Professional Documents
Culture Documents
Rao 2006
Rao 2006
Rao 2006
www.elsevier.com/locate/ijhmt
a
Department of Chemical Engineering, Andhra University, Visakhapatnam 530 003, India
b
Department of Mechanical Engineering, G.V.P. College of Engineering, Visakhapatnam, India
c
Department of Mechanical Engineering, JNTU, Kukatpally, Hyderabad, India
Abstract
The problem of natural convection heat transfer from a horizontal fin array is theoretically formulated by treating the adjacent inter-
nal fins as two-fin enclosures. A conjugate analysis is carried out in which the mass, momentum and energy balance equations for the
fluid in the two-fin enclosure are solved together with the heat conduction equations in both the fins. The numerical solutions by using
alternating direction implicit (ADI) method yield steady state temperature and velocity fields in the fluid, and temperatures along the fins.
Each end fin of the array is exposed to limited enclosure on one side and to infinite fluid medium on the other side. Hence a separate
analysis is carried out for the problem of end fin exposed to infinite fluid medium with appropriate boundary conditions. From the
numerical results, the heat fluxes from the fins and the base of the two-fin enclosure, and the heat flux from the end fin are calculated.
Making use of the heat fluxes the total heat transfer rate and average heat transfer coefficient for a fin array are estimated. Heat transfer
by radiation is also considered in the analysis. The results obtained for a four-fin array are compared with the experimental data available
in literature, which show good agreement. Numerical results are obtained to study the effectiveness for different values of fin heights,
emissivities, number of fins in a fixed base, fin base temperature and fin spacing. The numerical results are subjected to non-linear regres-
sion and equations are obtained for heat fluxes from the two-fin enclosure and single fin as functions of Rayleigh number, aspect ratio
and fin emissivity. Also regression equations are obtained to readily calculate the average Nusselt number, heat transfer rate and effec-
tiveness for a fin array.
2006 Elsevier Ltd. All rights reserved.
Keywords: Fin array; Natural convection; Radiation; Conjugate problem; Enclosure; Heat transfer
1. Introduction which runs several hundreds of miles. Heat pipes are also
used along with fins to enhance cooling rate. In various
Extended surfaces, which are popularly known as fins, applications heat from the fins is dissipated by natural as
are extensively used in air-cooled automobile engines and well as forced convection and radiation. It is observed that
in air-cooled aircraft engines. Fins are also used for the radiation contributes up to 20% of the total heat dissipa-
cooling of computer processors, and other electronic tion. Fins are used as arrays in all the applications. The heat
devices. Fins are used in the cooling of oil carrying pipe line transfer rates from fin arrays are found to be lower com-
pared to those predicted theoretically for the case of single
fin. The heat transfer rate from the fin array is found to
*
depend on the fin spacing.
Corresponding author. Tel.: +91 891 254 0391.
E-mail addresses: v.dharmarao@yahoo.com (V.D. Rao), svnayudu@
There has been continuous research on improving the
yahoo.com (S.V. Naidu), govindarao_budda@rediffmail.com (B.G. Rao), efficiency of heat exchange systems using fins involving nat-
ces_jntu@yahoo.com (K.V. Sharma). ural convection and forced convection. Early experimental
0017-9310/$ - see front matter 2006 Elsevier Ltd. All rights reserved.
doi:10.1016/j.ijheatmasstransfer.2006.03.010
3380 V.D. Rao et al. / International Journal of Heat and Mass Transfer 49 (2006) 3379–3391
Nomenclature
AB area of the horizontal base plate, (BW), m2 y position coordinate normal to the fin measured
AR aspect ratio for two-fin enclosure, (H/S) from the left fin, m
Aw half the cross-sectional area of the fin, (tFW), m2 y+ normalized position coordinate normal to the
B breadth of the horizontal base plate, m fin, yGr1/4/S
Cp specific heat, J kg1 K1
F radiation shape factor Greek symbols
g acceleration due to gravity, m s2 a thermal diffusivity of fluid, (k/qc)f, m2 s1
+
Gr Grashof number, gbðT w;0 T 1 ÞS 3 =m2f a aw/af
h heat transfer coefficient, W m2 K1 b isobaric coefficient or thermal expansion coeffi-
J radiosity, W m2 cient of fluid, K1
k thermal conductivity, W m1 K1 e emissivity
H height of the fin, m c temperature ratio parameter, (Tw,0 T1)/T1
L* characteristic length defined in Eq. (36) l dynamic viscosity, kg m1 s1
M fin parameter, kfPSGr1/4/(kwAw) m kinematic viscosity, m2 s1
N number of fins in the fin array q density, kg m3
rT 4 S
NR radiation parameter, k f ðT w;01T 1 Þ Gr11=4 W stream function
p pressure, Pascals W+ normalized stream function, W/(mfGr1/4)
P half-perimeter of the fin, (2tF + W), m r Stefan–Boltzman constant, (5.67 · 108 W m2
Pr Prandtl number, cpflf/kf K4)
q heat flux, W m2 hw Tw/T1
q+ normalized heat flux, qS/[kf(Tw,0 T1)]Gr1/4 f vorticity function, s1
Q heat transfer rate, W f+ normalized vorticity function, S2f/(mfGr3/4)
Ra Rayleigh number, gb(Tw,0 T1)S3/(mfaw)
Ra* modified Rayleigh number defined in Eq. (36), Subscripts
gb(Tw,0 T1)L*3/(mfaw) 1 fin in a two-fin enclosure
S spacing between adjacent fins, m 3 base in a two-fin enclosure
tF half-thickness of the fin, m 5 end fin defined in Eq. (26a)
t time, s B base of the fin array
t+ normalized time, (mftGr1/2)/S2 B, 0 base plate in the absence of fins
T temperature, K E end fins of the fin array
T+ normalized temperature, (T T1)/(Tw,0 T1) I internal fins of the fin array
u velocity component in x-direction, m s1 c convection
u+ normalized velocity component in x-direction, f fluid
Su/(mfGr1/2) m average
v velocity component in y-direction, m s1 r radiation
v+ normalized velocity component in y-direction, T total
Sv/(mfGr1/4) w fin surface
W width of the fin (also that of base plate), m w, 0 base of the fin
x position coordinate along the fin measured from 1 ambient medium
the base of the fin, m
x+ normalized position coordinate along the fin,
x/S
work on free convection in horizontal and vertical fin fins was 10–20% of total heat transfer. Sparrow and
arrays was conducted by Starner and McManus [1], Well- Acharya [10] presented the article on a natural convection
ing and Wooldridge [2], Harahap and McManus [3], Jones fin with a solution—determined non-monotonically vary-
and Smith [4], Donvan and Roher [5], Van de pol and Tier- ing heat transfer coefficient. A conjugate conduction–
ney [6], and Bar-Cohen [7]. Radiation heat transfer plays convection analysis has been made for a vertical plate fin
an important role in heat transfer from fin arrays. Edwards which exchanges heat with its fluid environment by natural
and Chaddock [8] showed that heat transfer by radiation convection. They showed that the local heat transfer coef-
from cylindrical fins of surface emissivity of 0.99 accounted ficient decreased at first, attained a minimum and then
to one-third of the total heat transfer. Chaddock [9] found increased with increasing down stream distance. Saikhed-
that heat transfer by radiation from polished aluminum kar and Sukhatme [11] observed that convective heat trans-
V.D. Rao et al. / International Journal of Heat and Mass Transfer 49 (2006) 3379–3391 3381
fer increased and radiative heat transfer decreased with an sition procedure to determine the temperature distribution
increase in the Grashoff number. Manzoor et al. [12] ana- within a convective, radiative longitudinal fin of variable
lyzed the heat lost from the fins by convection and radia- thermal conductivity subject to convection heat transfer.
tion through one-dimensional and 2D approach. Sparrow Dayan et al. [28] carried out an experimental and analytical
and Vemuri [13] conducted experiments on highly popu- study for a downward-facing hot fin array. They found that
lated horizontal pin fins fixed to a vertical base plate. They the array length, fin spacing and surface temperature effect
showed that with fins, heat transfer increased by six times the heat transfer coefficient mostly compared to the fin
more than the un-finned surface and radiation contribution height. They also showed that the optimal fin spacing var-
was 25–40% of the total heat loss. The same authors [14] ies within a narrow range that depends primarily on the
also studied the effect of orientation of the fin array and length of the array. It can be observed that any two adja-
heat transfer by radiation was estimated based on isother- cent fins in a fin array form a rectangular enclosure. In this
mal fins. Guglielmini et al. [15] conducted experiments on context, a theoretical treatment of natural convection heat
staggered vertical fins and showed that heat transfer per- transfer in enclosures is dealt with by some investigators
formance of staggered vertical fins is superior to that [29,30] for the cases of non-porous and porous media.
obtained from fins made up of U-shaped vertical channels The appropriate mass, momentum and energy balance
of the same bulk volume. Aihara et al. [16] conducted equations for the fluid can be found in these papers.
experiments on pin fin arrays with a vertical base plate to A survey of literature reveals only a few theoretical stud-
study the velocity distributions around pin fin. Radiation ies on horizontal fin arrays. The theoretical study of
was estimated based on the apparent emissivity concept Baskaya et al. [25] on a horizontal fin array considers nat-
and a formula for apparent emissivity was also presented. ural convection heat transfer only. The radiation mode of
According to Zografos and Edward Sunderland [17] a heat transfer is not taken into account. Further the compu-
pin fin array performs better than a plate fin array under tational domain in their paper includes a single fin and half
the same conditions. Sunil Reddy and Sobhan [18] studied of fin spacing (i.e., S/2), thus precluding the adjacent fin or
the performance of a rectangular fin with a line source at fins. The experimental study of Yuncu and Anbar [23] indi-
base. The interaction of natural convection with radiation cates the effect of the adjacent fins on heat transfer. This
and conduction in a slot was numerically studied by Balaji point is also revealed by the experimental data of Rao
and Venkateshan [19]. Rao and Venkateshan [20] carried and Venkateshan [20], who found that the heat transfer
out experiments on horizontal fin arrays. They showed that rate from the end fins is higher compared to that from
much larger heat fluxes are in short fins than long fins. Fur- the internal fins in a four-fin array. These observations
ther the convective heat transfer increased linearly with fin prompted the authors to consider the heat transfer to the
spacing, while the increase in radiation heat transfer fluid in a limited enclosure from adjacent fins in a fin array,
showed non-linear trend. Lina and Leela [21] applied the and also to include radiation mode of heat transfer in their
second law to a pin fin array under cross flow, and analysis.
observed that an increase in the cross flow fluid velocity Hence the problem of natural convection heat transfer
would enhance the heat transfer rate and hence reduce from a horizontal fin array is theoretically tackled by treat-
the heat transfer irreversibility. Abramzon [22] presented ing the adjacent internal fins as two-fin enclosures. The
a discussion on the estimation of radiation heat transfer problem is formulated as a conjugate of convection in
from rectangular fin arrays, where in radiation contributes the fluid in the enclosure and heat conduction in the fins,
20% of the total heat transfer. Yuncu and Anbar [23] con- and is solved by alternating direction implicit (ADI)
ducted experiments by mounting different numbers of fins method. The case of heat transfer from the end fins is also
on a heated horizontal base plate of width 250 mm. The formulated and solved. From these numerical results the
fin spacing decreased as the number of fins was increased. total heat transfer rate from an N-fin array is estimated.
They found that for a given base-to-ambient temperature Heat transfer by radiation is also considered in the
difference the convection heat transfer rate from arrays analysis.
reaches a maximum at a particular fin spacing and fin
height. de lieto Vollaro et al. [24] investigated on the opti- 2. Physical model and formulation
mal configuration of rectangular, vertical fins mounted on
a plate. Baskaya et al. [25] solved the three-dimensional 2.1. Internal fins
elliptic governing equations for horizontal rectangular fin
arrays by finite volume based CFD code. They investigated The two adjacent internal fins having a common base
the effects of fin spacing, fin height, length and temperature are shown in Fig. 1(a). The base is maintained at a constant
on performance of rectangular fin arrays. Guvenc and temperature Tw,0, and Tw,0 > T1, where T1 is ambient air
Yuncu [26] carried out an experimental investigation on temperature. The space between the fins is S. The height
performance of fin array and found that higher heat trans- and width of each fin are H and W respectively, and the
fer enhancement is obtained with vertically oriented base ratio (W/H) is by far greater than unity. The half-thickness
than with horizontally oriented base for fin arrays of same of the fin is tF. Heat is transferred from the fins and the
geometry. Chiu and Chen [27] used a Adomain‘s decompo- base to the ambient air by convection and radiation.
3382 V.D. Rao et al. / International Journal of Heat and Mass Transfer 49 (2006) 3379–3391
The u- and v-momentum balance equations, i.e., Eqs. (4) gested the following conditions of velocity and temperature
and (5) are coupled making use of vorticity f, which is to be prescribed on this boundary.
defined as follows:
ou ow oT
v¼ ¼ ¼ f ¼ 0 and ¼0 ð16Þ
ou ov ox ox ox
f¼ ð9Þ
oy ox
A discussion on the above boundary condition is given in
The terms in Eq. (4) are differentiated with respect to y and Roache [32]. There also exist two more alternate conditions
those in Eq. (5) are differentiated with respect to x. The that may be used at the top boundary instead of Eq. (16). A
resulting equations are subtracted one from the other to discussion on these conditions at the top boundary and
yield the following vorticity equation: their effect on the results is presented in Appendix A.
2 The values of vorticity f at the four boundaries are
of of of oT o f o2 f derived making use of the relevant boundary conditions
qf þu þv ¼ qf gb þ lf þ ð10Þ
ot ox oy oy ox2 oy 2 following Roache [32] and are shown below.
The boundary conditions for the inner as well as end fins 3.2. Heat transfer rate from inner fins and the base
are also written in normalized form, but are not shown here
to conserve space. The heat fluxes in dimensionless notation from the left
fin and base are given by the following equations:
Z AR " #
3. Method of solution þ 1 oT þ ðEþ þ
b1 J 1 Þ
q1 ¼ þ þ NR dxþ ð24aÞ
AR 0 oy y þ ¼0 R1
The procedure used to obtain the temperature and Z Gr1=4 þ
1 1 oT þ ðEþ b3 J 3 Þ
velocity fields as follows. The temperature fields in the qþ ¼ þ N dy þ
Gr1=4 oxþ xþ ¼0
3 R
two-fin enclosure and on the fins are obtained by the solu- Gr1=4 0 R3
tion of the normalized energy balance equation for the ð24bÞ
fluid, i.e., Eq. (18) in conjugation with the heat conduction
where Eþ þ
b1 and E b3 are the black body irradiations in dimen-
equations for fins, i.e., Eqs. (19) and (20). Alternating 4 4
sionless form. Eþ þ þ þ
b1 ¼ ðcT w;1 þ 1Þ and E b3 ¼ ðc þ 1Þ . T w;1
direction implicit (ADI) method [32] is used to solve the
is the normalized fin temperature, which is a function of
energy balance equation for the fluid. By this method T+,
x+. c is the temperature ratio parameter. J þ þ
1 and J 3 are
the dimensionless temperatures in the fluid are updated in
the radiosities in normalized form. The radiosities are to
x-direction in the first-half time (Dt+/2), and then in y-
be calculated using the expressions given in Eq. (1). The
direction in the second-half time (Dt+/2) by solving Eq.
first and second terms in the above integrals represent the
(18). Eqs. (19) and (20) are used in the ADI method to
convection and radiation components of the heat fluxes
update the normalized temperatures on the fins. Subse-
in normalized form. The heat transfer fluxes q1 and q3 from
quently the dimensionless vorticities, f+ at all x+ and y+
the fin and the base are given by the following equations:
are updated by solving Eq. (21) using ADI method. The
boundary conditions stated in Eqs. (13)–(17) in normalized q1 ¼ Gr1=4 k f ðT w;0 T 1 Þqþ
1 =S ð25aÞ
form are used in the solution of the energy balance and q3 ¼ Gr 1=4
k f ðT w;0 T 1 Þqþ ð25bÞ
3 =S
vorticity equations. The solution of these equations yields
T+ and f+ fields at all grid points in the enclosure at each It may be noted that q1 refers to the heat flux from one face
time step, Dt+. Making use of f+, normalized stream func- of the fin (either left or right) of thickness tF.
tions w+ are obtained by solving Eq. (22) by successive over
relaxation (SOR) method. The normalized velocity compo- 3.3. Heat transfer from the end fins
nents u+ and v+ at all x+ and y+ (i.e., at all grid points) are
calculated by solving Eq. (23). The procedure is continued A separate analysis is carried out to obtain the heat
for successive time steps until steady temperature and transfer rate from both the end (or, external) fins. This
velocity fields are obtained at all grid points in the two- model consists of a fin exposed to an infinite medium of
fin enclosure. air, which is maintained at a constant temperature T1.
V.D. Rao et al. / International Journal of Heat and Mass Transfer 49 (2006) 3379–3391 3385
In order to use the ADI method, it is considered to be an QT, the total heat transfer rate from the fin array is given
imaginary enclosure, in which the right boundary is an out- by
flow boundary. Hence u ¼ oy ov
¼ ow ¼ f ¼ 0. However care is
oy QT ¼ QI þ QB þ QE ð31Þ
taken in choosing the value of S, which now represents the
thickness of the boundary layer. Runs are obtained with
S = 10, 15 and 20 mm and it is verified that the results 4.1. Average Nusselt number for the fin array, Num
do not depend on the value of S. The left boundary, is
the fin and hence the boundary condition is the same as Total heat transfer area for the N-fin array is given by
Eq. (13). The right boundary is outflow. The same bound-
AT ¼ ðB 2NtF ÞW þ 2NH ðW þ 2tF Þ ð32Þ
ary conditions are prescribed at the top boundary as in Eq.
(16). For the bottom boundary u = v = w = 0 and oT ox
¼ 0. The average heat transfer coefficient, hm for the N-fin array
At the bottom fjx¼0 ¼ 2w x¼Dx
and at the top fjx¼H ¼ 2wðDxÞ 2 .
x¼H Dx for combined convection and radiation is defined by the
ðDxÞ2
following equation:
The temperatures and vorticities are obtained by solving
the energy balance and vorticity equations, Eqs. (18) and QT
hm ¼ ð33Þ
(21) using ADI method subject to the boundary conditions AT ðT w;0 T 1 Þ
for the end fin. Eq. (19) is used to update the temperatures
on the fin. The stream functions and velocities are updated The average Nusselt number for the N-fin array is defined
with the aid of Eqs. (22) and (23). This procedure is contin- as
ued at successive time steps until steady state temperature Num ¼ hm S=k f ð34Þ
and vorticites are obtained at all grid points.
The dimensionless heat flux from the end fin qþ 5 is given
by the following equation in normalized form: 4.2. Effectiveness of the fin array, E
Z AR "
þ
#
1 oT The effectiveness of the fin array is defined as the ratio of
þ
4
qþ
5 ¼ þ e1 N R fðcT þ
w þ 1Þ 1g dx
þ
AR 0 oy y þ ¼0 heat transfer with fins to that without fins over the same
area of the base plate of breadth B and width W. The area
ð26aÞ of the base plate is given by
where e5 = e1. AB ¼ BW ð35Þ
The heat flux from the end fin is given by
QB,0, the heat transfer rate from the base plate in the ab-
q5 ¼ Gr1=4 k f ðT w;0 T 1 Þqþ
5 =S ð26bÞ sence of fins is calculated as follows. The Nusselt number
Here also it may be noted that q5 refers to the heat flux for natural convection heat transfer from the horizontal
from the outer face of the end fin (either first or the last plate is calculated using the equation available in literature
in the array) of half-thickness tF. [31] as follows:
The heat fluxes q1, q3 and q5 for the two-fin enclosure hB;0 L
and single fin are computed from the numerical results ¼ 0:54Ra0:25 ð36Þ
kf
using Eqs. (24)–(26) respectively. These are further used
where Ra* = gb(Tw,0 T1)L*3/(mfaf) and L* = AB/{2(B +
to calculate the total heat transfer rate, QT, the average
W)}. The equation shown above is valid in the range
Nusselt number, Num and effectiveness, E for a fin array
104 6 Ra* 6 107.
containing N fins by the procedure outlined below.
QB,0, the heat transfer rate from the horizontal plate by
convection and radiation is
4. Total heat transfer rate from a fin array, QT
QB;0 ¼ AB ½hB;0 ðT w;0 T 1 Þ þ e3 rðT 4w;0 T 41 Þ ð37Þ
The breadth of the horizontal base plate, on which the
fins are vertically mounted, is given by Thus the effectiveness of the fin array is
E ¼ QT =QB;0 ð38Þ
B ¼ ðN 1ÞS þ 2NtF ð27Þ
where B is the breadth of the plate and N is the number of 5. Results and discussion
fins.
The heat transfer rates from the inner fins of the array Validation of the present theoretical model is made by a
(QI), the base of the fin array (QB) and the two end fins comparison with the experimental data of four different
(QE) are given by references available in literature. Experimental data of
Rao and Venkateshan [20] for a four-fin array are shown
QI ¼ 2ðN 2ÞH ðW þ 2tF Þq1 ð28Þ
in Fig. 2 for two different fin emissivities, viz., e1 = 0.05
QB ¼ ðB 2NtF ÞWq3 ð29Þ and e1 = 0.85. The experimental data of Jones and Smith
QE ¼ 2H ðW þ 2tF Þðq1 þ q5 Þ ð30Þ [4] are also shown in Fig. 2. Numerical results are obtained
3386 V.D. Rao et al. / International Journal of Heat and Mass Transfer 49 (2006) 3379–3391
Table 1
(i) Effect of fin spacing S on heat fluxes q1 and q3 for a two-fin enclosure, and (ii) effect of number of fins on QT, E and hm for fin-array for B = 66 mm
Two-fin enclosure, Q5 = 253 W m2 Fin array, B = 66 mm, (QB,0/AB) = 255 W m2,
S = (B 2NtF)/(N 1)
S, mm Ra q1, W m2 q3, W m2 N QT, W E hm, W m2 K1
H = 30 mm, Tw,0 = 70 C, tF = 0.75 mm
20.0 20,371 149 379 4 3.24 3.84 5.29
11.4 3773 96 214 6 2.83 3.35 3.31
7.7 1163 71 194 8 2.78 3.30 2.54
4.4 217 55 168 12 2.98 3.50 1.89
2.8 56 44 148 16 3.04 3.61 1.48
V.D. Rao et al. / International Journal of Heat and Mass Transfer 49 (2006) 3379–3391 3387
Table 2
Effects of fin base temperature (Tw,0), fin spacing (S) and fin height (H) on (i) heat fluxes in a two-fin enclosure, and (ii) total heat transfer rate (QT) and
effectiveness (E) of fin array
Tw,0 C Two-fin enclosure Four-fin array
q1, W m2 q3, W m2 q5, W m2 QI, W QB, W QE, W QT, W E
H = 30 mm, S = 20 mm
50 53 57 99 0.32 0.17 0.46 0.95 2.53
80 187 239 318 1.12 0.72 1.51 3.35 3.02
104 316 413 516 1.90 1.24 2.50 5.64 3.04
H = 30 mm, S = 5 mm
50 22 4.50 111 0.14 0.01 0.40 0.55 1.48
80 64 13.4 319 0.38 0.04 1.15 1.57 1.41
104 102 22.4 505 0.61 0.07 1.82 2.50 1.35
H = 70 mm, S = 20 mm
50 34 17 79 0.48 0.05 0.79 1.32 3.56
80 136 82 251 1.90 0.25 2.71 4.86 4.37
104 239 141 405 3.34 0.42 4.51 8.27 4.46
Fig. 7. Isotherms and streamlines for the fluid in the two-fin enclosure at H = 70 mm, and Tw,0 = 104 C—effect of fin spacing for S = 10 and 25 mm.
3390 V.D. Rao et al. / International Journal of Heat and Mass Transfer 49 (2006) 3379–3391
0:41 0:16
S 1 þ e1
E ¼ 0:022Ra0:33 N 0:9 ð43Þ
H 1 þ NR
Since Num ¼ kfðQ T =AT ÞS
ðT w;0 T 1 Þ
, Eq. (42) can be used to obtain QT as
well. Eqs. (42) and (43) agree with the numerical results
within a standard deviation of ±3.3% for the range of
parameters mentioned above.
6. Conclusions
ou ow [12] M. Manzoor, D.B. Inham, P.J. Heggs, The one dimensional analysis
v¼ ¼ ¼f¼0 and T ¼ T1 ð16aÞ of fin assembly heat transfer, ASME J. Heat Transfer 105 (1983) 645–
ox ox
651.
Further in Eq. (16) the velocity component u is not made [13] E.M. Sparrow, S.B. Vemuri, Natural convection–radiation heat
zero, and hence ou/ox = 0 from the equation of continuity. transfer from highly populated pin –fin arrays, ASME J. Heat
The following two possible conditions emanate, if the con- Transfer 107 (1985) 190–197.
[14] E.M. Sparrow, S.B. Vemuri, Orientation effects on natural convec-
dition that u = 0 is used at the top boundary.
tive/radiation pin-fin arrays, Int. J. Heat Mass Transfer 29 (1986)
oT 359–368.
u ¼ v ¼ w ¼ 0; and ¼ 0 for 0 6 y 6 S ð16bÞ [15] G. Guglielmini, E. Nannel, G. Tanda, Natural convection and
ox
radiation heat transfer from staggered vertical fins, Int. J. Heat Mass
u ¼ v ¼ w ¼ 0; and T ¼ T 1 for 0 6 y 6 S ð16cÞ Transfer 30 (1987) 1941–1948.
[16] T. Aihara, S. Maruyama, S. Kobayakawa, Free convective–radiative
The boundary conditions given by Eqs. (16b) and (16c) heat transfer from pin-fin arrays with a vertical base plate, Int. J. Heat
were used in literature for the cases of both open and closed Mass Transfer 33 (1990) 1223–1232.
top boundaries. Numerical results are obtained making use [17] Antonios I. Zografos, J. Edward Sunderland, Natural convection
of the above three conditions also, but are not shown here from pin fin arrays, Exp. Therm. Fluid Sci. 3 (1990) 440–449.
to conserve space. It is observed that slightly higher Nusselt [18] J.R.N.S. Sunil Reddy, C.B. Sobhan, Natural convection heat transfer
from a thin rectangular fin with a live source at the base—a finite
numbers are obtained particularly at low values of S with difference solution, Heat Mass Transfer 31 (1996) 127–135.
the boundary condition given by Eq. (16a) compared to [19] C. Balaji, S.P. Venkateshan, Combined conduction, convection and
the results obtained using Eq. (16). An optimum fin spacing radiation in a slot, Int. J. Heat Fluid Flow 16 (1995) 139–144.
does not appear if the condition of T = T1 from Eq. (16a) [20] Rammohan Rao, S.P. Venkateshan, Experimental study of free
is implemented at the top boundary. convection and radiation in horizontal fin arrays, Int. J. Heat Mass
Transfer 39 (1996) 779–789.
[21] W.W. Lina, D.J. Leela, Second-law analysis on a pin-array under
References cross flow, Int. J. Heat Mass Transfer 40 (1997) 1937–1945.
[22] B.M. Abramzon, A simple closed-form solution for evaluation of
[1] K.E. Starner, H.N. McManus Jr., An experimental investigation of radiative heat transfer from a rectangular fin array, IEEE Trans.
free convection heat transfer from rectangular fin arrays, J. Heat Compon., Pack. Manufact. Technol., Part A 20 (2) (1997) 225–
Transfer 85 (1963) 273–278. 229.
[2] J.R. Welling, C.V. Wooldridge, Free convection heat transfer co- [23] H. Yuncu, G. Anbar, An experimental investigation on performance
efficients from rectangular fin arrays, ASME J. Heat Transfer 87 of fins on a horizontal base in free convection heat transfer, Heat
(1965) 439–444. Mass Transfer 33 (1998) 507–514.
[3] F. Harahap, H.N. McManus, Natural convection heat transfer from [24] Andrea de lieto Vollaro, Stefano Grignaffini, Franco Gugliermetti,
horizontal rectangular fin arrays, ASME J. Heat Transfer 89 (1967) Optimum design of vertical rectangular fin arrays, Int. J. Therm. Sci.
32–38. 38 (1999) 525–529.
[4] Charles D. Jones, Lester F. Smith, Optimum arrangement of [25] Senol Baskaya, Mecit Sivrioglu, Murat Ozek, Parametric study of
rectangular fins on horizontal surfaces for free convection heat natural convection heat transfer from horizontal rectangular fin
transfer, ASME J. Heat Transfer 92 (1970) 6–10. arrays, Int. J. Therm. Sci. 39 (2000) 797–805.
[5] R.C. Donvan, W.M. Roher, Radiative and convective conducting fins [26] A. Guvenc, H. Yuncu, An experimental investigation on performance
on a plane wall, including mutual irradiation, ASME J. Heat Transfer of fins on a horizontal base in free convection heat transfer, Heat
93 (1971) 41–46. Mass Transfer 37 (2001) 409–416.
[6] D.W. Van de pol, J.K. Tierney, Free convective heat transfer from [27] Ching-Hung Chiu, Cha’o-Kuang Chen, Application of Adomian’s
vertical fin arrays, IEEE Trans. PHP-10 (4) (1974) 267–271. decomposition procedure to the analysis of convective–radiative fins,
[7] A. Bar-Cohen, Fin thickness for an optimized natural convection ASME J. Heat Transfer 125 (2003) 312–316.
array of rectangular fins, Trans. ASME 101 (1979) 564–566. [28] A. Dayan, R. Kushnir, G. Mittelman, A. Ullmann, Laminar free
[8] J.A. Edwards, J.B. Chaddock, An experimental investigation of the convection underneath a downward facing hot fin array, Int. J. Heat
radiation and free convection heat transfer from a cylindrical disk Mass Transfer 47 (2004) 2849–2860.
extended surface, Trans. Am. Soc. Heat. Refrig. Air-condit. Eng. 69 [29] D. Angirasa, R.L. Mahajan, Natural convection from L-shaped
(1963) 313–322. corners with adiabatic and cold isothermal horizontal walls, ASME J.
[9] J.B. Chaddock, Freeconvection heat transfer from vertical fin arrays, Heat Transfer 115 (1993) 149–157.
ASHRAE J. 12 (1970) 53–60. [30] G. Lauriat, V. Prasad, Non-Darcian effects on natural convection in a
[10] E.M. Sparrow, S. Acharya, A natural convection fin with a solution— vertical porous enclosure, Int. J. Heat Mass Transfer 32 (1989) 2135–
determined nonmonotonically varying heat transfer coefficient, 2148.
ASME J. Heat Transfer 105 (1981) 218–225. [31] F.P. Incropera, D.P. Dewitt, Fundamental of Heat and Mass
[11] N.H. Saikhedkar, S.P. Sukhatme, Heat transfer from rectangular Transfer, fourth ed., John Wiley & Sons, 1996.
cross-sectioned vertical fin arrays, in: Proceedings of the sixth [32] P.J. Roache, Computational fluid dynamics, Hermosa, Albuquerque,
national heat and mass transfer conference, HMT, 1981, pp. 9–81. N.M., 1985.