International Journal of Heat and Mass Transfer 53 (2010) 45364542

Contents lists available at ScienceDirect

International Journal of Heat and Mass Transfer

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

Simple method for estimation of unsteady state conduction heat ow

with variable surface temperature in slabs and spheres
Alireza Bahadori *, Hari B. Vuthaluru
Curtin University of Technology, School of Chemical and Petroleum Engineering, GPO Box U1987, Perth, WA 6845, Australia

a r t i c l e i n f o a b s t r a c t

Article history: When a solid is exposed to a hot gas or liquid, the resistance to heat transfer in the uid is usually sig-
Received 29 January 2009 nicant, and the surface temperature changes with time. In this work, an attempt has been made to for-
Received in revised form 15 June 2010 mulate a simple-to-use method for one-dimensional heat ow with variable surface temperatures in
Accepted 15 June 2010
slabs and spheres as a function of the Fourier number and the Biot number in order to arrive at the tem-
Available online 16 July 2010
perature distribution in the solid and the average solid temperature changes with time. The average abso-
lute deviation between the reported data and the proposed correlations are found to be 1.8%.
Keywords:
2010 Elsevier Ltd. All rights reserved.
Heat conduction
Fourier number
Biot number
Unsteady state heat transfer

1. Introduction 1.1. Theoretical background

In many applications, the temperatures are varying with time, One-dimensional unsteady state heat ow with constant sur-
and it is a necessary to understand the complete time history of face temperature for various shapes has been the subject of many
temperature variations. For example, in metallurgical applications, research studies [17,18].
the heat treating process can be controlled to directly alter the For example in a large slab, the governing equation for one-
characteristics of processed materials. Annealing (slow cooling) dimensional unsteady state heat ow with constant surface tem-
can soften metals and improve ductility. On the other hand, perature is [19]:
quenching (rapid cooling) can harden the strain boundary and
@2T @T
increase strength. In order to characterize this transient behavior, kA dxdt qcp Adx dt 1
the full unsteady equation is needed; therefore modeling physical @x2 @t
problems dealing with heat transfer unsteady state conduction Or after division by qcpAdxdt:
heat ow with variable surface temperature in solids are of great
@T k @2T @2T
interest in many technological applications. Considerable work, a 2: 2
dealing with rigorous and complicated analytical or semi-analyti- @t qcp @x2 @x
cal [19] modeling is found in the literature. Useful analytical solu- General solutions of unsteady-state conduction equations with
tions for fast transient conduction in spherical and cylindrical constant surface temperature are available in literature for various
geometries [10] and innite plate [11] subject to sudden and vio- shapes [1921]. For example the integration of Eq. (2) for the heat-
lent thermal effects on its surface are presented and available in ing or cooling of an innite slab of known thickness from both
the literature. Some studies have been devoted to the analysis of sides by a medium at constant surface temperature gives a con-
the effect of a surface lm on the thermal behavior of a medium verging solution [19]:
subjected to heat sources [12]. Some other authors studied the  
heat transfer within thermal contact resistance or other conditions T s  T 4 a1 Fo px 1 3px 1 25a1 Fo 5p x
e sin e9a1 Fo sin e sin  :
[1316]. To date, however, there is no simple-to-use predictive tool Ts  Ta p 2s 3 2s 5 2s
available for an accurate and rapid estimation of one-dimensional 3
heat ow with variable surface temperatures in slabs and spheres.
Eq. (3) can be used to determine the mid-plane temperature by
setting x = s. The average temperature is obtained by integrating
* Corresponding author. Tel.: +61 8 9266 1782; fax: +61 8 9266 2681. Eq. (3) to give [19]:
E-mail addresses: [email protected], bahadori.alireza@
gmail.com (A. Bahadori).

0017-9310/$ - see front matter 2010 Elsevier Ltd. All rights reserved.
doi:10.1016/j.ijheatmasstransfer.2010.06.040
 A. Bahadori, H.B. Vuthaluru / International Journal of Heat and Mass Transfer 53 (2010) 45364542 4537

Nomenclature

A tuned coefcient Ta initial temperature (C)

B tuned coefcient Tb average temperature at end of tT
Bi Biot number, dimensionless Tf bulk of uid temperature (C)
C tuned coefcient Ts surface temperature (C)
D tuned coefcient t time (s)
Fo Fourier number, dimensionless tT time required to heat or cool, s
h individual or surface heat transfer coefcient (W/ u coefcient of polynomial
(m C)) V Vandermonde matrix
i index x distance from surface
j index
k thermal conductivity (W/(m C)) Greek letters
2
m matrix row index for m  n matrix a thermal diffusivity (ms )
n matrix column index for m  n matrix h
T f T b
(average temperature change)
T f T a
q heat ow rate (W)
rm radius of solid sphere (m) w polynomial
s half thickness of slab (m) q density (kg/m3)
T temperature (C)

 
Ts  Tb 8 1 1 25a1 Fo engineers for the prediction of one-dimensional heat ow with
2 ea1 Fo e9a1 Fo e  ; 4
Ts  Ta p 9 25 variable surface temperature in slabs and spheres to arrive at the
temperature distribution in the solid and the average solid temper-
where T b average temperature of slab at time tT and Fo ast2T . For ature change with time. According to the authors knowledge,
a sphere of radius rm and Fo art2T the corresponding equation is there is no simple-to-use predictive tool existing in the literature
m
[19]: for an accurate estimation of temperature distribution in the slabs
Ts  Tb and spheres and the average solid temperature change with time.
0:608e9:87Fo 0:125e39:5Fo 0:0676e88:8Fo    5 In view of this necessity, our efforts have been directed at formu-
Ts  Ta
lating a simple-to-use predictive tool which can help engineers
However, Eqs. (1)(5) apply only when surface temperature is and researchers. Example is shown for the benet of engineers to
constant, which is nearly true when the source or sink for heat is clearly demonstrate the usefulness of the proposed predictive tool.
a metal of high conductivity [19,20]. When a solid is exposed to The present study discusses the formulation of a simple correlation
a hot gas or liquid, the resistance to heat transfer in the uid is usu- which can be of signicant importance for the practice engineers.
ally signicant [19,20] and the surface temperature changes with The proposed method in this work provides easier solution for
time as typically sketched in Fig. 1. the problem in comparison with perturbation method. Using the
For one-dimensional heat ow with variable surface tempera- perturbation method an approximate perturbation solution is ob-
ture in slabs and spheres where there is an external resistance, tained by truncating the series, usually by keeping only the rst
Eq. (2) is still applies and the heat ux into solid is equated to two terms, the initial solution and the rst order perturbation cor-
the rate of heat transfer through the uid layer as given by New- rection. In fact perturbation theory comprises mathematical meth-
tons law for convective heat transfer ods that are used to nd an approximate solution to a problem
 
q @T cannot be solved exactly, by starting from the exact solution of a
hT f  T s k : 6 related problem [21]. The present study discusses the formulation
A @x x0
of a simple correlation which can be of signicant importance for
In light of the above mentioned issues, there is an essential need the practice engineers. Example is also shown for the benet of
to develop a practical, reliable and easy-to-use method for practice engineers to clearly demonstrate the usefulness of the proposed
Ts 3 t predictive tool.
Tf

2. Methodology to develop predictive tool

In this work average temperature change (h) is correlated as a

Ts 2 t3 function of the Fourier number and the Biot number using Vander-
monde matrix in order to get the temperature distribution in the
T s1 Insulation
solid and then the average solid temperature change with time.
The Fourier number (Fo) is a dimensionless time and the Biot num-
ber, (Bi) is a measure of the relative importance of external and
internal resistances [20]
t2
Tf  Tb
h : 7
Tf  Ta

For a at slab
t1
hs
Bi : 8
Fig. 1. Temperature gradients in and near a slab being heated from one side. k
4538 A. Bahadori, H.B. Vuthaluru / International Journal of Heat and Mass Transfer 53 (2010) 45364542

And for a sphere Table 1

Tuned coefcients used in Eqs. (16)(19) for slab shapes.
hrm
Bi : 9 Parameter Tuned values
k
A1 1.0019315965808  101
B1 4.43577642143564  101
2.1. Vandermonde matrix C1 5.55331692119167  101
D1 1.78523759041059 101
A2 2.73626955975173
Vandermonde matrix is a matrix with the terms of a geometric B2 4.35750159483566
progression in each row, i.e., an m  n matrix [22] and a is matrix C2 3.14592647242833
elements in Eq. (10). D2 7.73611271915438  101
2 3 A3 2.27452953315178  101
1 a2 a21    an1
1
B3 1.16488526715623  101
6 7 C3 1.88610642721109  10-1
61
6
a3 a21    an1
1
7
7 D3 9.58083782173845  10-2
V 6. .. .. .. .. 7 10 A4 3.18994047068607  10-2
6 .. . . . . 7
4 5 B4 1.19571998808609  10-1
1 am a2m    an1 C4 2.30357038304352  10-1
m
D4 8.1617702407838  10-2
Or

V i;j aj1 11 In brief, the following steps are repeated to tune the coefcients
i
of equations.
for all indices I and j. The determinant of a square Vandermonde
matrix (where m = n) can be expressed as [18]: 2. Correlate the average temperature change (h) as a function of
Y Fourier number (Fo) for a given Biot number, (Bi) variable value.
detV aj  ai : 12
3. Repeat step 1 for other values of Biot number, (Bi).
1i<jn
4. Correlate corresponding polynomial coefcients, which are
The Vandermonde matrix evaluates a polynomial at a set of obtained in previous steps versus Biot number, (Bi) variable,
points; formally, it transforms coefcients of a polynomial a0 + so we will have a = f(Bi), b = f(Bi), c = f(Bi), d = f(Bi) (see Eqs.
a1x + a2x2 +  + an1xn1 to the values the polynomial takes at the (16)(19)).
points ai. The non-vanishing of Vandermonde determinant for dis-
tinct points ai shows that, for distinct points, the map from coef- Eq. (15) presents a new correlation in which four coefcients
cients to values at those points is a one-to-one correspondence, are used to correlate the average temperature change (h) in slabs
and thus the polynomial interpolation problem is solvable with un- as a function of Fourier number (Fo) and Biot number, (Bi)
ique solution; this result is called the unisolvence theorem [23]. variables.
They are thus useful in polynomial interpolation, since solving
the system of linear equations Vu = y for u with V an m  n Vander- lnh a bFo cFo2 dFo3 ; 15
monde matrix is equivalent to nding the coefcients uj of the
where:
polynomial(s) [23]
B1 C 1 D1
X
n1
a A1 3; 16
wx uj xj 13 Bi B2i Bi
j0
B2 C 2 D2
b A2 3; 17
of degree 6 n  1 which has (have) the property: Bi B2i Bi
B3 C 3 D3
wai yi for i 1; . . . ; m: 14 c A3 3; 18
Bi B2i Bi
The Vandermonde matrix can easily be inverted in terms of La- B4 C 4 D4
grange basis polynomials: each column referring to the coefcients d A4 3: 19
Bi B2i Bi
of the Lagrange basis polynomial, with terms in increasing order
going down. The resulting solution to the interpolation problem The tuned coefcients used in Eqs. (16)(19) are given in Table
is called the Lagrange polynomial [23]. 1 for slabs. The above methodology is applied to predict tempera-
ture change in spheres. For Biot number more than 10 Eqs. (15)
2.2. Development of predictive tool (19) and tuned coefcients in rst column of Table 2 are applied.
However, for Biot number less than 10, Eq. (15) along with Eqs.
Many reliable results and data have been presented in the liter- (20)(23) are applied to calculate temperature change (h) variable
ature and are available in the heat transfer texts and handbooks as a function of Fourier number (Fo) and Biot number, (Bi):
[19,20]. The required data to develop the rst correlation includes
the reliable data [19]. The following methodology [24,25] using a A1 B1 Bi C 1 Bi 2 D1 Bi 3 ; 20
MATLAB software [26] has been applied to develop the simple pre- 2 3
b A2 B2 Bi C 2 Bi D2 Bi ; 21
dictive tool for slabs. 2 3
c A3 B3 Bi C 3 Bi D3 Bi ; 22
Firstly, average temperature change (h) variable is correlated
using Vandermonede matrix as a function of Fourier number (Fo) d A4 B4 Bi C 4 Bi 2 D4 Bi 3 : 23
values for different Biot number, (Bi) values. Then, the calculated
coefcients for these equations are correlated using Vandermon-
ede matrix as a function of Biot number, (Bi) variable. The derived 3. Results
equations are applied to calculate new coefcients for Eq. (15) to
predict average temperature change (h). Table 1 shows the tuned Fig. 2 illustrates the results of proposed method for predicting
coefcients for Eqs. (16)(19) according to the data for slabs [19]. the average temperature change for slabs in comparison with some
 A. Bahadori, H.B. Vuthaluru / International Journal of Heat and Mass Transfer 53 (2010) 45364542 4539

Table 2 typical data obtained from the literature [19]. Fig. 3 shows the
Tuned coefcients used in equations for prediction of temperature changes in sphere results of proposed method for predicting the average temperature
shapes.
change for spheres in comparison with some typical data [19] for
Symbol Values for Biot number less Values for Biot number more wide range of conditions. As can be seen in Figs. 2 and 3, the results
than 10 used in Eqs. (20)(23) than 10 used in Eqs. (16)(19) of the new proposed method are accurate and acceptable. Figs. 4
for sphere shape for sphere shape
and 5 show the excellent performance of proposed method in pre-
A1 3.6581957638640  102 2.40922677378281  101 diction of temperature distribution in the solid slab and sphere
B1 6.56581888060303  102 1.59681981442401  101
C1 8.60990550844736  103 1.88314713411153  102
shape, respectively, and the average solid temperature change with
D1 1.85159882461873  104 5.8303267275467  102 time. Table 3 shows the accuracy of the new proposed predictive
A2 8.87346858156626  102 1.37817321749942  101 tool which is a mathematically compact and reasonably accurate
B2 2.53113274811367 2.22870180130477  101 correlation containing few tuned coefcients, in comparison with
C2 2.7731187874796  101 7.49520794982756  102
reported data [19] The average absolute deviation and maximum
D2 1.33821400631763  02 2.53577303927977  103
A3 1.3336500125053  101 1.78413220908218  101 deviation between the reported data and the proposed correlations
B3 1.48461701463394  101 2.90094852101887  102 are found to be 1.8% and 4.3%, respectively, demonstrating the
C3 3.1844049403057  101 5.00906190744855  103 good performance of proposed correlation for predicting the
D3 1.9881642466118  102 1.65532835500561  104 average temperature change for slabs. The proposed correlation
A4 5.0440251273362  102 2.19021901273536  101
is simple to use, employing basic algebraic equations that can
B4 5.4719204426561  102 4.57386110899493  102
C4 1.399662723727  101 8.64668886674005103 easily and quickly be solved by spreadsheet. In addition, the
D4 3.72976097685201  103 2.95896663056101 104 estimates are quite accurate, as evidenced by the comparisons with
literature data [19]. Example [19] is given below to demonstrate

0.9
Bi=0.5
Average Temperature Change, (Tf-Tb)/(Tf-Ta)

Data
0.8 Bi=1
Data
0.7 Bi=2
Data
0.6 Bi=5
Data
0.5 Bi=10
Data
0.4 Bi=20
Data
0.3 Bi=Infinity
Data
0.2

0.1

0
0 0.5 1 1.5 2 2.5 3
Fourier Number

Fig. 2. Prediction of the temperature distribution in the solid slab shape and the average solid temperature change with time in comparison with literature data [19].

1
Bi=0.50
0.9
Average Temperature Change, (Tf-Tb)/(Tf-Ta)

Data
Bi=1
0.8
Data
0.7 Bi=2
Data
0.6 Bi=5
Data
0.5 Bi=10
Data
0.4 Bi=50
Data
0.3 Bi=Infinity
Data
0.2

0.1

0
0 0.2 0.4 0.6 0.8 1 1.2 1.4 1.6 1.8
Fourier Number

Fig. 3. Prediction of the temperature distribution in the solid sphere shape and the average solid temperature change with time in comparison with literature data [19].
4540 A. Bahadori, H.B. Vuthaluru / International Journal of Heat and Mass Transfer 53 (2010) 45364542

Average Temperature Change, (Tf-Tb)/(Tf-Ta)

0.9

0.8

0.7

0.6

0.5 Biot number=0.70

0.4

0.3

0.2

0.1

0
0 0.2 0.4 0.6 0.8 1 1.2 1.4 1.6
Fourier Number

Fig. 4. Performance of proposed tool for prediction of the temperature distribution in the solid slab shape and the average solid temperature change with time.

1
Avergae Temperature Change, (Tf-Tb)/(Tf-Ta)

0.9

0.8

0.7

0.6

0.5
Biot number=0.5
0.4

0.3

0.2

0.1

0
0.5 1 1.5 2 2.5 3
Fourier Number

Fig. 5. Performance of proposed tool for prediction of the temperature distribution in the solid sphere shape and the average solid temperature change with time.

the simplicity of proposed predictive tool for estimating heat ow 3.2. Solution
with variable surface temperature in slabs and spheres as a func-
tion of the Fourier number and the Biot number. The proposed tool
appears to be superior owing to its accuracy and simple back- T f  T b 30  35
h 0:10 from Eq 7:
ground, wherein the relevant coefcients can be retuned quickly T f  T a 30  80
if more data are available in the future. It is expected that our ef- hrm 50  2:5  103
forts in this investigation will pave the way for arriving at an accu- Bi 0:96 from Eq 9:
k 0:13
rate prediction of heat ow with variable surface temperature in
slabs and spheres to get the temperature distribution in the solid Because Biot number is less than 10 we use the rst column coef-
and the average solid temperature change with time at various cients in Table 2 and Eqs. 15, 20, 21, (22) and (23):
conditions. Assume Fo is 1.06

a 1:86788323123  102 from Eq: 20;

3.1. Example
b 2:274891157621 from Eq 21;
Plastic spheres 5 mm in size and originally at 80 C are cooled in c 2:850429866  101 from Eq 22;
a uidized bed where the air temperature is 30 C. The density of d 1:27783247787  101 from Eq 23;
solid is 1100 kg/m3, the thermal conductivity is 0.43 W/(m C)
Calculated h 0:104 from Eq 15:
and the specic heat is 1700 J/(kg C). The external heat transfer
coefcient is 50 W/(m C). How long will it take for the average so- Assume Fo=1.08 because a,b,c and d are just as a function of Biot
lid temperature to reach 35 C? number, so their values do not change.
 A. Bahadori, H.B. Vuthaluru / International Journal of Heat and Mass Transfer 53 (2010) 45364542 4541

Table 3 perature in slabs and spheres. In particular, mechanical and process

Comparison of proposed correlation with data. engineers would nd the proposed approach to be user-friendly
Fourier Biot Reported average Calculated average Absolute involving no complex expressions with transparent calculations as
number number temperature change
 
temperature
 
deviation evident from the cited example. The paper is concerned at present
(Fo) Bi, T T T T percent with one-dimensional heat ow with variable surface temperature
h T ff T ba data[19] change h T ff T ba
value value
in slabs and spheres. It would be useful to extend to two- and
0 0.5 1 0.994 0.6 three-dimensional congurations in future studies. The proposed
0.11 1 0.9 0.891 1
method is superior owing to its accuracy and clear numerical back-
0.172 2 0.8 0.814 1.75
0.172 5 0.7 0.692 1.14 ground, wherein the relevant coefcients can be retuned quickly for
0.1875 10 0.6 0.612 2 various cases.
0.234 20 0.5 0.517 3.4
0.3 Innity 0.4 0.4059 1.475
2.875 0.5 0.3 0.2983 0.57 Acknowledgements
2.172 1 0.2 0.2005 0.25
1.25 5 0.1 0.1006 0.6 The lead author acknowledges the Australian Department of
1 20 0.09 0.0888 1.33 Education, Science and Training for Endeavour International Post-
2.172 2 0.08 0.0800 0
graduate Research Scholarship (EIPRS), the Ofce of Research &
1.06 Innity 0.06 0.062 3.33
1.81 5 0.04 0.0385 3.75 Development at Curtin University of Technology, Perth, Western
2.19 10 0.01 0.0102 2 Australia for providing Curtin University Postgraduate Research
0.75 2 0.4 0.417 4.25 Scholarship and the State Government of Western Australia for
0.93 20 0.1 0.103 3
providing top up scholarship through Western Australian Energy
0.53 10 0.3 0.288 4
Average absolute deviation percent (AADP) 1.80 Research Alliance (WA:ERA). Useful comments from anonymous
reviewers and the editor are also acknowledged which led to
improvements in the original version of the paper.

a 1:86788323123  102 from Eq: 20 References

b 2:274891157621 from Eq: 21
[1] H. Belghazi, M. El Ganaoui, J.C. Labbe, Analytical solution of unsteady heat
c 2:85042986624  101 from Eq: 22 conduction in a two-layered material in imperfect contact subjected to a
moving heat source, Int. J. Therm. Sci. 49 (2010) 311318.
d 1:27783247787  101 from Eq: 23 [2] R. Brockmann, K. Dickmann, P. Geshev, K.J. Matthes, Calculation of
temperature eld in a thin moving sheet heated with laser beam. R, Int. J.
Then; the calculated h 0:0:9985from Eq: 15 Heat Mass Transfer 46 (2003) 717723.
[3] M.N. Ozisik, Heat Conduction, second ed., Wiley, John & Sons, 1993.
Which is in good greement with given h = 0.10. [4] N. Laraqi, An exact explicit analytical solution of steady-state temperature in a
So, Fo = 1.08 is correct and acceptable. half space submitted to a circular moving heat source, ASME J. Tribol. 125 (4)
(2003) 859862.
Then, we calculate thermal diffusivity:
[5] S.H. Kang, H.S. Cho, Analytical solution for transient temperature distribution
in gas tungsten arc, welding with consideration of ller wire, Proc. Inst. Mech.
k 0:13
a 6:95  108 : Eng. 213 (Part B) (1999) 799811.
qc p 1100  1700 [6] S.M. Zubair, M.A. Chaudhry, Heat conduction in a semi-innite solid due to a
time-dependent laser source, Int. J. Heat Mass Transfer 39 (1996) 30673074.
The time for the average solid temperature to reach 35 C will [7] N. Laraqi, Velocity and relative contact size effect on the thermal constriction
be: resistance in sliding solids, ASME J. Heat Transfer 119 (1997) 173177.
[8] M. Amara, V. Timchenko, M. El Ganaoui, E. Leonardi, G. de Vahl Davis, A 3D
Fo  r 2m 1:08  2:5  103 2 computational model of heat transfer coupled to phase change in multilayer
t 97:12 s: materials with random thermal contact resistance, Int. J. Therm. Sci. 48 (2009)
a 6:95  108 421427.
[9] A. Bari, N. Alilat, J.G. Bauzin, N. Laraqi, Three-dimensional stationary thermal
This result is in good agreement with the reported data of 95 s of a bearing ball, Int. J. Therm. Sci. 43 (2004) 561568.
[19]. [10] A. Bari, N. Laraqi, Diagrams for fast transient conduction in sphere and long
cylinder subject to sudden and violent thermal effects on its surface, Appl.
Therm. Eng. 23 (2003) 13731390.
4. Conclusions [11] A. Bari, N. Laraqi, N. Alilat, Z. Zouaoui, Fast transient conduction in innite
plate subject to violent thermal effects, Appl. Therm. Eng. 24 (2004) 115.
[12] A. Bari, J.M. Garcia-de-Maria, N. Laraqi, Effect of thickness and thermal
In this paper, an attempt has been made to formulate a novel and properties of lm on the thermalr of moving rough interfaces, Eur. Phys. J.
simple-to-use method which is easier than existing approaches, less Appl. Phys. 26 (1) (2004) 2934.
complicated and with fewer computations for one-dimensional heat [13] M.K. El-Adawi, M.A. Abdel-Naby, S.A. Shalaby, Laser heating of two layer
system with constant surface absorption an exact solution, Int. J. Heat Mass
ow with variable surface temperature in slabs and spheres as a Transfer 38 (5) (1995) 947952.
function of Fourier number and Biot number in order to arrive at [14] A. Haji-Sheikh, J.V. Beck, Temperature solution in multi-dimensional multi-
the temperature distribution in the solid and the average solid tem- layer bodies, Int. J. Heat Mass Transfer 45 (2002) 18651877.
[15] N. Laraqi, Thermal impedance and transient temperature due to a spot of heat
perature change with time. Estimates from simplied approach on a half-space, Int. J. Therm. Sci. 49 (3) (2010) 529533.
were found to be quite reliable and accurate, as evidenced by the [16] J.G. Bauzin, N. Laraqi, A. Bari, Estimation of thermal contact parameters at the
comparisons with literature data [19] where the average absolute interface of two sliding bodies, J. Phys., Conf. Ser. 135 (1) (2008) 012015,
doi:10.1088/1742-6596/135/1/012015.
deviation and maximum deviation between the reported data and [17] L.C. Thomas, Heat Transfer, Prentice-Hall, Englewood Cliffs, NJ, 1992.
the proposed correlations are found to be around 1.80% and 4.3% [18] H. Grober, S. Erk, J. Jaeger, Conduction of Heat in Solids, Wiley, New York, 1969.
respectively. The predictive tool proposed in the present work is [19] W.L. McCabe, J.C. Smith, P. Harriot, Unit Operations of Chemical Engineering,
seventh ed., McGraw-Hill, NY, USA, 2005.
simple and unique expression which is non-existent in the
[20] M.P. Heisler, Trans. ASME 68 (1946) 493;
literature. In addition, the proposed correlations are smooth and M.P. Heisler, Trans. ASME 69 (1947) 227.
well-behaved (i.e. smooth and non-oscillatory) equations which [21] W.H. Cropper, Great Physicists: The Life and Times of Leading Physicists from
should allow for more accurate predictions. Our efforts in this Galileo to Hawking, Oxford University Press, 2004.
[22] W. Fulton, J. Harris, Representation Theory. A First Course, Graduate Texts in
investigation will denitely pave the way for arriving at accurate Mathematics, Readings in Mathematics, vol. 129, Springer-Verlag, USA, New
predictions of one-dimensional heat ow with variable surface tem- York, 1991.
4542 A. Bahadori, H.B. Vuthaluru / International Journal of Heat and Mass Transfer 53 (2010) 45364542

[23] R.A. Horn, C.R. Johnson, Topics in Matrix Analysis, Cambridge University Press, [25] A. Bahadori, H.B. Vuthaluru, Novel predictive tools for design of radiant
UK, 1991. Section 6.1. and convective sections of direct red heaters, Appl. Energy 87 (2010)
[24] A. Bahadori, H.B. Vuthaluru, Predictive tool for estimation of convection heat 21942202.
transfer coefcients and efciencies for nned tubular sections, Int. J. Therm. [26] MATLAB software, Version 7.6.0.324, The MathWorks, Inc, MA, USA,
Sci. 49 (2010) 14771483. 2008.