection and
diffusive convection in a vertical enclosure, Heat Mass Transfer 32 (1996) 115– entropy generation of nanofluid filled lid driven cavity under the influence of
125. inclined magnetic fields imposed to its upper and lower diagonal triangular
[11] M.A. Sheremet, I. Pop, A. Ishak, Double diffusive mixed convection in a porous domains, J. Magn. Magn. Mater. (2016).
open cavity filled with a nanofluid using Buongiorno’s model, Transp. Porous [40] M. Bouchmel, B. Nabil, A.M. Ammar, G. Kamel, O. Ahmed, Entropy generation
Media 109 (2015) 131–145. and heat transfer of Cu-water nanofluid mixed convection in a cavity, Int. J.
[12] D. Lin, Unsteady natural convection heat and mass transfer in a saturated Mech., Aerosp., Ind., Mechatron. Manuf. Eng. 8 (2014).
porous enclosure, Warne-und-stoffubertragung 28 (1993) 49–56. [41] A. Chamkha, M. Ismael, A. Kasaeipoor, T. Armaghani, Entropy generation and
[13] P. Nithiarasu, T. Sundararajan, K.N. Seetharamu, Double-diffusive natural natural convection of CuO-water nanofluid in C-shaped cavity under magnetic
convection in a fluid saturated porous cavity with a freely convecting wall, Int. field, Entropy (2016).
Commun. Heat Mass Transfer 24 (1997) 1121–1130. [42] M.Y.A. Jamalabadi, J.H. Park, C.Y. Lee, Optimal design of
[14] T.R. Mahapatra, D. Pal, S. Mondal, Effects of buoyancy ratio on double-diffusive magnetohydrodynamic mixed convection flow in a vertical channel with slip
natural convection in a lid-driven cavity, Int. J. Heat Mass Transfer 57 (2013) boundary conditions and thermal radiation effects by using an entropy
771–785. generation minimization method, Entropy 17 (2015) 866–881.
[15] H.T. Xu, T.T. Wang, Z.G. Qu, J. Chen, B.B. Li, Lattice Boltzmann simulation of the [43] A. Mahmoudi, I. Mejri, M.A. Abbassi, A. Omri, Analysis of the entropy
double diffusive natural convection and oscillation characteristics in an enclosure generation in a nanofluid-filled cavity in the presence of magnetic field and
filled with porous medium, Int. J. Heat Mass Transfer 81 (2017) 104–115. uniform heat generation/absorption, J. Mol. Liq. 198 (2014) 63–77.
[16] B. Goyeau, J.P. Songbe, D. Gobin, Numerical study of double-diffusive natural [44] F. Selimefendigil, H.F. Öztop, Natural convection and entropy generation of
convection in a porous cavity using the Darcy-Brinkman formulation, Int. J. nanofluid filled cavity having different shaped obstacles under the influence of
Heat Mass Transfer 39 (1996) 1363–1378. magnetic field and internal heat generation, J. Taiwan Inst. Chem. Eng. 00
[17] R. Bennacer, A. Tobbal, H. Beji, P. Vasseur, Double diffusive convection in a (2015) 1–15.
vertical enclosure filled with anisotropic porous media, Int. J. Therm. Sci. 40 [45] Z. Mehrez, A.E. Cafsi, A. Belghith, P.L. Quere, MHD effects on heat transfer and
(1996) 30–41. entropy generation of nanofluid flow in an open cavity, J. Magn. Magn. Mater.
[18] M. Mamou, P. Vasseur, E. Bilgen, Multiple solutions for double-diffusive (2014).
convection in a vertical porous enclosure, Int. J. Heat Mass Transfer 38 (1995) [46] F. Selimefendigil, H.F. Oztop, N.A. Hamdeh, Natural convection and entropy
1787–1798. generation in nanofluid filled entrapped trapezoidal cavities under the
[19] T. Basak, S. Roy, T. Paul, I. Pop, Natural convection in a square cavity filled with influence of magnetic field, Entropy (2016).
a porous medium: effects of various thermal boundary conditions, Int. J. Heat [47] K. Mehmood, S. Hussain, M. Sagheer, Numerical simulation of MHD mixed
Mass Transfer 49 (2006) 1430–1441. convection in alumina-water nanofluid filled square porous cavity using KKL
[20] K. Mehmood, S. Hussain, M. Sagheer, MHD mixed convection in alumina-water model: effects of non–linear thermal radiation and inclined magnetic field, J.
nanofluid filled square porous cavity using KKL model: effects of non-linear Mol. Liquid 238 (2017) 485–498.
thermal radiation and inclined magnetic field, J. Mol. Liq. 238 (2017) 485–498. [48] M. Sheikholeslami, M.G. Bandpy, D. Ganji, Numerical investigation of MHD
[21] K. Ghachem, L. Kolsi, C. Maatki, A. Alghamdi, H.F. Öztop, M.N. Borjini, H.B. Aissia, effects on Al2 O3 -water nanofluid flow and heat transfer in a semi-annulus
K. Al-Salem, Numerical simulation of three-dimensional double diffusive enclosure using LBM, Energy 60 (2013) 501–510.
convection in a lid-driven cavity, Int. J. Therm. Sci. 110 (2016) 241–250. [49] F. Selimefendigil, H.F. Öztop, Numerical study of MHD mixed convection in a
[22] N. Nithyadevi, A.S. Begum, H.F. Öztop, K. Al-Salem, Effects of inclination angle nanofluid filled lid driven square enclosure with a rotating cylinder, Int. J. Heat
and non-uniform heating on mixed convection of a nanofluid filled porous Mass Transfer 78 (2014) 741–754.
enclosure with active mid-horizontal moving, Int. J. Heat Mass Transfer 104 [50] S. Hussain, S. Ahmed, K. Mehmood, M. Sagheer, Effects of inclination angle on
(2017) 1217–1228. mixed convective nanofluid flow in a double lid-driven cavity with discrete
[23] A.M. Rashad, M.A. Ismael, A.J. Chamkha, M.A. Mansour, MHD mixed convection heat sources, Int. J. Heat Mass Transfer 106 (2017) 847–860.
of localized heat source/sink in a nanofluid-filled lid-driven square cavity with [51] K. Mehmood, S. Hussain, M. Sagheer, Mixed convection in alumina-water
partial slip, J. Taiwan Inst. Chem. Eng. 000 (2016) 1–14. nanofluid filled lid–driven square cavity with an isothermally heated square
[24] P. Biswal, M. Roy, S. Roy, T. Basak, Analysis of heatline based visualization blockage inside with magnetic field effect: introduction, Int. J. Heat Mass
for thermal management during mixed convection of hot/cold fluids Transfer 109 (2017) 397–409.
within entrapped triangular cavities, J. Taiwan Inst. Chem. Eng. 000 (2017) 1–20. [52] J. Koo, C. Kleinstreuer, Viscous dissipation effects in microtubes and
[25] M. Muthtamilselvan, P. Kandaswamy, J. Lee, Hydromagnetic mixed convection microchannels, Int. J. Heat Mass Transfer 47 (2004) 3159–3169.
in a lid-driven cavity filled with a fluid-saturated porous medium, Int. J. Appl. [53] K. Vafai, Handbook of Porous Media, second ed., Taylor & Francis, New York,
Math. Mech. 5 (2009) 28–44. 2005.
[26] H.F. Öztop, C. Sun, B. Yu, Conjugate-mixed convection heat transfer in a lid [54] D.A. Nield, A. Bejan, Convection in Porous Media, fourth ed., Springer, 2013.
driven enclosure with thick bottom wall, Int. Commun. Heat Mass Transfer 35 [55] M. Grott, J. Knollenberg, C. Krause, Apollo lunar heat flow experiment
(2008) 779–785. revisited: a critical reassessment of the in situ thermal conductivity
[27] M.A.R. Sharif, Laminar mixed convection in shallow inclined driven cavities determination, J. Geophys. Res. 115 (2010).
with hot moving lid on top and cooled from bottom, Appl. Therm. Eng. 27 [56] K.S. Reddy, D. Sreedhar, Thermal conductivity of natural fiber, glass fiber &
(2007) 1036–1042. CNTs reinforced epoxy composites, Int. J. Curr. Eng. Technol. 6 (2016).
[28] M.A. Mansour, R.A. Mohamed, M.M. Abd-Elaziz, S.E. Ahmed, Numerical [57] A.J. Chamkha, M.A. Ismael, Natural convection in differentially heated partially
simulation of mixed convection flows in a square lid-driven cavity partially porous layered cavities filled with a nanofluid, Numer. Heat Transfer, Part A 65
heated from below using nanofluid, Int. Commun. Heat Mass Transfer 37 (2014) 1089–1113.
(2010) 1504–1512. [58] A. Bejan, Entropy Generation Minimization, CRC Press, Boca Raton, Fl, 1982.
[29] N. Ouertatani, N. Ben Chiekh, B.B. Beyya, T. Lili, T. Campo, Mixed convection in [59] G. Matthies, L. Tobiska, The Inf-Sup condition for the mapped Q k
P disc
k
1 element
a double lid-driven cubic cavity, Int. J. Therm. Sci. 48 (2009) 1265–1272. in arbitrary space dimensions, Computing 69 (October) (2002) 119–139.
[30] M. Mamourian, K.M. Shirvan, R. Ellahi, A.B. Rahimi, Optimization of mixed [60] S. Hussain, F. Schieweck, S. Turek, Efficient Newton multigrid solution
convection heat transfer with entropy generation in a wavy surface square lid- techniques for higher order space time Galerkin discretizations of
driven cavity by means of Taguchi approach, Transp. Porous Media 102 (2016) incompressible flow, Appl. Numer. Math. 83 (2014) 51–71.
544–554. [61] M.A. Sheremet, I. Pop, Mixed convection in a lid driven square cavity filled by a
[31] A. Bejan, Second law analysis in heat transfer, Energy 5 (1980) 720–732. nanofluid: Buongiorno’s mathematical model, Appl. Math. Comput. 266 (2015)
[32] T. Basak, R. Anandalakshmi, S. Roy, I. Pop, Role of entropy generation on 792–808.
thermal management due to thermal convection in porous trapezoidal [62] L.K. Saha, M.C. Somadder, M.S. Uddin, Mixed convection heat transfer in a lid
enclosures with isothermal and non-isothermal heating of wall, Int. J. Heat driven cavity with wavy bottom surface, Am. J. Appl. Math. 1 (5) (2014) 92–101.
Mass Transfer 67 (2013) 810–828. [63] G.F. Zheng, M.Y. Ha, H.S. Yoon, Y.G. Park, A numerical study on mixed
[33] T.T. Naas, Y. Lasbet, C. Kezrane, Entropy generation analyze due to the steady convection in a lid-driven cavity with a circular cylinder, J. Mech. Sci. Technol.
natural convection of Newtonian fluid in a square enclosure, Int. J. Mech., 27 (1) (2013) 273–286.
Aerosp., Ind., Mechatron. Manuf. Eng. 9 (2015). [64] M.A.R. Sharif, Laminar mixed convection in shallow inclined driven cavities
[34] A.M. Bouchoucha, R. Bessaih, Natural convection and entropy generation of with hot moving lid on top and cooled from bottom, Appl. Therm. Eng. 27
nanofluids in a square cavity, Int. J. Heat Technol. 33 (2015) 1–10. (2007) 1036–1042.
[35] F. Selimefendigil, H.F. Öztop, N.A. Hamdeh, Natural convection and entropy [65] R. Iwatsu, J.M. Hyun, K. Kuwahara, Mixed convection in a driven cavity with a
generation in nanofluid filled entrapped trapezoidal cavities under the stable vertical temperature gradient, Int. J. Heat Mass Transfer 36 (1993)
influence of magnetic field, Entropy (2016). 1601–1608.
[36] G.R. Kefayati, Simulation of heat transfer and entropy generation of MHD [66] S. Hussain, K. Mehmood, M. Sagheer, MHD mixed convection and entropy
natural convection of non-Newtonian nanofluid in an enclosure, Int. J. Heat generation of water-alumina nanofluid flow in a double lid driven cavity with
Mass Transfer 92 (2016) 1066–1089. discrete heating, J. Magn. Magn. Mater. 419 (2016) 140–155.