International Journal of Thermofluids 26 (2025) 101114

Contents lists available at ScienceDirect

International Journal of Thermofluids

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

Magnetohydrodynamic nanofluids flow and heat transfer with radiative heat

flux and exothermic chemical reactions
Md. Mehedi Hasan a,b , M.J. Uddin a,c ,∗, Salah A. Faroughi d
a
Geo-Intelligence Laboratory, Ingram School of Engineering, Texas State University, San Marcos, TX, 78666, USA
b
Department of Mathematical Sciences, The University of Texas at Dallas, Richardson, TX, 75080, USA
c
Department of Mathematical and Physical Sciences, University of Nizwa, Sultanate of Oman
d
Department of Chemical Engineering, University of Utah, Salt Lake City, UT 84112, USA

ARTICLE INFO ABSTRACT

Keywords: Nanofluids are highly effective in optimizing thermal management in engineering systems. The complex
Nanofluid and multifaceted properties of nanofluids require in-depth exploration that transcends their immediate
Heat transfer technological and environmental applications. The exothermic chemical reactions and fundamental attributes
Thermal radiation
of nanofluids have intricate mechanisms to advance flow and heat transfer. To understand the mechanisms and
Magnetohydrodynamics
challenges associated with chemical reactions in nanofluids, this study investigates the flow dynamics and heat
Chemical reaction
transfer in a nanofluid-filled annulus formed between a square and a circle, considering the effects of radiative
heat flux, magnetohydrodynamics (MHD), and exothermic chemical reactions governed by Arrhenius kinetics.
The finite element method is employed to solve the governing equations, and the accuracy of the numerical
scheme is confirmed against published works. The distribution of velocity magnitude, isotherms, vorticity
function, and Nusselt number are examined across a wide range of critical parameters for the copper oxide-
water nanofluid. The current study also displays the heat transfer enhancement for 42 nanofluids. The results
indicate that, for the copper oxide-water nanofluid, both the thermal Rayleigh number and the exothermic
chemical reaction parameter significantly impact the convective flow. The average Nusselt number exhibits
an increasing trend with rising Frank–Kamenetskii and Rayleigh numbers but follows a decreasing pattern
with an increase in the radiation parameter. Higher Frank–Kamenetskii numbers, in conjunction with reduced
radiation parameters, significantly enhance heat transfer. The Nusselt number decreases as the magnetic field
intensity and the radius of the inner circle of the annulus increase. The optimal average Nusselt number is
achieved with a 45◦ − 45◦ − 90◦ magnetic field orientation and a nanoparticle volume fraction of 3%. Copper
oxide-water nanofluid shows a slightly higher average Nusselt number than the other nanofluids studied.

1. Introduction while maintaining operational efficiency [14]. Implementing efficient

cooling and heating techniques with nanofluids can potentially reduce
The flow enhancement and heat transfer intensification utilizing environmental impacts [15]. Nanofluids demonstrate outstanding fire
nanofluids within an annular space are essential in various engineering suppression and damage mitigation capabilities, taking advantage of
applications [1] such as nuclear power plants [2], oil recovery [3], their remarkable cooling properties [16].
drilling [4], chemical separation processes [5], chemical reactors [6], Scientists have conducted numerous experiments and numerical
combustion [7], electronic cooling systems [8,9], underground pollu- simulations on the natural convection of nanofluid flow [17–19] in
tant dispersion [10], and numerous natural and artificial heat transfer different geometric designs [20–22] to improve heat transfer efficiency.
processes [11]. Nanofluids contain nanoparticles and convectional flu- Their studies show that traditional fluids often face limitations in large-
ids, giving nanofluids better thermal conductivity. The higher thermal scale heat transfer requirements because of their minimal thermal
conductivity of nanofluids makes heat transfer much more efficient, conductivity. Adding solid particles to base fluids significantly im-
making them perfect for situations where cooling or heating needs proves thermal performance [23–26]. Nanoparticle concentration plays
to work simultaneously [12]. Convective nanofluids also contribute to a crucial role in improving heat transfer, and adjustments between
energy conservation by improving heat transfer [13]. Nanofluids en- base fluids and nanoparticles are required for optimal flow and heat
able effective thermal management, which supports device downsizing

∗ Corresponding author at: Department of Mathematical and Physical Sciences, University of Nizwa, Sultanate of Oman.
E-mail address: md.jashim@unizwa.edu.om (M.J. Uddin).

https://doi.org/10.1016/j.ijft.2025.101114
Received 24 July 2024; Received in revised form 14 January 2025; Accepted 29 January 2025
Available online 7 February 2025
2666-2027/© 2025 Published by Elsevier Ltd. This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/).
M.M. Hasan et al. International Journal of Thermofluids 26 (2025) 101114

transfer [27]. Ho et al. [28] empirically examined the impact of 0.1% 2. Model
alumina nanoparticles in water within a square enclosure, revealing an
approximate 18% enhancement in heat transfer. Natural convection in 2.1. Physical model
enclosures has been the subject of extensive research, involving heated
cylinders, various geometries, boundary conditions, working fluids, The nanofluids filled three-dimensional geometry of the problem
numerical techniques, and other factors that significantly affect heat is meticulously designed to reflect practical applications. Initially, a
transfer [29]. Acharya [30] investigated the optimal geometric shape two-dimensional domain is created to represent the annular region
of the cylinder to enhance heat transfer within a square enclosure, between a square and a circle. However, to better model the physical
considering diamond, square and circular cylinders. The results favored system, we extruded this 2D geometry in the 𝑧-direction by a small
the circular annulus for superior heat transfer performance, while the length of 0.5𝐿(0.1 m), creating a quasi-3D domain. To resolve the
diamond shape proved less effective. Khan et al. [31] examined the dimensional ambiguity and focus on the primary flow characteristics,
impact of a Y-shaped heated cylinder, showcasing its potential to im- we applied periodic boundary conditions on the 𝑥𝑦-planes (front and
prove heat transfer. Jakeer et al. [32] reviewed the cylinder’s position back surfaces). This allowed us to simulate a repeating flow pattern and
and found that its width significantly influenced the heat transfer. The effectively treat the problem as periodic along the 𝑧-direction, where
flow variables such as temperature and velocity repeat periodically.
convective flow of nanofluids under magnetohydrodynamics is impor-
Using this boundary condition, we could present the results in the
tant in various scientific and engineering sectors. Due to its importance,
𝑥𝑦-plane, capturing the essential flow features without including the
the effects of magnetohydrodynamics on convective nanofluid flow in
entire third dimension. The radius of the inner circular cylinder is
multiple settings and practically oriented geometries have been the sub-
𝐿 = 0.2 m representing the characteristic length of the geometry. The
ject of much research [33,34]. Seyyedi et al. [35] directly connect MHD
three-dimensional square disk has dimensions of 5𝐿(1 m) its length and
parameters and entropy generation. Kargarsharifabad [36] investigated
height while its width spans 0.5𝐿(0.1 m). The downward-pointing green
the influence of magnetic fields on the flow of a CuO-water nanofluid ( )
arrow signifies the gravitational direction, g m∕s2 . The 𝑥, 𝑦, and 𝑧
within a cubic domain. Their findings indicated that the presence of orientations on the right side indicate the orientation of these axes. A
a magnetic field slows down the flow of the nanofluid. Furthermore, zero fluid flow velocity is assumed at all walls, denoted as 𝐮 = 𝟎. The
the inclination angle of the magnetic field plays a significant role gravitational force 𝐠 = [0, 𝑔 , 0] acts in the downward direction parallel
in determining the flow directions of the nanofluid; As this angle to the 𝑦-axis. The configuration of the physical region takes the form
increases, the damping effect on the flow also intensifies [37]. of a square enclosure with the heated cylinder positioned at the point
Thermal radiation introduces an additional heat transfer mechanism (2.5𝐿, 2.5𝐿) where 𝐿 is the radius of the cylinder. The left surface of
that operates independently in fluid properties and has relevance in the geometry is considered sinusoidally heated, whereas the surface
various real-world applications [38]. Increased radiation parameters of the inner cylinder is considered uniformly heated (𝑇 = 𝑇ℎ ). The
improve heat transfer and intensify convective flow [39]. Reddy and right surface maintains a relatively colder temperature, set as 𝑇 = 𝑇𝑐 ,
Sridevi [40] studied the radiative flow behavior in a square duct using which ensures 𝑇ℎ > 𝑇𝑐 respectively, while the remaining surfaces are
carbon nanotubes and demonstrated that thermal radiation enhances considered as insulated boundaries (see Fig. 1).
heat transfer rate. Increasing the exothermic chemical reaction param-
eter led to a heat transfer increase of up to 22.26% at the center of 2.2. Mathematical formulation
the cavity [41]. The strength of the magnetic field and the presence
of the Arrhenius chemical reaction are also significant for controlling The investigation focuses on a three-dimensional laminar flow of
the flow of the nanofluids and the heat transfer that goes along with nanofluids, emphasizing its Newtonian, steady, and incompressible
it. In the convective nanofluid flow, both single-phase [42] and two- characteristics. The study uses Arrhenius kinetics to show the exother-
phase [43,44] models and nanofluid flow analysis by lattice Boltzmann mic chemical reaction, thermal radiative flux to determine the thermal
method [45,46] show that the chemical reactions can play a significant radiation, and the magnetic field to understand how the magnetic
positive impact on heat transfer. nature influences the total heat transfer. This study examines the
combined effects of chemical reactions, radiation, and MHD, essential
The literature review recommends using nanofluids in practical
in energy systems, aerospace, and environmental engineering. Exother-
heat transfer processes, yet advanced research is necessary to com-
mic reactions release heat, which is critical for industrial processes
prehensively understand their flow and heat transfer characteristics.
and power generation. Radiation supports energy transfer and medical
Due to the distinct properties of different nanofluids, generalization
therapies, while MHD is used in power generation and reactor cooling.
is not feasible. Further investigation is required into the hydrother-
Together, these phenomena enable applications such as rocket and
mal behavior of nanofluids in various geometric configurations, both plasma propulsion and optimizing energy release in nuclear fusion
with and without cylinders. Understanding convective nanofluid flow reactors, driving innovation across various industries. The consistent
necessitates examining natural convection with magnetohydrodynam- magnetic field 𝐁 is represented by,
ics (MHD), thermal radiation, and chemical reactions. The literature
𝐁 = 𝐢𝐵𝑥 + 𝐣𝐵𝑦 + 𝐤𝐵𝑧 , (1)
review highlights significant gaps in the research regarding the natural
( )
convection of CuO-water nanofluid in an annular space, particularly 𝐁 = 𝐵0 cos 𝛾1 , 𝐵0 cos 𝛾2 , 𝐵0 cos 𝛾3 , (2)
under the influence of a sloping magnetic field, thermal radiation, and √
chemical reactions. A key innovation of this study is to explore how where 𝐵0 = 𝐵𝑥2 + 𝐵𝑦2 + 𝐵𝑧2 and 𝛾1 , 𝛾2 , and 𝛾3 respectively signify the
MHD, thermal radiation, and exothermic chemical reactions driven inclination angles of the magnetic field in terms of the positive 𝑥, 𝑦,
by Arrhenius kinetics collectively affect flow and heat transfer. The and 𝑧-axes. The solid nanoparticles and the base fluid share the same
study investigates the impact of thermal Rayleigh number, Hartmann temperature, signifying thermal equilibrium. We neglect joule heating
number, nanoparticle concentration, magnetic field inclination angle, and viscous dissipation for simplicity in the present study. Based on the
assumptions, the governing equations are as follows [30,41,47],
and the characteristic length of the geometry on flow and heat transfer.
The purpose of this paper is to develop heat transfer coefficients for ∇⋅𝐮=0 (3)
CuO-water nanofluid flow by analyzing the factors that enhance heat ∇𝑝 𝜇𝑛𝑓 2 ( ) (𝜌𝛽)𝑛𝑓 1 ( )
𝐮 ⋅ ∇𝐮 = − + ∇ 𝐮 + 𝑇 − 𝑇𝑐 𝐠+ 𝐉 ∗ ×𝐁 (4)
transfer in practical applications. 𝜌𝑛𝑓 𝜌𝑛𝑓 𝜌𝑛𝑓 𝜌𝑛𝑓 𝐄

2
M.M. Hasan et al. International Journal of Thermofluids 26 (2025) 101114

Fig. 1. A schematic representation of the physical domain . (For interpretation of the references to color in this figure legend, the reader is referred to the web version of this
article.)

16𝜎 ∗ 𝑇𝑐3 𝑘 𝑎 𝜌𝑏𝑓 𝜇𝑛𝑓 Pr 𝜃(𝜌𝛽)𝑛𝑓

𝐮 ⋅ ∇𝑇 = 𝛼𝑛𝑓 ∇2 𝑇 + ( ) ∇2 𝑇 + 𝑄 0 (5) 𝐔 ⋅ ∇𝐔 = − ∇𝑃 + ∇2 𝐔 + Pr ⋅ Ra
3 𝜌𝑐𝑝 𝑛𝑓 𝑘∗ 𝑒(𝐸∕𝑅𝑇 ) 𝜌𝑛𝑓 𝑣𝑏𝑓 𝜌𝑛𝑓 𝛽𝑏𝑓 𝜌𝑛𝑓
𝜌𝑏𝑓 𝜎𝑛𝑓 2
where 𝐮, 𝑝, and 𝑇 represent the velocity component, pressure, and + Pr ⋅𝐻 𝑎 {(𝐔 × 𝐁) × 𝐁}, (12)
𝜌𝑛𝑓 𝜎𝑏𝑓
temperature, respectively. Furthermore, the expression 𝑎𝑘0 ⋅ 𝑒−(𝐸∕𝑅𝑇 ) ( )
( )−1 𝛼𝑛𝑓 𝜌𝑐𝑝 𝑏𝑓 4
characterize the Arrhenius equation; the expression 𝛼𝑛𝑓 = 𝜅𝑛𝑓 ⋅ 𝜌𝑐𝑝 𝑛𝑓 𝐔 ⋅ ∇𝜃 = 2
∇ 𝜃+ ( ) 𝑅𝑑 ∇2 𝜃 + 𝐹𝑘 𝑒𝜃 , (13)
denotes the thermal diffusivity of the nanofluid, while the meanings 𝛼𝑏𝑓 𝜌𝑐𝑝 𝑛𝑓 3
of all thermophysical coefficients are available in the Appendix. The
Lorentz force 𝐉𝐄∗ × 𝐁 is the output of the interaction between 𝐮 and and boundary conditions as
𝐁, while 𝐉𝐄∗ (electric current density) is defined by Ohm’s law as ⎧
⎪𝐔 = 𝟎, 𝜃 = sin 2𝜋(𝑌 + 𝑍), 0 ≤ 𝑌 ≤ 1, 𝑋 = 0, 0 ≤ 𝑍 ≤ 0.1,
𝐉𝐄∗ = 𝜎𝑛𝑓 (𝐄∗ + 𝐮 × 𝐁). The Lorentz force components due to 𝐄∗ = 0 ⎪𝐔 = 𝟎, 𝜃 = 0, 0 ≤ 𝑌 ≤ 1, 𝑋 = 1, 0 ≤ 𝑍 ≤ 0.1,
(neglecting any electric field), and cos2 𝛾1 + cos2 𝛾2 + cos2 𝛾3 = 1 in the ⎪
⎪𝐔 = 𝟎, ∇𝜃 = 0, 0 ≤ 𝑋 ≤ 1, 𝑌 = 0, 0 ≤ 𝑍 ≤ 0.1,
𝑥, 𝑦 and 𝑧-directions respectively yield, ⎨ (14)
( ( ) ) ⎪𝐔 = 𝟎, ∇𝜃 = 0, 0 ≤ 𝑋 ≤ 1, 𝑌 = 1, 𝑍 = 0.1,
𝜎𝑛𝑓 𝐵0 2 𝐰 cos 𝛾3 cos 𝛾1 − 𝐮 cos2 𝛾2 + cos2 𝛾3 + 𝐯 cos 𝛾1 cos 𝛾2 , (6) ⎪𝜃 = 1,
( ( ) ) 𝑋 = sin 𝛩 cos 𝛹 , 𝑌 = sin 𝛩 sin 𝛹 ,
⎪
𝜎𝑛𝑓 𝐵02 𝐰 cos 𝛾2 cos 𝛾3 − 𝐯 cos2 𝛾1 + cos2 𝛾3 + 𝐮 cos 𝛾1 cos 𝛾2 , (7) ⎪ 𝑍 = cos 𝛩, 0 ≤ 𝛩, 𝛹 ≤ 2𝜋 .
( ( 2 ) ) ⎩
2 2
𝜎𝑛𝑓 𝐵0 𝐮 cos 𝛾1 cos 𝛾3 − 𝐰 cos 𝛾1 + cos 𝛾2 + 𝐯 cos 𝛾2 cos 𝛾3 . (8)
( )−1 [ 2 ] ( )−1
where 𝛼 = 𝜅 𝜌𝑐𝑝 m ∕s is the thermal diffusivity, Pr = 𝑣𝑏𝑓 𝛼𝑏𝑓
Here are the boundary settings relevant to the current problem: ( )−1
is the Prandtl number, 𝑅𝑎 = 𝐿3 𝑔 𝛽𝑏𝑓 𝑅𝑇𝑐2 𝐸 𝛼𝑏𝑓 𝜈𝑏𝑓 is the Rayleigh
⎧ 0.5 −0.5
⎪𝐮 = 0, 𝑇 = 𝑇𝑐 + (𝑇ℎ − 𝑇𝑐 ) number, 𝐻 𝑎 = 𝐵0 𝐿𝜎𝑏𝑓 𝜇𝑏𝑓 is the magnetic field parameter (Hartmann
⎪ × sin 2𝜋(𝑦 + 𝑧), ( )−1 ∗ 3
0 ≤ 𝑦 ≤ 5𝐿, 𝑥 = 0, 0 ≤ 𝑧 ≤ 0.5𝐿, number), 𝑅𝑑 = 𝑘∗ 𝜅𝑏𝑓 4𝜎 𝑇𝑐 is the thermal radiation parameter, and
⎪ ( )
⎪𝐮 = 0, 𝑇 = 𝑇𝑐 , 0 ≤ 𝑦 ≤ 5𝐿, 𝑥 = 5𝐿, 0 ≤ 𝑧 ≤ 0.5𝐿, 𝐸 𝐿2 𝑄𝑘0 𝑎
⎪ 𝐹𝑘 = 𝑅𝑇 𝐸∕𝑅𝑇𝑐 is the exothermic chemical reaction parameter.
𝑇𝑐 𝜅𝑏𝑓 𝑒
𝑐
⎨ 𝐮 = 0, ∇𝑇 = 0, 0 ≤ 𝑥 ≤ 5𝐿, 𝑦 = 0, 0 ≤ 𝑧 ≤ 0.5𝐿, (9)
⎪ The dimensionless numbers allow for the comparison and scaling
⎪ 𝐮 = 0, ∇𝑇 = 0, 0 ≤ 𝑥 ≤ 5𝐿, 𝑦 = 5𝐿, 𝑧 = 0.5𝐿,
⎪𝑇 = 𝑇 , of systems, representing the balance of critical forces and mechanisms,
𝑥 = 𝐿 sin 𝛩 cos 𝛹 , 𝑦 = 𝐿 sin 𝛩 sin 𝛹 ,
⎪ ℎ
such as inertia, viscosity, and thermal effects, without relying on spe-
⎪ 𝑧 = 𝐿 cos 𝛩, 0 ≤ 𝛩, 𝛹 ≤ 2𝜋 .
⎩ cific units. The goal is to describe the flow behavior and heat transfer
in relation to the dominant forces or mechanisms. The primary dimen-
Next we transform the governing equations into a non-dimensional sionless number in this study is the Frank–Kamenetskii number, which
form [41]. To this end, we define, characterizes the rate of heat generation from exothermic chemical
( )
𝐱 𝐮𝐿 𝐿2 𝑝 𝑇 − 𝑇𝑐 𝐸 reactions. This parameter reflects the balance between heat produced
𝐗 = ,𝐔 = ,𝑃 = ,𝜃 = , (10)
𝐿 𝛼𝑏𝑓 2 𝜌 𝑅𝑇𝑐2
𝛼𝑏𝑓 𝑏𝑓 by reactions and heat lost through conduction, and it also indicates the
tendency for thermal instability when the Frank–Kamenetskii number
and substitute them into Eqs. (3), (4), (5), and (9), leading to the
is high.
non-dimensional form of governing equations as,
∇ ⋅ 𝐔 = 0, (11)

3
M.M. Hasan et al. International Journal of Thermofluids 26 (2025) 101114

Table 1
Thermophysical characteristics of fourteen nanoparticles and three base fluids at room temperature [48].
Items (p, bf) 𝜌 [kg/m3 ] 𝛽 [1/K] 𝜎 [1/Ω m] 𝑐𝑝 [J/kg K] 𝜅 [W/m K] 𝜇 [kg/ms]
Co 8900 1.3 × 10−5 16.02 × 106 420 100 –
Zn 7135 3.02 × 10−5 16.9 × 106 387 116 –
Cu 8933 1.67 × 10−5 59.6 × 106 385 401 –
Ag 10 500 1.8 × 10−5 63 × 106 233 429 –
Al 2701 2.31 × 10−5 3.5 × 107 902 237 –
Fe3 O4 5180 20.6 × 10−5 0.112 × 106 670 80.4 –
CuO 6320 1.8 × 10−5 59.6 × 106 531.8 76.5 –
ZnO 5610 4.7 × 10−5 16.9 × 106 495.04 29 –
TiO2 4250 0.9 × 10−5 2.6 × 106 686.2 8.9538 –
Al2 O3 3970 0.85 × 10−5 35 × 106 765 40 –
SiO2 2200 5.6 × 10−5 10.3 × 106 765 36 –
MWCNT 1600 4.2 × 10−5 105 796 3000 –
CoFe2 O4 4907 12.9 × 10−6 33.4 × 106 700 3.7 –
Mn–ZnFe2 O4 4900 1.2 × 10−6 18.3 × 106 800 5 –
Ethylene glycol (EG) 1114 57 × 10−5 1.07 × 10−6 2415 0.252 0.00304
Propylene glycol (PG) 1132 57 × 10−5 6 × 10−6 2349 0.258 0.0151
Water 997.1 21 × 10−6 5.5 × 10−6 4179 0.613 0.001003

Table 2
The mathematical correlations among the thermophysical properties of Nanofluid (CuO–H2 O)
Properties Nanofluid (CuO–H2 O)
Thermal diffusivity 𝛼nf = 𝜅nf ⋅ (𝜌𝑐𝑝 )−1
nf
Density 𝜌nf = −𝜙𝜌bf + 𝜙𝜌𝑝 + 𝜌bf
−1
[ ]2.8
Dynamic viscosity 𝜇nf ⋅ 𝜇bf = 1 + 5300 𝜙 ⋅ (1 − 𝜙)−1
Heat capacitance (𝜌𝑐𝑝 )nf = −𝜙(𝜌𝑐𝑝 )bf + 𝜙(𝜌𝑐𝑝 )𝑝 + (𝜌𝑐𝑝 )bf
Thermal expansion (𝜌𝛽)nf = (𝜌𝛽)bf + 𝜙(𝜌𝛽)𝑝 − 𝜙(𝜌𝛽)bf
Volumetric mass expansion (𝜌𝛽 ∗ )nf = (1 ∗ ∗
[ − 𝜙)(𝜌𝛽 )bf + 𝜙(𝜌𝛽 )𝑝 √ ]
Thermal conductivity 𝜅nf = 0.25 (3𝜙 − 1)𝜅𝑝 + (2 − 3𝜙)𝜅bf 𝛥 + 𝜅bf ;
here,
( )
−1 2
( )( −1
)
𝛥 = (3𝜙 − 1)2 𝜅𝑝 ⋅ 𝜅bf + (2 − 3𝜙)2 + 2 2 + 9𝜙 − 9𝜙2 𝜅𝑝 ⋅ 𝜅bf

2.3. Thermophysical properties Implementing the transformation as in (10), the dimensionless Nusselt
number (local) is,
𝜅𝑛𝑓 ( 4
)
The physical properties of both the nanoparticles and the base 𝑁 𝑢𝐿 = − 1 + 𝑅𝑑 ∇𝜃 (17)
fluids determine the thermophysical characteristics of nanofluids. These 𝜅𝑏𝑓 3
properties are inherent traits of the fluids and the materials employed. ( )
Hence, the average Nusselt number 𝑁 𝑢avg is written as,
It is worth noting that these properties may change with alterations 1
in state variables such as pressure and temperature, yet they retain 𝑁 𝑢𝑎𝑣𝑔 = 𝑁 𝑢𝐿 𝑑 𝐴 (18)
𝐴 ∫𝛺
their chemical identity. Table 1 depicts the thermophysical properties
The Eqs. (17) and (18) are used to calculate the local and average
of various nanofluids [48] are assessed at room temperature. The
Nusselt numbers, respectively, on the surface of the 𝑥𝑦-plane. The cal-
thermophysical properties presented in Table 1 are used to construct
culations are based on a cutline positioned directly on the surface rather
the 42 nanofluids. These properties are used to calculate the heat than within the domain. The geometry considered is three-dimensional,
transfer rate of nanofluids and find the significant nanoparticles and where the 𝑥𝑦-plane (front and back) is subjected to periodic boundary
nanofluids. The copper oxide-water nanofluid is employed to analyze conditions, and the Nusselt number is evaluated on the front surface of
the flow and heat transfer throughout the paper. the 𝑥𝑦-plane.
Table 2 provides a comprehensive overview of the interdependen-
cies between the thermophysical characteristics of both the nanofluid 3. Computational technique and validation
and the base fluid. We take these equations using empirical relations
that account for several factors and parameters. 3.1. Computational technique

2.4. Evaluation of the Nusselt number The Galerkin scheme of the finite element method (FEM) is used to
discretize the governing equations and obtain numerical solutions [51–
( ) 53]. The enclosed 3D domain is effectively divided into a non-uniform
The local Nusselt number 𝑁 𝑢𝐿 has been calculated for the current
triangular mesh suitable for accommodating random volume domains.
problem as [49,50],
The finite element equations utilize a six-node triangular element con-
𝐿ℎ 𝐿𝑞
𝑁 𝑢𝐿 = = , (15) nected to dependent variables such as velocities and temperature.
𝜅𝑏𝑓 𝜅𝑏𝑓 𝛥𝑇
The pressure gradients are aligned with lower-order multinomials and
where ℎ is the convective heat transfer coefficient, defined as the ratio momentum equations to satisfy the continuity equation. Consistent
of the heat flux (𝑞) to the temperature difference (𝛥𝑇 ), such that ℎ = pressure is assumed within each linear component, though it remains
𝑞∕𝛥𝑇 . Considering the Arrhenius kinetics of the chemical reactive flux discontinuous across elements. The Galerkin weighted residual method
( ) ( )
𝑞𝑐 and the presence of thermal radiative flux 𝑞𝑟 , we express the heat converts the dimensionless governing equations into integral form,
flux (𝑞) as follows, assuming equal weight and approximation functions. The nonlinear
( ) algebraic equations in Eq. (14) are solved using the Gauss quadra-
𝜅𝑛𝑓 𝐸 𝜅𝑛𝑓 𝐸 16𝜎1 𝑇𝑐3 𝜅𝑛𝑓 𝐸 16𝜎1 𝑇𝑐3 ture method. The iterative Newton–Raphson method is employed to
𝑞 = 𝑞𝑐 + 𝑞𝑟 = − ∇𝑇 − ∇𝑇 = − 1 + ∇𝑇
𝑅𝑇𝑐 𝑅𝑇𝑐 3𝑘∗ 𝑅𝑇𝑐 3𝑘∗ solve these nonlinear equations, with the convergence criterion set to
| 𝑛+1 |
|𝜉 − 𝜉 𝑛 | ≤ 10−5 (see Fig. 2).
(16) | |

4
M.M. Hasan et al. International Journal of Thermofluids 26 (2025) 101114

Fig. 2. Flowcharts of computational procedures. Panel (a) illustrates the step-by-step process flow of the Galerkin scheme within the finite element method, while panel (b) details
how the computer code operates to solve the final governing equations of the Galerkin scheme.

( )
Fig. 3. Grid refinement evaluation. The red line illustrates the velocity magnitude 𝑉𝑚 along the horizontal cut line (0 ≤ 𝑋 ≤ 1, 𝑌 = 0.27), and the blue line represents the mean
( ) ( ) ( )
Nusselt number 𝑁 𝑢avg on the heated left vertical wall with the number of grid elements. Panel (a) illustrates the velocity magnitude 𝑉𝑚 and mean Nusselt number 𝑁 𝑢avg
3 4
for various grid elements for the Rayleigh number 𝑅𝑎 = 10 , while panel (b) pertains to 𝑅𝑎 = 6 × 10 . (For interpretation of the references to color in this figure legend, the reader
is referred to the web version of this article.)

3.2. Grid independence test and code validation numbers, the results indicate that within a range of grid sizes up to
2 838 787, the average Nusselt number and the velocity magnitude
The grid test method is employed in the current study to achieve initially exhibit an increase before stabilizing. For the thermal Rayleigh
consistent results regardless of the number of grid constraints, thereby number Ra = 103 , the deviation in the mean heat transfer for the
demonstrating the reliability and accuracy of the numerical simula- last two grid sizes, compared to the grid of 280 968, is minimal,
tions. Optimal adjustment of the mesh allows the FEM to exhibit faster showing errors of only 0.102% and 0.108%, respectively. Similarly,
convergence. Therefore, a series of tests with multiple grid setups
the difference in velocity magnitude between the final two grid sizes
are conducted to determine the optimal grid configuration. A grid
and the 280 968 grid is negligible, with errors of merely 0.1% and
independence test is performed using the thermophysical properties of
0.04%, respectively. Similarly, for the Rayleigh number 𝑅𝑎 = 6 × 104 ,
the CuO-water nanofluid for Pr = 6.8377, 𝐻 𝑎 = 15, 𝐹𝑘 = 4, 𝑅𝑑 = 0.2,
𝜙 = 0.05, 𝛾1 = 30◦ , 𝛾2 = 90◦ , 𝛾3 = 60◦ , and 𝛥𝑇 = 25 K. the variation in average heat transfer is 0.07% and 0.04%, and in the
Fig. 3 demonstrates the grid independence test for 𝑁 𝑢𝑎𝑣𝑔 and 𝑉𝑚 for velocity magnitude, it is 0.02% and 0.01%. The observation suggests
two distinct values of the Rayleigh number, 𝑅𝑎 = 103 and 𝑅𝑎 = 6 × 104 . that the number of grid elements used within the domain, whether
We examined various grid numbers for this grid independence study, 280 968, 710 709, or 2 838 787, does not significantly affect the results.
specifically 7052, 14 884, 47 842, 95 661, 280 968, 710 709, and 2 838 787, Therefore, a grid configuration of 280 968 elements ensures the best
to establish the solution’s insensitivity. For both thermal Rayleigh performance in computational results and reduces computational time.

5
M.M. Hasan et al. International Journal of Thermofluids 26 (2025) 101114

Table 3
The comparison of our current results with the comparative data provided by Tiwari and Das [47] for 𝑅𝑎 = 103 to 𝑅𝑎 = 106 , 𝐻 𝑎 = 0, 𝜙 = 0.0,
and Pr = 0.71 (air).
𝑅𝑎 Tiwari and Das [47] This study Relative error
𝑈max 𝑉max 𝑁 𝑢avg 𝑈max 𝑉max 𝑁 𝑢avg 𝑈max 𝑉max 𝑁 𝑢avg
103 3.642 3.7026 1.0871 3.6026 3.8021 1.0973 0.01079 0.02685 0.00937
104 16.1439 19.6650 2.195 16.721 19.8260 2.2015 0.03574 0.00819 0.00296
105 34.30019 68.7646 4.450 34.5600 68.9261 4.6799 0.00758 0.00235 0.05164
106 65.5866 219.7361 8.803 65.8977 219.8283 8.9591 0.00474 0.00042 0.01773

Table 4
Compared the value of the mean Nusselt number (𝑁 𝑢avg ) of the current study with the study of Wan et al. [54], Davis et al. [55], and Ghasemi
et al. [50] for four distinct values of Ra as well as nanoparticle concentration (𝜙) when 𝛾 = 0◦ and 𝐻 𝑎 = 30.
𝑅𝑎 𝜙=0 𝜙 = 0.02
Wan et al. [54] Davis et al. [55] Ghasemi et al. [50] This study Ghasemi et al. [50] This study Relative error
103 1.117 1.118 1.002 1.114 1.060 1.211 0.1425
104 2.254 2.243 1.183 1.289 1.212 1.331 0.0982
105 4.598 4.519 3.150 3.878 3.138 3.352 0.0682
106 8.976 8.800 7.907 8.611 7.979 9.987 0.2517

The results are also validated with the data provided by [47,50, 4.1. Influence of exothermic chemical reaction parameter
54,55] for a particular condition, whereas the geometry and boundary
conditions remain consistent. The results have been compared with the Exothermic chemical reactions, generating heat as a byproduct,
data presented by Tiwari and Das [47] for a square cavity filled with significantly influence fluid velocity, temperature distribution, and heat
air. Comparative results, including maximum U and V velocities and transfer properties within the surrounding medium. Fig. 4 shows how
the average Nusselt number on the hotter surface, have been calcu- ( )
the Frank–Kamenetskii number 𝐹𝑘 directly impacts temperature dis-
lated. Table 3 presents these comparisons, demonstrating remarkable tribution, velocity, and vorticity function. Fig. 4(a) displays the velocity
agreement with the referenced studies and reinforcing confidence in magnitude with a chemical reaction parameter set to zero. In this case,
the numerical results of this study. various vortices with distinct flow patterns emerge at different locations
Further confirmation of the accuracy of the present code is es-
within the domain. The primary vortex, which exhibits a slight down-
tablished by evaluating its results with those obtained under specific
ward extension, forms in the upper left region of the domain, and the
conditions in steady-state investigations conducted by Wan et al. [54],
highest velocity occurs at its center. Two elongated vortices develop
Davis et al. [55], and Ghasemi et al. [50]. Table 4 details the average
near the inner circular cylinder along the comparatively colder wall.
Nusselt number, which expresses the comparison through quantitative
However, as the 𝐹𝑘 increases, the upper-left vortex and the opposing
data. Furthermore, comparing the output with Wan et al. [54] and
vortices progressively enlarge. When the 𝐹𝑘 value reaches 6, all vortices
Davis et al. [55], the present investigation is adapted for the water-
filled square domain (Pr = 0.71), and the mean Nusselt number is extend across the entire domain, indicating that the parameter 𝐹𝑘 is
computed for Ra = 103 to 106 and 𝜙 = 0. Additionally, Ghasemi dominant in controlling heat transfer and fluid flow patterns. As the
et al. [50] explored the behavior of alumina-water nanofluid within 𝐹𝑘 rises, there is a notable increase in the resultant velocity, suggesting
a square domain, specifically examining the influence of the magnetic that the chemical reaction intensifies the flow. In the horizontal cut line
field parameter and its inclination angle. Our current code replicates (0 ≤ 𝑋 ≤ 1, 𝑌 = 0.27), to be precise, for Frank–Kamenetskii numbers 1,
their findings under specific conditions, representing a particular case 3, and 6, the resultant velocity increases by 4.36%, 16.75%, and 48.72%,
with a magnetic field inclination angle 𝛾 = 0◦ and its parameter 𝐻 𝑎 = respectively. When 𝐹𝑘 = 6, the maximum velocity magnitude reaches
30, also considering the nanoparticle concentrations of 0% and 2%. The 33.09, indicating a 9.75% increment in the resultant velocity compared
present numerical code is in remarkable agreement with the findings to 𝐹𝑘 = 0. Panel (b) of Fig. 4 depicts the effect of three distinct 𝐹𝑘
of Wan et al. [54], Davis et al. [55], and Ghasemi et al. [50]. These values on the isothermal distribution. Other fluid parameters primarily
consistent agreements motivate further exploration of the numerical drive the distribution of isotherms at 𝐹𝑘 = 0. However, introducing
results in the present study. the exothermic chemical reaction results in a significantly convoluted
isotherm pattern throughout the enclosure, indicating that convection
4. Results and discussion is the dominant mode of heat transfer. Panel (c) of Fig. 4 vividly
illustrates the vorticity function for three values of 𝐹𝑘 , revealing the
The governing Eq. (11)–(13) are solved for a specified domain
enhanced efficiency of heat transport from the warmed surface to the
with the boundary conditions (14) for the velocity profiles, vorticity
colder surface as 𝐹𝑘 values increase. At 𝐹𝑘 = 6, the neighboring
functions, isotherms, and Nusselt numbers to understand the heat
vertex of the heated wall intensifies and extends vertically downward.
transfer and the flow of copper oxide-water nanofluid. The results have
Simultaneously, the adjacent vortex near the colder wall strengthens,
been analyzed from the physical and engineering point of view. All
and two smaller vortexes become distinctly visible within the primary
the contour plots are presented in the XY plane surface, represent-
vortex. In general, fluid flows become more robust and encompass
ing a two-dimensional front surface
( of the 3D
) geometry. These out-
comes include radiation factor 0 ≤ 𝑅𝑑 ≤ 0.4( , magnetic field the entire domain, indicating enhanced heat transport at higher 𝐹𝑘
) inclina- values. Specifically, for Frank–Kamenetskii numbers 1, 3, and 6, the
tion (0◦ ≤ 𝛾 ≤ 90◦ ), (Rayleigh )number 103 ≤ 𝑅𝑎 ≤ 105 , Frank-
Kamenetskii number 0 ≤ 𝐹𝑘 ≤ 6 , nanoparticle concentration (0 ≤ average Nusselt number on the vertically heated surface increases by
𝜙 ≤ 0.05), and Hartmann number (0 ≤ 𝐻 𝑎 ≤ 100). The standard 1.91%, 7.43%, and 27.64%, respectively.
parameter settings for displaying results are Pr = 6.8377, 𝑅𝑎 = 6 × 104 , Fig. 5 shows the local distribution of the Nusselt number on the
𝜙 = 0.05, 𝐻 𝑎 = 15, 𝐹𝑘 = 4, 𝑅𝑑 = 0.2, 𝛾1 = 30◦ , 𝛾2 = 90◦ , 𝛾3 = 60◦ , horizontal cutting line at 𝑌 = 0.27 (panel (a)) and the vertical cut line
and 𝛥𝑇 = 25 K, unless explicitly mentioned otherwise. It is important at 𝑋 = 0.72 (panel (b)) for numerous values of the Frank–Kamenetskii
to note that when addressing specific parameter variations, all other numbers. These visual representations enable us to identify areas where
parameters remain constant throughout the presentation of results in heat transfer is particularly efficient or deficient and how fluid ve-
this paper. locities vary throughout the system. Increasing the values of 𝐹𝑘 , a

6
M.M. Hasan et al. International Journal of Thermofluids 26 (2025) 101114

( ) ( )
Fig. 4. Impact of exothermic chemical reaction parameter 𝐹𝑘 . The panels (a), (b), and (c) depict the velocity magnitude 𝑉𝑚 [−] , isotherm distribution (𝜒[−]), and vorticity
function (𝜉[−]), respectively, for three distinct values of 𝐹𝑘 .

substantial reduction observed in the local Nusselt number within the right side. As the radiation parameter increases from 0 to 0.4, all
intervals 0.17 ≤ 𝑋 ≤ 0.21 and 0.45 ≤ 𝑌 ≤ 0.59. In contrast, in the vortices expand and occupy nearly the entire domain, signifying a
remaining sections of the cutting lines, the value of 𝑁 𝑢𝐿 experience a decrease in the resultant velocity. The minimum velocity magnitude
significant enhancement. Specifically, as 𝐹𝑘 rises from 0 to 6, the local of 30.2 occurred at the radiation parameter 𝑅𝑑 = 0.4, and there is a
Nusselt number decreases by 66.17% on the horizontal line 𝑋 = 0.19 reduction in the resulting velocity of 7.79% compared to the radiation
and experiences a maximum increase of 90.17% at the point 𝑋 = 0.65 parameter 𝑅𝑑 = 0. Radiation parameter enhancement primarily causes
along the horizontal cut line. On the vertical cut line, the Nusselt the deterioration in the resultant velocity due to the dominant cooling
number decreases by 8.08% at 𝑌 = 0.5 and exhibits a maximum influence on the fluid. This cooling effect tends to mitigate tempera-
increase of 70.19% at 𝑌 = 0.33. The velocity magnitude profiles shown ture gradients, diminish buoyancy-driven thermal flows, and introduce
in panels (c) and (d) of Fig. 5 exhibit a pattern similar to that of stability into the fluid. Fig. 6(b) provides a clear visual representation
the local Nusselt number. As the 𝐹𝑘 increase from 0 to 6, velocity of the isothermal performance for various 𝑅𝑑 values. The intensity of
magnitude decreases by 85.28% at 𝑋 = 0.2 and experiences a maximum radiation parameters is increased from 0 to 0.4, and they act as catalysts
increase of 26.55% at 𝑋 = 0.66 along the horizontal cut line. Likewise, for other parameters, leading to more efficient heat transfer. At the
on the vertical cut line, the velocity magnitude decreases by 8.42% at radiation parameter 𝑅𝑑 = 0.4, the isotherm of the system increased by
Y = 0.5 and exhibits a maximum increase of 18.7% at Y = 0.33. a maximum of 37.16% compared to the radiation parameter 𝑅𝑑 = 0.
Fig. 6(c) illustrates the vorticity function, unveiling the influence
4.2. Influence of thermal radiation parameter of radiation on the rotation of the fluid. Intricate vortex patterns are
visible in the vorticity contours in low- or no-radiation scenarios. As the
Panels (a), (b), and (c) of Fig. 6 show the effects of radiation radiation values increase, these vortices expand and weaken, signifying
( )
parameters 𝑅𝑑 on the velocity magnitude, isotherms, and vorticity the diminishing effect of radiation on convective motion and resulting
function, respectively. In the absence of the radiation parameter, a in more uniform vorticity patterns. The intensity of the vorticity func-
prominent vortex with the highest velocity magnitude is observed in tion decreases by 2% as the radiation parameter increases from 0 to
the upper-left corner of the enclosure, resulting in a maximum resultant 0.25. However, from 0.25 to 0.4, it increases by 5%. An initial 2%
velocity of 32.75. Two vertically elongated vortices appear on the decrease in vorticity intensity is observed as the radiation parameter

7
M.M. Hasan et al. International Journal of Thermofluids 26 (2025) 101114

( )
Fig. 5. Effects of the Frank–Kamenetskii Number 𝐹𝑘 . Panels (a) and (b) depict the local Nusselt number on the horizontal and vertical cutting lines at 𝑌 = 0.27 and 𝑋 = 0.72,
respectively, for various Frank–Kamenetskii numbers, as indicated in each panel. Panels (c) and (d) display the velocity magnitude on the corresponding cross-sectional lines for
the same Frank–Kamenetskii numbers.

increases from 0 to 0.25, indicating that enhanced radiation suppresses However, beyond 𝑋 = 0.22, the velocity magnitude remains relatively
convective motion. As radiation strengthens beyond 0.25, convective constant. As 𝑅𝑑 values increase from 0 to 0.4 within the range of
effects regain dominance, leading to a subsequent 5% increase in 0 ≤ 𝑋 ≤ 0.42, there is a 32.24% increase in heat transfer at 𝑋 = 0.41.
vorticity intensity. This transition at 0.25 marks a critical point in the After the previous range of the cut line, the maximum increase in heat
balance between radiation and convection. On the left vertical heated transfer, reaching 1.34%, is observed at 𝑋 = 0.67. As the radiation
surface, the average Nusselt number increases by 9.08% and 11.17% parameters increase, there is a gradual decrease in the magnitude of the
for 𝑅𝑑 = 0.25 and 𝑅𝑑 = 0.4, respectively, compared to 𝑅𝑑 = 0. velocity along the vertical cut line. Hence, the resultant fluid velocity
Fig. 7 shows the Nusselt numbers (local) and the resultant velocity reaches its minimum at Y = 0.35 and maximum at 𝑌 = 0.35. Changes
for the radiation parameter along both horizontal and vertical cutting in thermal conductivity and convective heat transfer characteristics,
lines. As the parameter 𝑅𝑑 increases, there is a gradual increase in influenced by the unique properties of nanofluids and the radiation
the Nusselt number (local) along both the horizontal and vertical cut parameter, impact the local Nusselt number and velocity magnitude of
lines. Along the horizontal cut line, the local Nusselt number exhibits a a nanofluid. As the radiation parameter increases, these factors alter
relatively moderate trend within the range of 0 ≤ 𝑋 ≤ 0.5 and 𝑋 > 0.8. the local Nusselt number and velocity magnitude along the horizontal
On the horizontal cut line at 𝑋 = 0.22, the minimum local Nusselt
and vertical cut lines.
number occurs, while at 𝑋 = 0.65, it reaches the maximum. The Nusselt
number exhibits variation along the vertical cutting line, reaching its
minimum and maximum values at 𝑌 = 0.16 and 𝑌 = 0.34, respectively. 4.3. Impact of 𝐹𝑘 and 𝑅𝑑 on temperature distribution
With an increase in radiation parameter values from 0 to 0.4, the local
Nusselt number at 𝑌 = 0.34 undergoes a substantial 61.36% increase. The panels (a) and (b) of Fig. 8 facilitate a visual inspection of how
The velocity magnitude along the horizontal and vertical cut lines is temperature changes along a specific vertical cut line in response to
shown in panels (c) and (d) of Fig. 7, respectively. In the horizontal variations in 𝐹𝑘 and 𝑅𝑑 values, respectively. This offers valuable insight
cut line, there is a noticeable increase in the velocity magnitude for into how these parameters affect the temperature distribution within
various 𝑅𝑑 values within the range of 0 ≤ 𝑋 ≤ 0.42, except at 𝑋 = 0.22. the system.

8
M.M. Hasan et al. International Journal of Thermofluids 26 (2025) 101114

( ) ( )
Fig. 6. Effect of radiation parameter 𝑅𝑑 on (a) velocity magnitude 𝑉𝑚 [−] , (b) isotherm distribution (𝜒[−]), and (c) vorticity (𝜉[−]) for 𝑅𝑑 = 0, 0.25, and 0.4. Increasing 𝑅𝑑
decreases velocity while enhancing isotherms distribution.

From panel (a) of Fig. 8, it is evident that the temperature decreases (c) display the velocity magnitude (first row), isotherms (second row),
linearly at lower 𝐹𝑘 values, but it takes on an upward parabolic and vorticity function (third row) for three distinct thermal Rayleigh
shape at higher 𝐹𝑘 values. When the value of 𝐹𝑘 reaches 2.7, the numbers. As the thermal Rayleigh number increases from 3 × 104 to 105 ,
isotherms become highly convoluted, exceeding the prevailing tem- there is a substantial 170% increase in the velocity. Additionally, the
perature conditions in the enclosure. The critical Frank–Kamenetskii maximum resultant velocity along the horizontal cut line 𝑌 = 0.27 ex-
( )
number 𝐹𝑘 = 2.7 serves as a pivotal point in the temperature profile periences a remarkable 195% increase. This underscores the significant
of the nanofluid heat transfer analysis, marking a transition from a impact of temperature difference and geometric length on the influence
linear temperature decrease (dominated by conduction) to an upward of the Rayleigh numbers on flow. The panel (b) of Fig. 9 describes that
parabolic shape (dominated by convection), indicating the influence each isotherm line within the cavity delineates the fluid’s ability to
of different heat transfer mechanisms at higher 𝐹𝑘 values. When the transfer thermal heat from the heated surface to the colder one. The
Frank–Kamenetskii number increases from 0 to 6, it leads to an approx- isotherms exhibit minimal impact near the heated and colder walls as
imate 57.1% increase in the average temperature along the vertical cut radiation is relatively low. They are predominantly evenly distributed
line (𝑋 = 0.7, 0.5 ≤ 𝑌 ≤ 1). The influence of radiation parameters on in the center of the domain, signifying a convective thermal transfer
heat transfer becomes evident when examining the temperature profile mode. Throughout the enclosure, the isotherms exhibit weak, consistent
illustrated in panel (b). Significantly, as the values of radiation param- patterns, except near the heated wall, where they run parallel. This
eters rise from 0 to 0.4, temperature distribution gradually increases. alignment means that the conduction mode of thermal transfer prevails
At the endpoint of the cut line (𝑌 = 1), the temperature measures 1.05 on the heated surface. At 𝑅𝑎 = 105 , the isotherms in the middle of
for 𝑅𝑑 = 0 and 1.47 for 𝑅𝑑 = 0.4, representing a 17.61% increase in the cavity become distorted, and they take on a pattern resembling
average temperature as the radiation parameter varies from 0 to 0.4. blooming flowers in the central part of the cavity. This distinct pattern
suggests that the convection mode of heat transfer predominates under
4.4. Influence of thermal Rayleigh numbers these conditions. The decline in isotherm values with an escalation in
the thermal Rayleigh number signifies the growing predominance of
The Rayleigh number significantly influences the flow and the natural convection. The variation in the vorticity function for three
distribution of isotherms and vorticity patterns. Figs. 9(a), (b), and distinct thermal Rayleigh numbers is also substantial, as depicted in

9
M.M. Hasan et al. International Journal of Thermofluids 26 (2025) 101114

( )
Fig. 7. Effects of the thermal radiation parameter 𝑅𝑑 . Panels (a) and (b) depict the local Nusselt number on the horizontal and vertical cutting lines at 𝑌 = 0.27 and 𝑋 = 0.72,
respectively, for various radiation parameters as indicated in each panel. Panels (c) and (d) display the velocity magnitude on the corresponding cross-sectional lines for the same
radiation parameters.

panel (c) of Fig. 9. Generally, as the intensity of Ra escalates, there is critical Rayleigh number for the present study. In panel 10(a), the local
a corresponding increase in the vorticity function. At a lower thermal Nusselt number values initially show an erratic trend, converging at
Rayleigh number, predominant vortices are close to the heated wall, two points along the cut line for all thermal Rayleigh number values.
resulting in lower vorticity functional values. As Ra increases to 3 × 104 , Beyond the midpoint of the cut line, the thermal Rayleigh number
these vortices begin to expand due to the influence of buoyancy forces, increases substantially, enhancing the maximum heat transfer value by
signifying an enhancement in heat transfer rates. Increasing the value 134% at X = 0.65. Heat transfer exhibits frequent fluctuations along
of Ra typically corresponds to a heightened temperature distinction the vertical cut line for all thermal Rayleigh numbers. As the thermal
between the heated and colder surfaces. This amplified temperature Rayleigh number values increase, the local Nusselt number takes on
gradient induces stronger natural convection currents within the fluid. a more regular sinusoidal shape, with the maximum heat transfer rate
These convective motions lead to the formation of vortices and in- occurring at Y = 0.35, amounting to 156%. A similar pattern is observed
creased vorticity within the fluid. As Ra rises from 3 × 104 to 105 , the for the velocity magnitude along the horizontal and vertical cut lines,
vorticity rate exhibits a maximum increase of 229%. Elevated vorticity as depicted in panels (c) and (d) of Fig. 10. The highest heat transfer
in the fluid ensures more efficient mixing of hot and cold regions rates are observed along the horizontal and vertical cut lines at X = 0.65
and improves overall heat transfer. In particular, for Rayleigh numbers
and Y = 0.35, with increases of 192% and 197%, respectively. As the
3 × 104 and 105 , the mean Nusselt number on the vertically heated
thermal Rayleigh number increases, buoyancy forces dominate fluid
surface increases by 8.45% and 46%, respectively, compared to the
flow, creating significant temperature differences within the system.
lower value of 𝑅𝑎 = 103 .
This dominance leads to the formation of convection cells or patterns
Fig. 10 shows the local Nusselt number and the velocity mag-
driven by density differences. The interaction between buoyancy-driven
nitude distribution for various thermal Rayleigh numbers. The local
convection and thermal gradients causes the observed fluctuations and
Nusselt number
( and velocity
) magnitude are weak for lower Rayleigh
gradual increases in local Nusselt numbers and velocity magnitude with
numbers 103 ≤ 𝑅𝑎 ≤ 104 and remain broadly consistent. When the
thermal Rayleigh number reaches 3 × 104 , the local Nusselt number varying thermal Rayleigh numbers.
and velocity become prominent; therefore, this value is considered the

10
M.M. Hasan et al. International Journal of Thermofluids 26 (2025) 101114

( ) ( )
Fig. 8. The temperature profiles for various Frank–Kamenetskii numbers 𝐹𝑘 and radiation parameters 𝑅𝑑 . Panels (a) and (b) display the temperature distribution along the
vertical cut line (𝑋 = 0.7, 0.5 ≤ 𝑌 ≤ 1) corresponding to different values of 𝐹𝑘 , and 𝑅𝑑 respectively.

( )
Fig. 9. Impact of Rayleigh numbers (Ra). The panels (a), (b), and (c) in the first, second, and third rows depict the velocity magnitude 𝑉𝑚 [−] , isotherm distribution (𝜒[−]), and
vorticity function (𝜉[−]), respectively, for three distinct values of 𝑅𝑎 = 103 , 𝑅𝑎 = 3 × 104 , and 𝑅𝑎 = 105 respectively.

11
M.M. Hasan et al. International Journal of Thermofluids 26 (2025) 101114

Fig. 10. Effects of thermal Rayleigh number (Ra). Panels (a) and (b) depict the local Nusselt numbers on the horizontal and vertical cutting lines at 𝑌 = 0.27 and 𝑋 = 0.72,
respectively, for various Rayleigh numbers as indicated in each panel. Panels (c) and (d) display the velocity magnitude on the corresponding cross-sectional lines for the same
Rayleigh numbers.

4.5. Influence of various parameters on 𝑁 𝑢𝑎𝑣𝑔 concerning 𝐹𝑘 and 𝑅𝑑 is identified at 𝑅𝑑 = 0.3. Consequently, at these two values of 𝑅𝑎, the
mean Nusselt number improved by 10.23% and 9.37%, respectively, up
Fig. 11 illustrates that the value of average Nusselt number exhibits to this critical point. Beyond the critical point, heat transfer decreased
a nearly consistent pattern for both 𝐹𝑘 and 𝑅𝑑 parameters
( with )the by 0.11% and 0.54%, for 𝑅𝑑 = 0.35 and 0.4, respectively. The heat
presence of numerous thermal Rayleigh numbers 103 ≤ 𝑅𝑎 ≤ 105 . A transfer rate is improved when the radiation parameter varies from
substantial change in 𝑁 𝑢avg is observed when 𝑅𝑎 varies from 5 × 0 to 0.3 because more heat radiates from the hotter surfaces to the
104 to 105 . As illustrated in Fig. 11(a), with the upsurge in the 𝐹𝑘 cooler surfaces. However, when radiation becomes dominant beyond
value, a modest elevation is observed in the heat transfer rate at the this threshold, it can reduce the effectiveness of convection, leading to
lower thermal Rayleigh number, whereas a noticeable increase in heat a gradual decline in heat transfer as 𝑅𝑑 rises.
transfer is evident at the higher value of 𝑅𝑎. Specifically, as 𝐹𝑘 escalates Fig. 12 shows the effects of the nanoparticle volume fraction on the
from 0 to 6, the heat transfer rate increases by 14.17% for 𝑅𝑎 = 103 average Nusselt number with varying 𝐹𝑘 and 𝑅𝑑 . Based on Fig. 12(a),
and 26.57% for 𝑅𝑎 = 105 . As the Frank–Kamenetskii number rises, the improvement in the heat transfer efficiency observed with the
signifying a greater rate of exothermic chemical reactions, it adds addition of the nanoparticle volume fraction up to 3% may be due
extra heat to the system. The combined impact of buoyancy-driven to the enhanced heat conductivity of nanofluids. However, increased
convection and exothermic chemical reactions collectively enhances viscosity, nanoparticle aggregation, and concentration polarization may
heat transfer within the nanofluids, increasing the average Nusselt reduce heat transport efficiency beyond this threshold.
number. There is a slight initial increase in thermal movement up to Therefore, in the context of the interaction between 𝐹𝑘 and 𝜙,
the radiation parameter, 𝑅𝑑 = 0.1, followed by a moderate decrease it is apparent that 3% of nanoparticle volume fraction is a critical
in heat transfer for lower thermal Rayleigh numbers. The heat transfer threshold in this study. As the value of 𝐹𝑘 upsurges from 0 to 6, the
rate increases to a critical point of 1.35%, after which it declines by value of the average Nusselt number changes by 33%, 31%, and 28%
2.85%. For 𝑅𝑎 = 5 × 104 , the average Nusselt number showed a for the nanoparticle volume fraction of 3%, 4%, and 5%, respectively.
gradual increase without noticeable fluctuations, reaching a peak of The heat transfer decreases for 𝜙 = 0.04 and 𝜙 = 0.05 by 0.61%
10.56%. In contrast, for 𝑅𝑎 = 8 × 104 and 𝑅𝑎 = 105 , a critical point and 3%, respectively, compared to the critical particle volume fraction

12
M.M. Hasan et al. International Journal of Thermofluids 26 (2025) 101114

Fig. 11. Effects of the thermal Rayleigh number on average Nusselt number. Panels (a) and (b) demonstrate the Nusselt number (average) along the left vertical surface concerning
( ) ( )
Frank–Kamenetskii numbers 𝐹𝑘 and radiation parameters 𝑅𝑑 , respectively, across the six distinct values of Rayleigh numbers.

Fig. 12. The nanoparticle volume fraction parameters affect the average Nusselt number. Panels (a) and (b) demonstrate the Nusselt number (average) along the left vertical
( ) ( )
surface concerning Frank–Kamenetskii numbers 𝐹𝑘 and radiation parameters 𝑅𝑑 , respectively, across the six distinct amounts of nanoparticle volume fraction.

and 𝐹𝑘 = 6. The relation between nanoparticle volume fraction and significant changes in the mean Nusselt number for 𝜙 = 0.05, with an
radiation parameters results in varying heat transfer, as displayed in increase of 11%.
Fig. 12(b). The enhanced thermal conductivity governs and increases Fig. 13 illustrates the impact of 𝐹𝑘 and 𝑅𝑑 on the mean Nusselt
heat transfer at lower nanoparticle volume fractions (0 ≤ 𝜙 ≤ 0.03) number concerning the six distinct values of the magnetic field pa-
)
and low radiation parameters ( 0 ≤ 𝑅𝑑 ≤ 0.25 . However, as radiation rameter. The Hartman number signifies the intensity of the magnetic
parameters increase, the dominance of radiation heat transfer can coun- field, with higher values indicating a more powerful magnetic field. A
teract the enhancement caused by nanoparticles. In this case, 𝑅𝑑 = 0.25 stronger magnetic field can generate a Lorentz force within the fluid.
is considered a critical thermal radiation number. For a nanoparticle This force restricts the movement of charged particles present in the
concentration of 3%, the highest heat transfer rate observed up to the fluid. This suppression of fluid motion, induced by the magnetic field,
critical point is 9%, and beyond this point, heat transfer decreases by leads to a noticeable decline in heat transfer, considering both the
0.77% for higher radiation parameters. Remarkably, at a higher value Frank–Kamenetskii numbers and the radiation parameters. As shown in
of 𝜙 (0.04 ≤ 𝜙 ≤ 0.05), the advantages of enhanced thermal conductivity Fig. 13, the value of 𝐹𝑘 increases from 0 to 6, and there is a significant
tend to outweigh the cooling impact of radiation, resulting in a renewed alteration in the mean Nusselt number, with a rate of change amounting
improvement in heat transfer. When 𝑅𝑑 ranges from 0 to 0.4, the most to 26.48% for 𝐻 𝑎 = 0 and 15.43% for 𝐻 𝑎 = 100, respectively. Likewise,

13
M.M. Hasan et al. International Journal of Thermofluids 26 (2025) 101114

Fig. 13. Effects of the magnetic field parameters on the average Nusselt number. Panels (a) and (b) demonstrate the Nusselt number (average) along the left vertical surface
( ) ( )
concerning Frank–Kamenetskii numbers 𝐹𝑘 and radiation parameters 𝑅𝑑 , respectively, across the six distinct values of magnetic field parameters.

with an increase in 𝑅𝑑 , there is a gradual rise in the value of 𝑁 𝑢avg effects, except when 𝛾1 = 0◦ for both 𝐹𝑘 and 𝑅𝑑 parameters. The highest
for lower 𝐻 𝑎 values. However, for higher 𝐻 𝑎 values, an initial slight mean Nusselt number (29%) is attained when using the combination of
upswing in the heat transfer rate is followed by a subsequent diminish. inclined magnetic angles 𝛾1 = 90◦ , 𝛾2 = 30◦ , and 𝛾3 = 45◦ , while varying
As 𝑅𝑑 increases within the range of 0 to 0.4, the most significant the 𝐹𝑘 value from 0 to 6. Considering the value of 𝑅𝑑 increasing from
increase in heat transfer, 11.17%, is observed at 𝐻 𝑎 = 15. Conversely, 0 to 0.4, the highest heat transfer rate (12%) is attained by employing
for 𝐻 𝑎 = 100, the most substantial reduction in heat transfer, 2.87%, the combination of inclined magnetic angles 𝛾1 = 90◦ , 𝛾2 = 30◦ , and
is observed when compared to the critical radiation parameter of 0.1. 𝛾3 = 45◦ .
Panels (a) and (b) in Fig. 14 elucidate how enlarging the radius of The results presented above pertain specifically to copper oxide-
a circular heated cylinder within a square cavity affects heat transfer water nanofluids. However, exploring the behavior of other nanopar-
while considering different values of 𝐹𝑘 and 𝑅𝑑 , respectively. Interest- ticles combined with various base fluids is crucial. Fig. 16 illustrates
ingly, it is observed that as the radius of the inner circular structure the average heat transfer performance, shown as the average Nusselt
increases, the average Nusselt number decreases significantly for both number, for 42 different nanofluids in a bar diagram. These calculations
𝐹𝑘 and 𝑅𝑑 . This phenomenon can be attributed to various factors, utilize the thermophysical properties provided in Table 1. The blue,
primarily disruptions in fluid flow patterns, the formation of boundary orange, and red colors represent water, ethylene glycol, and propylene
layers, and hindrances to perfect mixing. The fluid tends to flow glycol-based nanofluids, respectively. The average Nusselt number is
uniformly within the square enclosure without any obstructions. How- calculated at the bottom heated surface for the corresponding base
ever, introducing a circular heated geometry into this square enclosure fluids and their respective nanofluids. The enhancement is then cal-
disrupts natural flow patterns, leading to a gradual decrease in heat culated by comparing the base fluid with the corresponding nanofluid.
transfer. As the fluid circulates near the circular geometry, it forms It is observed that the CuO nanoparticles provide the maximum aver-
boundary layers on the surface of the geometry, leading to decreased age Nusselt number with the three base fluids, while CoFe2 O4 results
fluid velocity. The presence of these boundary layers and the disruption in the minimum average Nusselt number. The maximum and mini-
of the usual flow patterns can impede the adequate mixing of the fluid mum average Nusselt numbers, 10.66% and 4.23%, are obtained for
within the enclosure. Consequently, these factors can result in uneven CuO-water and Mn − ZnFe2 O4 -water nanofluids, respectively. In this
heating and hinder the efficient dispersion of heat within the enclosure, investigation, the maximum average Nusselt number is not significantly
ultimately decreasing the overall heat transfer efficiency. Increasing high compared to the published papers in the literature [17,19,22],
the intensity of 𝐹𝑘 from 0 to 6 yields a substantial heat transfer rate, due to the combined effects of MHD, thermal radiation, and exothermic
reaching 38% for 𝐿 = 0.1, and the minimal heat transfer rate, which chemical reactions. Water-copper oxide nanofluid exhibits the highest
is 8%, is noted for 𝐿 = 0.35. Additionally, for the lower values of heat transfer due to its high thermal conductivity, significant specific
the radius of the inner heated cylinder (0.1 ≤ 𝐿 ≤ 0.2), the value of heat capacity, and other parameters such as the combined effects
𝑁 𝑢avg progressively rose within the range of 𝑅𝑑 = 0 to 0.4. However, of MHD, thermal radiation, and exothermic chemical reaction. Other
beyond this range, the behavior of 𝑁 𝑢avg reversed; it is increasing for nanofluids, such as ZnO, TiO2 , and Al2 O3 , also display similar heat
lower 𝑅𝑑 values and decreasing for higher 𝑅𝑑 values. According to the transfer rates. This similarity can be attributed to their comparable
present investigation, the highest thermal transfer rate is achieved at thermal conductivities and specific heat capacities. Although these
𝐿 = 0.1, reaching 12%, while considering 𝑅𝑑 values ranging from 0 to nanoparticles have slightly lower thermal conductivities than CuO,
0.4. Conversely, at 𝐿 = 0.35, there is an initial 1.13% growth in 𝑁 𝑢𝑎𝑣𝑔 the differences do not create a significant disparity in heat transfer
at lower 𝑅𝑑 values, but as 𝑅𝑑 values rise, the heat transfer rate reduces performance. These nanofluids’ uniformity in nanoparticle dispersion
to an extreme of 12%. and stability further contributes to their similar heat transfer charac-
The effects of the magnetic field orientation angles on average teristics. Fig. 16 shows that ethylene glycol (EG) and propylene glycol
Nusselt number are depicted in Fig. 15. A consistent heat transfer (PG) are base fluids that yield results identical to water. The present
pattern is observed for various changes in inclined magnetic field investigation shows that CuO-water nanofluids can greatly increase

14
M.M. Hasan et al. International Journal of Thermofluids 26 (2025) 101114

Fig. 14. Effects of the magnetic field parameters on the average Nusselt number. Panels (a) and (b) demonstrate the Nusselt number (average) along the left vertical surface
( ) ( )
concerning Frank–Kamenetskii numbers 𝐹𝑘 and radiation parameters 𝑅𝑑 , respectively, across the six distinct values of magnetic field parameters.

Fig. 15. Effects of the inclined angle of the magnetic field on the average Nusselt number. Panels (a) and (b) demonstrate the Nusselt number (average) along the left vertical
( ) ( )
surface concerning Frank–Kamenetskii numbers 𝐹𝑘 and radiation parameters 𝑅𝑑 , respectively, across the six distinct angles of inclined magnetic field.

thermal performance in 3D convective systems, such as electronic cool- Incorporating temperature-dependent properties and particle shape fac-
ing and heat exchangers. Insights into the effects of thermal radiation, tors [58] into simulations could produce more precise heat transfer
MHD, and exothermic reactions highlight their potential for optimiz- and fluid flow analyzes, which we intend to examine in subsequent
ing energy generation, aerospace, and industrial processes, emphasiz- studies. Furthermore, future research can seek to use direct numerical
ing the benefits of nanofluids for efficient, high-performance applica- simulation at the particle level to more thoroughly explore the im-
tions. pact of particle–particle interactions, magnetic forces, and many-body
hydrodynamics on the heat transfer and flow behavior of nanofluids.
4.6. Future research

Experimental investigations [56,57] indicate that specific thermo- 5. Conclusions

physical characteristics of nanofluids, such as viscosity, are intimately
linked to temperature. Viscosity and thermal conductivity are also The flow of nanofluids in an annulus is numerically investigated
strongly influenced by the shape of particles, which can play a key using the finite element numerical method. The main results can be
role in improving the heat transfer and flow behavior in nanofluids. summarized as follows.

15
M.M. Hasan et al. International Journal of Thermofluids 26 (2025) 101114

Fig. 16. Enhancement of the average Nusselt number for 14 different nanoparticles compared to their base fluids (water, ethylene glycol, and propylene glycol), with a nanoparticle
volume fraction of 3% for each case . (For interpretation of the references to color in this figure legend, the reader is referred to the web version of this article.)

• For CuO-water nanofluids, the velocity magnitude tends to in- nanoparticles result in the lowest average Nusselt number.
crease with increases in 𝑅𝑎 and 𝐹𝑘 , while it decreases as the
thermal radiation parameters increase.
CRediT authorship contribution statement
• The isotherms intensified with increasing values of 𝐹𝑘 and 𝑅𝑑 ,
but gradually decreased as the thermal Rayleigh number 𝑅𝑎
Md. Mehedi Hasan: Writing – review & editing, Writing – original
increases.
draft, Validation, Methodology, Investigation, Formal analysis, Con-
• Increasing 𝐹𝑘 from 0 to 6 resulted in an approximately 57%
ceptualization. M.J. Uddin: Writing – review & editing, Visualization,
increase in the average temperature along the vertical cut line,
Validation, Supervision, Methodology, Investigation, Formal analysis,
leading to a 28% increase in the heat transfer rate along the left
Conceptualization. Salah A. Faroughi: Writing – review & editing,
vertical surface.
Visualization, Validation, Supervision, Methodology, Formal analysis,
• The vortex intensity was relatively weak with lower values of 𝐹𝑘
Conceptualization.
and 𝑅𝑑 , but became robust and extended throughout the entire
domain as 𝐹𝑘 and 𝑅𝑑 increase.
Declaration of competing interest
• The vorticity function becomes increasingly prominent with rising
𝑅𝑎, achieving the highest heat transmission rate of 46% at 𝑅𝑎 =
The authors declare that they have no known competing finan-
105 .
cial interests or personal relationships that could have appeared to
• The critical thermal Rayleigh number in this study has been
influence the work reported in this paper.
observed as 𝑅𝑎 = 3 × 104 .
• A substantial improvement in the local Nusselt number was ob-
Acknowledgments
served along the horizontal and vertical cut lines, with enhance-
ments of 134% and 156%, respectively, compared to the critical
S.A.F. would like to acknowledge supports by American Chemical
value.
Society Petroleum Research Fund, USA (award no. 66452-DNI9) and
• With an increase in 𝐹𝑘 values from 0 to 6, there was a noticeable
National Science Foundation Partnership for Research and Education
increase in the mean Nusselt number when the thermal Rayleigh
in Materials (PREM), USA (award no. DMR-2122041).
number and the volume fraction of nanoparticles increased. Con-
versely, the mean Nusselt number decreased as the characteristic
Appendix
length of the geometry and the Hartmann number increased.
• Optimal heat transfer is observed when the volume fraction of
CuO nanoparticles in the base fluid is 3%, the characteristic Nomenclature
length of the geometry is 0.1 m, and magnetic field angle ori- [ ]
entations of 𝛾1 = 45◦ , 𝛾2 = 45◦ , and 𝛾3 = 90◦ . 𝐴 Area of the geometry m2
• The heat transfer rate gradually increased when the radiation [ ]
parameters were relatively low but decreased as the thermal 𝑎 Reactant density mol∕m3
radiation value increased. √ √
• At the critical Rayleigh number 𝑅𝑎 = 105 , a 9.37% increase in 𝐵 Applied magnetic field [ Ω k g∕m s]
heat transfer is observed as the radiation parameter 𝑅𝑑 increases
𝐸 Energy [J∕mol]
from 0 to 0.3. However, beyond this point, there is a slight 0.54%
decrease in heat transfer at 𝑅𝑑 = 0.4. [ ]
𝑔 Gravitational force m∕s2
• Among the 42 nanofluids investigated for optimal heat transfer,
CuO (Copper Oxide) nanoparticles exhibit the highest average 𝐿 Characteristic geometric length [m ]
Nusselt number across the three base fluids (water, ethylene [ ]
glycol, and propylene glycol), while CoFe2 O4 (Cobalt Ferrite) 𝑝 Pressure k g∕ms2

16
M.M. Hasan et al. International Journal of Thermofluids 26 (2025) 101114

√ √
𝐵0 Magnetic field strength [ Ω k g∕m s] Greek symbols

MHD Magnetohydrodynamics 𝛾1,2,3 Sloping angle of magnetic field

[ ]
𝑐𝑝 Specific heat [J∕k g K ] 𝜌 Density k g∕m3
[ ]
𝑞 Heat flux W∕m2 𝜅 Conductivity of fluid [J∕m K s] or [W∕m K ]
[ ]
𝑄 Chemical reaction exothermicity [J∕mol] 𝛼 Diffusivity parameter (thermal) m2 ∕s

[ ] [ ]
𝑅 Constant of fluid J K −1 mol−1 𝑣 Viscosity (kinematic) m2 ∕s

𝜙 Nanoparticle’s concentration
𝑇 Fluid heat [K ]
𝜌𝛽 ∗ Expansion of mass (volumetric)
𝐮 Component of velocity [𝐮(𝐮, 𝐯, 𝐰)] [m∕s]
𝛽 Expansion coefficient (thermal) [1∕K ]
𝐱 Coordinates [𝐱(𝐱, 𝐲, 𝐳)] [m]
𝜎 Electrical conductivity [1∕Ω m]
𝐄∗ Electric field [V∕m]
[ ]
[ ] 𝜎 ∗ Stefan–Boltzmann constant W m−2 K −4
𝐉𝐄∗ Electric current density A∕m2
𝜃 Non-dimensional heat
k0 Pre-exponential coefficient [1∕s]
𝜇 Viscosity [k g∕ms]
𝑘∗ Average absorption factor [1∕m] [ ]
𝜌𝑐𝑝 Capacitance of heat J∕m3 K
𝑇𝑐 Low temperature [K ]
𝜌𝛽 Expansion of thermal (volumetric)
𝑇ℎ Warmed temperature [K ]
𝜆 Conductivity of fluid [W∕m K ]

𝜉 Vorticity (dimensionless)
Non-dimensional parameters
𝜒 Isotherm (dimensionless)
𝐹𝑘 Frank–Kamenetskii number
𝛩, 𝛹 Angle of the parametric curve
Ha Hartmann number

𝑁 𝑢avg Nusselt number (average) Data availability

Data will be made available on request.

𝑁 𝑢𝐿 Nusselt number (local)

𝑃 Pressure References

𝑅𝑎 Rayleigh number [1] Sajjad Porgar, Gabriela Huminic, Angel Huminic, Roznaz Najibolashrafi, So-
mayeh Salehfekr, Application of nanofluids in heat exchangers-a state-of-the-art
𝑅𝑑 Radiation parameter review, Int. J. Thermofluids (2024) 100945.
[2] Sascha Becker, Eckart Laurien, Three-dimensional numerical simulation of flow
Pr Prandtl number and heat transport in high-temperature nuclear reactors, Nucl. Eng. Des. 222
(2–3) (2003) 189–201.
[3] Kaiyuanwen Liu, Yue Sun, Yuxuan Liu, Lin Li, Caili Dai, Preparation of carbon-
𝐕𝐦 Velocity magnitude based active nanofluids with temperature and salt resistance for enhanced oil
recovery, J. Mol. Liq. 394 (2024) 123685.
𝐔 Velocity component [𝐔(𝐔, 𝐕, 𝐖)] [4] Gao Li, Mou Yang, Yingfeng Meng, Zhiming Wen, Yanmin Wang, Zhongtao Yuan,
Transient heat transfer models of wellbore and formation systems during the
𝐗 Cartesian coordinates [𝐗(𝐗, 𝐘, 𝐙)] drilling process under well kick conditions in the bottom-hole, Appl. Therm.
Eng. 93 (2016) 339–347.
[5] Md Jashim Uddin, Md Mehedi Hasan, Salah A. Faroughi, Convective transport
in a nanofluid-filled annulus with an eccentric hypocycloid within a square disk,
Subscripts Int. J. Thermofluids 22 (2024) 100664.
[6] S.A. Logtenberg, M. Nijemeisland, Anthony G. Dixon, Computational fluid
dynamics simulations of fluid flow and heat transfer at the wall–particle contact
ℎ High
points in a fixed-bed reactor, Chem. Eng. Sci. 54 (13–14) (1999) 2433–2439.
[7] Wen Ao, Zhimin Fan, Yi Gao, Yang Wang, Peijin Liu, Larry KB Li, Ignition and
𝑐 Low combustion characteristics of boron-based nanofluid fuel, Combust. Flame 254
(2023) 112831.
𝑠 Solid particle [8] K.R. Aglawe, R.K. Yadav, S.B. Thool, Preparation, applications and challenges
of nanofluids in electronic cooling: A systematic review, Mater. Today: Proc. 43
𝑛𝑓 Nanofluid (2021) 366–372.
[9] Hakan F. Öztop, Patrice Estellé, Wei-Mon Yan, Khaled Al-Salem, Jamel Orfi,
Omid Mahian, A brief review of natural convection in enclosures under localized
𝑏𝑓 Base fluid heating with and without nanofluids, Int. Commun. Heat Mass Transfer 60 (2015)
37–44.

17
M.M. Hasan et al. International Journal of Thermofluids 26 (2025) 101114

[10] D Thenmozhi, M Eswara Rao, PD Selvi, RLV Renuka Devi, S Kiranmaiye, [32] Shaik Jakeer, P BalaAnki Reddy, AM Rashad, Hossam A Nabwey, Impact of
Ch Nagalakshmi, The significance of magnetized thermal radiation on the heated obstacle position on magneto-hybrid nanofluid flow in a lid-driven porous
magnetohydrodynamic (MHD) behavior of Williamson hybrid ferrofluids over cavity with Cattaneo-Christov heat flux pattern, Alex. Eng. J. 60 (1) (2021)
a stretching sheet, Int. J. Thermofluids 25 (2025) 100997. 821–835.
[11] Mojtaba Sepehrnia, Ghanbarali Sheikhzadeh, Golnoush Abaei, Mahdi Motame- [33] Mohsen Sheikholeslami, Houman B. Rokni, Simulation of nanofluid heat transfer
dian, Study of flow field, heat transfer, and entropy generation of nanofluid in presence of magnetic field: a review, Int. J. Heat Mass Transfer 115 (2017)
turbulent natural convection in an enclosure utilizing the computational fluid 1203–1233.
dynamics-artificial neural network hybrid method, Heat Transfer— Asian Res. [34] Wasim Jamshed, Numerical investigation of MHD impact on Maxwell nanofluid,
48 (4) (2019) 1151–1179. Int. Commun. Heat Mass Transfer 120 (2021) 104973.
[12] Rahman Saidur, K.Y. Leong, Hussein A. Mohammed, A review on applications [35] Seyyed Masoud Seyyedi, AS Dogonchi, M Hashemi-Tilehnoee, M Waqas, DD
and challenges of nanofluids, Renew. Sustain. Energy Rev. 15 (3) (2011) Ganji, Investigation of entropy generation in a square inclined cavity using con-
1646–1668. trol volume finite element method with aided quadratic Lagrange interpolation
functions, Int. Commun. Heat Mass Transfer 110 (2020) 104398.
[13] O Anwar Beg, FT Zohra, MJ Uddin, AIM Ismail, Saratha Sathasivam, Energy
[36] Hadi Kargarsharifabad, Experimental and numerical study of natural convection
conservation of nanofluids from a biomagnetic needle in the presence of stefan
of Cu-water nanofluid in a cubic enclosure under constant and alternating
blowing: Lie symmetry and numerical simulation, Case Stud. Therm. Eng. 24
magnetic fields, Int. Commun. Heat Mass Transfer 119 (2020) 104957.
(2021) 100861.
[37] Chemseddine Maatki, Kaouther Ghachem, Lioua Kolsi, Ahmed Kadhim Hussein,
[14] R.D. Jilte, Ravinder Kumar, Mohammad H. Ahmadi, Cooling performance of
Mohamed Naceur Borjini, Habib Ben Aissia, Inclination effects of magnetic field
nanofluid submerged vs. nanofluid circulated battery thermal management
direction in 3D double-diffusive natural convection, Appl. Math. Comput. 273
systems, J. Clean. Prod. 240 (2019) 118131.
(2016) 178–189.
[15] Lin Chen, Xin-Rong Zhang, Natural convection supercritical fluid systems for [38] Pablo Marín, Fernando V Díez, Salvador Ordóñez, Reverse flow reactors
geothermal, heat transfer, and energy conversion, in: Energy Solutions to Combat as sustainable devices for performing exothermic reactions: Applications and
Global Warming, Springer, 2017, pp. 391–433. engineering aspects, Chem. Eng. Process.- Process. Intensif. 135 (2019) 175–189.
[16] Ahmad Hajatzadeh Pordanjani, Saeed Aghakhani, Masoud Afrand, Mohsen Shar- [39] Mikhail A. Sheremet, Ioan Pop, Alin V. Rosca, The influence of thermal radiation
ifpur, Josua P Meyer, Huijin Xu, Hafiz Muhammad Ali, Nader Karimi, Goshtasp on unsteady free convection in inclined enclosures filled by a nanofluid with
Cheraghian, Nanofluids: Physical phenomena, applications in thermal systems sinusoidal boundary conditions, Internat. J. Numer. Methods Heat Fluid Flow
and the environment effects-a critical review, J. Clean. Prod. 320 (2021) 128573. 28 (8) (2018) 1738–1753.
[17] S.O. Giwa, M. Sharifpur, M.H. Ahmadi, J.P. Meyer, A review of magnetic field [40] P. Sudarsana Reddy, P. Sreedevi, Effect of thermal radiation and volume fraction
influence on natural convection heat transfer performance of nanofluids in square on carbon nanotubes based nanofluid flow inside a square chamber, Alex. Eng.
cavities, J. Therm. Anal. Calorim. 145 (2021) 2581–2623. J. 60 (1) (2021) 1807–1817.
[18] Ahmed Kadhim Hussein, Mohamed M Awad, Lioua Kolsi, Farshid Fathinia, [41] Md Mehedi Hasan, M.J. Uddin, R. Nasrin, Exothermic chemical reaction of
IK Adegun, A comprehensive review of transient natural convection flow in magneto-convective nanofluid flow in a square cavity, Int. J. Thermofluids 16
enclosures, J. Basic Appl. Sci. Res. 4 (11) (2014) 17–27. (2022) 100236.
[19] Zainab Kareem Ghoben, Ahmed Kadhim Hussein, Three-dimensional analysis of [42] Farnaz Sanei, Dheyaa J Jasim, Soheil Salahshour, Omid Ali Akbari, Nafiseh
the thermal behavior of alumina-water nanofluid inside hexagonal and octagonal Emami, et al., Numerical simulation of the nanofluid flow and heat transfer in
enclosed domain, J. Adv. Res. Fluid Mech. Therm. Sci. 103 (1) (2023) 40–63. porous microchannels with different flow path arrangements using single-phase
[20] Sara Rostami, Saeed Aghakhani, Ahmad Hajatzadeh Pordanjani, Masoud Afrand, and two-phase models, Int. J. Thermofluids 24 (2024) 100846.
[43] Mohammad N Fares, AL-Saad Mohammad, Heider HJ Almutter, Dheyaa J Jasim,
Goshtasp Cheraghian, Hakan F Oztop, Mostafa Safdari Shadloo, A review on the
Mohammad Ali Fazilati, Soheil Salahshour, Sh Baghaei, Two-phase analysis of
control parameters of natural convection in different shaped cavities with and
heat transfer of nanofluid flow in a wavy channel heat exchanger: A numerical
without nanofluid, Processes 8 (9) (2020) 1011.
approach, Int. J. Thermofluids 23 (2024) 100786.
[21] Dheyaa J Jasim, Seyed Hossein Hashemi Karouei, Thamer J Mohammed,
[44] Ali Azad, Ali Akbar Abbasian Arani, Ali Arefmanesh, Rahim Shamsoddini,
Ahmed Salah Al-Shati, Hydrothermal behaviour of hybrid nanofluid flow in two
Analysis of nanofluid flow and heat transfer inside a channel with smoothed
types of shell and helical coil tube heat exchangers with a new design. Numerical
particle hydrodynamics, Int. J. Thermofluids 25 (2025) 100995.
approach, Int. J. Thermofluids 24 (2024) 100902.
[45] Saba Y Ahmed, Qusay Rasheed Al-Amir, Hameed K Hamzah, Farooq H Ali,
[22] Bouchmel Mliki, Mohamed Ammar Abbassi, Ahmed Omri, Zeghmati Belkacem, Azher M Abed, Ahmed Al-Manea, Karim Egab, Raed Al-Rbaihat, Khalid Saleh,
Lattice Boltzmann analysis of MHD natural convection of CuO-water nanofluid in Ali Alahmer, Investigation of natural convection and entropy generation of
inclined C-shaped enclosures under the effect of nanoparticles Brownian motion, non-Newtonian flow in molten polymer-filled odd-shaped cavities using finite
Powder Technol. 308 (2017) 70–83. difference lattice Boltzmann method, Numer. Heat Transfer B (2024) 1–26.
[23] Ammar I Alsabery, Ali S Abosinnee, Saleem K Al-Hadraawy, Muneer A Ismael, [46] Zaid Al-Dulaimi, Hakim T Kadhim, Malik F Jaffer, Ahmed Al-Manea, Raed Al-
Mehdi A Fteiti, Ishak Hashim, Mikhail Sheremet, Mohammad Ghalambaz, Ali J Rbaihat, Ali Alahmer, Enhanced conjugate natural convection in a corrugated
Chamkha, Convection heat transfer in enclosures with inner bodies: A review on porous enclosure with Ag-MgO hybrid nanofluid, Int. J. Thermofluids 21 (2024)
single and two-phase nanofluid models, Renew. Sustain. Energy Rev. 183 (2023) 100574.
113424. [47] Raj Kamal Tiwari, Manab Kumar Das, Heat transfer augmentation in a two-sided
[24] Ammar I Alsabery, Muneer A Ismael, Engin Gedik, Ali J Chamkha, Ishak lid-driven differentially heated square cavity utilizing nanofluids, Int. J. Heat
Hashim, Transient nanofluid flow and energy dissipation from wavy surface Mass Transfer 50 (9–10) (2007) 2002–2018.
using magnetic field and two rotating cylinders, Comput. Math. Appl. 97 (2021) [48] Elaine Fabre, S.M. Sohel Murshed, A comprehensive review of thermophysical
329–343. properties and prospects of ionanocolloids in thermal energy applications, Renew.
[25] Ammar I Alsabery, Muneer A Ismael, Ali J Chamkha, Ishak Hashim, Effect of Sustain. Energy Rev. 151 (2021) 111593.
nonhomogeneous nanofluid model on transient natural convection in a non-Darcy [49] Hakan F. Oztop, Eiyad Abu-Nada, Numerical study of natural convection in
porous cavity containing an inner solid body, Int. Commun. Heat Mass Transfer partially heated rectangular enclosures filled with nanofluids, Int. J. Heat Fluid
110 (2020) 104442. Flow 29 (5) (2008) 1326–1336.
[26] Muneer A. Ismael, Fatih Selimefendigil, Ali J. Chamkha, Mixed convection in [50] B. Ghasemi, S.M. Aminossadati, A. Raisi, Magnetic field effect on natural
a vertically layered fluid-porous medium enclosure with two inner rotating convection in a nanofluid-filled square enclosure, Int. J. Therm. Sci. 50 (9)
cylinders, J. Porous Media 20 (6) (2017). (2011) 1748–1756.
[51] Olgier Cecil Zienkiewicz, Robert Leroy Taylor, Finite Element Method: Vol. 3:
[27] Kamel Milani Shirvan, Rahmat Ellahi, Mojtaba Mamourian, Mohammad Moghi-
Fluid Dynamics, Elsevier Science & Technology Books, 2000.
man, Effects of wavy surface characteristics on natural convection heat transfer
[52] Ramon Codina, Comparison of some finite element methods for solving the
in a cosine corrugated square cavity filled with nanofluid, Int. J. Heat Mass
diffusion-convection-reaction equation, Comput. Methods Appl. Mech. Engrg. 156
Transfer 107 (2017) 1110–1118.
(1–4) (1998) 185–210.
[28] C.J. Ho, W.K. Liu, Y.S. Chang, C.C. Lin, Natural convection heat transfer of
[53] Md Jashim Uddin, M.M. Rahman, Finite element computational procedure for
alumina-water nanofluid in vertical square enclosures: An experimental study,
convective flow of nanofluids in an annulus, Therm. Sci. Eng. Prog. 6 (2018)
Int. J. Therm. Sci. 49 (8) (2010) 1345–1353.
251–267.
[29] A. Tanase, D.C. Groeneveld, An experimental investigation on the effects of flow [54] C. Wan, B.S.V. Patnaik, G.W.D. Wei, A new benchmark quality solution for the
obstacles on single phase heat transfer, Nucl. Eng. Des. 288 (2015) 195–207. buoyancy-driven cavity by discrete singular convolution, Numer. Heat Transfer:
[30] Nilankush Acharya, On the flow patterns and thermal control of radiative natural Part B: Fundam. 40 (3) (2001) 199–228.
convective hybrid nanofluid flow inside a square enclosure having various shaped [55] Graham de Vahl Davis, Natural convection of air in a square cavity: a bench mark
multiple heated obstacles, Eur. Phys. J. Plus 136 (8) (2021) 889. numerical solution, Internat. J. Numer. Methods Fluids 3 (3) (1983) 249–264.
[31] Z.H. Khan, W.A. Khan, M. Hamid, Hongtao Liu, Finite element analysis of hybrid [56] Muhammad Tawalbeh, Ibrahim Shomope, Amani Al-Othman, Comprehen-
nanofluid flow and heat transfer in a split lid-driven square cavity with Y-shaped sive review on non-Newtonian nanofluids, preparation, characterization, and
obstacle, Phys. Fluids 32 (9) (2020). applications, Int. J. Thermofluids (2024) 100705.

18
M.M. Hasan et al. International Journal of Thermofluids 26 (2025) 101114

[57] Sarit Kumar Das, Nandy Putra, Peter Thiesen, Wilfried Roetzel, Temperature [58] A. Yahyaee, Influence of nanoparticle shapes in nanofluid film boiling on vertical
dependence of thermal conductivity enhancement for nanofluids, J. Heat Transf. cylinders: A numerical study, Int. J. Thermofluids 22 (2024) 100631.
125 (4) (2003) 567–574.

19