1 s2.0 S0167732223008061 Main
1 s2.0 S0167732223008061 Main
1 s2.0 S0167732223008061 Main
A R T I C L E I N F O A B S T R A C T
Keywords: Heat transport induced by buoyancy caused natural convection inside a segmented or selectively heated
TiO2-water nanofluid enclosure has grown in importance over the decades due to its significance in many fields in industrial imple
Natural convection mentations such as heat source cooling, food sterilizers, crystal growth, geothermal power systems, solar thermal
Square cavity
collectors, melting and recrystallization procedures, microelectronic and nuclear industries, biomedical, and so
Partially active walls
Numerical analysis
forth. The novelty of current modern analysis is to scrutinize the natural convection of Titanium Oxide-water
Finite element method nanofluid in the cavity with partially active walls simulated by the finite element (FEM) method. The fluid
flow and the thermal transportation within the enclosure are scrutinized. Two cases are carried out in this novel
study, heat sink with constant temperature Tc consists in case 1 from the left and right partially active cavity
walls and isothermal heat source having temperature Th correspondingly, where(Th > Tc ). While upper and
lower walls with inactive portions of the left and right walls are kept insulated. In case 2 the left and right walls
of the cavity are heated while a partially active bottom wall is cold with other inactive parts insulated by the
cavity walls. The finite element method is applied to tackle the governing equations. The Significance of Rayleigh
number (100⩽Re⩽1e6) and volume fractions of nanopowders (0.01⩽ϕ⩽0.06) against fluid flow and heat transfer
is demonstrated through streamlines and isotherms. From the results it is concluded that velocity amount is
increased with escalating the values of Rayleigh number. The thermal results are enhanced with Rayleigh
number. Furthermore the heat transfer is rised with volume fraction of nanoparticles. Additionally, from the
analysis reveals that local Nusselt number is enhanced with Rayleigh number and nanoparticle volume fraction.
* Corresponding author.
E-mail address: hassan@ujs.edu.cn (H. Waqas).
https://doi.org/10.1016/j.molliq.2023.122003
Received 8 January 2023; Received in revised form 14 April 2023; Accepted 1 May 2023
Available online 6 May 2023
0167-7322/© 2023 Elsevier B.V. All rights reserved.
S.A. Khan et al. Journal of Molecular Liquids 382 (2023) 122003
Fig. 1(a-b. ). physical description of natural convection in Cavity together with (a) Case 1: Left and Right partially active walls are cold and heated, inactive portions
are adiabatic (b) Case 2: sided partially active walls are heated; partially active portion of lower walls is cold while the portions which inactive of the whole cavity are
kept adiabatic.
induced by any external source. Such a mechanism can be found in a [17] disclosed the nanofluidic flow on nanochannel. Mosavi et al. [18]
range of engineering fields, with cooling implementations, confinement scrutinized the Ar/Cu nanofluid flow. Shang et al. [19] examined H2O/
building cooling, room ventilation, cooling systems, reservoirs, solar CuO nanofluid with dynamic simulation. Aman et al. [20] investigated
collectors, etc. Because of nanofluids’ outstanding and adjustable the carbon nanotubes Maxwell nanofluids.
characteristics, such as greater thermal efficiency, clogging in flow The microelectronic devices cooling have been a critical problem in
passages due to tiny size and shape and very high specific surface areas recent decades as a result of technological advances, attracting the
of submicron particles, many researchers have been interested in using attention of scientists. Many recent researches have demonstrated that
them in natural convection flow. Islam et al. [5] analyzed the natural employing nanofluids, which have good thermal stability, can enhance
convectional flow of nanoliquid through enclosure. The convectional heat transference performance. The investigation of heat transfer inside
flow across dual plate was investigated by Kumar & Premachandran [6]. cavities of various shapes, including as square-shaped, U-shaped, T-
Nazir et al. [7] introduced the magnetized nanofluid in the cavity shaped, C-shaped, L-shape, and incinerator-shaped, has been done in
problem. Ghalambaz et al. [8] illustrated the hybrid nanofluid inside a recent years, and the results reveal that employing a nanofluid than a
square cavity. Noghrehabadi et al. [9] dissected the nanofluid convec regular fluid improves convection heat flow. Hatami et al. [21] disclosed
tional flow across cone-shaped geometry. The convectional flow of the nanofluidic through a T-shaped cavity. The obtained outcomes
water-based nanofluid flow across a wavy cone was scrutinized by designated that heat transfer enhances by introducing nanoparticles in
2
S.A. Khan et al. Journal of Molecular Liquids 382 (2023) 122003
Case 1:
Fig. 1c. Normal element size Mesh distribution of the square cavity. It is considered that on the left and right sides of the cavity the heat
sinks with constant temperature Tc and isothermal heat source with
temperature Th respectively. Consider heat source temperature is higher
than the heat sinkTh > Tc . The length of the heat sink and the heat
source isL/2. It is supposed that the above and lowest walls of the cavity
with inactive parts of the side walls are insulated.
Case 2:
In this case partially active side walls (left & right) of the enclosure
are heated walls with temperature Th and the partially active bottom
wall is cooled with temperatureTc , as shown in Fig. 2. The above cavity
wall is considered thermally insulated.
Case 1:
3
S.A. Khan et al. Journal of Molecular Liquids 382 (2023) 122003
On the side wall of cavity at partially active portion:u = 0, v = 0, Here the thermally expansion coefficient for mono nanoliquid can be
T = Tc written as
On active right cavity wall:u = 0, v = 0, T = Th
(ρβ)nf = (1 − ϕ)(ρβ)f + ϕ(ρβ)s . (7)
On insulated portions of walls:u = 0, v = 0, Tn = 0
Where βf represents the base fluid coefficient of thermal expansion and
Case 2:
βs indicates the solid particle coefficient of thermal expansion.
Nanofluid thermal diffusivity can be expressed as:
On active wall portions of cavity along left and right sides:u = 0,v =
0, T = Th knf
αnf = ( ) . (8)
On the lower partially active wall of cavity:u = 0, v = 0, T = Tc ρc p nf
On insulated walls:u = 0, v = 0, Tn = 0
The nanofluid thermal conductivity is define as
2.2.1. Thermophysical characteristics of nanofluid (
knf ks + 2kf − 2ϕ kf − ks
)
Many formulations for the nanofluid properties are developed in = ( ). (9)
kf ks + 2kf + ϕ kf − ks
literature. In the current problem, the relation depends on volume
friction of nanoparticles to obtain the properties of nanofluid is obeying The nanofluid dynamic viscosity is
[27,29].
knf
The nanofluid density can be evaluated by following relation μnf = . (10)
(1 − ϕ)2.5
ρnf = (1 − ϕ)ρf + ϕρs . (5)
The presented thermophysical features of base fluid and particles are
Here, ϕ denotes the volume fraction, ρf be the base fluid density and ρs described in Table 2.
represents the particle density.
The heat capacitance is defined as 2.2.2. Dimensionless scales
( ) ( ) ( ) Dimensionless variables for converting the above-mentioned gov
ρcp nf = (1 − ϕ) ρcp f + ϕ ρcp s . (6) erning equations:
4
S.A. Khan et al. Journal of Molecular Liquids 382 (2023) 122003
Fig. 2b. Streamlines and Isotherms outcomes matches with Sheikhzadeh et al. [26].
( )
Table 1 ∂V * ∂V * ∂P* μnf ∂2 V * ∂2 V *
Grid adaptation study atϕ = 0.04, Ra = 1E5 andPr = 6.7.
U* *
+ V * * = − ρf /ρnf * + +
∂X ∂Y ∂Y αρnf ∂X *2 ∂Y *2 (14)
Level Mesh Mini. Average Nuavg CPU
+(ρβ)nf /βf ρnf RaPrθ,
elements Quality Quality Times (s)
X * = x/L, Y * = y/L, U * = uL/αf , V * = vL/αf , On the partially active left wall:U* = V * = 0, θ = 0 (heat sink)
* 2
P = pL /ρnf α 2
= θ(Th − Tc ) + Tc , Pr = νf /αf ,
f,T (11) On the partially active right wall:U* = V * = 0, θ = 1 (heat source)
Ra = gρf βf (Th − Tc )L3/μf αf On insulated walls: U* = V * = 0, ∂θ/∂n = 0.
Table 2
Thermophysical features of host fluid with nanoparticle.
( ) ( ) ( ) ( )
Properties μ Ns/m2 ρ kg/m3 cp (J/kg.K) K(W/m.K) β K− 1 α m2 /s
5
S.A. Khan et al. Journal of Molecular Liquids 382 (2023) 122003
Fig. 3. Streaamlines and Isotherms at different Rayleigh number when ϕ = 0.04 for Case 1.
6
S.A. Khan et al. Journal of Molecular Liquids 382 (2023) 122003
Fig. 3. (continued).
Case 2: ∫ Y2
⃒
hL knf ∂θ ⃒⃒
Nuavg = =2 dY. (18)
On the partially active left and right walls of cavity:U* = V * = 0, kf kf Y1 ∂X ⃒X=0
θ = 1 (heat source) Here Y1 = y1 andY2 = y2 .
L L
On the partially active lower wall of cavity:U* = V * = 0, θ =
0 (heat sink)
3. Numerical approach
On insulated wall: On insulated walls:U* = V * = 0, ∂θ/∂n = 0.
In the current novel Scrutinization, the nonlinear coupled partial
2.2.3. Nusselt number calculation
differential equations (PDE’s) with subjective boundary restrictions are
The Nusselt number which is associated with thermal transportation
numerically tackled utilizing FEM (finite element method). To achieve
of square cavity is mathematically defined as:
this, we transmit the nonlinear PDEs into a linear system using Newton’s
Lqw linearization method. Because to the fact that a big system is divided
Nuloc = ,
kf (Th − Tc ) into a finite number of smaller components, known as finite elements,
FEM computations are capitalised in the majority of complex domains.
where
Discretizing the full flow regime or solution zone into numerous smaller
( )
∂T triangular components is the first step. The pressure of Equations (13)–
qw = − knf (16) ( * *) ( )
∂y (14) is addressed by the equation P = (γ ∂∂UX* + ∂∂UY* ), where γ ≈ 107 is
⃒ denoted as a penalty factor. The solution procedure’s convergence limit
knf ∂θ ⃒⃒ is10− 6 . The outline of the finite element method is shown in Fig. 2a. To
Nuloc = − . (17)
kf ∂Y ⃒Y * =0
*
validate our model, we have made a comparative analysis with analysis
of Sheikhzadeh et al. [26]. The streamlines and Isotherms in Fig. 2b of
The average Nusselt number (Nuavg ) is computed through the following our results and Sheikhzadeh et al. [26] are mentioned. Here it is
expression: observed a same pattern for our current result with published work. In
comparative analysis we develop a streamlines isotherms forRa = 103 .
7
S.A. Khan et al. Journal of Molecular Liquids 382 (2023) 122003
Fig. 4. Streaamlines and Isotherms at different Rayleigh number when ϕ = 0.04 for Case 2: when active part of Left and Right wall of cavity are hot but active part of
cavity lower wall part is cold.
Table 3 depicts the comparison analysis of current results with other to larger buoyancy forces. The Rayleigh numbers are influenced by the
investigators results for average Nusselt number. There is fair agreement rise in buoyancy force and convection intensity, which enhances the
between present result with findings of Sheikhzadeh et al. [26] and strength of the flow rotation. The Rayleigh numbers are influenced by
Khanafer et al. [27]. the rise in buoyancy force, which improves the power of the flow rota
tion. The strength of the streamlines, as well as the number and size of
4. Result’s interpretation the cells, rises as the Rayleigh number (RN) improves owing to the
strong buoyancy forces contrasted to the viscous forces. Furthermore, it
Current pagination analyzes convection flow of mono nanofluid via is analyzed that escalating the Rayleigh number improves the flow ve
square cavity. The current portion performs numerical results via locity of the nanofluid, which is owing to an increment in the influence
streamlines, temperature contour and isotherm against different values of buoyancy force inside the enclosure, particularly near the walls.
of Rayleigh number parameters at ϕ = 0.04 for both cases. For both Similarly, the impact of the buoyancy force in the cavity caused by a
cases the heat sink and heat source length are equals, is half of length of augmentation in the Rayleigh number is magnified by a modification in
square cavity. the temperature contours and isotherm distribution.
The significance of Rayleigh number on streamlines and isotherms
for the TiO2 − water nanofluid in cavity with side walls partially active
4.1. Significance of Rayleigh number for case 1 at selected amount Ra = 1000 at ϕ = 0.04 is shown in Fig. 3 (b). The
influence of convection thermal transfer becomes increasingly sub
Fig. 3 (a-e) illustrates the decoration of streamlines and isotherms stantial as the Rayleigh number grows. The influence of convection
inside the partially active walls of square enclosure for Rayleigh number thermal transportations becomes increasingly substantial as the Ray
when . Fig. 3 (a) reveals the consequence of Rayleigh number for . An leigh number grows, and different boundary layers form along the
antic clockwise vertex are appears in inside the cavity for this case. It is enclosure partially active walls. By increasing Rayleigh number, the
observed the streamlines intensity escalates with Rayleigh number due
8
S.A. Khan et al. Journal of Molecular Liquids 382 (2023) 122003
Fig. 4. (continued).
isotherms close to energy source and sink region. At larger Rayleigh cavity. However, one can significantly raise the buoyancy forces and
number the flow nature and thermal distribution are appreciably outweigh the viscous forces by raising the Rayleigh number. High
exaggerated as display in Fig. 3 (c-d). From these figures we concluded Rayleigh numbers favor convection currents. Additionally, the iso
that isotherms intensity improves with escalating Rayleigh number. therms exhibit distortion at larger Rayleigh numbers due to the better
Here streamlines and isotherms are more crowded at active surface of convection impacts brought on by an increment in the cavity’s flowing
cavity at larger Rayleigh number, as shown Fig. 3 (e). At Ra = 1e6 the velocity.
streamlines are close to partially heat source wall appear on vertical of
9
S.A. Khan et al. Journal of Molecular Liquids 382 (2023) 122003
4.2. Significance of Rayleigh number for case 2 isotherms plot are plotted at. From this we concluded that here two
contour rotating vortices are appeared with a reverse way within in the
The significance of Rayleigh number (RN) for streamlines contours enclosure of cavity. Left vortex move with clockwise while right move
and isotherm lines for cavity filled by nanofluid. In this case the partially with anti-clockwise. At lower Rayleigh number floatability forces im
active side walls (left & right) and partially bottom wall are heated and proves, the streamlines condense near the wall. From the figure it is
cold respectively. scrutinized that by growing the variations of Rayleigh number the
Fig. 4 (a-c) presents represents streamlines (LHS) and isotherms floatability effect improved as a result the vortex circulation speed de
(RHS) for dissimilar magnitudes of Rayleigh number in cavity filled by creases within the enclosure. The convective heat transmission becomes
nanofluid with volume fraction. In Fig. 4 (a) the streamlines and more noteworthy by mounting the Rayleigh number. From the figure we
10
S.A. Khan et al. Journal of Molecular Liquids 382 (2023) 122003
Fig. 8. Local Nusselt number (LNN) estimation for solid nanoparticle fraction.
observed that alongside the active walls of cavity the streamlines and condensed near the active walls. From Fig. 4 (e) on critical values of
isotherms examine the larger velocities and strong convective heat Rayleigh number we observed isotherms are much packed on partially
transfer in these portions of cavity. Furthermore, it can be observed that active cold walls. Hence the convection heat transfer is improved. From
by escalating the Rayleigh number the streamlines and isotherms are the figure it is concluded that the streamlines are highly packed next to
more crowded at partially active walls. From the Fig. 4(b-e) it can be the active hot wall of the enclosure. Here two oval shape vortices are
analyzed from the collection of Rayleigh number to the streamlines are seen in the cavity. Physically, the symmetry state effluents and the
closely packed on active walls such as sided heated walls and lower centralization of the streamlines in the cavity’s core tend to the hot wall
partially cold wall of length. It is analyzed the isotherms are more as the Rayleigh number rises. Rayleigh number improvement leads to a
11
S.A. Khan et al. Journal of Molecular Liquids 382 (2023) 122003
4.3. Nusselt number The authors declare that they have no known competing financial
interests or personal relationships that could have appeared to influence
Fig. 7 examines the consequence of Rayleigh number (RN) via local the work reported in this paper.
Nusselt number (Heat transfer rate). It is concluded that the Heat
transfer rate is improved by growing the values of the Rayleigh number. Data availability
The consequence of nanoparticle fraction on the Heat transfer rate is
depicting in Fig. 8. Here, it is analyzed that local Nusselt number (Heat Data will be made available on request.
transfer rate) is declined by collective the magnitude of volume fraction.
The correlation for the average Nusselt number in case 1 of paramters Acknowledgments
Rayleigh number and nanoparticles volume fraction is addressed as:
(
Nuavg = 0.4568ϕ + 0.3343Ra0.25
)
(19) The authors extend their appreciation to the Deanship of Scientific
Research at King Khalid University, Abha, Saudi Arabia for funding this
work through Large Groups Project under grant number RGP.2/51/44.
5. Tables
The researchers would like to acknowledge Deanship of Scientific
Research, Taif University for funding this work.
See Tables 1-3.
6. Conclusions References
12
S.A. Khan et al. Journal of Molecular Liquids 382 (2023) 122003
nanofluid flow: the effects of spherical barriers size, J. Mol. Liq. 319 (2020) [24] U. Rashid, D. Lu, Q. Iqbal, “Nanoparticles impacts on natural convection nanofluid
114183. flow and heat transfer inside a square cavity with fixed a circular obstacle”, Case
[19] Y. Shang, R. Balali Dehkordi, S. Chupradit, D. Toghraie, A. Sevbitov, Studies in Thermal Engineering 44 (2023) 102829.
M. Hekmatifar, W. Suksatan, R. Sabetvand, The computational study of [25] A. Hassan, A. Hussain, U. Fernandez-Gamiz, M. Arshad, H. Karamti,
microchannel thickness effects on H2O/CuO nanofluid flow with molecular J. Awrejcewicz, F.M. Alharbi, A. Elfasakhany, A.M. Galal, Computational
dynamics simulations, J. Mol. Liq. 345 (2022) 118240. investigation of magneto-hydrodynamic flow of newtonian fluid behavior over
[20] S. Aman, I. Khan, Z. Ismail, M.Z. Salleh, Q.M. Al-Mdallal, Heat transfer obstacles placed in rectangular cavity, Alex. Eng. J. 65 (2023) 163–188.
enhancement in free convection flow of CNTs Maxwell nanofluids with four [26] G.A. Sheikhzadeh, A. Arefmanesh, M.H. Kheirkhah, R. Abdollahi, Natural
different types of molecular liquids, Sci. Rep. 7 (1) (2017) 2445. convection of Cu–water nanofluid in a cavity with partially active side walls,
[21] M. Hatami, J. Zhou, J. Geng, D. Song, D. Jing, Optimization of a lid-driven T- European Journal of Mechanics-B/Fluids 30 (2) (2011) 166–176.
shaped porous cavity to improve the nanofluids mixed convection heat transfer, [27] K. Khanafer, K. Vafai, M. Lightstone, Buoyancy-driven heat transfer enhancement
J. Mol. Liq. 231 (2017) 620–631. in a two-dimensional enclosure utilizing nanofluids, Int. J. Heat Mass Transf. 46
[22] A.S. Dogonchi, M.A. Ismael, A.J. Chamkha, D.D. Ganji, Numerical analysis of (19) (2003) 3639–3653.
natural convection of Cu–water nanofluid filling triangular cavity with [28] Patankar, S. V. (1980). Numerical heat transfer and fluid flow, Hemisphere Publ.
semicircular bottom wall, J. Therm. Anal. Calorim. 135 (6) (2019) 3485–3497. Corp., New York, 58, 288.
[23] T. Armaghani, M.S. Sadeghi, A.M. Rashad, M.A. Mansour, A.J. Chamkha, A. [29] H.C. Brinkman, The viscosity of concentrated suspensions and solutions, J. Chem.
S. Dogonchi, H.A. Nabwey, MHD mixed convection of localized heat source/sink in Phys. 20 (4) (1952) 571.
an Al2O3-Cu/water hybrid nanofluid in L-shaped cavity, Alex. Eng. J. 60 (3)
(2021) 2947–2962.
13