Research Article Heat Transfer and Performance Enhancement of Porous Split Elliptical Fins

Hindawi
International Journal of Energy Research

Volume 2023, Article ID 9206017, 26 pages
https://doi.org/10.1155/2023/9206017
Research Article
Heat Transfer and Performance Enhancement of Porous Split
Elliptical Fins
Alok Ranjan,1 Ranjan Das ,2 Debabrata Barik ,3 Sagnik Pal ,4 Arindam Majumder,4
Madhujit Deb,4 and Milon Selvam Dennison 5
1
Department of Mechanical Engineering, Government Engineering College Khagaria, Bihar 848203, India
2
Department of Mechanical Engineering, Indian Institute of Technology Ropar, Punjab 140001, India
3
Department of Mechanical Engineering, Karpagam Academy of Higher Education, Coimbatore, India 641021
4
Department of Mechanical Engineering, National Institute of Technology Agartala, Tripura 799046, India
5
Department of Mechanical Engineering, Kampala International University Western Campus, Kampala 20000, Uganda
Correspondence should be addressed to Debabrata Barik; debabrata93@gmail.com

and Milon Selvam Dennison; milon.selvam@kiu.ac.ug
Received 8 March 2023; Revised 30 June 2023; Accepted 10 July 2023; Published 24 July 2023
Academic Editor: Guojun Yu
Copyright © 2023 Alok Ranjan et al. This is an open access article distributed under the Creative Commons Attribution License,
which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.
To augment the heat transfer phenomenon, the infusion of various fin geometry over the heated plate is being investigated by various
researchers. Solid fins of the porous medium can enhance the convective heat transfer, providing a higher surface area-to-volume ratio
for heat transfer. In this work, the fluid flow pattern and thermodynamic analysis of porous-based split elliptical fins mounted
staggered over a heated base plate is numerically studied with the Reynolds numbers in the range of 783 to 1839, which is
dependent on the fin dimension. The variable parameters were dimensionless transverse offset (TO∗ = transverse offset/diameter),
which varied from 0 to 0.5; dimensionless longitudinal offset (LO∗ = longitudinal offset/diameter), which varied from 0 to 0.25;
porosity (ɸ), which varies from 0.8 to 0.92; pores per inch (PPI), which was 10; permeability (Pn ); and inertial parameter (F). To
count the viscous and inertial effect inside the porous zone, the Forchheimer–Brinkman extended Darcy model was adopted. The
associated parameters, the Nusselt number (Nu), frictional coefficient (C f ), and performance evaluation criteria (PEC), are
thoroughly analyzed over the TO∗ and LO∗ combination. The results of the investigation revealed that the highest value of Nu and
PEC was obtained by TO∗ = 0:5 and LO∗ = 0 at ɸ = 0:92, which were approximately 54% and 79% higher than the solid circular
fin at Re = 1839. Additionally, a response function based on Nu was obtained using the response surface method, and cuckoo
optimization was assessed to identify the optimal Nu. The optimal Nu is established at TO∗ = 0:4141 and LO∗ = 0 (ɸ = 0:92 and
Re = 1839) and was validated with the present investigation with an accuracy of 1.20%.
1. Introduction surface area to volume ratio, the tortuous path for better
intermixing fluid, and are lighter in weight, which may
Due to the major advancement in manufacturing sciences, alleviate the convective heat transfer physics. Jeng et al. [1]
porous medium-based extended surfaces may fulfill the have performed an analytical study on porous heat sinks
requirement of quick thermal dissipation. Certain thermally using fin theory. The local thermal equilibrium condition
linked devices, such as heat exchangers, passive heat sinks, is chosen and suitable for the porosity range, the Reynolds
automotive radiators, EGR (exhaust gas recirculation) units, number, and height of the porous fins. Orhan [2] has
and the shell of lithium-ion battery packages need to quickly performed a numerical investigation on the porous ring attached
release heat to the surrounding. In that manner, fins are con- to the circular conduit. Results show that using a porous-based
sidered as it is a cheaper and easier way to dissipate thermal ring can generate a higher level of turbulence, which augments
energy. The porous medium-based fins can provide a higher the heat transfer phenomenon. Leong and Jin [3] have carried
2 International Journal of Energy Research
out an experimental study for a metal foam-based heat sink for the microscopic method provides the optimum diameter to
the enhanced heat transfer rate. It has been found that higher alleviate the heat transfer rate. Madhavan et al. [14] investi-
pore-density metal foam can only be used where pumping gated the thermal and flow behavior of high porosity-based
power is not a big constraint; otherwise, lower pore-density metal foam coupled with flow impingement. Results reveal
metal foam is recommended for a higher transfer of heat. To that the coefficient of heat transfer for fully foam condition
estimate the impact of the porous structure in the double- is much higher than the smooth case, and further upon cou-
pipe heat exchanger, Targui and Kahalerras [4] carried out pling, impinging jet with metal foam alleviates the coefficient
numerical research. It has been observed that, at lower perme- heat transfer in a much more extensive way for the given range
ability, the heat transfer rate is higher than that of a higher of the Reynolds number (Re). A rectangular channel has been
permeability-based porous structure. It is mainly due to a employed in the numerical analysis by Ranjan et al. [15] to
higher amount of recirculation phenomenon attached to the examine the performance of drop- and split-drop-shaped fins.
low permeability-based porous medium. Shuja et al. [5] con- The results indicate that drop-shaped fins are better at defer-
ducted a numerical study to gauge the effect of increasing ring or repressing flow pass-through, whereas splitting causes
porosity at different heat flux values for a porous block the supplemental area to increase and strengthen the convec-
attached to a square cavity enclosure. It has been seen that tive heat transfer phenomenon. To comprehend the thermal
Nu gets alleviated with incrementing the porosity value of and flow performance of split elliptic-shaped fins, Ranjan
the porous block, i.e., upon increasing porosity, the flow et al. [16] carried out a numerical study. Results reveal the aug-
emerging out of the porous block modifies the flow field and mentation in Nu upon splitting the fin in dimensionless trans-
contributes to the convection phenomenon. verse (TO∗ ) and longitudinal offset (LO∗ ). Further cuckoo
Li et al. [6] performed a numerical investigation to search optimization technique is used to find out the optimum
analyze the thermal-cum-flow behavior of porous blocks Nu, and a correlation has been developed to find out the value
arranged in a staggered manner inside a conduit. Results of the Nu number corresponding to the dimensionless offsets
reveal that lower value permeability associates more vortices and Re.
at the rear end of the fins, which enhances the rate of heat Kim et al. [17] have analyzed the temperature distribu-
dissipation. Mehrizi et al. [7] carried out numerical work tion in the porous medium-based heat sink using one and
to gauge the impact of fin position with an inserted porous two-equation models under local equilibrium conditions. It
medium in a heat exchanger. It has been observed that, upon has been perceived that, as the value of the Darcy number
increasing the porosity of the porous medium, the increment decreases or the effective thermal conductivity (k f ) ratio
and decrement of the average Nusselt number (Nu) depend increases, the fluid temperature approaches the solid tem-
upon the fin position. Khaled et al. [8] experimented to ana- perature and one equation model with local thermal equilib-
lyze the thermal impact of metal foam encased over a cylin- rium assumption becomes valid. Li et al. [18] investigated
der. Results show that a higher porosity-based porous layer the performance of metal foam, solid pin-fin, and combined
provides a higher value of the Nusselt number. Khaled [9] metal foam-solid pin-fin heat sinks numerically. The local
has done analytical studies on pin-fin with a core filled with nonequilibrium assumption has been preferred, and the
porous media. It has been found that thermal efficiency results show that the hybrid, i.e., metal foam with solid
increases with an increase in porosity value. Hao et al. [10] pin-fin, outperforms the other heat sinks. Zhao et al. [19]
have done an experimental investigation for elongated circu- have performed a numerical study on porous-based fins
lar pores with a porous copper plate mounted over a heated mounted over heat exchangers with varying pore density
surface. It has been demonstrated that foam with higher and porosity values. It has been seen that, as the value of
porosity values exhibits higher heat dissipation capabilities pore density increases from 5 to 50 pores per inch (PPI),
than foam with lower porosity values. Morkos et al. [11] the value of the performance evaluation factor also improves
conducted an experimental investigation to visualize the and gets reduced over the increase in porosity from 0.8 to
effect of porosity, pore size, and ligament size for metal foam 0.975. Huang and Vafai [20] have numerically analyzed the
under fluid flow conditions in a wind tunnel. It has been thermal and fluid flow behavior of porous blocks placed in
noticed that below a threshold velocity, the fluid cannot a rectangular channel. The Brinkman-Forchheimer-Darcy
penetrate through the metal foam, and the downstream model has opted to account for the inertial and viscous effect
velocity of the metal foam increases with an increase in of the flow through the porous block. The results demon-
porosity value. strate that the value of Nu increases upon decreasing the
Ramakrishna et al. [12] performed an experimental study Darcy number and increasing the value of both Re and
to gauge the effect of metal foam orientation and pore density inertial parameter (F).
(at 93% porosity) over the thermal performance factor under Yang et al. [21] used numerical analysis to determine the
buoyancy-induced convective conditions. Results show that effect of changes in permeability and pin-fin cross-section
at lower pore density, the metal foam yields higher thermal shape in a rectangular channel. It has been perceived that,
performance compared to higher pore density metal foam. as the value of permeability decreases, the value of heat flux
Upalkar et al. [13] conducted an experimental and numerical and pressure drop increases inside the porous zone. It is
study over a z-shaped corrugated perforated plate heat sink mainly due to the quick increase in the viscosity effect upon
using the macroscopic and microscopic approaches. The decreasing permeability. Calmidi and Mahajan [22] have
macroscopic method provides the optimum number of the performed both experimental and numerical work to see
perforated plate under pressure drop constraints, whereas the effect of thermal dispersion in high porosity (0.89-0.97)
International Journal of Energy Research 3
Wd
21.21 mm
Flow out
Left side
Ld 10.60 mm
Hd
ϕ𝛿
e
sid
om
Bott
hf
e
sid
Tip lb
de tb
m
t si
118 m
z h
y Rig
Flow in
x
118 mm
(a) (b)
Transverse offset
Longitudinal offset
118 mm
(c)
Figure 1: Schematic diagram of the (a) computational domain, (b) porous-based elliptical fin-base plate setup, and (c) elliptical fin with
longitudinal and transverse offset.
based metal foam under forced convection conditions. The Table 1: Computational boundary conditions.
results reveal that the thermal dispersion is lower in foam- !
air combination, whereas it is highly accountable with bulk Flow in Flow inlet velocity, V isothermal
transportation for foam-water combination. Fu et al. [23] temperature T 1
performed a numerical investigation to see the thermal Flow outlet pressure P; isothermal
energy and flow transport for porous blocks under the lam- Flow out temperature T 2
inar flow regime. It has been found that the augmentation in !
Left, right, tip, and No slip on the wall; V = 0; isothermal
thermal performance increases with higher porosity and
bottom wall sides temperature T 1
pore diameter. Ko and Anand [24] conducted experiments
! !
to investigate the porous-based baffle effect on heat transfer Heated base plate Heat flux; q z = constant Qz
in a rectangular channel. The result for heat transfer with a
baffle is 300% greater than the case without a baffle. It is also
Table 2: Grid independency test for circular and porous based 40

elliptical shaped fin geometry.
Number of Average fin surface 35

Fin shape
FVM grids temperature (K)
1231224 339.238 30
Circular 2136784 338.892
Nu
2690545 338.884
25
679192 321.851
921137 321.905
20
Porous-based elliptical 1183903 321.932
fin at Ø = 0:92 2002784 321.971
2824070 321.991
0.5 1 1.5 2 2.5
3216973 321.999 Inlet flow velocity (m/s)
Numerical (current work)

worth noting that for both low and high flow rates, enhance- Khan et al., [29]
ment of heat transfer is inversely related to permeability. Colburn [42]
Hausen [43]
Kuo and Tien [25] have done numerical studies for aug-
menting heat transfer rate in a duct filled with metal foam. Figure 2: Comparison of Nu for circular fin geometry at different
The role of dispersion conductivity (directly proportional inlet flow velocities.
to the velocity and square root of permeability) on heat
transfer is done. It has been perceived that the dispersion
conductivity subsumes the stagnant conductivity for the 326
higher flow rate and results in Nu improvement. Kiwan 324
et al. [26] experimented with porous fins attached to a verti-
Average fin temperature (K)
cally heated cylinder under free convection. Two different 322

porous materials were considered, viz., lower porosity with
320
higher permeability and higher porosity with lower perme-
ability. The results clarify that the Nusselt number is a strong 318
permeability function, leading to higher heat transfer from
the fin consisting of lower porosity with higher permeability. 316
Ndlovu and Moitsheki [27] have compared the thermal per- 314
formance of porous fins with solid fins. It has been noted
that the average temperature value is lower in porous fins 312
compared to solid fins. Kundu et al. [28] have analyzed four
fins (porous) profiles, viz., rectangular, convex, and two expo- 0.8 1 1.2 1.4 1.6 1.8 2 2.2 2.4
nential in nature. It has been seen that, upon using higher Inlet flow velocity (m/s)
porosity, the overall thermal conductivity gets reduced, but k-ϵ-RNG k-𝜔-Standard
at the same time, the convective rate of heat transfer improves k-ϵ-Standard k-𝜔-SST
and results in decreasing average fin temperature.
Khan et al. [29] have analytically determined the role of Figure 3: Comparison of average fin temperature vs. inflow
standard and staggered arrangement of circular cross- velocity for various turbulence models.
section fin over the heat transfer and pressure drop. Nowa-
days, the work on using nature-inspired optimization
techniques for optimizing various geometrical parameters, (PSO) for inverse heat transfer problems. They found that
cost, and efficiency of heat transfer devices has been done by the ant colony optimization performed better for inverse heat
several researchers. Kamath et al. [30] have used (ANN-GA) transfer problems, followed by particle swarm optimization
coupled hybrid optimization technique to get the optimal tem- and the CAS. Rao and Saroj [33, 34] have compared Jaya
perature variation in a vertical conduit filled with metal foam and elitist-Jaya algorithm results based on the economical
for the given input velocity and heat condition. Sridharan design of shell and tube heat exchangers with other optimiza-
et al. [31] have used ANN to obtain the objective function tion techniques like GA and PSO. Ayala et al. [35] have done
based on input and output variable relations and multiobjec- design optimization of heat exchangers upon combining the
tive particle swarm optimization to get an optimum set of geo- free search algorithm with the differential evolutionary algo-
metrical variables for phase change material- (PCM-) based rithm and compared the based numerical results with a non-
cylindrical heat sink. Udayraj et al. [32] have studied the three dominated sorting genetic algorithm, version II. Masoud
optimization techniques, viz., ant colony optimization, cuckoo et al. [36] have done economical design optimization of shell
search algorithm (CSA), and particle search optimization and tube heat exchangers using CSA, PSO, and GA. They
70
Elliptical fins
60
Nu 50
40
30
20
ϕ = 0.8 ϕ = 0.9 ϕ = 0.92
Porosity ϕ
Re
783
1174
1839
Figure 4: Comparison of Nu value for elliptical fin for varying porosity value at a given flow velocity.
75
At 𝜙 = 0.92
10 PPI
65
55
Nu
45
35
25
15
Solid Elliptical 𝛿/8 𝛿/6 𝛿/4 𝛿/2
circular
Transverse offset
Re
783
1174
1839
Figure 5: Comparison of Nu for solid circular fin and porous-based elliptical fins with varying TO’s∗ from 0 (ellipse), δ/8, δ/6, δ/4, and δ/2
at ɸ = 0:92 for different inlet flow velocities.
found that the operating cost was reduced by 78% and 48% heat transfer may be augmented using porous-based fins.
upon using CSA compared to PSO and GA. Raja et al. [37] Porous medium provides a higher surface area to volume
opted for the HTS- (heat transfer search-) based algorithm ratio, high tortuosity, and better interfluid mixing, which
to optimize fin and tube heat exchanger geometrical parame- alleviates heat transfer. The present numerical study has
ters. Further, the outcomes of HTS were compared with GA, been conducted on solid-based split elliptical fins to analyze
PSO, and ant colony optimization techniques. the fluid flow and thermal behavior of variable Reynolds
Based on the study of the literature survey, it has been number with air as a working medium (and Pr = /0:71),
noted that the requirement of quick heat dissipation from and obtained results were validated with the CSA technique.
devices showing thermal behavior is essential. The generated Here, the consideration is given to study the effect of varying
heat can be released into the atmosphere using extended porosity (0.8 to 0.92), PPI = 10, permeability resistance, iner-
surfaces. The work on solid-based fins is desirable, but the tial parameter, and the Reynolds number (based upon fin)
75
At 𝜙 = 0.9
10 PPI
65
55
Nu
45
35
25
15
circular
Transverse offset
Re
783
1174
1839
75
At 𝜙 = 0.8
65 10 PPI
55
Nu
45
35
25
15
circular
Transverse offset
Re
783
1174
1839
from 783 to 1839 on the rate of heat transfer using porous- is associated with the circular and elliptical-shaped fins with
based split elliptical fins. Further, the Cuckoo search algo- further modification in terms of splits. Figures 1(a)–1(c)
rithm is used to determine the maximum Nusselt number show the schematic diagram of the computational domain,
value based upon the ɸ, TO∗ , LO∗ , and Re. porous-based elliptical fin-base plate setup, and elliptical
fin with longitudinal and transverse offset.
Initially, 17 numbers of circular fins (diameter δ = 15 mm
2. Geometry and Formulation and height h f = 105 mm) were arranged over a rectangular
The phenomenon encompassing heat transfer through its base plate of dimension 118 × 118 × 5 mm3 in a computa-
surface has always shown its great impact upon using fins tional domain of square cross-section duct with dimension
and their geometrical configuration. The present work here 111 × 120 × 658 mm3 , as shown in Figure 1(a). In order to
Temperature:
296 311.667 327.333 343 358.667 374.333 390
𝛷 𝛿/8 𝛿/6 𝛿/4 𝛿/2
0.06 0.06 0.06
0.06
0.04 0.04 0.04
0.04
0.02 0.02 0.02
0.02
0.92
Y
Y
0
Y
0 0
0
−0.02 −0.02 −0.02
−0.02
−0.04 −0.04 −0.04
−0.04
−0.06 −0.06 −0.06
−0.06
−0.06 −0.04 −0.02 0 0.02 0.04 0.06 −0.06 −0.04 −0.02 0 0.02 0.04 0.06 −0.06 −0.04 −0.02 0 0.02 0.04 0.06
−0.06 −0.04 −0.02 0 0.02 0.04 0.06
X X X
X
0.06 0.06 0.06 0.06
0.04 0.04 0.04 0.04
0.02 0.02 0.02 0.02

0.9
Y
Y
Y
Y
0 0 0 0
−0.02 −0.02 −0.02 −0.02
−0.04 −0.04 −0.04 −0.04
−0.06 −0.06 −0.06 −0.06

−0.06 −0.04 −0.02 0 0.02 0.04 0.06 −0.06 −0.04 −0.02 0 0.02 0.04 0.06 −0.06 −0.04 −0.02 0 0.02 0.04 0.06 −0.06 −0.04 −0.02 0 0.02 0.04 0.06
X X X X
0.06 0.06 0.06 0.06

0.04 0.04 0.04 0.04
0.02 0.02 0.02 0.02
0.8
Y
Y
Y
0 0 0 0
−0.02 −0.02 −0.02 −0.02
−0.04 −0.04 −0.04 −0.04
−0.06 −0.06 −0.06 −0.06

−0.06 −0.04 −0.02 0 0.02 0.04 0.06 −0.06 −0.04 −0.02 0 0.02 0.04 0.06 −0.06 −0.04 −0.02 0 0.02 0.04 0.06 −0.06 −0.04 −0.02 0 0.02 0.04 0.06
X X X X
Figure 8: Temperature variation plot for varying TO’s∗ and porosity at Re = 1839.
extend the work, the elliptical-shaped fin with the same vol- !

ð2Þ
∇∙V = 0,
ume and height as of former fin is considered, as shown in
Figure 1(b), and the extended modification is done in terms ! ! −∇P !
of LO∗ and TO∗ , as shown in Figure 1(c). The major axis of V ∇∙V = + ϑ ∇2 V : ð3Þ
ρ
the elliptical-shaped fin is assumed to be double the minor
axis, and thus, upon equating the volume of the elliptical to Initially, for work validation, the empirical formula
the cylindrical fin, one of the dimensions is obtained (on keep- related to Nu developed [29] for circular cross-section fin
ing the same height for both fins). Here, the consideration is is shown in Equation (4). Similarly, the other validation-
pronounced only for the conjugate physics of conduction associated formulas for accounting U max (used for obtaining
and convection, whereas the radiative form of thermal energy Re which is based on fin dimension) and coefficient constant
is discarded. For solving the present problem, the finite vol- (C 1 ) (used for obtaining Nu) are [29] shown in Equation (5)
ume method-based ANSYS Fluent solver is used. To assign and Equation (6).
the physical term of the computational domain, the boundary
condition is mentioned in Table 1. h∙D
Nu = = C 1 Re0:5 Pr0:34 , ð4Þ
For the phenomenon involving heat transfer, the denota- kf
tion of the temperature distribution in the considered com-
putational domain is solved using an energy equation [38], 0:61S0:091 S0:053
as shown in Eq. (1). Similarly, for obtaining the velocity C1 = t l
, ð5Þ
½1 − 2 exp ð−1:09Sl Þ
field, the mass and momentum conservation equations are

solved [38], as shown in Equations (2) and (3). St St
U max = Max: U , U : ð6Þ
St − 1 ∞ 2ðSd − 1Þ ∞
! À Á However, the flow physics involved for the porous
V ∙∇T = α ∇2 T , ð1Þ
medium work differently than the former conservation
Temperature:
296 316.8 337.6 358.4 379.2 400
𝛷 𝛿/8 𝛿/6 𝛿/4 𝛿/2
0.06 0.06 0.06 0.06
0.04 0.04 0.04 0.04
0.02 0.02 0.02 0.02

0.92
Y
Y
Y
0 0 0 0
−0.02 −0.02 −0.02 −0.02
−0.04 −0.04 −0.04 −0.04
−0.06 −0.06 −0.06 −0.06

−0.06 −0.04 −0.02 0 0.02 0.04 0.06 −0.06 −0.04 −0.02 0 0.02 0.04 0.06 −0.06 −0.04 −0.02 0 0.02 0.04 0.06 −0.06 −0.04 −0.02 0 0.02 0.04 0.06
X X X X
0.06 0.06 0.06 0.06
0.04 0.04 0.04 0.04
0.02 0.02 0.02 0.02

0.9
Y
Y
Y
Y
0 0 0 0
−0.02 −0.02 −0.02 −0.02
−0.04 −0.04 −0.04 −0.04
−0.06 −0.06 −0.06 −0.06

−0.06 −0.04 −0.02 0 0.02 0.04 0.06 −0.06 −0.04 −0.02 0 0.02 0.04 0.06 −0.06 −0.04 −0.02 0 0.02 0.04 0.06 −0.06 −0.04 −0.02 0 0.02 0.04 0.06
X X X X
0.06 0.06 0.06 0.06
0.04 0.04 0.04 0.04
0.02 0.02 0.02 0.02

0.8
Y
Y
0 0 0 0
−0.02 −0.02 −0.02 −0.02
−0.04 −0.04 −0.04 −0.04
−0.06 −0.06 −0.06 −0.06

−0.06 −0.04 −0.02 0 0.02 0.04 0.06 −0.06 −0.04 −0.02 0 0.02 0.04 0.06 −0.06 −0.04 −0.02 0 0.02 0.04 0.06 −0.06 −0.04 −0.02 0 0.02 0.04 0.06
X X X X
equations due to the involvements of terms like ɸ, F, and For obtaining the viscous and inertial effect in the porous
permeability (Pn ). So, the governing flow physics involved zone, the values of permeability and the Forchheimer
in the mass and momentum conservation [39] for finding constant [41] are obtained using Equations (10) and (11).
out the velocity of fluid throughout the porous zone of the Similarly, the formulas for obtaining the dp , d f , and χ are
computational domain is mentioned in Equations (7) and mentioned in Equations (12)–(14).
(8). Similarly, the value temperature field in the porous zone
is obtained using the conservation of energy [39], shown in 0:00926:ɸ:d2p
Equation (9). Here, the local thermal equilibrium assump- Pn = , ð10Þ
χ−1
tion is opted upon considering the same temperature field
for both the solid and fluid. However, Vazifeshenas et al. sffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi rffiffiffiffiffiffiffiffiffiffiffiffi !−1
[40] have employed local thermal equilibrium for analyzing −0:8 ɸ 1−ɸ 1
F = 0:0095G 1:18 , ð11Þ
the forced convection in metal foam with water as the work- 3ðχ − 1Þ 3π G
ing fluid.
whereas G is the shape function based on the porosity value.
∂ρ !
ɸ +∇∙ ρV = 0, ð7Þ 0:025417
∂t dp = , ð12Þ
PPI
1 ! ! ∇P ϑ 2 ! ϑ ! F !! rffiffiffiffiffiffiffiffiffiffiffiffiffiffi
V ∙∇ V = − + ∇ V− V − pffiffiffiffiffi V V , ð8Þ ð1 − ɸ Þ d p
ɸ 2 ρ ɸ Pn Pn d f = 1:18 , ð13Þ
3π G
∂T ! " rffiffiffiffiffiffiffiffiffiffiffi !2 !#−1
ɸðρcÞ f + ð1 − ɸÞðρcÞs + ðρcÞ f V ∙∇T π 1−ɸ 1
À ÁÀ ∂t Á ð9Þ χ= 1 − 1:18 : ð14Þ
= ɸk f + ð1 − ɸÞks ∇2 T : 4 3π G
Temperature:
296 306.571 317.143 327.714 338.286 348.857 359.429 370
𝛷 𝛿/8 𝛿/6 𝛿/4 𝛿/2
0.06 0.06 0.06 0.06
0.04 0.04 0.04 0.04
0.02 0.02 0.02 0.02

0.92
Y
Y
Y
0 0 0 0
−0.02 −0.02 −0.02 −0.02
−0.04 −0.04 −0.04 −0.04
−0.06 −0.06 −0.06 −0.06

−0.06 −0.04 −0.02 0 0.02 0.04 0.06 −0.06 −0.04 −0.02 0 0.02 0.04 0.06 −0.06 −0.04 −0.02 0 0.02 0.04 0.06 −0.06 −0.04 −0.02 0 0.02 0.04 0.06
X X X X
0.06 0.06 0.06 0.06
0.04 0.04 0.04 0.04
0.02 0.02 0.02 0.02

0.9
Y
Y
Y
0 0 0 0
−0.02 −0.02 −0.02 −0.02
−0.04 −0.04 −0.04 −0.04
−0.06 −0.06 −0.06 −0.06

−0.06 −0.04 −0.02 0 0.02 0.04 0.06 −0.06 −0.04 −0.02 0 0.02 0.04 0.06 −0.06 −0.04 −0.02 0 0.02 0.04 0.06 −0.06 −0.04 −0.02 0 0.02 0.04 0.06
X X X X
0.06 0.06 0.06 0.06
0.04 0.04 0.04 0.04
0.02 0.02 0.02 0.02

0.8
Y
Y
Y
0 0 0 0
−0.02 −0.02 −0.02 −0.02
−0.04 −0.04 −0.04 −0.04
−0.06 −0.06 −0.06 −0.06

−0.06 −0.04 −0.02 0 0.02 0.04 0.06 −0.06 −0.04 −0.02 0 0.02 0.04 0.06 −0.06 −0.04 −0.02 0 0.02 0.04 0.06 −0.06 −0.04 −0.02 0 0.02 0.04 0.06
X X X X
hlc
Nu = , ð15Þ
kf
4 × Volume
lc = : ð16Þ
Total wetted area of the fin
0.04
Y
Similarly, for measuring the losses (friction) in the compu-

tational domain due to the projection of fin, the value of the
frictional coefficient [10] is obtained using Equation (17). In
order to compare the porous fin with the solid fin in terms of
usability, the performance evaluation criterion is evaluated
using Equation (18). Further, to capture generated turbulence
in the computational domain, the k-ϵ-RNG model has been
0 opted for, which is consisting of turbulence kinetic energy (k)
X and turbulence dissipation term (ϵ), as mentioned in Equation
(19) and Equation (20).
Figure 11: Top view of flow around and within the porous-based
single fin.
2τS
Cf = , ð17Þ
ρU 2∞
Now, for the present work, the rate of convection to the
conduction for the given fin geometry, i.e., the Nusselt number, Nu/Nuo
is calculated using Equation (15). The value of the characteristic PEC = À Á1/3 , ð18Þ
dimension used for obtained Nu is shown in Equation (16). C f /C f o
0.07
At 𝜙 = 0.92
10 PPI
0.06
0.05
0.04
Cf
0.03
0.02
0.01
circular
Transverse offset
Re
783
1174
1839
Figure 12: Comparison of frictional losses at ɸ = 0:92 with varying TO’s∗ for different flow velocities.
0.07
At 𝜙 = 0.9
10 PPI
0.06
0.05
Cf
0.04
0.03
0.02
0.01
circular
Transverse offset
Re
783
1174
1839

! μt geometry and formulation section. Initially, seventeen num-
div ρk u = div grad k + 2μt Eij ∙Eij − ρε, ð19Þ bers of circular cross-section fins infused upon a heated plate
σk
in a staggered way are considered, as shown in Figure 1(b).

! μt ε ε2 Further, to obtain a higher degree of accuracy in the
div ρεu == div grad ε + C1ε 2μt Eij ∙Eij − C2ε ρ :
σε k k result, grid independency is performed using the tetrahedral
meshing approach. The meshing is done individually for the
ð20Þ
circular and porous-based elliptical-shaped fins. It is found
3. Results and Discussions that the variation of average fin surface temperature is
negligible as the number of elements for both the circular
The current work proceeds with the given fin setup in a and porous-based elliptical fin is above 2:0 × 106 , as given
computational fluid domain discussed in the above in Table 2.
0.07
At 𝜙 = 0.8
0.06 10 PPI
0.05
Cf 0.04
0.03
0.02
0.01
circular
Transverse offset
Re
783
1174
1839
Y velocity:
0.2 0.8 1.4 2 2.6 3.2 3.8 4.4 5
𝛷 𝛿/8 𝛿/6 𝛿/4 𝛿/2

0.06 0.06 0.06 0.06
0.04 0.04 0.04 0.04
0.02 0.02 0.02 0.02

0.92
Y
Y
Y
0 0 0 0
−0.02 −0.02 −0.02 −0.02
−0.04 −0.04 −0.04 −0.04
−0.06 −0.06 −0.06 −0.06

−0.06 −0.04 −0.02 0 0.02 0.04 0.06 −0.06 −0.04 −0.02 0 0.02 0.04 0.06 −0.06 −0.04 −0.02 0 0.02 0.04 0.06 −0.06 −0.04 −0.02 0 0.02 0.04 0.06
X X X X
0.06 0.06 0.06 0.06
0.04 0.04 0.04 0.04
0.02 0.02 0.02 0.02

0.9
Y
0 0 0 0
−0.02 −0.02 −0.02 −0.02
−0.04 −0.04 −0.04 −0.04
−0.06 −0.06 −0.06 −0.06

−0.06 −0.04 −0.02 0 0.02 0.04 0.06 −0.06 −0.04 −0.02 0 0.02 0.04 0.06 −0.06 −0.04 −0.02 0 0.02 0.04 0.06 −0.06 −0.04 −0.02 0 0.02 0.04 0.06
X X X X
0.06 0.06 0.06 0.06
0.04 0.04 0.04 0.04
0.02 0.02 0.02 0.02

0.8
Y
0 0 0 0
−0.02 −0.02 −0.02 −0.02
−0.04 −0.04 −0.04 −0.04
−0.06 −0.06 −0.06 −0.06

−0.06 −0.04 −0.02 0 0.02 0.04 0.06 −0.06 −0.04 −0.02 0 0.02 0.04 0.06 −0.06 −0.04 −0.02 0 0.02 0.04 0.06 −0.06 −0.04 −0.02 0 0.02 0.04 0.06
X X X X
Figure 15: Circulation phenomenon plot for varying TO’s∗ and porosity at Re = 1839.
Y velocity:
0.5 1 1.5 2 2.5 3
𝛷 𝛿/8 𝛿/6 𝛿/4 𝛿/2

0.06 0.06 0.06 0.06
0.04 0.04 0.04 0.04
0.02 0.02 0.02 0.02

0.92
Y
Y
Y
0 0 0 0
−0.02 −0.02 −0.02 −0.02
−0.04 −0.04 −0.04 −0.04
−0.06 −0.06 −0.06 −0.06

−0.06 −0.04 −0.02 0 0.02 0.04 0.06 −0.06 −0.04 −0.02 0 0.02 0.04 0.06 −0.06 −0.04 −0.02 0 0.02 0.04 0.06 −0.06 −0.04 −0.02 0 0.02 0.04 0.06
X X X X
0.06 0.06 0.06 0.06
0.04 0.04 0.04 0.04
0.02 0.02 0.02 0.02

0.9
Y
Y
0 0 0 0
−0.02 −0.02 −0.02 −0.02
−0.04 −0.04 −0.04 −0.04
−0.06 −0.06 −0.06 −0.06

−0.06 −0.04 −0.02 0 0.02 0.04 0.06 −0.06 −0.04 −0.02 0 0.02 0.04 0.06 −0.06 −0.04 −0.02 0 0.02 0.04 0.06 −0.06 −0.04 −0.02 0 0.02 0.04 0.06
X X X X
0.06 0.06 0.06 0.06
0.04 0.04 0.04 0.04
0.02 0.02 0.02 0.02

0.8
Y
Y
0 0 0 0
−0.02 −0.02 −0.02 −0.02
−0.04 −0.04 −0.04 −0.04
−0.06 −0.06 −0.06 −0.06

−0.06 −0.04 −0.02 0 0.02 0.04 0.06 −0.06 −0.04 −0.02 0 0.02 0.04 0.06 −0.06 −0.04 −0.02 0 0.02 0.04 0.06 −0.06 −0.04 −0.02 0 0.02 0.04 0.06
X X X X
After the grid testing, the given computational setup is The coupled scheme has opted for pressure-velocity cou-
analyzed over the range of Re (783 to 1839) in the laminar pling, and for spatial discretization, the pressure staggered
flow regime and gets validated with the available empirical option (PRESTO) scheme (for pressure) and second-order
correlation [29, 42, 43] developed for circular pin-fin, as upwind scheme (for momentum and energy) are used. The
shown in Figure 2. effect of porosity on split and unsplit elliptical fins is dis-
Upon confirmation of the work with [29] a maximum cussed in further sections.
and minimum error percentage of 6.22% and 0.54% (at
maximum and minimum accounted Re number), the work 3.1. Effect of Porosity with Transverse Offsets
is directed toward analyzing the solid and porous-based
elliptical fins with modifications in terms of dimensionless 3.1.1. The Nusselt Number. The comparison of the Nu value
split offsets. In order to capture the generated turbulence, for elliptical fin for varying porosity values at a given flow
the k-ϵ-RNG turbulence model is opted for; however, the velocity is shown in Figure 4.
other related model has also been examined and mentioned It can be seen in Figure 4 that, as the porosity (ɸ)
in Figure 3. increases from 0.8 to 0.92, the Nu increased, due to an
Upon analyzing the various turbulence model, the increase in tortuosity within the fin, which leads to better
k-ϵ-RNG model is a model found to be suitable and accurate. solid-fluid interaction. The trend of increasing Nu with the
The k-ϵ-RNG model is based upon the two equations, i.e., k increase in porosity value has been found for porous layers
(turbulent kinetic) and ϵ (turbulent dissipation) out of which wrapped over a solid cylinder [8] and similarly for porous
some extra term is encompassed in the ϵ equation, which is blocks [5]. In the present work, the primary form of analysis
absent in the k-ω-Standard. The extra term helps in capturing is accounted for the split-shaped fin. Initially, upon changing
the swirling phenomenon affecting the turbulence and signif- the porosity value, the effect of varying transverse offset
icantly affects the accuracy. associated with the elliptical shape of fins is investigated. It
Y velocity:
0.2 0.6 1 1.4 1.8 2.2
𝛷 𝛿/8 𝛿/6 𝛿/4 𝛿/2

0.06 0.06 0.06 0.06
0.04 0.04 0.04 0.04
0.02 0.02 0.02 0.02

0.92
Y
Y
Y
0 0 0 0
−0.02 −0.02 −0.02 −0.02
−0.04 −0.04 −0.04 −0.04
−0.06 −0.06 −0.06 −0.06

−0.06 −0.04 −0.02 0 0.02 0.04 0.06 −0.06 −0.04 −0.02 0 0.02 0.04 0.06 −0.06 −0.04 −0.02 0 0.02 0.04 0.06 −0.06 −0.04 −0.02 0 0.02 0.04 0.06
X X X X
0.06 0.06 0.06 0.06
0.04 0.04 0.04 0.04
0.02 0.02 0.02 0.02

0.9
Y
Y
0 0 0 0
−0.02 −0.02 −0.02 −0.02
−0.04 −0.04 −0.04 −0.04
−0.06 −0.06 −0.06 −0.06

−0.06 −0.04 −0.02 0 0.02 0.04 0.06 −0.06 −0.04 −0.02 0 0.02 0.04 0.06 −0.06 −0.04 −0.02 0 0.02 0.04 0.06 −0.06 −0.04 −0.02 0 0.02 0.04 0.06
X X X X
0.06 0.06 0.06 0.06
0.04 0.04 0.04 0.04
0.02 0.02 0.02 0.02

0.8
Y
Y
0 0 0 0
−0.02 −0.02 −0.02 −0.02
−0.04 −0.04 −0.04 −0.04
−0.06 −0.06 −0.06 −0.06

−0.06 −0.04 −0.02 0 0.02 0.04 0.06 −0.06 −0.04 −0.02 0 0.02 0.04 0.06 −0.06 −0.04 −0.02 0 0.02 0.04 0.06 −0.06 −0.04 −0.02 0 0.02 0.04 0.06
X X X X
is found that, as the value of transverse offset increases as the value of transverse offset increases from 0 to δ/2, the
from δ/8 to δ/2, the value of the Nu increases for all the value Nu number increases. It is mainly because, at a higher
respective Re. transverse offset width gap, the fluid will easily pass through
The comparison of Nu for solid circular fin and porous- the split gap, take up the heat from the split gap area, and
based elliptical fins with varying TO’s∗ from 0 (ellipse), δ/8, then move forward, i.e., without creating much hot spot at
δ/6, δ/4, and δ/2 at ɸ = 0:92 for different inlet flow velocities the fin wake or trailing region. And thus, as the gap increases
is shown in Figure 5. The comparison of Nu for solid circu- from δ/8 to δ/2, the average fin temperature gets lowered
lar fin and porous-based elliptical fins with varying TO’s∗ and indirectly alleviates the Nu number, as shown in
from 0 (ellipse), δ/8, δ/6, δ/4, and δ/2 at ɸ = 0:9 for different Figures 8–10 for ɸ = 0:92, 0.9, and 0.8 (at respective Re
inlet flow velocities is shown in Figure 6. The comparison of values). Moreover, at fixed Re, and porosity value, augmen-
Nu for solid circular fin and porous-based elliptical fins with tation in heat transfer occurred as the transverse offset width
varying TO’s∗ from 0 (ellipse), δ/8, δ/6, δ/4, and δ/2 at increases from δ/8 to δ/2. With an increase in the Nu value,
ɸ = 0:8 for different inlet flow velocities is shown in the velocity between the gaps increases, which assists in
Figure 7. dragging the fluid present inside the porous fin and thus
Figure 5 illustrates that at ɸ = 0:92 and Re = 1839, the lessen frictional losses. However, a front view of flow within
value of Nu for the transverse offset of 0 (ellipse), δ/8, δ/6, and around the porous-based fins is shown in Figure 11.
δ/4, and δ/2 are approximately 62.56, 64.75, 66.57, 67.72,
and 68.87, respectively. Similarly, for other values of Re, 3.1.2. Frictional Coefficient. It is essential to account for the
Nu is accounted at ɸ = 0:92, as shown in Figure 5. Similar skin friction coefficient value because this affects the perfor-
trend for Nu is found for the ɸ = 0:9 and 0.8 at respective mance evaluation criteria of a particular fin shape. Due to
values of Re, as shown in Figures 6 and 7. Out of all the con- the change in porosity, the value of the frictional coefficient
sidered porosity values, the highest value of Nu is obtained gets changed, which creates flow resistance to the upcoming
by TO∗ = δ/2 at ɸ = 0:92 and Re = 1839. As discussed above, fluid. Figure 11 shows that as the porosity value decreases
1.8
At 𝜙 = 0.92
10 PPI
1.6
PEC 1.4
1.2
1.0
0.8
circular
Transverse offset
Re
783
1174
1839
Figure 18: Comparison of performance evaluation criteria at ɸ = 0:92 with varying TO’s∗ for different flow velocities.
1.8
At 𝜙 = 0.9
10 PPI
1.6
1.4
PEC
1.2
1.0
0.8
circular
Transverse offset
Re
783
1174
1839
from 0.92 to 0.8, the skin friction coefficient increases for the fluid velocity in between the split gap gets improved,
porous-based elliptical fins. which results in lowering the friction by accelerating the
A similar trend for friction factor has been obtained for fluid attached to the porous zone to channelize toward the
the metallic porous structure filled in a vertical duct for higher fluid velocity region. At the smaller gap width, i.e.,
experimental analysis [44]. It is because, as the porosity at transverse offset = δ/8, more circulation phenomenon
values decrease from 0.92 to 0.8, the diameter of the solid can be seen at the fin wake or trailing region (leading to
fiber increases, leading to an increase in the interfacial area. higher frictional loss), which diminishes with an increase
As the transverse offset width gap increases, the value of C f in the transverse offset width, as shown in Figure 15–17.
also gets reduced for all the considered porosity (0.92, 0.9,
and 0.8) values, as shown in Figures 12–14. It is because as 3.1.3. Performance Evaluation Criteria. The performance
the transverse offset width gap increases from δ/8 to δ/2, evaluation criterion (PEC) is an essential parameter for
1.8
At Ф = 0.8
10 PPI
1.6
PEC 1.4
1.2
1.0
0.8
circular
Transverse offset
Re
783
1174
1839
75
At Ø = 0.92
65 10 PPI
55
45
Nu
35
25
15
5
Longitudinal offset 𝛿/8 𝛿/6 𝛿/4 𝛿/8 𝛿/6 𝛿/4 𝛿/8 𝛿/6 𝛿/4 𝛿/8 𝛿/6 𝛿/4
Transverse offset 𝛿/8 𝛿/6 𝛿/4 𝛿/2
Re
783
1174
1839
Figure 21: Comparison of Nu for porous-based split elliptical fins with varying LO’s∗ from δ/8, δ/6, δ/4, and δ/2 for particular TO’s∗ at
ɸ = 0:92 for different inlet flow velocities.
indicating the performance and usability of a fin shape in generalizing the PEC value. One is the porous nature of the
compared to the other. Initially, PEC is accounted for the fin, which provides a higher surface area-to-volume ratio and,
solid-based circular fin, and this value of PEC w.r.t to its in turn, augments the value of the Nu number, but at the same
Re is considered a reference value. Figure 18 shows that time, the losses associated with friction also rise in the system.
the value of PEC for all the considered porous-based split And, further, to recede the frictional losses, the elliptic-shaped
and unsplit fins is more than unity, which authenticates that fin gets split in terms of transverse offset, and it reduces upon
these taken shapes are better than the solid circular cross- increasing the offset width.
section fin. The highest value of PEC (=/1.78) is obtained
by transverse offset = δ/2 at 10 PPI and ɸ = 0:92, which is 3.2. Effect of Porosity with Split Offset Combination
78.1% higher than the solid circular fin at its respective Re
number, shown in Figure 18. Similarly, the value of PEC 3.2.1. The Nusselt Number. The work on transverse offset,
for ɸ = 0:9 and 0.8 at their respective Re number is men- i.e., changing it from δ/8 to δ/2, is being accounted for and
tioned in Figures 19 and 20. Two major concepts are involved examined. It is found that the heat transfer phenomenon
75
At Ø = 0.9
65 10 PPI
55
45
Nu
35
25
15
5
Re
783
1174
1839
75
At Ø = 0.8
65 10 PPI
55
45
Nu
35
25
15
5
Re
783
1174
1839
gets augmented upon splitting the elliptic-shaped fin in a in the Nu. Except for the case transverse offset = δ/2, the Nu
transverse direction to the fluid flow. Here, the focus is given value increases as the longitudinal offset changes from δ/8 to
to varying the split offset in the longitudinal direction (at a δ/4. It is mainly because at transverse offset = δ/2, as the lon-
fixed value of transverse offset). It is found that, at ɸ = 0:8 gitudinal offset increases from δ/8 to δ/4, the fluid-solid inter-
to 0.92 for all the Re values, as the value of longitudinal offset action gets alleviated with higher velocity in between the gap
changes from δ/8 to δ/4 at fixed values of transverse offset, and increasing flow resistance from the transverse direction
i.e., from δ/8 to δ/4, the value of Nu decreases, as shown of the fin, i.e., the gap in between the fins from either side
in Figures 21–23. It is because as the fin splits, the gap in get reduces in more extent. Thus, upon varying longitudinal
the flow direction increases, the fluid-solid interaction in offsets from δ/8 to δ/4 at a fixed value of transverse offsets
between the split gets reduced, i.e., fluid will not pertain for (i.e., δ/8 to δ/4), the average fin temperature increases,
more time between the split gap, and this leads to a decrease whereas, for transverse offset = δ/2, the value of average fin
Temperature (K)
296 311.667 327.333 343 358.667 374.333 390
LO 𝛿/8 𝛿/6 𝛿/4
0.06 0.06 0.06
TO
0.04 0.04 0.04
0.02 0.02 0.02

𝛿/8
Y
Y
Y
0 0 0
−0.02 −0.02 −0.02
−0.04 −0.04 −0.04
−0.06 −0.06 −0.06

−0.06 −0.04 −0.02 0 0.02 0.04 0.06 −0.06 −0.04 −0.02 0 0.02 0.04 0.06 −0.06 −0.04 −0.02 0 0.02 0.04 0.06
X X X
0.06 0.06 0.06
0.04 0.04 0.04
0.02 0.02 0.02

𝛿/6
Y
Y
0 0 0
−0.02 −0.02 −0.02
−0.04 −0.04 −0.04
−0.06 −0.06 −0.06

−0.06 −0.04 −0.02 0 0.02 0.04 0.06 −0.06 −0.04 −0.02 0 0.02 0.04 0.06 −0.06 −0.04 −0.02 0 0.02 0.04 0.06
X X X
0.06 0.06 0.06
0.04 0.04 0.04
0.02 0.02 0.02

𝛿/4
Y
Y
Y
0 0 0
−0.02 −0.02 −0.02
−0.04 −0.04 −0.04
−0.06 −0.06 −0.06

−0.06 −0.04 −0.02 0 0.02 0.04 0.06 −0.06 −0.04 −0.02 0 0.02 0.04 0.06 −0.06 −0.04 −0.02 0 0.02 0.04 0.06
X X X
0.06 0.06 0.06
0.04 0.04 0.04
0.02 0.02 0.02

𝛿/2
Y
Y
Y
0 0 0
−0.02 −0.02 −0.02
−0.04 −0.04 −0.04
−0.06 −0.06 −0.06

−0.06 −0.04 −0.02 0 0.02 0.04 0.06 −0.06 −0.04 −0.02 0 0.02 0.04 0.06 −0.06 −0.04 −0.02 0 0.02 0.04 0.06
X X X
Figure 24: Temperature plot for porous-based split elliptical fins with varying LO’s∗ at particular TO’s∗ for ɸ = 0:92 at Re = 1839.
temperature gets reduced over the change in longitudinal off- formance. It is accounted that for all the cases except at trans-
set from δ/8 to δ/4, as shown in Figure 24 for ɸ = 0:92 and verse offset = δ/2 with varying longitudinal offset from δ/8 to
Re = 1839. δ/4, the value of skin friction factor decreases and vice-versa
for the former at every value of porosity (0.8 to 0.92) and its
3.2.2. Frictional Factor. The friction factor’s value is evaluated respective Re, as shown in Figures 25–27. It is mainly because
to decode the friction role in determining the fin setup’s per- as the split gap in the flow direction changes from δ/8 to δ/4
0.07
At Ø = 0.92
10 PPI
0.06
Cf
0.05
0.04
0.03
Re
783
1174
1839
Figure 25: Comparison of C f for porous-based split elliptical fins with varying LO’s∗ from δ/8, δ/6, δ/4, and δ/2 for particular TO’s∗ at
0.07
At Ø = 0.9
10 PPI
0.06
Cf
0.05
0.04
0.03
Re
783
1174
1839
Figure 26: Comparison of C f for porous-based split elliptical fins with varying LO’s∗ from δ/8, δ/6, δ/4, and δ/2 for particular TO’s∗ at
for transverse offset = δ/2, the solid-fluid interaction from the Figures 29–31 that for a given value of ɸ (0.8, 0.9, and
transverse side of the fin increases, and this leads to providing 0.92), the value of PEC is more than one, which signifies
the greater extent of circulation phenomenon at the wake and the applicability of concerned fins is better than the solid cir-
from either side of the fin, as shown in Figure 28, for ɸ = 0:92 cular fin for all the considered Re numbers. The highest
and Re = 1839. value of PEC (=/1.73) in the case of both longitudinal offset
and transverse offset combination is obtained at ɸ = 0:92
3.2.3. Performance Evaluation Criteria. The value of PEC is and Re = 1839 for longitudinal offset = δ/4 (at fixed trans-
evaluated for the range of Re (783 to 1839) using split ellip- verse offset = δ/4). It is mainly due to the higher rate of con-
tical fins in both the transverse offset and longitudinal offset vection to the conduction obtained concerning the relative
combination at ɸ = 0:8, 0.9, and 0.92. It can be seen in frictional losses shown by longitudinal offset = δ/4 (at fixed
0.07
At Ø = 0.8
10 PPI
0.06
Cf 0.05
0.04
0.03
Re
783
1174
1839
Figure 27: Comparison of C f for porous-based split elliptical fins with varying LO’s∗ from δ/8, δ/6, δ/4, and δ/2ɸ for particular TO’s∗ at
transverse offset = δ/4) compared to other combinations. It is nique for the design of the experiment to obtain an
also noted that the value of PEC decreases on decreasing poros- approximation of the response function concerning
ity value for any combination of considered fins. It is mainly the Nusselt number (Nu), that is, corresponding to
because as the porosity value changes from 0.92 to 0.8, the the second-order model
interfacial area increases, which offers higher frictional losses.
(3) Thirdly, construct an ANOVA table (to evaluate the
factors’ influences on the response function) and
3.3. Mathematical Model Using RSM. The response surface examine multiple experiment model designs. Assess
method (RSM), a popular and important mathematics tech- the value of P and determination coefficient R2 , in
nique, uses numerical or experimental trials to determine an more detail because these parameters give insight
appropriate objective function. It is frequently utilized when into the accuracy of the response variable’s (Nu’s)
numerous factors are influencing the response function. Pri- prediction when the provided independent varia-
marily, it is effective for first- and second-order polynomial ble—i.e., dimensionless terms—is executed
functions. If the response is specified by the linear function of
the variable, the first-order model is effective; alternatively, the In order to develop a mathematical correlation with Nu as
second-order model is developed. Here, in this current investi- an objective function with input variables as TO∗ , LO∗ , Re, and
gation, the variables, i.e., dimensionless terms, can impact the ɸ, the RSM response surface method is opted for. The response
response parameter in an increasing or decreasing form; conse- surface method accounts for both the experimental design and
quently, the second-order model is constructed. The following regression analysis to obtain correlation; the detailed procedure
mentioned points provide detailed information about RSM. is outlined in mentioned references. The result of the numerical
evaluation is employed in the MINITAB 17 to fetch a second-
(1) First, obtain details from the computational experi- order polynomial function, and the developed function is
ment regarding the input variables (dimensionless shown in Equation (21). The value of determination coefficient
parameters, such as TO, LO, and Re), as well as the R2 is obtained as 99.72%, which shows the higher accuracy for
intended output (Nusselt’s number). Additionally, the developed mathematical model.
the response surface plot is generated using the
Nu = 57:6 + 0:02015 × Re − 114 × ɸ + 25:27 × TO∗ − 12:1
response function and independent variable data
× LO∗ − 0:000001 × Re2 + 61:6 × ɸ2 − 36:58 × TO∗2
(2) Build a second-order mathematical model employ- + 3:77 × LO∗2 + 0:01453 × Re × ɸ + 0:004780 × Re ð21Þ
ing the input variables (which include a first-order ∗ ∗ ∗
× TO − 0:00057 × Re × LO − 4:23 × ɸ × TO + 0:8
model, a quadratic term, and a cross-product term),
followed by implementing the full factorial tech- × ɸ × LO∗ + 17:86 × TO∗ × LO∗ ,
Velocity (m/s)
0.5 1 1.5 2 2.5 3 3.5 4 4.5 5 5.5
LO 𝛿/8 𝛿/6 𝛿/4

0.06 0.06 0.06
TO
0.04 0.04 0.04
0.02 0.02 0.02

𝛿/8
Y
Y
0 0 0
−0.02 −0.02 −0.02
−0.04 −0.04 −0.04
−0.06 −0.06 −0.06

−0.06 −0.04 −0.02 0 0.02 0.04 0.06 −0.06 −0.04 −0.02 0 0.02 0.04 0.06 −0.06 −0.04 −0.02 0 0.02 0.04 0.06
X X X
0.06 0.06 0.06
0.04 0.04 0.04
0.02 0.02 0.02

𝛿/6
Y
Y
0 0 0
−0.02 −0.02 −0.02
−0.04 −0.04 −0.04
−0.06 −0.06 −0.06

−0.06 −0.04 −0.02 0 0.02 0.04 0.06 −0.06 −0.04 −0.02 0 0.02 0.04 0.06 −0.06 −0.04 −0.02 0 0.02 0.04 0.06
X X X
0.06 0.06 0.06
0.04 0.04 0.04
0.02 0.02 0.02

𝛿/4
Y
Y
Y
0 0 0
−0.02 −0.02 −0.02
−0.04 −0.04 −0.04
−0.06 −0.06 −0.06

−0.06 −0.04 −0.02 0 0.02 0.04 0.06 −0.06 −0.04 −0.02 0 0.02 0.04 0.06 −0.06 −0.04 −0.02 0 0.02 0.04 0.06
X X X
0.06 0.06 0.06
0.04 0.04 0.04
0.02 0.02 0.02

𝛿/2
Y
Y
Y
0 0 0
−0.02 −0.02 −0.02
−0.04 −0.04 −0.04
−0.06 −0.06 −0.06

−0.06 −0.04 −0.02 0 0.02 0.04 0.06 −0.06 −0.04 −0.02 0 0.02 0.04 0.06 −0.06 −0.04 −0.02 0 0.02 0.04 0.06
X X X
Figure 28: Circulation phenomenon for porous-based split elliptical fins with varying LO’s∗ at particular TO’s∗ for ɸ = 0:92 at Re = 1839.
1.8
At Ø = 0.92
10 PPI
1.6
1.4
PEC
1.2
0.8
Re
783
1174
1839
Figure 29: Comparison of performance evaluation criteria at ɸ = 0:92 with varying LO’s∗ and at fixed TO’s∗ for different flow velocities.
1.80
At Ø = 0.9
10 PPI
1.60
1.40
PEC
1.20
1.00
0.80
Re
783
1174
1839
whereas 783 ≤ Re ≤ 1839, TO∗ = TO/δ, LO∗ = LO/δ and search algorithm is further employed to optimize the
Re = ðU inlet × δÞ/ϑ: problem. The cuckoo search method is a nature-based opti-
mization technique, i.e., based upon the cuckoo bird’s repro-
3.3.1. Cuckoo Search Optimization Algorithm. The response ductive cum different hatching natures. The cuckoo bird
surface method-based single-objective function in this case does not lay their egg in their own nest but in the host bird
has been executed via the cuckoo search algorithm. A broad house nest. The nature of laying the egg in the host bird’s
range of engineering-based optimization problems can be nest and getting hatched by the host bird has inspired the
resolved via cuckoo search optimization. While executing researcher to read and do research. The simple steps
extensive global optimization, it utilizes the full advantage followed by the cuckoo bird for their eggs are shown below:
of the switching parameter in order to maintain local and
global variables stable and consistent. The current work (1) Each cuckoo will place its egg in the randomly opted
relies on the single objective function (Nu), and the cuckoo host bird’s nest
1.8
At Ø = 0.8
10 PPI
1.6
1.4
PEC
1.2
0.8
Re
783
1174
1839
(2) The best quality of the eggs is carried out for a fur- 69.5
ther generation, and those nests will be treated as
69
the best nest
(3) Finally, upon defining the number of host nests 68.5
properly, the host bird nest may identify a different
Nu
68
egg with a probability of Pa ϵ (0, 1), and then that
egg may get destroyed, or the bird will leave the pres- 67.5
ent nest and move away 67
The establishment of the cuckoo bird offspring in the 66.5
algorithm is carried out using Levy’s flight equation, men- −10 0 10 20 30 40 50
tioned in the following equation. Number of iteration
Figure 32: Convergence curve for Nu variation using the cuckoo

xi ðt + 1Þ = xi ðt Þ + βt –λ , ð22Þ search algorithm.
where β > 0 and 1 ≤ λ ≤ 3.

The geometrical optimization using CSA is carried out Step 4: now, using Equation (21), establish the fitness
for the split elliptical-shaped fins to maximize the Nu and value (Nu) for the entire nest.
obtain the optimum value of the taken variables. The CSA Step 5: now, examine the whole nest and look for the
is coded in MATLAB software and performed on a highly nest showing the maximum Nu.
configured workstation system available at the Institute lab- Step 6: use the Levy flight Equation (22) to obtain a ran-
oratory. The detailed steps for carrying out the CSA with domly generated new nest.
input variables TO∗ , LO∗ , Re, and ɸ are mentioned below. Step 7: now, again, mark out the fitness value (Nu) of the
Step 1: set up the population size (N = 20) and a maxi- new nest, i.e., xi ðt + 1Þ.
mum number of iterations (t max = 50), along with λ = 2 Step 8: further, compare the obtained Nu of the new nest
and Pa = 0:25. xi ðt + 1Þ with the randomly opted nest from the population,
Step 2: secondly, create random nests (in N number). say, xi ðtÞ.
Now, the desired location of each randomly created nest is Step 9: further, if the value of Nu for the newly opted
marked off with the randomly opted input variables as nest is greater than the Nu value of xi ðtÞ, then replace it with
TO∗ , LO∗ , Re, and ɸ within limits. xi ðt + 1Þ.
Step 3: now, initiate the iterative step, i.e., with t = 0 and Step 10: now, arrange the nest in decreasing order of
further move with t = t + 1. their desired Nu value.
Table 3: Model validation with optimal parameter settings.
Parameters Re ɸ TO∗ LO∗ Nu

Predicted 1839 0.92 0.4141 0 69.3354
Computational experiment 1839 0.92 0.4141 0 70.1810
Error percentage (%) 1.20
70 because tortuosity within the fin increases with

+5% porosity, thus improving solid-fluid interaction
60
(2) The value of Nu increases after splitting the fin in the
perpendicular flow direction
−5%
Nu Predicted
50
(3) The highest Nu and performance evaluation criteria
40
value is obtained at TO∗ = 0:5 and LO∗ = 0 at ɸ =
0:92 (at PPI = 10), which are approximately 54%
and 79% higher than the solid circular fin at Re =
30
1839
20 (4) The value of C f reduces as the porosity increases
20 30 40 50 60 70 from 0.8 to 0.92. It is mainly because of the reduction
Nu Numerical in the solid-based fiber upon increasing porosity
Figure 33: Comparison plot between predicted and numerically (5) Modified split-based elliptical fins have shown their
obtained Nu values. influence on C f . Upon increasing the width of the
split in the transverse direction, the C f reduces for
all the considered values of porosity
Step 11: now, from the population of the randomly opted
nests, deduce the fraction (Pa) of the worst nest with the (6) The highest value of PEC (=/1.78) is obtained by
newly generated nest showing a higher value of Nu than transverse offset = δ/2 at 10 PPI and ɸ = 0:92, which
the worst nests. is 78.1% higher than the solid circular fin at its
Step 12: further, move to the next step 3 if the examined respective Re number
number of iterations is lesser than the maximum; else, pro- (7) RSM is used to develop a correlation to determine
ceed to the next. the value of Nu depending on the input variables
Step 13: at last, procure the final best solution and its
related value of Nu. (8) The optimal Nu is accounted for using cuckoo opti-
The convergence curve of the opted algorithm, while mization, with an error of 1.20% compared to
completing the present work is shown in Figure 32. Result numerical work
reveals that at TO∗ = 0:4141 and LO∗ = 0 with porosity as
0.92 and Re = 1839, the optimum value of Nu comes out to Nomenclature
be 69.33, and the comparison with the numerical work is
shown in Table 3, with an error of 1.2%. At last, a compar- ANN: Artificial neural network
ative analysis is done between the predicted and numerically RSM: Response surface method
procured Nu. The obtained results are within ±5%, as shown Cf : Frictional factor
in Figure 33. Cf o: Friction factor of base case
cf : Specific heat of fluid (J/kg K)
4. Conclusions cs : Specific heat of solid (J/kg K)
CSA: Cuckoo search algorithm
Initially, circular fins mounted over the base plate in a stag- C1 : Coefficient constant
gered manner are validated [29, 42, 43]. The maximum and D: Diameter of channel (mm)
minimum errors are 6.22% and 0.54% at the highest and df : Solid fiber diameter (mm)
lowest examined inlet flow velocities [29]. Further, work is dp : Pore diameter (mm)
forayed toward analyzing the porous nature of elliptical fins F: Forchheimer’s constant/inertial parameter
and their modification in terms of splits. The following point G: Shape function
is drawn out from the numerical study: GA: Genetic algorithm
h: Convective heat transfer coefficient (W/m2 K)
(1) The Nu number increases as the porosity changes HTS: Heat transfer search algorithm
from 0.8 to 0.92 for the considered fins. It may be Hd: Duct height (mm)
hf : Height of fin (mm) F: Fin

kf : Effective thermal conductivity of fluid (W/m K) i: Inlet
ks : Thermal conductivity of solid (W/m K) l: Longitudinal
lb : Length of base plate (mm) max: Maximum
lc : Characteristic length, (mm) o: Base case
Ld : Duct length (mm) p: Pore
LO∗ : Dimensionless longitudinal offset S: Solid
N: Population size T: Transverse.
Nu: Nusselt number
Nuo : Nusselt number of base cases Data Availability
P: Outlet pressure (Pa)
Pa : Probability factor used in CSA The data used to support the findings of this study are
PCM: Phase change material included within the article.
PEC: Performance evaluation criteria
Pr: Dimensionless Prandtl number Conflicts of Interest
PRESTO: Pressure staggering option
Pn : Permeability The authors declare that they have no conflict of interest.
PPI: Pores per inch
Pr: Prandtl number Acknowledgments
PSO: Particle search optimization
qz : Heat flux (W/m2) The authors honestly thank the Indian Institute of Technol-
Re: Reynolds’ number ogy Ropar, the National Institute of Technology Agartala,
Sl : Longitudinal pitch (mm) and the Karpagam Academy of Higher Education, India,
St : Transverse pitch (mm) for providing the facilities to carry out the research work.
T: Temperature (K)
T 1: Constant inlet section temperature (K)
References
tb: Base plate thickness (mm)
t: Number of iterations [1] T. M. Jeng, S. C. Tzeng, and Y. H. Hung, “An analytical study
TO∗ : Dimensionless transverse offset of local thermal equilibrium in porous heat sinks using fin the-
t max : Maximum number of iterations ory,” International Journal of Heat and Mass Transfer, vol. 49,
U max : Maximum velocity of air (m/s) no. 11-12, pp. 1907–1914, 2006.
U ∞: Inlet Free stream velocity of air (m/s) [2] A. S. Orhan, “Heat transfers and pressure drops for porous-
! ring turbulators in a circular pipe,” Applied Energy, vol. 83,
V: velocity vector
no. 3, pp. 280–298, 2006.
Wd: Duct width (mm)
[3] K. C. Leong and L. W. Jin, “Effect of oscillatory frequency on
x: Nest in CSA
heat transfer in metal foam heat sinks of various pore densi-
X, Y, Z: Fin coordinates (m). ties,” International Journal of Heat and Mass Transfer,
vol. 49, no. 3-4, pp. 671–681, 2006.
Greek Symbols
[4] N. Targui and H. Kahalerras, “Analysis of fluid flow and heat
α: Thermal diffusivity (m2/s) transfer in a double pipe heat exchanger with porous struc-
β: Step size in CSA tures,” Energy Conversion and Management, vol. 49, no. 11,
pp. 3217–3229, 2008.
λ: CSA parameter
[5] S. Z. Shuja, B. S. Yilbas, and M. Kassas, “Flow over porous
μ: Dynamic viscosity (Pa-s)
blocks in a square cavity: influence of heat flux and porosity
ϑ: Kinematic viscosity (m2/s) on heat transfer rates,” International Journal of Thermal Sci-
ρ: Density (kg/m3) ences, vol. 48, no. 8, pp. 1564–1573, 2009.
τs : Shear stress (N/m2) [6] H. Y. Li, K. C. Leong, L. W. Jin, and J. C. Chai, “Analysis of
χ: Tortuosity fluid flow and heat transfer in a channel with staggered porous
ɸ: Porosity blocks,” International Journal of Thermal Sciences, vol. 49,
δ: Diameter of cylindrical fin (mm) no. 6, pp. 950–962, 2010.
∇: Gradient [7] A. A. Mehrizi, M. Farhadi, K. Sedighi, and M. A. Delavar,
ϵ: Error term. “Effect of fin position and porosity on heat transfer improve-
ment in a plate porous media heat exchanger,” Journal of the
Subscripts Taiwan Institute of Chemical Engineers, vol. 44, no. 3,
pp. 420–431, 2013.
b: Base [8] A. Khaled, H. F. Oztop, and S. Kiwan, “Effects of porosity and
c: Characteristics thickness of porous sheets on heat transfer enhancement in a
d: Duct cross flow over heated cylinder,” International Communica-
f: Fiber tions in Heat and Mass Transfer, vol. 38, no. 9, pp. 1279–
f: Fluid 1282, 2011.
[9] A.-R. A. Khaled, “Heat transfer enhancements using pin fins- of Heat and Mass Transfer, vol. 46, no. 22, pp. 4191–4199,
porous-media systems,” Journal of Porous Media, vol. 14, 2003.
no. 12, pp. 1059–1075, 2011. [25] S. M. Kuo and C. L. Tien, “Heat transfer augmentation in a
[10] H. Du, D. Lu, J. Qi et al., “Heat dissipation performance of porous foam-material filled duct with discrete heat sources,” in Inter-
copper with elongated cylindrical pores,” Journal of Materials Society Conference on Thermal Phenomena in the Fabrication
Science and Technology, vol. 30, no. 9, pp. 934–938, 2014. and Operation of Electronic Components. I-THERM '88,
[11] B. Morkos, S. V. S. Dochibhatla, and J. D. Summers, “Effects of pp. 87–91, Los Angeles, CA, USA, 1988.
metal foam porosity, pore size, and ligament geometry on fluid [26] S. Kiwan, H. Alwan, and N. Abdelal, “An experimental inves-
flow,” Journal of Thermal Science and Engineering Applica- tigation of the natural convection heat transfer from a vertical
tions, vol. 10, no. 4, article 041007, 2018. cylinder using porous fins,” Applied Thermal Engineering,
[12] T. Ramakrishna, S. Ghosh Moulic, and A. Bhattacharya, vol. 179, article 115673, 2020.
“Experimental investigation of buoyancy-induced convection [27] P. L. Ndlovu and R. J. Moitsheki, “Steady state heat transfer
in high-porosity open-cell aluminum metal foams under dif- analysis in a rectangular moving porous fin,” Propulsion and
ferent orientations,” Journal of Thermal Science and Engineer- Power Research, vol. 9, no. 2, pp. 188–196, 2020.
ing Applications, vol. 13, no. 6, article 061003, 2021. [28] B. Kundu, D. Bhanja, and K. S. Lee, “A model on the basis of
[13] S. A. Upalkar, S. Gakhar, and S. Krishnan, “Analysis of fluid analytics for computing maximum heat transfer in porous
flow and heat transfer in corrugated perforated plate fin heat fins,” International Journal of Heat and Mass Transfer,
sinks,” Journal of Thermal Science and Engineering Applica- vol. 55, no. 25-26, pp. 7611–7622, 2012.
tions, vol. 14, no. 6, article 061013, 2022. [29] W. A. Khan, J. R. Culham, and M. M. Yovanovich, “Modeling
[14] S. Madhavan, P. Singh, and S. V. Ekkad, “Jet impingement heat of cylindrical pin-fin heat sinks for electronic packaging,” IEEE
transfer enhancement by packing high-porosity thin metal Transactions on Components and Packaging Technologies,
foams between jet exit plane and target surface,” Journal of vol. 31, no. 3, pp. 536–545, 2008.
Thermal Science and Engineering Applications, vol. 11, no. 6, [30] P. M. Kamath, C. Balaji, and S. P. Venkateshan, “Heat transfer
article 061016, 2019. enhancement with discrete heat sources in a metal foam filled
[15] A. Ranjan, S. S. Yadav, and K. Das, “Thermal and flow analysis vertical channel,” International Communications in Heat and
of split drop-shaped pin fins for improved heat transfer rate,” Mass Transfer, vol. 53, pp. 180–184, 2014.
Journal of The Institution of Engineers (India): Series C, [31] S. Sridharan, R. Srikanth, and C. Balaji, “Multi objective geo-
vol. 101, no. 2, pp. 375–382, 2020. metric optimization of phase change material based cylindrical
[16] A. Ranjan, R. Das, S. Pal, A. Majumder, and M. Deb, “Use of heat sinks with internal stem and radial fins,” Thermal Science
cuckoo search algorithm for performance evaluation of Split and Engineering Progress, vol. 5, pp. 238–251, 2018.
elliptic shaped fins for enhanced rate of heat transfer,” Journal [32] K. Udayraj, P. Mulani, A. D. Talukdar, and R. Alagirusamy,
of Heat Transfer, vol. 143, no. 6, article 063301, 2021. “Performance analysis and feasibility study of ant colony opti-
[17] S. J. Kim, D. Kim, and D. Y. Lee, “On the local thermal equilib- mization, particle swarm optimization and cuckoo search
rium in microchannel heat sinks,” International Journal of algorithms for inverse heat transfer problems,” International
Heat and Mass Transfer, vol. 43, no. 10, pp. 1735–1748, 2000. Journal of Heat and Mass Transfer, vol. 89, pp. 359–378, 2015.
[18] Y. Li, L. Gong, M. Xu, and Y. Joshi, “Hydraulic and thermal [33] R. V. Rao and A. Saroj, “Constrained economic optimization
performances of metal foam and pin fin hybrid heat sink,” of shell-and-tube heat exchangers using elitist-Jaya algorithm,”
Applied Thermal Engineering, vol. 166, article 114665, 2020. Energy, vol. 128, pp. 785–800, 2017.
[19] T. Zhao, H. Tian, L. Shi et al., “Numerical analysis of flow char- [34] R. V. Rao and A. Saroj, “Economic optimization of shell-and-tube
acteristics and heat transfer of high–temperature exhaust gas heat exchanger using Jaya algorithm with maintenance consider-
through porous fins,” Applied Thermal Engineering, vol. 165, ation,” Applied Thermal Engineering, vol. 116, pp. 473–487, 2017.
article 114612, 2020. [35] H. V. Ayala, P. Keller, M. de Fátima Morais, V. C. Mariani,
[20] P. C. Huang and K. Vafai, “Analysis of forced convection L. dos Santos Coelho, and R. V. Rao, “Design of heat
enhancement in a channel using porous blocks,” Journal of exchangers using a novel multiobjective free search differential
Thermophysics and Heat Transfer, vol. 8, no. 3, pp. 563–573, evolution paradigm,” Applied Thermal Engineering, vol. 94,
1994. pp. 170–177, 2016.
[21] J. Yang, M. Zeng, and Q. Wang, “Numerical study on forced [36] A. Masoud, Y. Song, B. Sunden, and G. Xie, “Economic opti-
convective heat transfer in porous pin fin channels,” in Pro- mization design of shell-and-tube heat exchangers by a
ceedings of the ASME 2008 Heat Transfer Summer Conference cuckoo-search-algorithm,” Applied Thermal Engineering,
collocated with the Fluids Engineering, Energy Sustainability, vol. 73, no. 1, pp. 1032–1040, 2014.
and 3rd Energy Nanotechnology Conferences. Heat Transfer: [37] B. D. Raja, V. Patel, and R. L. Jhala, “Thermal design and opti-
Volume 2, pp. 11–21, Jacksonville, Florida, USA, 2008. mization of fin-and-tube heat exchanger using heat transfer
[22] V. V. Calmidi and R. L. Mahajan, “Forced convection in high search algorithm,” Thermal Science and Engineering Progress,
porosity metal foams,” Journal of Heat Transfer, vol. 122, vol. 4, pp. 45–57, 2017.
no. 3, pp. 557–565, 2000. [38] T. L. Bergman, T. L. Bergman, F. P. Incropera, D. P. Dewitt,
[23] W. S. Fu, H. C. Huang, and W. Y. Liou, “Thermal enhance- and A. S. Lavine, Fundamentals of Heat and Mass Transfer,
ment in laminar channel flow with a porous block,” Interna- John Wiley & Sons, USA, 7th edition, 2011.
tional Journal of Heat and Mass Transfer, vol. 39, no. 10, [39] M. Mesgarpour, A. Heydari, and S. Saedodin, “Numerical
pp. 2165–2175, 1996. analysis of heat transfer and fluid flow in the bundle of porous
[24] K. H. Ko and N. K. Anand, “Use of porous baffles to enhance tapered fins,” International Journal of Thermal Sciences,
heat transfer in a rectangular channel,” International Journal vol. 135, pp. 398–409, 2019.
[40] Y. Vazifeshenas, K. Sedighi, and M. Shakeri, “Heat transfer in

PEM cooling flow field with high porosity metal foam insert,”
Applied Thermal Engineering, vol. 147, pp. 81–89, 2019.
[41] J. Yang, M. Zeng, Q. Wang, and A. Nakayama, “Forced con-
vection heat transfer enhancement by porous pin fins in rect-
angular channels,” Journal of Heat Transfer, vol. 132, no. 5,
article 051702, 2010.
[42] A. P. Colburn, “A method of correlating forced convection
heat transfer data and a comparison with fluid friction,” Trans-
actions of the American Institute of Chemical Engineers, vol. 29,
pp. 174–210, 1933.
[43] H. Hausen, Heat Transfer in Counter Flow Parallel Flow and
Cross Flow, McGraw-Hill, Inc, New York, 1983.
[44] G. Venugopal, C. Balaji, and S. P. Venkateshan, “Experimental
study of mixed convection heat transfer in a vertical duct filled
with metallic porous structures,” International Journal of
Thermal Sciences, vol. 49, no. 2, pp. 340–348, 2010.

Research Article Heat Transfer and Performance Enhancement of Porous Split Elliptical Fins

Uploaded by

Copyright:

Available Formats

Research Article Heat Transfer and Performance Enhancement of Porous Split Elliptical Fins

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Research Article Heat Transfer and Performance Enhancement of Porous Split Elliptical Fins

Uploaded by

Copyright:

Available Formats

Hindawi

International Journal of Energy Research

Correspondence should be addressed to Debabrata Barik; debabrata93@gmail.com

Academic Editor: Guojun Yu

Table 2: Grid independency test for circular and porous based 40

Number of Average ﬁn surface 35

Numerical (current work)

cally heated cylinder under free convection. Two diﬀerent 322

0.06 0.06 0.06 0.06

0.04 0.04 0.04 0.04

0.02 0.02 0.02 0.02

−0.02 −0.02 −0.02 −0.02

−0.04 −0.04 −0.04 −0.04

−0.06 −0.06 −0.06 −0.06

0.06 0.06 0.06 0.06

−0.02 −0.02 −0.02 −0.02

−0.04 −0.04 −0.04 −0.04

−0.06 −0.06 −0.06 −0.06

extend the work, the elliptical-shaped ﬁn with the same vol- !

0.04 0.04 0.04 0.04

0.02 0.02 0.02 0.02

−0.02 −0.02 −0.02 −0.02

−0.04 −0.04 −0.04 −0.04

−0.06 −0.06 −0.06 −0.06

0.06 0.06 0.06 0.06

0.04 0.04 0.04 0.04

0.02 0.02 0.02 0.02

−0.02 −0.02 −0.02 −0.02

−0.04 −0.04 −0.04 −0.04

−0.06 −0.06 −0.06 −0.06

0.06 0.06 0.06 0.06

0.04 0.04 0.04 0.04

0.02 0.02 0.02 0.02

−0.02 −0.02 −0.02 −0.02

−0.04 −0.04 −0.04 −0.04

−0.06 −0.06 −0.06 −0.06

0.04 0.04 0.04 0.04

0.02 0.02 0.02 0.02

−0.02 −0.02 −0.02 −0.02

−0.04 −0.04 −0.04 −0.04

−0.06 −0.06 −0.06 −0.06

0.06 0.06 0.06 0.06

0.04 0.04 0.04 0.04

0.02 0.02 0.02 0.02

−0.02 −0.02 −0.02 −0.02

−0.04 −0.04 −0.04 −0.04

−0.06 −0.06 −0.06 −0.06

0.06 0.06 0.06 0.06

0.04 0.04 0.04 0.04

0.02 0.02 0.02 0.02

−0.02 −0.02 −0.02 −0.02

−0.04 −0.04 −0.04 −0.04

−0.06 −0.06 −0.06 −0.06

Similarly, for measuring the losses (friction) in the compu-

𝛷 𝛿/8 𝛿/6 𝛿/4 𝛿/2

0.04 0.04 0.04 0.04

0.02 0.02 0.02 0.02

−0.02 −0.02 −0.02 −0.02

−0.04 −0.04 −0.04 −0.04