1

COPPER OXIDE-WATER NANOFLUID FLOW WITHIN AN ANNULUS SHAPED

CAVITY: A NUMERICAL STUDY ON NATURAL CONVECTIVE HEAT TRANSFER
M. J. Uddin1*, M. A. Halim1, M. Mohiuddin1, Shalauddin1

1 Faculty of Science and Information Technology, Daffodil International University, Dhaka, Bangladesh
Abstract- The heat transfer for copper oxide-water nanofluid flow inside a concentrical annulus
between a colder square and het up elliptical cylinder using nonhomogeneous dynamic model has
been investigated numerically. The uniform temperature is applied for the elliptic cylinder and
square wall. An unvarying magnetic field is enforced within an enclosure. The momentum, energy
and concentration equations along with the continuity equations of nanofluids are strongly coupled
and nonlinear and solved using the Galerkin finite element method. The flow, thermal and
concentration fields have been displayed to recognize the heat transfer for copper oxide-water
nanofluid. The nature of the heat transfer is justified for pertinent parameters of the problem. The
results show that the flow, thermal field, and concentration field are strongly controlled by the
applied magnetic field. The heat transfer increases significantly for the increase of nanoparticle
volume fraction, thermal Rayleigh number and slightly for the magnetic field inclination angle
whereas it decreases remarkably for an increase of the nanoparticle diameter and the magnetic field
parameter. The similar patterns but opposite effects of heat transfer distribution occur for the
increment of the magnetic field and the buoyancy force parameter.
Keywords:Finite element method, nanofluid, nanoparticles, solar collector, heat transfer

NOMENCLATURE

concentration of nanofluid
reference concentration
specific heat
Brownian diffusion coefficient
diameter of nanoparticle
thermal diffusion coefficient
acceleration due to gravity
the Hartmann number
Boltzmann constant
length of the bottom wall
Lewis number
Nusselt number
average Nusselt number
dynamic diffusion parameter
dynamic thermo-diffusion parameter

________________________________________________________________________________
Learned to share G. VACQUIER, LPCM, University of Provence, 13331 Marseille Cedex 3, France.
 2

the empirical shape factor of nanoparticles

dimensional pressure (Pa)
dimensionless pressure
Prandtl number
local thermal Rayleigh number
local solutal Rayleigh number
Schmidt number
dimensional time
temperature
the temperature of the hot wall
the temperature of the cold wall
dimensional velocity components
dimensionless velocity components
dimensional coordinates
dimensionless coordinates
Greek letters
thermal diffusivity
the coefficient of volume expansion
the coefficient of mass expansion
the electric conductivity
nondimensional time
density
dynamic viscosity
the kinematic viscosity
dimensionless temperature
nanoparticles volume fraction
dimensionless concentration
thermal conductivity
stream function
the sphericity of the nanoparticles
concentration drop
temperature drop

Subscripts
average
base fluid
nanofluid
solid nanoparticle
 3

1. INTRODUCTION
The several types of design in the industrial application have been given an enormous
attention due to its importance in heat transfer covering for the past few decades. The multiple
designs of the annulus-shaped of the enclosures are testing in the multidisciplinary research for the
optimal results. There are huge potentialities as well as applications of the annulus shape cavities.
As for example, the heat passes through the vacuums inside, of a vessel surrounded by the fluid so
that the heat can be transferred smoothly from the source to the outer surface. In the nuclear power
plant, the huge generating heat passes through the tube and this can be damaged immediately when
the produced heat from the cradle pass across the susceptible surface of the pipe if there is no fluid
surrounded over it. Also in the geothermal energy extraction, the design of the pipe filled with fluid
is an annulus shaped. Nowadays, different shapes of the several applications such as batteries,
computer appliances, industrial gadgets, and solar thermal technological instrumentation are
accessible commercially for heat transfer and cooling intent. To find the maximum heat transfer and
cooling operation, the investigations are happening in great numbers and the research on this field
increases over the decades in the scientific and engineering communities. The foremost aim is to
accomplish the highest heat transfer rate in the applications.
Also, scientists have given huge standing on the convection mode of heat transfer because it
occurs in the real-life phenomena as well as in the industrial applications. Specifically, the
buoyancy-driven convection is a central focus of many issues since it occurs in many applications
such as in the vessels filled with fluids has been applied to cool the radiator, nuclear reactor tor,
electronic machineries and devices (Davis [1], Omri et al. [2], Tzeng et al. [3]). On the other hand,
the conventional fluids such as air, water, acetone, ethylene glycol, and kerosene are using in the
practical applications cannot fulfill the requirements to get the finest outcomes. The conventional
fluids have low thermal conductivity and other thermophysical properties of them are not optimal.
Nowadays, the fluids of higher thermal conductivity are the basic requirement in the applications.
So, to adjust with prerequisite the mathematicians, physicists and engineers have suggested
nanofluids which are formed by steadily mixing by means of the nanoparticles with the
conventional fluids. Nanofluids have higher thermal conductivity and optimal thermal properties
which can help to get the optimal heat transfer rate. The exhibition of optimal heat transfer rate
using nanofluids in many experiments is available in the literature, for example, Xuan and Li [4],
Wen and Ding [5], Buongiorno [6], Tiwari and Das [7]. The basic understandings along with many
advantages, rheological applications and researches of nanofluids are accessible in Uddin et al. [8].
There are many inquiries available in the literature regarding the buoyancy-driven natural
convective heat transfer using nanofluids. Using aluminum-water nanofluid, Lai and Yang [9]
conducted an investigation of the natural convective heat transfer in a square-shaped cavity. They
investigated the efficiency of this specific nanofluid for two different conditions such as the fixed
Rayleigh number and the fixed temperature difference. The outcomes of their investigation are that
the typical heat transfer rate enhances for the upsurge of the amount of nanoparticles and the
thermal Rayleigh number. However, the deviations of the Nusselt number has been shown as a
nonlinear distribution in the work of Tiwari and Das [7]. They have explored the heat transfer
proliferation in a square cavity employing nanofluids. Also, they have shown the circumstances and
regimes of where the flow of nanofluid changed from the natural convection to forced convection
due to the addition of the nanoparticles. The natural thermal cooling of a local thermal origin at the
backside of the triangular form occupied with packed nanoparticles is studied by Ghasemi and
Aminossadati [10], and experienced that the add-on of nanoparticles to unpolluted water makes
better cooling operation, particularly in a low thermal Rayleigh figure. Using copper-water
nanofluid, Soleimani et al. [11] have studied the natural convection in a semi-annulus shaped
geometry. Their result shows that there is an optimal angle for the geometry to get the maximum
heat transfer. Also, using alumina-water nanofluid, Sheikholeslami et al. [12] have analyzed the
natural convection heat transfer in a half-annulus shaped geometry and found that the heat transfer
 4

enhancement gains for enhances of the magnetic field parameter and the opposite result are found
for the increases of the buoyancy parameter. Also, Sheikholeslami and Abelman [13] studied the
heat transfer in the two horizontal coaxial cylinders using two-phase model and found that the heat
transfer is controlled by the different parameters and the width and the hight of the geometry of the
problem.
The huge effort has been given to theorizing nanofluid. The first modeling of nanofluids is
given by Xuan and Roetzel [14] to theoretically analyze the convection heat transfer. Another
nanofluid model called particle migration modeling has established excellently by Ding and Wen
[15]. The two-component mathematical model describing nanofluid with nanoparticle and fluid
continuity equations has been given by Buongiorno [6]. Xu et al. [16] have studied buoyancy-driven
heat transfer numerically using one-component in a concentrical annulus where the inner wall was
the triangular shaped. They proposed the correlations of the mean Nusselt number using the curve
matching analysis. Their result shows that the fixed radius ratio, as well as the tilted geometry angle
of the annulus, has an insignificant effect on the heat transfer. Sheremet and Pop [17] have the
studied two-component model. It is found that the Brownian motion of nanoparticles and the
particle migration is very important to describe nanofluids. One-component model is also used to
describe the behaviors of nanofluids widely. Khanafer et al. [18] have inspected mathematically the
consequence of nanofluid supposing that the nanofluid is in solo phase and conveyed that heat
transmission rate has been upgraded because of the upsurge of nanomaterial. Sunu et al. [19]
inspected experimentally the heat transfer in an annulus where the two-heat exchanger with a four-
sided corrugated wall has been considered. Throughout their experiment, the shape of the inner tube
is tested and found that the grooved wall has a significant effect on the heat transfer. Also, it is
found that the heat transfer is remarkably higher for higher Reynolds number compared to the lower
one. Siavashi [20] have used the two-phase mixture model to conduct the numerical study of the
alumina-water nanofluid flow features in an annulus. In this investigation, two novel parameters,
configuration number, and performance number are introduced. The performance number is decided
as the proportion of heat transfer to the pressure departure. According to the configuration number,
the performance number is tested for Renolds number and found a significant effect. Using dynamic
nonhomogeneous model, Uddin et al. [21-24] have studied the heat transfer rate in nanofluids filled
different shapes of enclosures. They have shown the uniform level of concentrations for 1-20 nm
diameter of nanoparticles in the mixture. The better heat transfer coefficients are calculated and
compared with experimental investigations in their studies.
From the aforementioned consequences, it becomes clear that the annulus-shaped vessel is
significant in the performance. In addition, various types of vessels (Bhagoria et al. [25] and
Sheikholeslami et al [26]) have been studied on the basis of the many advantages of using
nanofluids in the annulus.
The main purpose of the paper is to check the effect of nanofluid filled annulus on the
operational geometry where the inner wall is elliptic. Obviously, the dynamically-modeled CuO-
water nanofluid-filled wirelessly-shaped vessel, which is not considered to be an unvarying frontier
state, is still not studied yet. This numerical modeling is put through with a precise finite element
technique, and the actual performance results are illustrated in the form of streamlines, isotherms,
and isoconcentrations, where the local Nusselt number dispersions for the pertinent constants of the
problem is displayed graphically.
 5

Figure 1. Schematic of the problem

1. PROBLEM FORMULATIONS

2.1 Physical modeling

The physical model of the problem is displayed in Fig. 1, where x and y are Cartesian coordinates.
This is designed according to the problem of practical applications. The annulus between a square
an ellipse containing copper oxide-water nanofluid has been placed in the coordinates with a minor
axis of an ellipse b and semi-major axis a where the length of the square-shaped vessel is, L . The
no-slip conditions for velocity are declared on the boundary walls. Also, the inner elliptical wall of
the cavity is a hotter wall and modeled as T  Th while the outer wall of the annulus is a
comparatively colder wall, postured as T  TC , wherever Th  TC . The existence of the nanofluid
concentration is within the entire enclosure whereas there is no concentration available at the
boundaries nevertheless avoiding the singularities of the solution, it is assumed that C  Ch on all
the boundaries. However, primarily, it is presumed that nanofluid concentration is set aside at
stumpy concentration CC but for t  0 , it is assumed as Ch in the whole domain such that Ch  CC . A
uniform magnetic field, B  Bx eˆx  By eˆy is considered within the experiment where the unit vectors eˆx ,
eˆ y along with the coordinate axes and B02  Bx2  By2 . The magnetic field’s direction creates an angle
 with the positive x -axis. The variations of the thickness of the semi-circular annulus in presence
of the magnetic field will be analyzed in the present problem.

1.2 Mathematical modeling

2.2.1 Conservation equations for nanofluid

The nonhomogeneous dynamic model is applied to describe the flow mechanisms of copper water-
nanofluid inside the annulus. In the present problem, the radiation and chemical reaction are not
included for minimizing the computational complexities whereas the Brownian, thermophoretic
diffusion, and temperature and concentration differences due to natural gravitational effects are
included through the model. It is assumed that the molecules of conventional fluid and the
nanoparticles are an equilibrium in the mixture. The molecules and nanoparticles are the species of
the nanofluid because nanoparticles are very tiny, simply can be fluidized. The two-dimensional
 6

nonhomogeneous dynamic governing (continuity, momentum in x -direction, momentum in y -di-

rection energy and concentration) equations are as follows:
ux  vy  0 (1)
 nf (ut  uu x  vu y )   p x   nf (u xx  u yy )
(2)
 nf B02  v sin   cos    u sin 2  
 nf (vt  uvx  vv y )   p y   nf (vxx  v yy )  (  ) nf g (T  Tc )
(3)
(  * ) nf g (C  Cc )   nf B02 u sin   cos    v cos 2  
Tt  uTx  vTy   nf (Txx  Tyy ) + DB C 1 (CxTx  C yTy )
(4)
 DT T 1 [Tx   Ty  ]
2 2

Ct  uCx  vC y  DB Cxx  C yy   C DT T 1 Txx  Tyy 

(5)
 DT T 1 C xTx  C yTy 
The equations (1-5) respectively are the nanofluid continuity equation, momentum equation in x -
direction, momentum equation in y - direction, the nanofluid energy equation, and the nanofluid
concentration equation. Here, u and v , p , g , T , TC , C and CC respectively are the velocities,
pressure, gravity, temperature, reference temperature, concentration and the reference concentration
of nanofluid. Also, nf ,  nf  nf   nf / (  c p ) nf , (  c p ) nf ,  nf , (  ) nf and (  )nf are the dynamic
*

viscosity, density, thermal diffusivity, thermal conductivity, heat capacity, volumetric thermal
expansion and volumetric mass expansion of nanofluid. The applied magnetic field is included with
the x - momentum and y - momentum equations. The last term of equations (2) and (3) is the
expression of the magnetic field where the nanofluid velocity ( v) and the functional magnetic field
(B) generates a Lorentz force, J E × B , where J E   nf (E + v × B) , E is the electric field and  nf is the
electrical conductivity. We have a two-dimensional magnetic field, B  ( B0 cos  , B0 sin  , 0) where the
orientation angle for the direction of the magnetic field is  with the horizontal axis and a constant
magnitude is B0 where B0  Bx2  By2 . Let us consider the nanofluid velocities v  (u, v, 0) and an
electric field E = 0 where we assumed that the surface is not able to conduct electricity or sound.
Therefore the Lorentz forces for x -direction and the y -direction can be calculated respectively are
as follows:
 nf B02  v sin   cos    u sin 2   (6a)
 nf B02  u sin   cos    v cos 2  
(6b)
The special electrical conductibility equation for a nanofluid is found from the popular electric
conductivity relation of the mixture which is given by Maxwell [27] is as follows:
2 bf   p  2 ( bf   p )
 n f   bf (6c)
2 bf   p   ( bf   p )
Since there is no clear evidence on which correlation of the nanofluidic properties gives precise
results, the most difficult task in the nanofluid problem simulation is the calculation of
thermophysical properties, nevertheless, the viscosity of nanofluid is assessed by the standing
association for the mixture of two components. Even though there are lots of research about the
thermophysical properties of nanofluidic (Uddin et al. [23]), the representative thermophysical
properties equations are as:
 nf  bf (1   ) 2.5 (7)
 nf  (1   ) bf   p (8)
 c 
p nf   (  c p ) p  (1   )(  c p )bf (9)
  nf  (1   )(  )bf   (  ) p (10)
 7

  *
nf
 (1   )(  * )bf   (  * ) p (11)
(n  1) bf   p  (n  1) ( bf   p )
 nf   bf
(n  1) bf   p   ( bf   p )
(12)
 p c p 1
 2k BTC (3nf d p )
2
k BTC
DB  (13)
3 nf d p
 nf  bf  nf
DT  0.126 (14)
 bf  nf
where,  is the solid volume fraction, where, k B is the Boltzmann constant, d p is the diameter of a
nanoparticle, bf is the thermal expansion of the bulk fluid and the symbol  is the correction fac-
tor depends on the size and shape of the particles as   0.0002d p  0.1537 . Here, n  3 /  , where 
is the sphericity defined as the ratio between the surface area of the sphere and the surface area of
the real particle with equal volumes. Also, experimentally proven that n  3 n  3.7, n  4.9 , n  5.7
and n  8.6 represents sphere, brick, cylinder, platelet, and blade shape nanoparticle respectively.
Also, the appropriate initial and boundary conditions of the annulus-shaped enclosure are as fol-
lows:
For t  0 , entire domain: u  v  0 , T  TC , C  Ch (15a)
For t  0 , On the inner elliptical wall:
u  v  0 , T  Th , C  Ch (15b)
On the outer square wall: u  v  0 , T  TC , C  C h (15c)

2.2.2 Non-dimensional governing equations

The alteration of the equations of dimensional governments in a non-dimensional set of coupled

equations is a common practice in the investigation of fluid dynamics. This is also important
because transport tools can be easily described by this process. There are some benefits to make
them non-dimensional, such as flow monitoring parameters of the problem can be easily identified,
there is no dimensional barrier and the results can include any shape and magnitude of the
geometry. Therefore, by obtaining the advantages mentioned above, equations (1) - (5) can be
transformed into non-dimensional equations, for the following relations without dimension:
uL vL x y T  TC
U V
 bf ,  bf ,
X  ,Y  ,  ,
L L T
pL2  t C  CC
 ,   bf2 ,   (16)
bf  2bf L C
where,  represents the non-dimensional time, t is the dimensional time,  bf is the base fluid’s
diffusivity, L is the positional measurement of the geometry, T  Th  TC is the temperature
difference, C  Ch  CC is the concentration difference within the nanofluid, TC is applied nominal
temperature and CC is the nominal reference concentration for nanofluid. Using the equation (16)
and be familiar with (C / CC ) 1 and (T / TC ) 1 , the governing equations (1)-(5) can be written
dimensionless form respectively as:
Continuity equation:
U X  VY  0 (17)
 8

bf  nf
U   UU X  VU Y   PX  Pr(U XX  U YY 
 nf  bf  nf
 nf
(18)
 Pr Ha 2 V sin   cos    U sin 2  
 bf
bf  nf
V  UVX  VVY   PY  Pr(VXX  VYY )
 nf  bf  nf
(  )nf
 RaT Pr   RaC Pr  (19)
 bf  nf
 nf
 Pr Ha 2 U sin   cos    V cos 2  
 bf
  U  x  V  y   nf ( bf ) 1  XX  YY 
(20)

 Le 1   X  X  Y Y   Pr NTBT Sc 1  X   Y 
2 2

  U  X  V Y  Pr Sc 1   XX  YY 
(21)
+ Pr Sc 1 NTBTC  XX  YY   Pr Sc 1 NTBT   X  X  Y Y 
Here, equations (17)-(21) are respectively called dimensionless nanofluid continuity equation,
momentum equation in X -component, momentum equation in Y - component, energy equation,
and molar concentration equation. The non-dimensionless parameters appeared in the above
3 1
equations are called, RaT  L bf g T ( bf  bf ) is thermal Rayleigh number,
RaC  L3 g C (  * ) nf ( bf  bf nf ) 1 is solutal Rayleigh number, Pr   bf ( bf ) 1 , is Prandtl number,
Ha 2  L2 bf B02 ( bf bf ) 1 is the Hartmann number, Le   bf CC ((  c p )bf DB C ) 1 is the modified Lewis num-
ber, NTBT  DT T ( DBTC ) 1 is the dynamic thermo-diffusion parameter, NTBTC  DT ( DB CTC )1 TCC is the
1
dynamic diffusion parameter, Sc  bf ( bf DB ) is the Schmidt number.
Also, the initial and boundary conditions in the dimensionless form for the present problem along
with the above-stated model can be written as:
For   0 , entire domain: U  V  0,   0,   1 (22a)
For   0 ,
On the inner elliptical wall:
U  V  0,   1,   1 (22b)
On the outer square walls: U  V    0,   1 (22c)

2. COMPUTATIONAL PROCEDURES

The Galerkin Moderate Element Method weighing residual scheme was used in conjunction with
the boundary conditions (22) for the management of key equations (17) to (21). Analytical
descriptions of this arithmetical method can be found in the book of Zienkiewicz et al. [28] and the
Codina's study [29]. The step-by-step procedure of this digital method on the main equations of this
problem can be found in the study of Uddin and Rahman [30]. With this method, the domain is
restrained in the finite elements consisting of non-triangular elements. Six knots triangular elements
are used for the equations of finite element. Then, non-linear partial differential equations are
transported to the system of integral equations by enforcing the Galerkin weighted enduring
technique. The integration intricate to each of the calculations is completed by the exact integration
method. Using the boundary and initial conditions, the obtained non-linear equations must be
changed. The iteration technique of Newton-Raphson has been adapted to the term derivative of the
Newton Raphson equation to solve the set of global nonlinear algebraic equations as a matrix, and
the Euler implicit method to solve the global array. A PDE solver linked with the interface of MAT-
 9

LAB is used to solve the whole physical system along with the governing equations. The results
convergence criteria is set to     10 here  represents the variables U , V ,  ,  and m repre-
m 1 m 5

sents the iteration number.

3.1 Grid independence test

It is important to find the number of grids which indicates the results are the independent of a num-
ber of the elements for the geometry. Using the thermophysical properties of copper-oxide-water
nanofluid and fixing the following values of the parameter: RaC  104 , RaT  106 , Ha  100 ,   0 ,
d p  10 nm ,   0.05 , n  3 and   1 , the several number elements in an annulus in terms the average
Nusselt number have been tested. We have examined six unlike non-undifferentiated control grid
structures with the following number of elements: 1228, 1784, 2612, 7784, 21738 and 26668. The
average Nusselt number for the aforesaid number of grids is calculated from the heated elliptic in-
ner wall of the annulus which is shown in Figure 2. It is observed from the figure that the value of
the average Nusselt number for the element, 21738 is almost the same as that of the elements, 2668.
In this case, we have taken 21738 elements to display our results and the results are independent of
a number of grids.

Figure 2. Grids of the enclosure vs average Nusselt number

Table 1. Comparisons of the present numerical data in terms of average Nusselt number with that
of the results of Saghir et al. [31] and Ho et al. [32].

Nuave
Studies   1%   2%   3%
Ho et al. [31] 31.8633 31.6085 28.2160
Saghir et al.
32.2037 31.0905 29.0769
[32]
Present result 32.1900 31.0354 28.9580

3.2 Code Validation

Two benchmark studies such as the numerical study of Saghir et al. [31] as well as the experimental
work of Ho et al. [32] have been considered to test the present code of the problem. The compar-
isons are displayed in Table 1. We have used our present code for an absolute benchmark study of
 1
0

Saghir et al. [32]. The governing equations, the physical relations, the type of nanofluid and the ge-
ometry of this study are used for the present code to display the comparative results. The results are
calculated by the following empirical equation of the mean Nusselt number:
Nuave  0.069 [  nf , h (  nf ) 1 ]0.404 [Prnf ,h (Prnf ) 1 ]0.333 (23)
1
Here,  nf , h ( nf ) , represents the proportion of the thermal expansion of nanofluid at hot temperature
1
and that of the mean temperature. Also, Prnf , h (Prnf ) represents the ratio of the Prandtl number which
is quantified for alumina oxide -water nanofluid at the hot temperature to the Prandtl number of the
same nanofluid at the average temperature of 26 degree Celsius and the temperature difference of 12
degree Celsius. The equation (23) is used to compare the benchmark studies with the present inves-
tigation. The outcomes in terms of the mean Nusselt number for different solid particle volume frac-
tion have been compared and shown in Table 1. It is found that the values of mean Nusselt number
for several solid volume fractions from our numerical code have reflected the similar results with
benchmark numerical studies and with the experimental outcomes. These analogous consequences
are heartened us to show and deliberate more solutions to the problem.

3. RESULTS AND DISCUSSIONS

The thermophysical properties of copper-oxide nanoparticle and the base fluid like water is imple-
mented for discussing the results. In the problem design, some parameters are defined by the values
of the physical characteristics. The thermophysical belongings of water are:  bf  0.613 W m-1 K -1 ,
bf  997.1 kg m -3 , (c p )bf  4179 Jkg -1 K -1 ,  bf  5.5 105 Sm 1 ,  bf  21 105 K -1 , bf  0.001003 kg m 1 s 1 .

The properties of copper-oxide nanoparticle are:  p  76.5 W m -1 K -1 ,  p  6320 kg m-3 , (c p ) p  531.8

6 1 5
Jkg -1 K -1 ,  p  59.6 10 Sm ,  p  1.8  10 K -1 . Also, to define the values of the parameters, we have
well-thought-out some standard values such as: C  0.01 , T  10 K , TC  300 K , d p  10 nm , CC  1 ,
n  3 , and   0.05 . Using the above-mentioned values, some sample values of the physical parame-
ters arrived into the dimensionless governing equations for the copper oxide-water nanofluids are as
follows: Sc  26111 , NTBT  0.0075495 , NTBTC  0.75495 , Le  3.82  105 , DT  8.725  1012 and
11
DB  3.8525  10 . On the other hand, the Hartmann number, the thermal Rayleigh number, the solu-
tal Rayleigh number, and the magnetic field inclination angle are changed to investigate the heat
transfer of the current problem. Nanofluid is a mixture of a normal fluid and the solid particles. So
the characteristics of nanofluids are absolutely depended on the characteristics of base fluids and
types of nanoparticles. Also, the parameters entered into the problem depend on that of the types,
amount of base fluids, solids as well as the size of the solids.
The numerical results in terms of the free convective heat transfer in an annulus containing
nanofluid are discussed in this division. In the simulation, the properties of copper-oxide and water
are used to explain the thermal, flow, concentration fields and the local Nusselt number distribution.
The important features like the effects of the nanoparticle volume fraction, the nanoparticle length
and the shape of nanoparticles using copper-oxide-water nanofluid on the heat transfer application
are the main focus of the current investigations. The ratio of two different buoyancy-parameter used
in the study is secured to 100. The flow fields in terms of streamlines, the thermal fields in terms of
isotherms, and uniformity of the solid particles in terms of the isoconcentrations for several perti-
nent parameters are displayed to analyze the problem. The distinction of the local Nusselt number at
the heated internal surface of the vessel for the Hartmann number, Rayleigh number, nanoparticle
sizes, volume fraction as well as, magnetic field tilted angle and shape factors of nanoparticle have
been displayed graphically. The concise but adequate clarification tracks apiece result.
 1
1

Figure 3 displays the Streamlines (left column) and isotherms (right column) distributions for
RaT  105 (top row), RaT  106 (middle row) and RaT  107 (bottom row) for fixed values of   0.05 ,
d p  10 nm , Ha  100 , n  3 ,   00 at   1 (steady state). As can be seen from the left column of the
figure that that two rotational big vortices are significantly strong inside the enclosure throughout
the cases.

Streamlines Isotherms

RaT  105 RaT  105

RaT  106 RaT  106

RaT  3  107 RaT  3  107

Figure 3. Streamlines (left column) and isotherms (right column) distributions for RaT  105 (top
row), RaT  106 ( middle row) and RaT  3  107 (bottom row) for fixed values of   0.05 , d p  10 nm ,
Ha  100 , n  3 ,   00 at   1 .

There are two small but strong vortices also visible inside each of big vortex which indicates the
strong flow of nanofluid in the annulus. Also, it is important to note that the directions of the vor-
tices are opposite to each other. The cells of vortices are formed in a symmetric manner showing
very normal nature of convection. This intends that the flow decorations are symmetrically distrib-
uted in the objects. This happens due to the presence of the horizontal elliptical inner walls inside
the square enclosure and the heat transfers through the elliptical shape of heat source uniformly to
nanofluid in the annulus. Fluid takes heat energy from the inner elliptic wall uniformly and becomes
 1
2

lighter to prompt a convective current to the square walls. Lighter fluid goes up and the colder por-
tion of fluid follows the line traced by the colder wall due to the symmetric nature of the square
colder walls. As a consequence, symmetric cells within the two vortices are obtained. Also, as the
thermal Rayleigh number increases, the strength of the flow field upsurges. For the lower value of
thermal Rayleigh number, the densities of streamlines are not that much pronounced. Also, the val-
ues of maximum surface velocities are very low in this case showing poor convective heat transfer.
However, for higher values of thermal Rayleigh number, streamlines are vigorously visible in the
entire enclosure and maximum surface velocities are optimal.

Isoconcentrations at   0.1 Isoconcentrations at   1.0

RaT  105 RaT  105

RaT  106 RaT  106

RaT  3  107 RaT  3  107

Figure 4. Isoconcentrations at   0.1 (left column) and isoconcentrations at   1 (right column) for
RaT  105 (top row), RaT  106 ( middle row) and RaT  107 (bottom row) for fixed values of   0.05 ,
d p  10 nm , Ha  100 , n  3 ,   00 .
This indicates the advanced flow fields. This is due to the fact that the higher temperature differ-
ences occurred for higher values of thermal Rayleigh number which makes the optimal random col-
lations, migrations, and motion of the nanoparticles in the annulus. From the right column of Figure
3, we have seen that the isotherms are very strong and significantly distorted for the higher values
of thermal Rayleigh number. The heat goes to upwards and spreads almost in the entire enclosure
 1
3

excellently in this case. The thickness of the thermal boundary layer on the walls is not that much
significant due to the high-temperature gaps. This indicates strong convection mode of heat transfer
in the applications. Whereas for the lower values of the thermal Rayleigh number, the isotherms are
not considerable distorted, they are compressed to the elliptical wall, and they are almost parallel to
the heated wall. The thickness of the thermal layer on the heated boundary is significant in this case.
This indicates that the conduction mode of heat transfer controls in the vessel. This occurs because
of the lesser temperature differences change within the annulus.

Streamlines Isotherms

Ha  0 Ha  0

Ha  20 Ha  20

Ha  50 Ha  50

Figure 5. Effect of magnetic field on streamlines (left column) and isotherms (right column) for
  300 (top row),   600 (middle row) and   900 (bottom row) with d p  10 nm ,   0.05 , Ha  100 ,
n  3 at   1 .

Figure 4 illustrates the effect of the thermal Rayleigh number on the isoconcentrations of nanofluid
for different values of thermal Rayleigh number at   0.1 (left column) and at   1 (right column).
It is clear that the pattern and the state of the isoconcentrations change pointedly over the periods of
time. The strengths and levels of isoconcentrations increase for the time   1 than that of the time,
  0.1 . At the beginning of the mixing, the nanoparticle has a tendency to cluster on the hotter
boundary. This means that initially, nanoparticles take time to settle down in the base fluid. The
particle migrations and the random collisions along with the molecules of the base fluid in this case
not in the equilibrium state. Also, the terminal velocity of the particle is not negligible in this case.
 1
4

Over the course of time, the loop of the isoconcentrations become very strong and accelerated the
flow that creates the full-bodied loop of isoconcentrations within the annulus. Also, as the thermal
Rayleigh number increases, the strength and the levels of the isoconcentrations increases noticeably,
the densities of the lines of isoconcentrations in the loops upsurges, and the lines of isoconcentra-
tions spread all the way in the cavity throughout the different stages of times. These happen because
of the higher temperature differences in the cavity results in the higher thermal Rayleigh number for
which the particle diffusions and random interactions of particles are maximized. Hence, the even
dissemination of the tracing of nanoparticles occurs for the higher value of Rayleigh number which
indicates the better convection in the enclosure.

(a) (b)

(c) (d)

(e) (f)
Figure 6. The local Nusselt number versus different values of (a) Rayleigh number (b) Hartmann
number, (c) magnetic field inclination angle (d) nanoparticle volume fraction (e) diameter of the
particle and (f) shape factors of the particle for the fixed values of   0.05 , RaT  1.7 106 , d p  10 nm ,
Ha  100 , n  3 ,   00 at   1 .

Figure 5 represents the effects of the magnetic field parameter on the isotherms (right column) and
streamlines (left column) for fixing different parameters. It is observed from the figure that the mag-
netic field parameter namely the Hartmann number significantly suppresses and controls the flow
field and thermal field. The configurations of the streamlines and isotherms are remarkably influ-
enced by the Hartmann number. This is due to the hydro-magnetized cooper-oxide nanoparticles
which have crystallographic construction and demonstrating ferromagnetism behavior. The vortices
inside the enclosure oppositely rotate each other. The rotations within the annulus are comparative
lesser in presence of the magnetic field parameter. Also, the pattern of the rotations completely un-
changed in presence of the magnetic field compared with the flow field where the magnetic field is
not present. However, the densities are increased due to the increased values of the Hartmann num-
ber. This is due to the higher intensities of the constant magnitude of the magnetic field. As the
value of the magnetic field parameter are increased from Ha  0 to Ha  50 , the strengths of the
 1
5

streamlines and isotherms are decreased, the intensities of the rotations of the flow inside the enclo-
sure dwindled and the number of vortices and the thermal plumes of isotherms stepped-down. The
compact streamlines and suppressed isotherms to the heated elliptical wall within the annulus are
observed for Ha  50 . Also, the level of the isotherms near the boundaries of the colder walls is de-
creased for the increased values of the Hartmann number. These indicate that the heat transfer in the
applications can be controlled and decreased by the magnetic field parameter.
The heat transfer in terms of the local Nusselt number dissemination at the heated elliptical shape of
a heat source for the different pertinent parameters fixing others related parameters of the problem
is displayed in Figure 6. As can be seen from the figure that the heat transfer distribution signifi-
cantly molded with the changes of the parameters of the problem. Specifically, from Figure 6(a), we
have seen that the local Nusselt number is increased when the thermal Rayleigh number is in-
creased. The three different values of the thermal Rayleigh number is tested. It is found the higher
values of the thermal Rayleigh number exhibits significantly higher heat transfer distribution. Also,
the pattern of the heat transfer distribution changes with the changes of the thermal Rayleigh num-
ber. As Rayleigh number increases the pattern of the local Nusselt number distribution becomes
necklace shaped and more panoptic whereas for the intermediate value of the thermal Rayleigh
number, it is almost a triangular and for lower values of the thermal Rayleigh number, it is a zigzag-
shaped. These indicate that the highest average heat transfer rate can be found for the higher values
of the Rayleigh number.
Nevertheless, the almost reverse pattern of the local Nusselt number distribution can be compre-
hended by the Figure 6(b) where it represents the heat transfer distribution for different values of
the magnetic field parameter. The heat transfer distribution on the heated wall of the vessel is dis-
played for Ha  0  50 . As can be observed that as the Hartmann number decreases, the distribution of
the local Nusselt number diminutions significantly. Therefore, it is clear that the heat transfer in an
application can be modified by the setup of the magnetic field. Also, the pattern of the heat transfer
distribution can be controlled using the magnetic field setup.
The effect of the magnetic field inclination angle on the heat transfer distribution is an interesting
result in the natural convective heat transfer research. Figure 6(c) represents the effect of the direc-
tions of the magnetic field. It is found that the local Nusselt number dissemination augments
slightly with the augmentation of the magnetic field tilted angle. In this case, the effects are not that
much distinguishable. Explicitly, the swelling tendency of the local Nusselt number distribution is
not that much pronounced for 00    450 whereas, for   90 , the heat transfer distribution is notice-
ably higher than that of the other angles.
The nanoparticle volume fraction versus the local Nusselt number is illustrated in Figure 6(d). As
can be found that as nanoparticle volume fraction increases, the local Nusselt number increases no-
ticeably. Also, the heat transfer distribution on the heated wall for nanofluid is significantly higher
than that of the base fluid. This is due to the resultant effect of the nanoparticle and the base fluid. A
pivotal notice is that the distribution of heat transfer for base fluid where the solid particles are ab-
sent is much squeezed, almost a straight line although the heated wall is an elliptical shaped. Never-
theless, for nanofluid, the heat transfer distribution is very wide on the heated elliptical wall and it
looks a cone-shaped. This may be due to the movement of the nanoparticles in the fluid which en-
hances the normal heat transfer distribution. This represents that the stable and average heat transfer
can be found utilizing nanofluid compared to that of the base fluid.
Figure 6(e) depicts the heat transfer distribution on the heated wall for different size of the solid
particles in the mixture. The significant effect is noticed in this regard. As can be seen from the fig-
ure that the heat transfer distribution is seriously declined with the increase in the size of the
nanoparticle diameter. This is due to the diffusion nature of the nanoparticles. The Brownian diffu-
sion and the thermophoretic diffusion are reciprocal of the size of the solid particles in the solution.
It is important to the mansion that several sizes of the nanoparticles are also tested and the conspic-
 1
6

uous decreasing trend in the heat transfer distribution is perceived for d p  1  60 nm. Also, the prom-
inent modification (declining tendency) of the heat transfer for different size of the nanoparticle is
enhanced if the nanoparticle volume fraction is increased. Hence, in the application of nanofluid, it
is suggestive to mix up as much as smaller solid particles with the base fluid.
The effects of the shape factors of solid particles of base fluid over the heat transfer distribution are
depicted in Figure 6(f). Using the shape factor, the original shape of the solid nanoparticle in the ap-
plication can be determined. Here, the shape factor 3, 4.9 and 8.6 represents the spherical,
cylindrical and blade-shaped of nanoparticles. It is observed that the heat transfer distribution is
increased as the shape factor of the nanoparticle is increased. In this case, the spherical shape of
nanoparticle shows lowest heat transfer distribution whereas the blade shape of solid particle dis-
plays the maximum heat transfer distribution. So it is essential to note that whenever nanofluids are
produced for the heat transfer application, it is better to stably mix up the solid nanoparticles of hav-
ing highest shape factor with the conventional fluid.

4. CONCLUSIONS

The numerical simulation on the natural convective heat transfer flow of nanofluid filled an annu-
lus shaped geometry between a square and ellipse-shaped boundaries. The temperature of the ellip-
tical inner wall is assumed to be uniformly distributed whereas the outer square wall is assumed to
be considerably colder. The numerical results have been demonstrated that the flow field, thermal
field and concentration level of nanofluid in the geometry are extremely stricken by the parameters
of the problem, the time as well as the shape of the heat source. The magnetic field can control the
heat transfer rate and the pattern of the flow. The effect of the direction of the magnetic filled is not
much pronounced in this study. The stability of nanoparticles in the solution occurred after certain
periods of time. Also, the nanoparticles and the base fluid have taken time to reach the stable and
equilibrium state. Also, a significant level of buoyancy force is needed to make a thermal equilib-
rium solution. In the initial periods of time, the isotherm was unstable and after a certain time, it
became stable. The heat transfer distributions are calculated for the different pertinent parameters of
the problem. The heat transfer distribution is increased when the thermal Rayleigh number, the
nanoparticle volume fraction, the shaped factors of the nanoparticle are increased. On the other
hand, the heat transfer distribution is decreased when the applied magnetic field (Hartmann number)
and the size of the nanoparticle are increased. The effect of the direction of the magnetic field on
the heat transfer distributions is quite positive.

CONFLICT OF INTEREST
None

REFERENCES

[1] de Vahl Davis G. (1983). Natural convection of air in a square cavity: a benchmark numerical
solution. International Journal for numerical methods in fluids. 3(3): 249-264.
[2] Omri A, Orfi J, Nasrallah SB. (2005). Natural convection effects in solar stills. Desalination.
183(1-3): 173-178.
[3] Tzeng SC, Liou JH, Jou RY. (2005). Numerical simulation-aided parametric analysis of natural
convection in a roof of triangular enclosures. Heat Transfer Engineering. 26(8):69-79.
[4] Xuan Y, Li Q. (2003). Investigation of convective heat transfer and flow features of nanofluids.
Journal of Heat transfer. 125(1):151-155.
 1
7

[5] Wen D, Ding Y. (2004). Experimental investigation into convective heat transfer of nanofluids
at the entrance region under laminar flow conditions. International journal of heat and mass
transfer. 47(24):5181-5188.
[6] Buongiorno J. (2006). Convective transport in nanofluids. Journal of heat transfer. 128(3):240-
250.
[7] Tiwari RK, Das MK. (2007). Heat transfer augmentation in a two-sided lid-driven differentially
heated square cavity utilizing nanofluids. International Journal of Heat and Mass Transfer. 50
(9-10):2002-2018.
[8] Uddin MJ, Al Kalbani KS, Rahman MM, Alam MS, Al-Salti N, Eltayeb I. (2016). Fundamen-
tals of nanofluids: evolution, applications and new theory, International Journal of Biomathe-
matics and Systems Biology. International Journal of Biomathematics and Systems Biology.
2(1):1-32.
[9] Lai FH, Yang YT. (2011). Lattice Boltzmann simulation of natural convection heat transfer of
Al2O3/water nanofluids in a square enclosure. International Journal of Thermal Sciences.
50(10):1930-1941.
[10] Ghasemi B, Aminossadati SM. (2010). Brownian motion of nanoparticles in a triangular en-
closure with natural convection. International Journal of Thermal Sciences. 49(6): 931-940.
[11] Soleimani S, Sheikholeslami M, Ganji DD, Gorji-Bandpay M. (2012). Natural convection
heat transfer in a nanofluid filled semi-annulus enclosure. International Communications in Heat
and Mass Transfer. 39(4): 565-574.
[12] Sheikholeslami M, Gorji-Bandpy M, Ganji DD. (2013). Numerical investigation of MHD
effects on Al2O3-water nanofluid flow and heat transfer in a semi-annulus enclosure using LBM.
Energy. 60: 501-510.
[13] Sheikholeslami M, Abelman S. (2015). Two-phase simulation of nanofluid flow and heat
transfer in an annulus in the presence of an axial magnetic field. IEEE Transactions on
Nanotechnology. 14(3):561-569.
[14] Xuan Y, Roetzel W. (2000). Conceptions for heat transfer correlation of nanofluids. Interna-
tional Journal of Heat and Mass Transfer. 43(19):3701-3707.
[15] Ding Y, Wen D. (2005). Particle migration in a flow of nanoparticle suspensions. Powder
Technology. 149(2-3):84-92.
[16] Xu X, Sun G, Yu Z, Hu Y, Fan L, Cen K. (2009). Numerical investigation of laminar
natural convective heat transfer from a horizontal triangular cylinder to its concentric cylindrical
enclosure. International Journal of Heat and Mass Transfer. 52(13-14):3176-3186.
[17] Sheremet MA, Pop I. (2015). Mixed convection in a lid-driven square cavity filled by a
nanofluid: Buongiorno’s mathematical model. Applied Mathematics and Computation. 266:792-
808.
[18] Khanafer K, Vafai K, Lightstone M. (2003). Buoyancy-driven heat transfer enhancement in
a two-dimensional enclosure utilizing nanofluids. International journal of heat and mass trans-
fer. 46(19):3639-3653.
[19] Sunu PW, Sutarna IN, Arsawan I. (2017). Experimental Investigations of Heat Transfer
Enhancement in a Double Tube Heat Exchanger with Rectangular Grooved. Advanced Science
Letters. 23(12):12190-12193.
 1
8

[20] Siavashi M, Bahrami HR, Saffari H. (2015). Numerical investigation of flow characteristics,
heat transfer and entropy generation of nanofluid flow inside an annular pipe partially or
completely filled with porous media using two-phase mixture model. Energy. 93:2451-2466.
[21] Uddin MJ, Rahman MM. (2017). Numerical computation of natural convective heat trans-
port within nanofluids filled semi-circular shaped enclosure using nonhomogeneous dynamic
model. Thermal Science and Engineering Progress. 1:25-38.
[22] Uddin MJ, Rahman MM, Alam MS. (2017). Analysis of natural convective heat transport in
homocentric annuli containing nanofluids with an oriented magnetic field using nonhomoge-
neous dynamic model. Neural Computing and Applications.
https://doi.org/10.1007/s00521-017-2905-z
[23] Uddin MJ, Alam MS, Rahman MM. (2017). Natural convective heat transfer flow of
nanofluids inside a quarter-circular enclosure using nonhomogeneous dynamic model. Arabian
Journal for Science and Engineering. 42(5):1883-1901.
[24] Uddin MJ, Alam MS, Al-Salti N, Rahman MM. (2016). Investigations of Natural convec-
tion heat transfer in nanofluids filled horizontal semicircular-annulus using nonhomogeneous
dynamic model. Am J Heat Mass Transf. 3(6):425-452.
[25] Bhagoria JL, Saini JS, Solanki SC. (2002). Heat transfer coefficient and friction factor
correlations for rectangular solar air heater duct having a transverse wedge-shaped rib roughness
on the absorber plate. Renewable Energy. 25(3):341-69.
[26] Sheikholeslami M, Hayat T, Alsaedi A.(2017). Numerical study for external magnetic
source influence on water-based nanofluid convective heat transfer. International Journal of
Heat and Mass Transfer. 106:745-55.
[27] Maxwell, JA. (1873). A treatise on electricity and magnetism. Vol. I-II, Clarendon Press,
Oxford.
[28] Zienkiewicz OC, Taylor RL, Nithiarasu P. (2005). The Finite Element Method for Fluid Dy-
namics, 6th edition, Elsevier.
[29] Codina R. (1998). Comparison of some finite element methods for solving the diffusion-
convection-reaction equation. Computer Methods in Applied Mechanics and Engineering.
156(1-4):185-210.
[30] Uddin MJ, Rahman MM. (2018). Finite Element Computational Procedure for Convective
Flow of Nanofluids in an Annulus. Thermal Science and Engineering Progress. 6: 251-267
[31] Ho CJ, Liu WK, Chang YS, Lin CC. Natural convection heat transfer of alumina-water
nanofluid in vertical square enclosures: an experimental study. International Journal of Thermal
Sciences. 49(8):1345-1353.
[32] Saghir MZ, Ahadi A, Mohamad A, Srinivasan S. (2016). Water aluminum oxide nanofluid
benchmark model. International Journal of Thermal Sciences 109: 148-158.