Int. J. of Applied Mechanics and Engineering, 2015, vol.20, No.4, pp.

733-756
DOI: 10.1515/ijame-2015-0048

NUMERICAL STUDY OF THE HEAT TRANSFER PHENOMENON OF A

RECTANGULAR PLATE INCLUDING VOID, NOTCH USING FINITE
DIFFERENCE TECHNIQUE
S.K. DEB NATH*
Computational Materials Research Initiative
Institute for Materials Research
Tohoku University, JAPAN
E-mail: sankar_20005@yahoo.com

N.K. PEYADA
Department of Aerospace Engineering
Indian Institute of Technology
Kharagpur, INDIA

In the present study, we have developed a code using Matlab software for solving a rectangular aluminum
plate having void, notch, at different boundary conditions discretizing a two dimensional (2D) heat conduction
equation by the finite difference technique. We have solved a 2D mixed boundary heat conduction problem
analytically using Fourier integrals (Deb Nath et al., 2006; 2007; 2007; Deb Nath and Ahmed, 2008; Deb Nath,
2008; Deb Nath and Afsar, 2009; Deb Nath and Ahmed, 2009; 2009; Deb Nath et al., 2010; Deb Nath, 2013) and
the same problem is also solved using the present code developed by the finite difference technique (Ahmed et
al., 2005; Deb Nath, 2002; Deb Nath et al., 2008; Ahmed and Deb Nath, 2009; Deb Nath et al., 2011; Mohiuddin
et al., 2012). To verify the soundness of the present heat conduction code results using the finite difference
method, the distribution of temperature at some sections of a 2D heated plate obtained by the analytical method is
compared with those of the plate obtained by the present finite difference method. Interpolation technique is used
as an example when the boundary of the plate does not pass through the discretized grid points of the plate.
Sometimes hot and cold fluids are passed through rectangular channels in industries and many types of technical
equipment. The distribution of temperature of plates including notches, slots with different temperature boundary
conditions are studied. Transient heat transfer in several pure metallic plates is also studied to find out the
required time to reach equilibrium temperature. So, this study will help find design parameters of such structures.

Key words: transient, steady state, finite-difference method, analytical method, rectangular plate, void, notch

1. Introduction

Applications involving thermal conduction occur in many areas of science and be design (Siegel and
Howell, 1972; zisik, 1989), in power generation (Burmeister, 1993), glass manufacturing (Siegel and
Howell, 1972) and astrophysics (Jaeger, 1950; Battaner, 1996). Simulations of heat conduction are generally
classified into direct heat conduction problem (DHCP) and inverse heat conduction problem (IHCP)
estimation. DHCP analysis is commonly used to assess the temperature distribution within conductive media
when the existing boundary conditions, thermo-physical properties of the material body, or the intensity of
the heat source within it is known. IHCPs were first proposed by Stolz (1960), who also provided numerical
solutions. Unlike the batch methods, sequential methods have been widely used in the prediction of
parameters for time dependent problems (Beck et al., 1985; Beck, 1970; 1968). Examples are the control-
theory method proposed by Zongrui and Zhongwu (1994) to estimate parameters of inverse transient heat
conduction problems. Ching-China and Hong-Yung (1998) estimated unknown parameters using

*
To whom correspondence should be addressed

Unauthenticated
Download Date | 10/4/17 10:54 PM
734 S.K.Deb Nath and N.K.Peyada

experimental data in linear systems relating to heat conduction using the Kalman filter theory and technique.
The Kalman filter technique was proposed by Kalman (1960). It has been widely used in engineering,
scientific, and administrative applications including inertial navigation (Grewal, 1986). It was first applied to
IHPCs by Scarpa and Milano (1995) who took the estimated parameters as new variables and combined
them with the state variables of original equations to form a new set of state equations. Computer programs
for transient and steady-state heat conduction in two and three dimensions, and in cylindrical coordinates,
have been developed. These are called HEAT2 (Blomberg, 1990; 1991), HEAT3 (Blomberg, 1993; 1994)
and HEAT2R (Blomberg, 1994).
A hybrid numerical technique which combines the differential transformation and finite difference
(FD) approximation (Peng and Chen, 2011) is employed to predict the laser heating problem. Blackwell (1990)
used the Laplace transformation method to obtain an analytical solution for the temperature profile in a semi-
infinite body with exponentially decaying (with position) heat source and convective boundary condition.
Jordan (2003) constructed a nonstandard finite difference scheme to solve an initial-boundary value problem
involving a quartic nonlinearity that arises in heat transfer involving conduction with thermal radiation. Lo
(2011) presents a numerical approach using the hybrid differential transform finite difference method to study
heat transfer in a thin film exposed to ultrashort-pulsed lasers based on the hyperbolic two-step model. The
differential transformation method (Chen and Lin, 1998; Yu and Chen, 1998), a function transformation
technique based on Taylors series is applied mainly to solve the initial value problem by Lo (2011).
In the literature survey the finite difference technique is applied to solve the temperature distribution
in only regular shape bodies without defects. We developed a MATLAB code considering the interpolation
technique to solve plates with regular and arbitrary boundaries. The interpolation and discretization
technique of the stressed structures as mentioned in the references (Ahmed et al., 2005; Deb Nath, 2002; Deb
Nath et al., 2008; Ahmed and Deb Nath, 2009; Deb Nath et al., 2011; Mohiuddin et al., 2012) is used to
discretize the present 2D heat conduction problem in the present MATLAB code. But all the practical
problems are not regular in shapes. Besides, we introduce a notch and rectangular slot on the boundary and
inside the heat conduction plate and obtain temperature distribution throughout the plate. In the present
paper, the flow charts of our MATLAB code developed for transient heat conduction and steady state heat
conduction problems are given for a better understanding of the program philosophy. To verify the
soundness of the present simulation results, we solved a mixed boundary heat conduction problem
analytically using the Fourier integrals (Deb Nath et al., 2006; 2007; 2007; Deb Nath and Ahmed, 2008; Deb
Nath, 2008; Deb Nath and Afsar, 2009; Deb Nath and Ahmed, 2009; 2009; Deb Nath et al., 2010; Deb Nath,
2013) and the distribution of the temperature of some sections of the plate obtained by the analytical method
is compared with those of the present FD method which shows good agreement with the analytical method.

2. Theoretical formulations
2.1a. Finite difference form of the governing equation for the steady state heat transfer problems
using the central difference technique

The governing equation for steady two-dimensional heat conduction with constant properties and
zero sources is

2T 2T
0. (2.1)
x 2 y 2

Using the central difference formula, we obtain

2T ( x , y ) T ( x h, y ) T ( x h, y ) 2T ( x, y )
2
O ( h) . (2.1a)
x h2

Unauthenticated
Download Date | 10/4/17 10:54 PM
Numerical study of the heat transfer phenomenon of a rectangular 735

Here h = x and (i, j) stands for x and y, respectively in such a way that i+1 = x+h, i-1 = x-h. The
term O (h) indicates that the error in the omission of the term is of the order of h.

2 T (i , j ) T (i 1, j ) T (i 1, j ) 2T (i, j )
2
O ( h) , (2.1b)
x h2

2T ( x, y ) T ( x, y k ) T ( x, y k ) 2T ( x, y )
2
O( k ) . (2.1c)
y k2

Here k = y and (i, j) stands for x and y, respectively in such a way that j+1 = y+k, j-1 = y-k.
The term O (k) indicates that the error in the omission of the term is of the order of k.

2T (i, j ) T (i, j 1) T (i, j 1) 2T (i, j )

2
O(k ) . (2.1d)
y k2

The finite difference form of Eq.(2.1) is

T (i 1, j ) T (i 1, j ) 2T (i, j ) T (i, j 1) T (i, j 1) 2T (i, j )

2
0. (2.1e)
h k2
Assuming finite difference mesh lengths in x, y direction are same, i.e., h = k.
The central-difference approximation of the second derivative in space is
T (i 1, j ) T (i 1, j ) 4T (i, j ) T (i, j 1) T (i, j 1) 0 . (2.2)

This equation is applied at each interior grid point of the meshed structure and a set of algebraic
equations is obtained after applying proper boundary conditions.

2.1b. Finite difference form of the boundary condition of steady state heat transfer problem
considering the interpolation technique

When the finite difference mesh does not match the physical boundary of the problem, then the
distribution of the temperature is obtained by four point interpolation. The details of the four point
interpolation technique is given in Ahmed et al. (2005), Deb Nath (2002), Deb Nath et al. (2008), Ahmed
and Deb Nath (2009), Deb Nath et al. (2011), Mohiuddin et al. (2012). The following equation is used to
interpolate the temperature on the arbitrary boundary of the heat conduction problem as shown in Fig.12

1 aT (i 1, j 1) bT (i 1, j ) aT (i, j 1) bT (i, j )
T (is, ja ) s. t
st ab ab

1
T (is, ja ) asT (i 1, j 1) bsT (i 1, j ) atT (i, j 1) btT (i, j ) (2.3)
hk
where T(is,ja) is the temperature after interpolation using four points.
Referring to Fig.12 for top-left boundary, some actual boundary points are P, Q, R and S. For the
boundary point P, the reference natural field grid point is R1. The position of P with respect to R1 is defined
by a and s where a and s are the small distances in the positive i and j direction, respectively. For point P, the
surrounding grid points are R1, R2, R3 and R4, whose field references are {i, j+1}; {i, j};{i+1, j} and {i+1,
j+1}, respectively. The above equation can be used only on the reference boundary when the interpolation
technique is needed; on the other hand when there is no need of the interpolation technique, the above
equation is applied on the physical boundary of the problem.

Unauthenticated
Download Date | 10/4/17 10:54 PM
736 S.K.Deb Nath and N.K.Peyada

3. Theories for transient 2D heat conduction

Conservation of energy to a nodal region of volume as shown in Fig.5 considering explicit method is
applied, where

E in E g E st (3.1)
where

Eg 0

rate of heat entering into the region = E
in

rate of heat generated into the region = E
g

rate of heat stored in the region = E
st

Using Fig.5 and Eq.(3.1), the transient heat conduction equation can be written as

qa qb qc qd E st . (3.2)

Using Fig.5, Eq.(3.2) can be expanded in the following form

Ti ,pj 1 Ti ,pj Ti ,pj 1 Ti ,pj Ti p1, j Ti ,pj

kAc kAc kAc
y y x
(3.3)
Ti p1, j Ti ,pj Ti ,pj1 Ti ,pj
kAc CAc x ,
x t

let x y .

Equation (3.3) can be simplified in the following form

k.t Ti ,pj 1 Ti ,pj 1 Ti p1, j Ti p1, j 4Ti ,pj Ti ,pj1 Ti ,pj , (3.4)
C x
2

F Ti ,pj 1 Ti ,pj 1 Ti p1, j Ti p1, j 4Ti ,pj Ti ,pj1 Ti ,pj . (3.5)

Using Fig.5 and Eq.(3.5), the final form of the transient heat conduction equation is as follows

Ti ,pj1 Ti ,pj 1 4F0 F0 Ti p1, j Ti p1, j Ti ,pj 1 Ti ,pj 1 ,

(3.6)
k t
F0 ,
C x
2

k is thermal conductivity, is the density of a material, C is the specific heat capacity, t is the time step in
second, x is the mesh length in the x direction, y is the mesh length in the y direction, Ac is the cross

Unauthenticated
Download Date | 10/4/17 10:54 PM
Numerical study of the heat transfer phenomenon of a rectangular 737

sectional area of the plate. qa is the rate of heat flow in the AE direction, qb is the rate of heat flow in the BE
direction, qc is the rate of heat flow in the DE direction, qd is the rate of heat flow in the CE direction as
shown in Fig.5.
To solve the transient phenomena in a 2D heat conduction plate, the plate is discretized into grids. At
the boundary grids, the real boundary temperature of the plate is applied. Inside the plate the above Eq.(3.6)
is applied at all grids except the boundary grids of the plate. The initial temperature at all internal grid points
of the plate is are considered to be 300C. Using boundary temperatures, initial temperatures at all internal
grids and applying the above Eq.(3.6) in all internal grid points of the plate, the temperature at all grids is
updated with time till the temperature of all grids reaches the equilibrium temperature.

4. Analytic solution of a mixed boundary heat conduction problem

The governing equation of the 2D conduction heat transfer is

2T 2T
0. (4.1)
x 2 y 2

To solve the problem shown in Fig.9 satisfying the boundary conditions mentioned in Tab.1
analytically, we assume

T Ym sin x , (4.2)

T
Ym ..cos x , (4.2a)
x

2T
2
Ym . 2 .sin x , (4.2b)
x

T
Ym ..sin x , (4.2c)
y

2T
Ym.sin x . (4.2d)
y 2

Combining Eqs (4.1), (4.2b) and (4.2d), we obtain the following relationship

Ym . 2 sin x Ym sin x 0 , (4.2e)

Ym . 2 Ym 0 . (4.3)

Let us assume

Ym e my , (4.4)

Ym m. e my , (4.4a)

Ym m 2 . e my . (4.4b)

Combining Eqs (4.3), (4.4) and (4.4b), we obtain

Unauthenticated
Download Date | 10/4/17 10:54 PM
738 S.K.Deb Nath and N.K.Peyada

m 2 .e my 2 .e my 0 , (4.5)

m2 2 0 , (4.5a)

m . (4.6)

The general solution of the Eq.(4.3) is

Ym Aey Be y . (4.7)

Combining Eqs (4.2) and (4.7) we obtain the following relation which express the temperature
distribution throughout the plate

T Aey Bey sin x . (4.8)

The above Eq.(4.8) can be used to solve the problem as shown in Fig.9 (Deb Nath et al., 2006; 2007;
2007; Deb Nath and Ahmed, 2008; Deb Nath, 2008; Deb Nath and Afsar, 2009; Deb Nath and Ahmed, 2009;
2009; Deb Nath et al., 2010; Deb Nath, 2013)

T Aey Bey sin x E0 Em sin x , (4.9)

E 0 ,
a
2 4T
Em
a
T0 sin x 0
m
where m 1, 3, 5......
0

when y 0, T T0 10000 C ,

A B Em ,

when y b, T T0 10000 C ,

A.e b B.e b Em ,

A

Em 1 e b ,
e b e b

B Em

Em 1 e b .
b b
e e

After substituting A and B in the above Eq.(4.8), we obtain the distribution of temperature
throughout the plate.

5. Results and discussion

The distribution of temperature of a plate having mixed boundary conditions can be solved using
analytical and numerical methods. Although the analytical method gives accurate result, only few simple
problems can be solved analytically, whereas complex practical problems cannot be solved analytically. So

Unauthenticated
Download Date | 10/4/17 10:54 PM
Numerical study of the heat transfer phenomenon of a rectangular 739

an alternative approach to solve complex practical heat conduction problems is numerical. Among the
numerical approaches, finite element and finite difference methods are most popular. The finite element
method is widely used to solve the heat conduction problem because the finite difference method is
sometimes unable to deal with the problems with complex boundary conditions. If the boundary conditions
of the heat conduction problem are satisfied accurately by the finite difference method, it gives accurate
results like the analytical approach. Here we study different types of practical and complex heat conduction
problems by the finite difference method using our MATLAB code applying the algorithm mentioned in
flowcharts 1 and 2. Before analyzing some heat conduction problems, we have solved a simple mixed
boundary heat conduction problem analytically as well as by the finite difference method and we compare
the distribution of temperature at some sections of the plate obtained by the analytical approach to that of
some sections of the plate obtained by the finite difference method. In the present analysis, we show the
temperature distribution in different practical problems. Figure 1 shows a simple 2D heat conduction plate
problem in which the temperature of top side of the plate is 1000C and on the bottom side, it is 00C. The
temperature of the left and right side of the plate is 300C. The length and width of the plate are 10 unit. To
solve the problem using the finite difference technique, the plate is discretized in different mesh sizes. The
four different mesh sizes are taken into account in the present study (h = 0.5, k = 0.5; h = 0.25, k = 0.25; h =
0.2, k= 0.2 and h = 0.125, k = 0.125) and in every case mesh lengths in x and y directions are assumed same.
Effects of mesh length on the temperature distribution at sections x = 8 and y = 8 of the plate as shown in
Fig.1 are illustrated in Figs 2 and 3. From Figs 2 and 3 it is clearly observed that there is a negligible effect
of mesh length on the temperature distribution of the plate. For this reason, we choose the mesh length h =
0.5 and k = 0.5 for the discretization of all other problems which are studied here. Figure 4 illustrates the
contour plot of the temperature distribution throughout the plate as observed in Fig.1 and from the contour
plot, we can understand the temperature distribution at any section of the plate roughly which shows the
same temperature. From Fig.4, it is observed that two different temperature profiles generate inside the plate
as shown in Fig.1. Figure 6 shows a mixed boundary heat conduction plate problem in which the temperature
of the top boundary is 10000C and the temperature of the bottom boundary is 300C and the temperature of the
left and right boundary is 300C. Using this problem as shown in Fig.6, the transient heat conduction
phenomenon is solved using Eq.(3.6) and flow chart 1. Figures 7 and 8 show the distribution of temperature
at the sections y= 10 and x = 10 of the plate as a function of x and y respectively with time. With the increase
of time temperature increases and after a certain time, temperature does not increase, i.e., the system reaches
equilibrium temperature. The shape of the distribution of temperature at each interval of time is similar. To
verify the soundness of the present analysis of the temperature distribution of 2D heat conduction problems,
we have solved the problem as shown in Fig.9 with the boundary conditions mentioned in Tab.1 analytically
using Fourier integrals (Deb Nath et al., 2006; 2007; 2007; Deb Nath and Ahmed, 2008; Deb Nath, 2008;
Deb Nath and Afsar, 2009; Deb Nath and Ahmed, 2009; 2009; Deb Nath et al., 2010; Deb Nath, 2013). The
same problem is also solved using the finite difference technique by our code using the flowchart 2. Figure
10 shows a comparative study of the distribution of temperature at some sections of the plate obtained by the
analytical method to that obtained by the finite difference method. The present temperature distribution at
some sections such as y/a = 0, 0.25 and 0.5 of the plate obtained by the finite difference method totally
coincides with the temperature of those sections of the plate obtained by the analytical method. From the
comparative study it is clear that the present solution of different problems using the finite difference
technique is as accurate as the analytical approach. The rectangular plate with its aspect ratio 2 (the ratio of
the dimension of the plate in x and y directions) as shown in Fig.11 is solved considering steady state heat
transfer using the finite difference technique and the flowchart 2. Using the interpolation technique based on
Eq.(2.3) as shown in Fig.12, the same problem is solved when one of the boundaries of the plate does not
pass through the grids of the domain as shown in Fig.11. Figure 13 shows the contour plot of the distribution
of temperature of the plate as shown in Fig.11 by the finite difference technique using the steady state heat
conduction concept. From Fig.13, it is observed that the temperature distribution of the boundary is equal to
the applied boundary conditions which ensures the reliability of the solutions. The same problem as shown in
Fig.1 is extended introducing a notch on the top side of the plate and hot fluid with temperature 1000C is
passed through this notch as shown in Fig.14. Except the notch the temperature of the top boundary of the

Unauthenticated
Download Date | 10/4/17 10:54 PM
740 S.K.Deb Nath and N.K.Peyada

plate is 00C and the temperature of the left and right side boundary of the plate is 00C. The temperature of the
bottom boundary of the plate is 200C. Figure 15 shows the distribution of temperature of the plate as shown
in Fig.14 which satisfies the physical temperature boundary conditions of the plate and inside the plate two
different temperature profiles are observed. At left and right portions of the plate two symmetric temperature
distributions are observed. Along the middle portion of the plate, another type of temperature profile is
observed. From the top side of the plate the temperature decreases towards the bottom side of the plate.

Start

Automotive grid
generation and grid
numbering

Apply grid id 1 for

physical boundary and
2 for interior grid of the
structure

if(grid id equal 1

Yes No

Temperature of the Initial applied temperature

boundary grid of the of the interior grid of the
plate is applied plate is equal to
environmental temperature

Transient heat conduction Eq.(3.6) is applied in all

internal grid points and with time temperature of the
internal grid is recorded till the system is reached in
equilibrium

Flowchart 1. Flow chart of different steps to determine the temperature profile of 2-D transient heat
conduction problem in our developed MATLAB code 1.

Unauthenticated
Download Date | 10/4/17 10:54 PM
Numerical study of the heat transfer phenomenon of a rectangular 741

Start

Automotive grid
generation and grid
numbering

Apply grid id 1 for

physical boundary and
2 for interior grid of the
structure

if(grid id equal 1

Yes No

Temperature of the Governing 2D heat

boundary grid of the conduction equation is
plate is applied applied in all of the interior
grids of the plate

Temperature of all grid points of the meshed plate is

obtained after solving a set linear algebraic equations

Flow chart 2. Flow chart of different steps to determine the temperature profile of 2-D steady state heat
conduction problem in our developed MATLAB code 2.

Unauthenticated
Download Date | 10/4/17 10:54 PM
742 S.K.Deb Nath and N.K.Peyada

T=1000C W= 10 unit

section y = 12
T=300C

section x = 8
section x = 2
y j
section x = 12
T=300C
i

section y = 8

Governing equation of
2D heat conduction

O x T = 00C

section y = 2

Fig.1. Application of the governing equation of 2D heat conduction of a rectangular plate at different
temperatures.
100

90 h=0.5,k=0.5
h=0.25,k=0.25
h=0.2,k=0.2
h=0.125,k=0.125
80

70

60
Temparature oC

50

40

30

20

10

0
2 3 4 5 6 7 8 9 10 11 12
Y-Coordinate

Fig.2. Effects of mesh size on the temperature distribution as a function of y coordinate of a rectangular
plate at the section, x = 8 unit as shown in Fig.1.

Unauthenticated
Download Date | 10/4/17 10:54 PM
Numerical study of the heat transfer phenomenon of a rectangular 743

32.5

h=0.5,k=0.5
h=0.25,k=0.25
h=0.2,k=0.2
h=0.125,k=0.125

32

31.5
Temparature C
o

31

30.5

30
2 3 4 5 6 7 8 9 10 11 12
X-Coordinate

Fig.3. Effects of mesh size on the temperature distribution as a function of x coordinate of a rectangular
plate at the section, y = 8 unit as shown in Fig.1.

12
100
100

95100

95
95
90
85
80

75 90

4560
85

55

0
540
90 85

40 55
750 65

80
7

0
65 7
80

35 50
35
75

10
70 65

55 60

45
60 55

45
50

40
50

40
8
45

35
35
40
Y Axis

35

6
30
30

30

25
25

20

4
15

20

30 20 30
10

25
15 1

25
5 10 15

5 1e-005

0
1e-005

1e-005
5

0
0 2 4 6 8 10 12
X Axis

Fig.4. Contour plot of the distribution of temperature throughout the plate according to the problem
mentioned in Fig.1.

Unauthenticated
Download Date | 10/4/17 10:54 PM
744 S.K.Deb Nath and N.K.Peyada

A i,,j+1

qa
y

C E i,,j D
x
i-1,,j qd qc i+1,,j

qb

i,,j-1
x

Fig.5. Explicit method to solve the transient heat transfer phenomenon using the mentioned meshes.

T=10000C W= 10 unit

section y = 12
T=300C

section x = 2 section x = 10
y j
section x = 12
T=300C
i

section y = 10

Governing equation of
2D heat conduction

O x T =300C

section y = 2

Fig.6. Application of the transient heat conduction phenomenon of a rectangular plate with its boundaries at
different temperatures.

Unauthenticated
Download Date | 10/4/17 10:54 PM
Numerical study of the heat transfer phenomenon of a rectangular 745

1000

900

800

700
Temparature oC
600 8.33 minutes
25 minutes
500 41.7 minutes
58.3 minutes
75 minutes
400 150 minutes
300 minutes
300 600 minutes
1200 minutes
200 2400 minutes
4800 minutes
9600 minutes
100

0
-5 0 5 10
X-Coordinate

Fig.7. Distribution of temperature at the section y = 10 unit considering the transient heat conduction
phenomenon as shown in the problem given in Fig.6.

1000
8.33 minutes
900 25 minutes
41.7 minutes
800
58.3 minutes
700 75 minutes
150 minutes
Temparature C

600 300 minutes

o

600 minutes
500 1200 minutes
2400 minutes
400 4800 minutes
9600 minutes
300

200

100

0
2 4 6 8 10 12 14
Y-Coordinate

Fig.8. Distribution of temperature at the section x = 10 unit considering the transient heat conduction
phenomenon as shown in the problem given in Fig.6.

Unauthenticated
Download Date | 10/4/17 10:54 PM
746 S.K.Deb Nath and N.K.Peyada

T=10000C
D C

a T=00C
0
T=0 C
y/a=0.5

y/a=0.25

A x
y/a=0 T=10000C B

Fig.9. A square thin aluminum plate with different thermal boundary conditions.

1200 y/a = 0 (Analytical solution)

y/a = 0 (FD solution)
1000

800 y/a = 0.25 (Analytical solution)

Temperature, K

y/a = 0.25 (FD solution)

600

400

200
y/a = 0.50 (Ananlytical solution)
0 y/a = 0.50 (FD solution)

-200
0.0 0.2 0.4 0.6 0.8 1.0
Normalized position, x/b

Fig.10. Comparative study of the distribution of temperature at different sections of the plate as shown in
Fig.9 by FD and analytical approach.

Unauthenticated
Download Date | 10/4/17 10:54 PM
Numerical study of the heat transfer phenomenon of a rectangular 747

Boundary conditions with interpolation

section y = 9.0
0
T=1000 C

Boundary condition
T=10000C
T=300C

Section x = 16.5
section x = 1.0
T=300C
y

Governing equation of 2D
heat conduction (2)

x T=00C
O section y = 1.0

Fig.11. Application of the governing equation of 2D heat conduction and boundary conditions using the
interpolation technique of a rectangular plate at different temperatures.

R4(i+1, j+1)
R1 (i, j+1) s
b
a P
t
R2(i, j) R3(i+1, j)

Fig.12. Locators a, b, s and t of the boundary point P with respect to its reference point R1, R2, R3, and R4.

Unauthenticated
Download Date | 10/4/17 10:54 PM
748 S.K.Deb Nath and N.K.Peyada

9 88950 0
1000 1000 1000
000 0
95 0 95 0 8505 0

656442055 0 20 0
807

00
75

00 000
950

30 00305 0 5505
0

0
70

70

100 0
0
900 900
8

2015
0
85 0 850 0

4
65

35 0
80 0 80 0 0

50

50
60
75 0 75 0
7

30 0
70 0 70 0

55
65 0

15 0

0
50 60 0
0 65 0
0
55
6 45
60 0
50
0
0
0 45
55 0

40

25 0
10 0

0
0

25
500
40 0
35

0
Y Axis

5 35

15 0
450

20 0
0

0
0

20

10 0
30
30 400
0
50

50
4 350

300
15

25 0
0

25 0

3 20 0 20 0
10 15 0
0
150
10 0
2 100
50 50
50

1 0.0001 0.0001 0.0001

0
0 2 4 6 8 10 12 14 16
X Axis
Fig.13. Contour plot of the distribution of temperature throughout the plate according to the problem
mentioned in Fig.11 using the interpolation technique on the top boundary.
2D heat conduction governing
Eq.(2.2)
j
T=1000C

T=0OC
i

T=00C

T=0OC
T=00C
T=0OC
y

x, i T=200C

Fig.14. Application of the governing equation of 2D heat conduction and boundary conditions with different
temperatures of a rectangular plate having a notch.

Unauthenticated
Download Date | 10/4/17 10:54 PM
Numerical study of the heat transfer phenomenon of a rectangular 749

10

9 1e-005 2
4 2 1e-005
64

0 250
3235802104
10

005

80 0 9
8206345
22

70

1e -005
4050
9800
8

10

254
1e -

14
8 12
8

16
6

22

45
18 16

60
2

10
26

2
70
60

18

12
4
7

40

4
24

30

6
14
10
6

20
50

20

8
28
26

14
45
1e-005

22
30
28

24
Y Axis

1e-005
35

35
12

40
5

16
6

22
16

10
2

26
18

2
12
4

18
4

30

4
24

3 28

6
10

14

26
20

20
24

8
14
22
1e-005

2
8
12

1e-005
22

16
6

16

10 12
18 18
2

2
20 20

0
0 5 10 15
X Axis
Fig.15. Contour plot of the distribution of temperature throughout the plate according to the problem
mentioned in Fig.14.

Table 1. Boundary conditions of the 2D heat conduction problem as shown in Fig.9.

AB 1000 C
Temperature at the CD 1000 C
boundary of the plate AD 0 C
BC 0 C
A TA 0 C
Temperature of the B TB 0 C
corner points of the
C TC 0 C
plate
D TD 0 C

The problem defined in Fig.14 is extended including a slot inside the plate and a hot fluid having
temperature 10000C is passed through this rectangular slot and a hot fluid having temperature 1000C is
passed through the triangular channel and the rest of the boundaries of the plate, the temperature is
maintained at 300C as shown in Fig.16. Figure 17 describes the temperature distribution of the plate
according to the defined problem as shown in Fig.16. Away from the rectangular slots, the temperature
gradually decreases and is equal to the applied temperature on the boundaries when it approached the

Unauthenticated
Download Date | 10/4/17 10:54 PM
750 S.K.Deb Nath and N.K.Peyada

boundary. A rectangular plate having three slots is meshed in which a hot fluid having temperature 10000C is
passed through the middle slot and cold fluid having temperature 00C is passed through the rest of other two
slots as shown in Fig.18. The distribution of temperature of the problem defined in Fig.18 is shown in Fig.19.
Three different symmetric temperature distributions are observed in the plate. Between the hot and cold
zones, the temperature changes abruptly. A rectangular plate having three slots is meshed in which a cold
fluid having temperature 00C is passed through the middle slot and a hot fluid having temperature 10000C is
passed through the rest of other two slots as shown in Fig.20. The distribution of temperature of the problem
defined in Fig.20 is shown in Fig.21. Two different symmetric temperature distributions are observed in the
plate. Between the hot and cold zone, the temperature changes abruptly. Using the transient phenomenon, the
time required to reach the steady state temperature is obtained for the present problem as shown in Fig.6
considering different materials which are shown in Table 2. From Table 2, it is observed that the plates made
of silver and stainless steel take the lowest time and the longest time, respectively, to reach the steady state
temperature.

T=300C T=300C

T=1000C

T=300C

y
T=10000C T=300C

x T=300C
O

Fig.16. Boundary conditions of the aluminum plate having a notch on the top surface and a rectangular slot
inside the plate.

Unauthenticated
Download Date | 10/4/17 10:54 PM
Numerical study of the heat transfer phenomenon of a rectangular 751

10

9 30.000097 4050
70 70
40 30.000097

10 0
10 0 150 50

40
50

30.000097
20 0
8

40
150

70
30.000097
300
25 20 0
0
40 0

0
25
7

70
50 0

150

100
30
0
70 0 60

40
100 0
6 80 0

150
900

60 0

250
200

500

50
300
10 00
Y Axis

20 0
40
50

30.000097
40
70

400
30.000097

250

700

4
1000
500

700
800

70
900
80 0
150

100
30 0
250 0

90
80 0

60 0

40 0
3
20

1000
300

0
10 0

0
20 250
60 0
50

70 0

15
50
0

50
0
2 15
40

0 40 0
50

70 40

97
250 300

00
200 10 0 70
30

.00
.00 10 0 150
70 50
1 009

30
50
4030.000097 40 30.000097
7

0
0 2 4 6 8 10 12 14 16 18
X Axis

Fig.17. Contour plot of the distribution of temperature throughout the plate according to the problem
mentioned in Fig.16.
T=300C

T=30OC

T=300C

T=300C
T=00C T=10000C T=00C
y

x T=300C

Fig.18. Boundary conditions of the aluminum plate having three slots inside the plate (Case 1).

Unauthenticated
Download Date | 10/4/17 10:54 PM
752 S.K.Deb Nath and N.K.Peyada

14

12

30

30
35

80
40
40

60
30

20 0
0 35
10
30 25
60 10 0
25
40 0

0
50
30 0
10 20
80
30 0
20

80 0
25 70 0 80
35

0. 0001
30

1000
10 5 15 20040 0

10
5
60 0 60 0 15 25 30
20
15
20
40 900 20060 15
30 40
500 35
8 30
100
10 700 25
0.0001 0.0001
60 100 10
Y Axis

25 900
5 500
800 400
300 5
20 300 20
0.0001
80 800
1000
25 80
1000700 5
6 0.0001
105
35
15200400600 600
30
10
15
5 20 25 30
15 20
40
900 20060
40 15

0.0001
30 10
40
200 0
100 50 0 35

900
25
10

5 7
5

60 30 0600000

80 0
15 15
100
25 0
4 20 40 00 0
25 3 20
80 25
80
60 30
80 20 0
3540 30
100

35 60
40
2 30
30

0
0 5 10 15 20 25 30
X Axis

Fig.19. Contour plot of the distribution of temperature of the plate as shown in Fig.18.

T=300C

T=30OC

T=300C

y
T=10000C T=00C T=10000C
T=300C

T=30OC

x T=300C

Fig.20. Boundary conditions of the aluminum plate having three slots inside the plate (Case 2).

Unauthenticated
Download Date | 10/4/17 10:54 PM
Numerical study of the heat transfer phenomenon of a rectangular 753

14

12

40

40
60
100

100
60

80
80

200
60

20
80 40
40 80 30
30 0

500
60 0 0 10 0
50
10

60
0
20
40 0 100 40 0
00 0 0 0
30 4 40 0

800
40 6 20 0 60
100 70 0 40 200
30 0 70 0
80 0
10 00 100
900 900
200 60 0 60 20080
1000 50 0
80
8 80 700
500 0.0001
0.0001
80
70 0 60
40
900 500 40 600
800
900 1000 300
Y Axis

300 500
60 1000 600 100
800 20 20
800
40 300
400600 400 100 400 400
20060 200
6 100
700 40 300 700
800
1000 00
40 0 900
100 900605040 0

1
60 10 20080

0.000
20 0 50 0 60 0 00
60 0 80 30 0
900

60

900
80 7080
00 40
60 0 70 0

80
40 20 80
50 0

0
80
4 60
80
20 03
00
30 0 50
0 20 0

10 0
0

40 30
10 0
40

40 0 20

0 0
20 60 40 0 60
10 60

1060
0

80

80
2

40
40

0
0 5 10 15 20 25 30 35
X Axis

Fig.21. Contour plot of the distribution of temperature of the plate as shown in Fig.20.

Table 2. Required time to reach the steady state considering the transient phenomena as shown in Fig.6.

Materials Time taken to rich steady state (hr)

Aluminum 60
Stainless steel 504
Copper 52
Lead 211
Tin 134
Tungsten 85
Silver 36
Chromium 179
Zinc 127

6. Conclusion

To verify the soundness of the solution of a 2D heat conduction problem obtained by the present
finite difference technique, the analytical solutions of some sections of the plate with mixed boundary
conditions are compared with the distribution of temperature of those sections of the plate obtained by the
present finite difference method and the temperature distributions of some sections obtained by both of the

Unauthenticated
Download Date | 10/4/17 10:54 PM
754 S.K.Deb Nath and N.K.Peyada

methods agree well. So, the temperature distributions of some practical 2D heat conduction problems
obtained by the present finite difference method are reliable. Firstly, a 2D aluminum plate with mixed
boundary conditions is solved using the transient heat conduction approach to understand the time dependent
temperature distribution at different sections of the plate. The transient heat conduction phenomenon in a
plate is studied considering different materials. Secondly, an optimum mesh size is determined by varying
the mesh density of a 2D rectangular heat conduction plate to solve the present problems. Thirdly, the same
rectangular heat conduction plate is solved considering without the interpolation technique when the
boundary of the plate passes through the meshing of the plate. Then this problem is solved considering
interpolation technique when the boundary of the plate does not match the discretized grids of the plate. The
temperature distribution on the boundary of an aluminum plate does not change if the interpolation technique
is used to solve it when the boundary does not match with the regular grid network. Then a notch is
introduced on the top of the plate and hot fluid is passed through it and effects of hot fluid on the temperature
distribution near the notch are discussed. Then this problem is extended to introduce a rectangular slot inside
it. Effects of slots and notches on the temperature distribution of the plate are clearly observed. Besides,
three rectangular slots are introduced inside the rectangular plate. Effects of the hot and cold fluid passing
through it on the temperature distribution of the plate are also analyzed.

Nomenclature

A constant
B constant
b length of the plate
E0 constant
Em constant
FD finite difference
h finite difference mesh length in the x direction
k finite difference mesh length in the y direction
m 1, 3, 5..
T temperature distribution as a function of x and y directions
Ym function of y
Ym first derivative with respect to y
Ym second derivative with respect to y
m

b

References

Ahmed S.R. and Deb Nath S.K. (2009): A simplified analysis of the tire-tread contact problem using displacement
potential based finite difference. Computer Modelling in Engineering and Sciences, vol.44, pp.35-64.
Ahmed S. Reaz, Deb Nath S.K. and Uddin M.W. (2005): Optimum shapes of tire treads for avoiding lateral slippage
between tires and roads. Vol.64, pp.729-750.
Battaner E. (1996): Astrophysical Fluid Dynamics. Cambridge: Cambridge University Press.
Beck J.V. (1968): Determination of undisturbed temperatures from thermocouple measurements using correction
Kernels. Nucl. Eng. Design, vol.7, pp.9-12.
Beck J.V. (1970): Nonlinear estimation applied to non-linear inverse heat conduction problem. Intl. J. Heat Mass
Transfer, vol.13, pp.703-716.
Beck J.V., Blackwell B. and Clair C.R.S. (1985): Inverse Heat Conduction. New York: Wiley.
Blackwell B.F. (1990): Temperature profile in semi-infinite body with exponential source and convective boundary
condition. J. Heat Transfer, vol.112, pp.567-571.

Unauthenticated
Download Date | 10/4/17 10:54 PM
Numerical study of the heat transfer phenomenon of a rectangular 755

Blomberg T. (1990): HEAT2- A heat transfer PC- program. Proceeding of the 2nd Conference on Building Physics in
the Nordic Countries, Division of Building Technology, Department of Civil Engineering, The Norwegian Institute
of Technology, The University of Tronheim, Alfr, Getz vei3, N-7034 Trondheim, Norway.
Blomberg T. (1991): HEAT2- A heat transfer PC-program. Manual for HEAT2, Department of building physics,
Lund University, P.O. Box 118, S-221 00 Lund, Sweden, CoDEN: LUTVDG/(TVBH-7122).
Blomberg T. (1993): HEAT3-A three-dimensional heat transfer computer program. Proceeding of the 3rd conference
on building physics in the Nordic countries, Buiding Physics 93 (Bjarne Saxhof, editor), page 339, Thermal
Insulation laboratory, Lyngby, Denmark, ISBN 87-984610-0-1 volume 1.
Blomberg T. (1994): HEAT3-A three-dimensional heat transfer computer program Manual for HEAT3. Department
of building physics, Lund University, P.O.Box. 118, S-221 00 Lund, Sweden. CODEN: LUTVDG/(TVBH-7169).
Blomberg T. (1994): HEAT2R-A PC-Program for heat conduction in cylindrical coordinates r and z. Department of
Building Physics, LUND University. P.O. BOX 118, S-221 00 Lund, Sweden CODEN: LUTVDG/(TVBH-7178).
Burmeister L.C. (1993): Convective Heat Transfer. 2nd Ed., New York: Wiley.
Chen C.L. and Lin Y.C. (1998): Solution of two boundary problems using the differential transform method. J. Optim.
Theory, Appl., vol.99, pp.23-35.
Ching-China J. and Horng.-Yung J. (1998): Experimental investigation in inverse heat conduction problem.
Numerical Heat Transfer, Part A, vol.34, pp.75-91.
Jaeger J.C. (1950): Conduction of heat in a solid with a power law of heat transfer at its surface. Proc. Camb. Phil.
Soc., vol.46, pp.634-641.
Deb Nath S.K. (2002): A study of wear of tire treads. MSc. Thesis, Department of Mechanical Engineering,
Bangladesh University of Engineering and Technology, Dhaka, Bangladesh.
Deb Nath S.K. (2008): Displacement potential approach to solution of stiffened composite cantilever beams under
combined loading. International Journal of Applied Mechanics and Engineering, vol.13, No.1, pp.2141.
Deb Nath S.K. (2013): Effects of fiber orientation and material isotropy on the analytical elastic solution of a stiffened
orthotropic panel subjected to a combined loading. Advances in Materials Science and Engineering, vol.2013,
Article ID 710143,13 pages.
Deb Nath S.K., Ahmed S.R. and Afsar A.M. (2006): Displacement potential solution of short stiffened flat composite
bars under axial loading. International Journal of Applied Mechanics and Engineering, vol.11, No.3, pp.557575.
Deb Nath S.K., Afsar A.M. and Ahmed S.R. (2007): Displacement potential solution of a deep stiffened cantilever
beam of orthotropic composite material. Journal of Strain Analysis for Engineering Design, vol.42, No.7, pp.529-
540.
Deb Nath S.K., Afsar A.M. and Ahmed S.R. (2007): Displacement potential approach to the solution of stiffened
orthotropic composite panels under uniaxial tensile load. Proceedings of the Institution of Mechanical Engineers,
Part G: Journal of Aerospace Engineering, vol.221, No.5, pp.869-881.
Deb Nath S.K. and Ahmed S.R. (2008): Analytical solution of short guided orthotropic composite columns under
eccentric loading using displacement potential formulation. Proceedings of the Institution of Mechanical
Engineers, Part G: Journal of Aerospace Engineering, vol.222, No.4, pp.425-434.
Deb Nath S.K. and Afsar A.M. (2009): Analysis of the effect of fiber orientation on the elastic field in a stiffened
orthotropic panel under uniform tension using displacement potential approach. Mechanics of Advanced
Materials and Structures, vol.16, No.4, pp.300-307.
Deb Nath S.K. and Ahmed S.R. (2009): Displacement potential solution of stiffened composite struts subjected to
eccentric loading. Applied Mathematical Modelling, vol.33, No.3, pp.1761-1775.
Deb Nath S.K. and Ahmed S.R. (2009): Elastic analysis of short orthotropic composite columns subjected to uniform
load over a part of the tip. Proceedings of the Institution of Mechanical Engineers, Part G: Journal of Aerospace
Engineering, vol.223, No.2, pp.95105.

Unauthenticated
Download Date | 10/4/17 10:54 PM
756 S.K.Deb Nath and N.K.Peyada

Deb Nath S.K., Ahmed S.R. and Kim S.-G. (2010): Analytical solution of a stiffened orthotropic plate using alternative
displacement potential approach. Proceedings of the Institution of Mechanical Engineers, Part G: Journal of
Aerospace Engineering, vol.224, No.1, pp.8999.
Deb Nath S.K., Akanda M.A.S., Ahmed S.R. and Uddin M.W. (2008): Numerical investigation of bond-line stresses of
tire tread section. International Journal of Applied Mechanics and Engineering, vol.13, pp.43-61.
Deb Nath S.K., Ahmed S.R., Kim S-G. and Wong C.H. (2011): Effect of tire material on the prediction of optimum tire
tread sections. International Journal for computational methods in Engineering Science and Mechanics, vol.12,
290-302.
Grewal M.S. (1986): Application of Kalman filtering to the calibration and alignment of inertial navigation systems.
presented at PLANS 86- position location and Navigation Symposium, Las Vegas, Nevada.
Jordan P.M. (2003): A nonstandard finite difference scheme foe non-linear heat transfer in a thin finite rod. Journal of
Difference Equations and Applications, vol.9, pp.1015-1021.
Kalman R.E. (1960): A new approach to linear filtering and prediction problems. ASME J. Basic Eng., ser. 82d,
pp.35-45.
Lo C-Y (2011): A study of two-step heat conduction in laser heating using the hybrid differential transform method.
Numerical Heat Transfer, Part B, vol.59, pp.130-146.
Mohiuddin M., Uddin M.W., Deb Nath S.K. and Ahmed S.R. (2012): An alternative numerical solution to a screw-
thread problem using displacement-potential approach. International Journal for Computational Methods in
Engineering Science and Mechanics, vol.13, pp.254-271.
zisik M.N. (1989): Boundary Value Problems of Heat Conduction. New York: Dover.
Peng H-S and Chen C.L. (2011): Application of hybrid differential transformation and finite differential transformation
and finite difference method on the laser heating problem. Numerical Heat Transfer, Part A, vol.59, pp.28-42.
Scarpa F. and Milano G. (1995): Kalman smoothing technique applied to the inverse heat conduction problem.
Numer. Heat Transfer, part B, vol.28, pp.79-96.
Siegel R. and Howell J.H. (1972): Thermal Radiation Heat Transfer. New York: McGraw-Hill.
Stolz Jr.G. (1960): Numerical solutions to an inverse problem of heat conduction for simple shapes. ASME J. Heat
Transfer, vol.82, pp.20-26.
Zongrui L. and Zhongwu L. (1994): A control theory method for solutions of inverse transient heat conduction
problems. Trans. ASME, vol.116, pp.228-230.
Yu L.T. and Chen C.K. (1998): The solution of the Blasius equation by the differential transform method. Math.
Comput. Model, vol.28, pp.101-111.

Received: July 17, 2014

Revised: September 17, 2015

Unauthenticated
Download Date | 10/4/17 10:54 PM