Available online at www.sciencedirect.

com

International Journal of Heat and Mass Transfer 51 (2008) 865–872

www.elsevier.com/locate/ijhmt

Temperature distribution in a rectangular plate heated

by a moving heat source
J. Kidawa-Kukla
Institute of Mechanics and Machine Design, Cze˛stochowa University of Technology, Da˛browskiego 73, 42-200 Cze˛stochowa, Poland

Available online 29 October 2007

Abstract

In this paper, a solution to the problem of heat conduction in a rectangular plate subjected to the activity of a moving heat source is
presented. The temperature of the plate changes because a limited area on the plate surface is heated by a heat source. The heat source
moves along an elliptical trajectory which always remains within the plate area. An exact solution to the problem in an analytical form is
obtained by applying the Green’s function method. Exemplary results of numerical calculations to determine the temperature distribu-
tion in the plate are presented.
Ó 2007 Elsevier Ltd. All rights reserved.

Keywords: Heat conduction; Moving heat source; Temperature distribution in a plate; Green’s function method

1. Introduction moving heat source with a Gaussian distribution. It was

assumed that the heat source moved with a constant veloc-
Mathematical models of heat conduction with moving ity along a line which was parallel to one edge of the rect-
heat sources have practical applications in numerous man- angular plate. A circular Gaussian heat source moving at a
ufacturing processes such as welding, metal cutting, and constant relative velocity over the surface of a solid with a
the grinding and drilling of metals. The problem is the sub- finite depth and width was also assumed by Manca et al.
ject of many publications (for instance, Refs. [1–6]), and [3]. An analytical form of the solution to the three-dimen-
authors often use analytical methods to solve these types sional problem was derived using a Green’s function
of problems. The application of analytical models of heat method (GFM). The Green’s function describes the tem-
conduction is limited in regular domains, but their use is perature distribution caused by an instantaneous, local
more profitable in the analysis of the process. energy pulse. The specific use of the GFM in solving vari-
An analytical solution to the problem of the three- ous heat conduction problems was widely discussed by
dimensional temperature distribution generating by a mov- Beck et al. [4]. Examples of applications of the GFM in
ing laser beam source in a finite domain was presented by solving heat conduction problems in beams and a plate
Araya and Gutierrez [1]. The heat source is modeled as a with moving heat sources are presented by Kidawa-Kukla
laser beam with a Gaussian distribution or as a spatially [5–7]. In Refs. [5,6] analytical solutions to the temperature
uniform plane. The results and discussion concern the distribution in rectangular beams heated by a moving heat
boundary effects on the temperature of a workpiece. An source which moves harmonically around a fixed point on
analytical model which describes a three-dimensional tem- the beam surface are presented. These solutions were then
perature fields in a finite thickness plate was investigated by used to determine the displacement of the beam induced by
Cheng and Lin [2]. The considered plate was heated by a cyclic changes in the temperature. A parabolic equation [5]
or a hyperbolic equation [6] of the heat conduction was
assumed in the mathematical model. The temperature dis-
E-mail address: [email protected] tribution in a rectangular plate subjected to the activity of a

0017-9310/$ - see front matter Ó 2007 Elsevier Ltd. All rights reserved.
doi:10.1016/j.ijheatmasstransfer.2007.04.011
866 J. Kidawa-Kukla / International Journal of Heat and Mass Transfer 51 (2008) 865–872

Nomenclature

T ðx; y; z; tÞ temperature of a plate (°C) e size of the quadratic element on the plate surface
gðx; y; tÞ volumetric energy of heat source (W m3) heated by the heat source (m)
xðtÞ; y ðtÞ functions describing the movement of the heat d( ) Dirac delta function
source (m) a0 ; a1 heat transfer coefficients (W m1 K1)
T 0 ; T 1 temperatures of a surrounding medium (°C) k thermal conductivity (W m1 K1)
a; b; h dimensions of a plate (m) u angular velocity of the moving heat source
t time (s) (rad s1)
x0 ; y 0 coordinates of fixed point of a plate (m)
Gðx; y; z; t; n; g; 1; sÞ Green’s function Subscripts
i, j, k, l, m, n indices
Greek symbols
j thermal diffusivity (m2 s1)
H heat flux of source (W m2)

heat source is presented in Ref. [7]. It was assumed that the as a solution to a three-dimensional heat conduction prob-
heat source moves harmonically around a fixed point along lem solved using the Green’s function method.
a segment which always remains on the plate.
The analytical form of the solution to the non-homoge- 2. Problem formulation
neous heat conduction problem usually involves an infinite
series which is characterized by slow convergence. For Consider a rectangular plate of uniform thickness h with
example, low convergence can be observed in the steady edge lengths a and b, as shown in Fig. 1. The temperature
state part of solutions to the partial heating of solids. T ðx; y; z; tÞ of the plate satisfies the differential equation of
The verification of solution methods to the heat transfer heat conduction:
problem for the partial heating of a rectangular solid is
1 oT 1
the subject of the paper by Beck et al. [8]. In order to r23 T  þ gðx; y; tÞ ¼ 0 ð1Þ
improve the convergence of the solution, the authors of j ot k
papers [4,8] recommend the use of the GFM in conjunction where r23  ðo2 =ox2 Þ þ ðo2 =oy 2 Þ þ ðo2 =oz2 Þ, j is thermal
with a time-partition method. In this approach the large diffusivity, k is thermal conductivity and gðx; y; tÞ denotes
and the small-time forms of the Green’s functions are a volumetric energy generation. In this study, it is assumed
applied. that the thermal energy is provided by a heat source which
The first step in solving a linear heat transfer problem moves along a trajectory on the plate surface. Therefore,
using GFM consists in to derive the Green’s function. An the function gðx; y; z; tÞ takes the form
auxiliary initial-boundary problem should be solved to 8H h
determine the function. The solution to this problem can < 4e2 dðz  2Þ
>
gðx; y; tÞ ¼ for xðtÞ  e < x < xðtÞ þ e; y ðtÞ  e < y < y ðtÞ þ e
be obtained, for example, by using a method for the sepa- >
:
0 otherwise
ration of variables, the Laplace transform or a method
using images [4]. Alternative representations of Green’s ð2Þ
functions for two-dimensional heat conduction problems where H characterises the stream of heat, 2e is the size of
are presented by Melnikov [9], where the functions were the quadratic element on the plate surface heated by the
obtained by means of a combination of the Laplace trans-
form, the eigenfunction expansion method and the varia-
tion of parameters method. A set of Green’s functions, z
Laser beam
useful in solving various heat conduction problems, is
given in Ref. [4]. The form of the Green’s function is of h y
great significance because the above mentioned slow con-
vergence of the solution can effect the accuracy of numeri-
cal calculations.
This paper presents an analytical solution to the heat a
conduction problem in a plate which is subjected to a mov-
ing heat source. The temperature of the plate changes
because a limited area on the plate surface is heated by a x
b
heat source which moves along an elliptical trajectory.
The temperature field in the rectangular plate is obtained Fig. 1. Scheme of the considered plate.
 J. Kidawa-Kukla / International Journal of Heat and Mass Transfer 51 (2008) 865–872 867

heat source and d() is the Dirac delta function. The func- – Integration of both sides of the equation obtained in the
tions xðtÞ; y ðtÞ, which describe the elliptical trajectory of second step, with respect to n, g, f, s in the intervals (0,
the heat source are: aÞ, (0, bÞ, (0, hÞ, (0, tÞ, respectively.
xðtÞ ¼ x0 þ A cos ut; y ðtÞ ¼ y 0 þ B sin ut ð3Þ
As a result one obtains
where x0, y0, A, B are constants and u is the angular veloc- Z Z b Z h Z t   
a
ity of the moving heat source. So that the heat source al- 1 oT 1
r23  T ðn; g; 1; sÞ þ gðn; g; 1; sÞ
ways remains within the plate area, it is assumed that the 0 0 0 0 j os k
constants occurring in Eq. (3) satisfy the following inequal-  Gðx; y; z; t; n; g; f; sÞ df ds dg dn ¼ 0 ð12Þ
ities: A þ e < x0 < a  A  e, B þ e < y 0 < b  B  e.
The differential equation (1) is complemented by the fol- Next, the integral in Eq. (12) is integrated by parts: the
lowing initial and boundary conditions: terms which include derivatives of the function T with re-
T ðx; y; z; 0Þ ¼ 0 ð4Þ spect to n, g, f, are integrated by parts twice and the term
including the derivative with respect to s is integrated once.
T ð0; y; z; tÞ ¼ T ða; y; z; tÞ ¼ 0; T ðx; 0; z; tÞ ¼ T ðx; b; z; tÞ ¼ 0 After utilizing the initial and boundary conditions (4)–(7)
ð5Þ and (9),(10), the following equation is obtained:
koT ðx; y; h; tÞ=oz ¼ a1 ½T 1  T ðx; y; h; tÞ ð6Þ Z a Z b Z h Z t    
2 1 oT
koT ðx; y; 0; tÞoz ¼ a0 ½T 0  T ðx; y; 0; tÞ ð7Þ r3 þ Gðx; y; z; t; n; g; f; sÞ
0 0 0 0 j os
where a0 and a1 are the heat transfer coefficients, and T 0 , T 1 
1
are known temperatures of the surrounding medium.  T ðn; g; 1; sÞ þ gðn; g; f; sÞGðx; y; z; t; n; g; f; sÞ
k
3. Solution to the problem  ds df dg dn þ B:c: ¼ 0 ð13Þ

where
The solution to the initial-boundary problem, which is Z Z Z
a b t
given by Eq. (1) and conditions (4)–(7), is determined by B:c: ¼ ½l1 T 1 Gjn¼h þ l0 T 0 Gjn¼0  ds dg dn
using Green’s function method. The Green’s function 0 0 0
Gðx; y; z; t; n; g; 1; sÞ is a solution to the differential equation:
  Finally, using (11) in Eq. (13) and using the properties of
2 1 o the Dirac delta function, one obtains:
r3  Gðx; y; z; t; n; g; 1; sÞ
j ot
T ðx; y; z; tÞ
¼ dðx  nÞdðy  gÞdðz  nÞdðt  sÞ ð8Þ Z Z Z Z
1 a b h t
and satisfies the homogeneous initial-boundary conditions, ¼ gðn; g; f; sÞGðx; y; z; t; n; g; f; sÞ
k 0 0 0 0
analogous to the initial and boundary conditions (4)–(7):
 ds df dg dn þ B:c: ð14Þ
Gjt¼0 ¼ 0; Gjx¼0 ¼ Gjx¼a ¼ 0; Gjy¼0 ¼ 0; Gjy¼b ¼ 0; ð9Þ
Substituting the function gðn; g; f; sÞ given by (2) into Eq.
ðkoG=oz  a0 GÞjz¼0 ¼ 0; ðkoG=oz þ a1 GÞjz¼h ¼ 0; ð10Þ
(14), the temperature of the plate, T ðx; y; z; tÞ, is expressed
The Green’s function for the considered heat conduction as
problem as a solution to the homogeneous differential Z t Z xðsÞþe Z y ðsÞe
H
problem (8)–(10) is presented in Appendix A. The applica- T ðx; y; z; tÞ ¼ Gðx; y; z; t; n; g; h; sÞ dg dn ds
4e2 k
tion of a reciprocity relation [4]: Gðx; y; z; t; n; g; f; sÞ ¼ 0 xðsÞe y ðsÞe

Gðn; g; f; s; x; y; z; tÞ, in Eq. (8), yields þ B:c: ð15Þ

 
1 oT The Green’s function Gðx; y; z; t; n; g; f; sÞ given by (A.1) is
r23 þ Gðx; y; z; t; n; g; f; sÞ
j os now utilized in Eq. (15). After evaluation of the integrals
¼ dðx  nÞdðy  gÞdðz  nÞdðt  sÞ ð11Þ with respect to g, n, one obtains the temperature
2 2 2 2 2
T ðx; y; z; tÞ in the form
where r23 2
 ðo =on Þ þ ðo =og Þ þ ðo =of Þ. Eq. (11) is then
used to solve the problem. 4jH X 1 X 1 X 1
wn ðzÞwn ðhÞ
T ðx; y; z; tÞ ¼ K jmn ðtÞ
The Green’s function is applied to determine the temper- p2 e2 k j¼1 m¼1 n¼1 jmQn
ature T in the plate. To this end the following steps should jxp myp jpe mpe
be performed:  sin sin sin sin þ B:c: ð16Þ
a b a b
– Replacement of variables x; y, z, t in Eq. (1) by n, g, f, s, where wn ðzÞ; Qn are defined in Appendix A; and
respectively. Z t
jpxðsÞ mpy ðsÞ
– Multiplication of both sides of the equation obtained in K jmn ðtÞ ¼ sin sin expðjc2jmn ðt  sÞÞ ds
0 a b
the first step, by the Green’s function Gðx; y; z; t; n;
g; 1; sÞ. ð17Þ
868 J. Kidawa-Kukla / International Journal of Heat and Mass Transfer 51 (2008) 865–872
 2  2
with c2jmn ¼ b2n þ jpa þ mp b
, bn are roots of equation numerically calculated. An alternative way is to present
(A.11). the integrand in the form of an infinite series, and integrate
The integral in Eq. (17), after taking into account the the series ‘‘term by term”. First, the function K jmn ðtÞ is writ-
functions xðsÞ and y ðsÞ which are given in Eq. (3), can be ten as

T [°C]
300
0.4
200

100 0.3

0 b [m]
0 0.2

0.2
0.1
a [m]

0.4
0

Fig. 2. Temperature distribution on the plate surface for t = 3600 s, when the heat source moves along a section parallel to an edge of the plate: a ¼ 0:5 m,
b ¼ 0:4 m, h ¼ 0:02 m, A ¼ 0:2 m, B = 0.

0.4
0.75
a [m] b [m]

0.5
0.2

0.25
0 T [°C]
0 800
600
400
200
0

Fig. 3. Temperature distribution on the plate surface for t = 3600 s (heat source moves along an elliptical trajectory), a ¼ 1:0 m, b ¼ 0:5 m, h ¼ 0:01 m,
A ¼ 0:3 m, B ¼ 0:2 m.
 J. Kidawa-Kukla / International Journal of Heat and Mass Transfer 51 (2008) 865–872 869

Z
jpx0 mpy 0 t  expðjc2jmn ðt  stÞÞ ds
K jmn ðtÞ ¼ sin sin cosðlj cos usÞ cosðmm sin usÞ Z
a b 0 jpx0 mpy 0 t
 expðjc2jmn ðt  sÞÞ ds þ cos cos sinðlj cos usÞ sinðmm sin usÞ
a b 0
Z
jpx0 mpy 0 t  expðjc2jmn ðt  sÞÞ ds ð18Þ
þ sin cos cosðlj cos usÞ sinðmm sin usÞ
a b 0
 expðjc2jmn ðt  sÞÞ ds where lj ¼ jpA
a
and mm ¼ mpB
b
. The integrals occurring in Eq.
Z (18) were evaluated and the results are presented in Appen-
jpx0 mpy 0 t
þ cos sin sinðlj cos usÞ cosðmm sin usÞ dix B.
a b 0

600
T [°C]
400
200

0.75
1

0.75 0.5 b [m]

a [m] 0.5
0.25
0.25

b 1

0.8

0.6

b [m]

0.4

0.2

0
0 0.2 0.4 0.6 0.8 1
a [m]

Fig. 4. Temperature distribution T ðx; y; z; tÞ on the square plate surface for z ¼ h at time t = 3600 s (heat source moves along an circular trajectory): (a) 3D
plot, (b) isothermal lines; a ¼ 1:0 m, b ¼ 1:0 m, h ¼ 0:01 m, A ¼ B ¼ 0:25 m.
870 J. Kidawa-Kukla / International Journal of Heat and Mass Transfer 51 (2008) 865–872

4. Numerical examples source which moves over its surface was established. The
moving heat source causes cyclic heating of various plate
The solution to the considered three-dimensional heat areas. The temperature distribution in the considered plate
conduction problem is used in the numerical investigation in an analytical form was obtained using the time-depen-
of the temperature distribution in a rectangular steel plate dent Green’s function. The advantage of this approach is
which is heated by a moving heat source. For each calcula- that a solution without any additional simplification can
tion, dimensional and physical properties of the plate and be obtained. Numerical calculation of the temperature dis-
the heat source are: k = 51.4 W m1 K1, j = 1.29  tribution were performed using the analytical form of the
105 m2 s1, u = 0.2 p rad s1, a0 = a1 = 100 W m1 K1, solution. The changes in temperature on the plate surface
x0 ¼ a=2, y 0 ¼ b=2 and H = 105 W m2. The temperature are shown for a steel plate subjected to the activity of a
of the surrounding medium is assumed as T 0 ¼ moving heat source. The results are presented for the heat
T 1 ¼ 100 °C. The calculations were performed with use source moving along a section parallel to one edge of the
the Mathematica software. plate, moving along an elliptical trajectory, and moving
In the first example, it was assumed that the heat source in a circular trajectory. The temperature is highest at the
moves harmonically along a section parallel to one edge of point of heat source location, but the temperature
the plate (it was assumed B = 0 in Eq. (3)). In this case, the decreases behind the source. In the data assumed for the
calculations were performed for a plate with: a ¼ 0:5 m, numerical calculations, in particular for the assumed veloc-
b ¼ 0:4 m, h ¼ 0:02 m and A ¼ 0:2 m. The temperature dis- ity of the heat source, the temperature field of the plate
tribution on the plate surface z ¼ h, for time t = 3600– area changes insignificantly over time apart from in a lim-
3610 s, is shown in Fig. 2. The time interval between obser- ited area around the heat source. Moreover, the tempera-
vations of the temperature of the plate is set at 10 s because ture is considerably lower in the area around the centre
this is the time the heat source takes to execute a complete of the circular or elliptical trajectory than near the
cycle of movement. A relative long time overall observation trajectory.
time (3600–3610 s) was selected in order that the process
can be treated practically as in a steady state during this Appendix A
time. The location of the heat source is clearly visible in
the figure and is shown by the temperature peak. The tem- The Green’s function Gðx; y; z; t; n; g; 1; sÞ of the three-
perature is lower in front of heat source than behind the dimensional heat conduction problem ((10) and (11)) can
location of the heat source. be expressed in a form of the product [6]
The temperature distribution on the plate surface when
the heat source moves along an elliptical trajectory is illus- Gðx; y; z; t; n; g; 1; sÞ
trated in Fig. 3. The geometrical dimensions of the consid- 4j
¼ H ðt  sÞGX ðx; n; t  sÞGY ðy; g; t  sÞGZ ðz; f; t  sÞ
ered plate are: a ¼ 1:0 m, b ¼ 0:5 m, h ¼ 0:01 m. The ab
elliptical trajectory of the movement of the heat source is ðA:1Þ
characterized by Eq. (3) with A ¼ 0:3 m and B ¼ 0:2 m.
The temperature distribution on the plate surface is pre- where H ðt  sÞ is the Heaviside function, GX ðx; n; t  sÞ,
sented for t = 3600 s, when the heat source is located at GY ðy; g; t  sÞ and GZ ðz; f; s  tÞ are Green’s functions of
the vertex of the ellipse. The temperature increases dramat- one-dimensional problems: in x-direction, y-direction and
ically around the point of the plate at the moment of heat z-direction. The formulations of the considered one-dimen-
source transition, and the temperature changes over a low sional initial- boundary problems are as follows:
range only in the area of the middle of the ellipse. x-direction:
In the third example, the temperature distribution on a  2 
o 1 o
square plate heated by a source which moves along a circu-  GX ðx; n; t  sÞ ¼ dðx  nÞdðt  sÞ ðA:2Þ
ox2 j ot
lar trajectory was determined. The centre of the circle coin-
cides with the middle of the square. The following data GX jt¼0 ¼ 0; GX jx¼0 ¼ GX jx¼a ¼ 0 ðA:3Þ
were assumed for the calculations: a ¼ 1:0 m, b ¼ 1:0 m, y-direction:
h ¼ 0:01 m, A ¼ B ¼ 0:25 m. The temperature distribution  
on the plate surface z ¼ h, for t = 3600 s as a function of o2 1 o
 GY ðy; g; t  sÞ ¼ dðy  gÞdðt  sÞ ðA:4Þ
two variables x, y, is presented in Fig. 4(a), and the isother- oy 2 j ot
mal lines are shown in Fig. 4(b). Isotherms behind the heat GY jt¼0 ¼ 0; GY jy¼0 ¼ GY jy¼b ¼ 0 ðA:5Þ
source are much thicker than in the front and the position
of the moving heat source can be clearly seen (Fig. 4(b)). z-direction:
 
o2 1 o
5. Conclusions  GZ ðz; f; t  sÞ ¼ dðz  fÞdðt  sÞ ðA:6Þ
oz2 j ot
GZ jt¼0 ¼ 0; ðkoGZ =oz  a0 GZ Þjz¼0 ¼ 0; ðkoGZ =oz þ a1 GZ Þjz¼h ¼ 0;
In this paper, an analytical model to describe the three-
dimensional temperature field for a finite plate with a heat ðA:7Þ
 J. Kidawa-Kukla / International Journal of Heat and Mass Transfer 51 (2008) 865–872 871

The Green’s functions GX ðx; n; t  sÞ, GY ðy; g; t  sÞ and Hence the function K jmn ðtÞ in Eq. (18) may be then written
GZ ðz; f; s  tÞ, are given by Beck et al. in Ref. [4]. The func- as
tions can be written in the form X1 X 1 
 2 ð1Þi sin
jpx0 mpy 0 ð1Þ
2X 1
jp jpx jpn K jmn ðtÞ ¼ 4
a
sin
b i k
v v J 2i ðlj ÞJ 2k ðmm ÞI jmn ðt; 2i; 2kÞ
GX ðx; n; t  sÞ ¼ expðj ðt  sÞÞ sin sin i¼0 k¼0
a j¼1 a a a jpx0 mpy 0 ð2Þ
þ sin cos v J 2i ðlj ÞJ 2kþ1 ðmm ÞI jmn ðt; 2i; 2k þ 1Þ
ðA:8Þ a b i
  jpx0 mpy 0
2X 1
mp 2 jpy jpg þ cos sin
ð1Þ
v J 2iþ1 ðlj ÞJ 2k ðmm ÞI jmn ðt; 2i þ 1; 2kÞ
GY ðy; g; t  sÞ ¼ exp j ðt  sÞ sin sin a b k
b m¼1 b b b 
jpx0 mpy 0 ð2Þ
þ cos cos J 2iþ1 ðlj ÞJ 2kþ1 ðmm ÞI jmn ðt; 2i þ 1; 2k þ 1Þ
ðA:9Þ a b
2X 1
1 ðB:9Þ
GZ ðz; f; t  sÞ ¼ expðjb2n ðt  sÞÞwn ðzÞwn ðfÞ
h n¼1 Qn
where
ðA:10Þ
Z t
where wn ðzÞ ¼ bn cos bn z þ l0 sin bn z, Qn ¼ ðb2n þ l20 Þ ð1Þ
I jmn ðt; r; sÞ ¼ cosðrusÞ cosðsusÞ expðjc2jmn ðt  sÞÞ ds
2
þl0 l1
1 þ hbb2nðl sin2 bn h , for n = 1, 2, . . . , and bn are roots 0
n 0 þl1 Þ
of the equation: ðB:10Þ
Z t
ð2Þ
bn ðl0 þ l1 Þ cos bn h  ðb2n  l0 l1 Þ sin bn h ¼ 0: ðA:11Þ I jmn ðt; r; sÞ ¼ cosðrusÞ sinðsusÞ expðjc2jmn ðt  sÞÞ ds
0
ðB:11Þ
Appendix B
After evaluation the integrals (B.10), (B.11) are:
An analytic form of the integrals in equation (18) can be 1
ð1Þ
obtained by using the following relationships [10]: I jmn ðt; r; sÞ ¼ expðtc2jmn jÞc2jmn jðc4jmn j2 þ ðr2 þ s2 Þu2 Þ
C jmn ðr  sÞC jmn ðr sÞ
X
1
1
cosðr sin uÞ ¼ 2 vi J 2i ðrÞ cos 2iu ðB:1Þ þ ½c2 j cos½ðr  sÞtu þ ðr  sÞu sin½ðr  sÞtu
2C jmn ðr  sÞ jmn
i¼0
1
X
1 þ ½c2 j cos½ðr þ sÞtu þ ðr þ sÞu sin½ðr þ sÞtu
sinðr sin uÞ ¼ 2 J 2iþ1 ðrÞ sinð2i þ 1Þu ðB:2Þ 2C jmn ðr þ sÞ jmn
i¼0 ðB:12Þ
and the relationships which can be obtain from (B.1, B.2): ð2Þ 1
X
1 I jmn ðt; r; sÞ ¼ expðtc2jmn jÞsuðc4jmn j2 þ ðr2 þ s2 Þu2 Þ
C jmn ðr  sÞC jmn ðr þ sÞ
cosðr cos uÞ ¼ 2 ð1Þi vi J 2i ðrÞ cos 2iu ðB:3Þ
1
i¼0 þ ½ðr  sÞu cos½ðr  sÞtu  c2jmn j sin½ðr  sÞtu
2C jmn ðr  sÞ
X
1
sinðr cos uÞ ¼ 2 ð1Þi J 2iþ1 ðrÞ cosðð2i þ 1ÞuÞ ðB:4Þ þ
1
½ðr þ sÞu cos½ðr þ sÞtu þ c2jmn j sin½ðr þ sÞtu
i¼0 2C jmn ðr þ sÞ
where J m ðrÞ are Bessel functions and v0 ¼ 0:5, vi ¼ 1 for ðB:13Þ
i ¼ 1; 2; . . . . From Eqs. (B.1)–(B.4) one obtains
cosðlj cos utÞ cosðmm sin utÞ with C jmn ðuÞ ¼ c4jmn j2 þ u2 u2 .
X1 X 1
i
¼4 ð1Þ vi vk J 2i ðlj ÞJ 2k ðmm Þ cos 2iut cos 2kut References
i¼0 k¼0

ðB:5Þ [1] G. Araya, G. Gutierrez, Analytical solution for a transient three-

cosðlj cos utÞ sinðmm sin utÞ dimensional temperature distribution due to a moving laser beam,
Int. J. Heat Mass Transfer 49 (2006) 4124–4131.
X1 X 1
¼4 ð1Þi vi J 2i ðlj ÞJ 2kþ1 ðmm Þ cos 2iut sinð2k þ 1Þut [2] P.J. Cheng, S.C. Lin, An analytical model for the temperature field in
i¼0 k¼0 the laser forming of sheet metal, J. Mater. Process. Technol. 101
(2000) 260–267.
ðB:6Þ
[3] O. Manca, B. Morrone, V. Naso, Quasi-steady three-dimensional
sinðlj cos utÞ cosðmm sin utÞ temperature distribution induced by a moving circular Gaussian heat
X1 X 1
i
source in a finite depth solid, Int. J. Heat Mass Transfer 38 (7) (1995)
¼4 ð1Þ vk J 2iþ1 ðlj ÞJ 2k ðmm Þ cosð2i þ 1Þut cos 2kut 1305–1315.
i¼0 k¼0 [4] J.V. Beck, K.D. Cole, A. Haji-Sheikh, B. Litkouhi, Heat Conduction
ðB:7Þ Using Green’s Functions, Hemisphere, New York, 1992.
[5] J. Kidawa-Kukla, Application of the Green functions to the problem
sinðlj cos utÞ sinðmm sin utÞ
of the thermally induced vibration of a beam, J. Sound Vibr. 262
X1 X 1
i (2003) 865–876.
¼4 ð1Þ J 2iþ1 ðlj ÞJ 2kþ1 ðmm Þ cosð2i þ 1Þut sinð2k þ 1Þut [6] J. Kidawa-Kukla, Analysis of beam vibration induced by moving
i¼0 k¼0
heat source, in: H.A. Mang, F.G. Rammerstorfer, J. Eberhardsteiner,
ðB:8Þ J. (Eds.), Proceedings of the Fifth World Congress on Computational
872 J. Kidawa-Kukla / International Journal of Heat and Mass Transfer 51 (2008) 865–872

Mechanics (WCCM V), July 7–12, 2002, Vienna, Austria. Vienna [8] J.V. Beck, A. Haji-Sheikh, D.E. Amos, D. Yen, Verification solution
University of Technology, Austria, ISBN 3-9501554-0-6, <http:// for partial heating of rectangular solids, Int. J. Heat Mass Transfer 47
wccm.tuwien.ac.at>. (2004) 4243–4255.
[7] J. Kidawa-Kukla, Temperature distribution in a rectangular plate [9] Y.A. Melnikov, An alternative construction of Green’s functions for
subjected to the activity of a mobile heat source, in: XII International the two-dimensional heat equation, Eng. Anal. Bound. Elem. 24
Science and Engineering Conference. Machine-Building and Techn- (2000) 467–475.
osphere of the XXI Century. 12–17 September 2005, Sevastopol, vol. [10] H.S. Carslaw, J.C. Jaeger, Conduction of Heat in Solids, Clarendon
5, 2005, pp. 111–115. Press, London, 1959.