International Journal of Discoveries and

Innovations in Applied Sciences

| e-ISSN: 2792-3983 | www.openaccessjournals.eu | Volume: 3 Issue: 10

Computer Modeling of Two-Dimensional Non-Stationary Heat

Conduction Problems Considering Point Heat Sources Using the FEM

Jumaniyozov Sardor Pirnazarovich

National University of Uzbekistan named after Mirzo Ulugbek

Abstract:
The process of heat exchange in mechanical systems is discussed in the article. The solution to two-
dimensional non-stationary heat conduction problems in the presence of point heat sources is
studied based on the finite element method. To solve a plane heat conduction problem, a triangular
finite element with rectilinear sides is used. The basic relations for linear triangular finite elements
are presented, taking into account the influence of point heat sources. A source set at a specific
point on the plane describes a heater in the form of a string. This point on the plane is the trace of
the string and is set by the heat generation power per unit length. Based on the developed
computational algorithm and software, a numerical solution to a specific problem is given and the
influence of point heat sources on the distribution of the temperature field of the structure is
studied.
Keywords: heat transfer, heat source, convective heat transfer, non-stationarity, FEM.

In many engineering problems, knowledge of the temperature distribution in a structure is an

important aspect. The amount of heat supplied to it or lost by it can be calculated provided that the
temperature distribution is known. In the article, the temperature field is formed by considering the
influence of a point heat source set at a specific point on the plane and describes the presence of a
heater in the form of a string. The trace of the string is a point on the plane, determined by the
power of heat generation.
Formulation of the problem
The problem of temperature distribution at different time points is solved in a plane formulation [1,
2]. To find the temperature field in the rectangular calculation domain (a plate) ABCD (Fig. 1), a
two-dimensional non-stationary heat conduction problem is solved based on equation [4]:
T  2T  2T
  K xx 2  K yy 2  Q, (1)
t x y
where   c is the specific volumetric heat capacity; c is the specific heat capacity of the
material;  is the density; K xx , K yy are the heat conductivity coefficients in the corresponding
directions; Q is the power of the heat source inside the body.

Published under an exclusive license by open access journals under Volume: 3 Issue: 10 in Oct-2023
Copyright (c) 2023 Author (s). This is an open-access article distributed under the terms of Creative Commons
Attribution License (CC BY).To view a copy of this license, visit https://creativecommons.org/licenses/by/4.0/

35
 International Journal of Discoveries and
Innovations in Applied Sciences
| e-ISSN: 2792-3983 | www.openaccessjournals.eu | Volume: 3 Issue: 10

Fig. 1. Scheme of the computational domain

Symmetry conditions are set on lines AD and BC (the so-called natural boundary conditions [3]):
T T
 0, 0. (2)
x AD x BC

A constant temperature (Dirichlet condition) is set on line AB, equal to the ambient temperature T∞:
T AB  T . (3)
Conditions that determine the convective exchange of heat with the medium can be set on line AB:
T
K yy  h(T  T )  0 , (4)
y
where h is the heat transfer coefficient.
The temperature field throughout the entire computational domain at time t = 0:
T ( x, y,0)  T 0 ( x, y) is set as the initial condition.
In this case, the solution to the system consisting of equation (1) and boundary conditions (2) - (4)
is reduced to minimizing the functional [4]:

1 T 
2
 T   T 
2
h
    K xx    K yy    2QT  2 T  dV   (T  T )2 , (5)
V
2   x   y  t  S1
2

where S1 is the area of surface where convective heat exchange occurs.

Finite element solution to the problem
The temperature in the finite element is specified as the product of two independent functions:
T (e)  N ( x, y)T (t )
or in a matrix form:
Ti (t ) 
 
T (e)
  Ni ( x, y ) N j ( x, y ) N k ( x, y)  T j (t )  . (6)
T (t ) 
 k 
For finite element e , the condition for the extremum of the functional leads to the following
system of differential equations:
Published under an exclusive license by open access journals under Volume: 3 Issue: 10 in Oct-2023
Copyright (c) 2023 Author (s). This is an open-access article distributed under the terms of Creative Commons
Attribution License (CC BY).To view a copy of this license, visit https://creativecommons.org/licenses/by/4.0/

36
 International Journal of Discoveries and
Innovations in Applied Sciences
| e-ISSN: 2792-3983 | www.openaccessjournals.eu | Volume: 3 Issue: 10
 T 
c e    k e  T    f e   0 , (7)
t
here
[c e ]   [ N ][ N ]
T
dV , (8)
Ve

[k e ]   [ Be ][ De ][ Be ]T dV   h[ N ][ N ]T dS , (9)
Ve S2

{f (e) }    (e)
Q[N(e) ]T dV   h  [N (e) ]T dS , (10)
V S(2e )

where
Ve is the volume of the finite element;
[N ] is the shape function matrix;

[ B e ] is the matrix of shape function derivatives;

[ D e ] is the matrix of material properties containing thermal conductivity coefficients.
Integral (8) when solving a two-dimensional heat conduction problem has the following form
(damping matrix):
2 1 1
 Aat 
[c ] 
e
 1 2 1  , (11)
12
1 1 2 
where
А is the area of the finite element; at is the element thickness.
The thermal conductivity matrix (in the absence of convection) has the following form:
 bi bi bi b j bibk   ci ci ci c j ci ck 
k xx at   k a  
[k ] 
e
b j bi bjbj b j bk   yy t c j ci c jc j c j ck  ,
4A 
(12)
4A
bk bi bk b j bk bk  ck ci ck c j ck ck 
 
where bi  y j  yk ; ci  x j  xk ; (the remaining quantities are obtained by circular permutation of
indices i, j, k).
If side i of the j finite element is subject to convection, then the second integral in (9) is:
2 1 0
hat Lij  
S h[ N ][ N ] dS  6 1 2 0 ,
T
(13)
2
0 0 0 

where Lij is the length of the side of the element between nodes i and j.

Published under an exclusive license by open access journals under Volume: 3 Issue: 10 in Oct-2023
Copyright (c) 2023 Author (s). This is an open-access article distributed under the terms of Creative Commons
Attribution License (CC BY).To view a copy of this license, visit https://creativecommons.org/licenses/by/4.0/

37
 International Journal of Discoveries and
Innovations in Applied Sciences
| e-ISSN: 2792-3983 | www.openaccessjournals.eu | Volume: 3 Issue: 10
If heat is lost by convection between sides with nodes j, k or k, i, then matrix (13) is transformed
[4].
We assume that the value of Q is constant inside the element, then:
 L1  1
  QV  
Q  [N] dV  Q  L 2  dV 
T
1 . (14)
3 
V V
 L3  1
If convective heat exchange occurs on the side with nodes i j, then the “load vector” in solving the
heat conduction problem is:
1 
hT ij t  
L a
S hT [ N ] dS  2 1 .
T

0 
 
2

For a finite element mesh, a system of ordinary differential equations is written as:
{T }
[C ]  [ K ]{T }  {F}  0 , (15)
t
where
[C ]  [ce ]; [ K ]  [k e ]; [ F ]  [ f e ] .
e e e

Replacing the time derivative in equation (15) with its finite-difference analog, we obtain an
implicit difference scheme for solving the heat equation using the finite element method [3-6]:
 [C ]   [C ] 
  [ K ] {T }n1    [ K ] {T }n  {F}n1 (16)
 t   t 
Thus, if the temperature vector T  at time point t n is known, then the temperature of the plate at
n

time point tn1  tn  t is obtained as a result of solving the system of linear algebraic equations
(16).
Computational experiment and analysis of results
The following problem is considered as a test example [7]:
Three cables pass through a heat-conducting medium, as shown in Fig. 2. The thermal conductivity
W
coefficients of the medium is K x  K y  50 . The heat transfer coefficient on the surface of
cm  K
W
the medium is h  10 . The medium under consideration is limited on the sides by a thick
cm2  K
layer of insulation. Air temperature on the surface of the medium is Т   40 0C . Temperature of
the lower layer of the medium is T0  20 0C . The heat radiation power of each cable is Q  200W
. Copper has the following thermophysical characteristics (1, in Fig. 2):
  384 W / (mº C ),   8800 kg / m3 , с  381 J / (kgº C) . Thermophysical parameters of the material of
inclusion are: steel (2, in Fig. 2): 2  46 W / (m  С ), 2  7800 kg / m , с2  460 J / (kg  С ) .
3

Published under an exclusive license by open access journals under Volume: 3 Issue: 10 in Oct-2023
Copyright (c) 2023 Author (s). This is an open-access article distributed under the terms of Creative Commons
Attribution License (CC BY).To view a copy of this license, visit https://creativecommons.org/licenses/by/4.0/

38
 International Journal of Discoveries and
Innovations in Applied Sciences
| e-ISSN: 2792-3983 | www.openaccessjournals.eu | Volume: 3 Issue: 10
Dimensions of the material of inclusion: n=18 cm, m=2.375 cm. Plate dimensions: a=5 cm, b=4
cm, c=5 cm, d=5 cm, e=4 cm.

а b
Fig. 2. Arrangement of cables in a heat-conducting medium

The temperature distribution in a homogeneous material (Fig. 2.a) at t = 6, 120, and 300 seconds is
presented in Fig. 3. At the initial stage of the process (t = 6 sec), there is no influence of sources,
then at (t = 120 and 300 sec), the influence of sources is observed, characterized by thermal
concentration in the vicinity of sources [9-12].

t = 6 sec t = 120 sec t = 300 sec

Fig.3. Temperature distribution in a homogeneous material

In the presence of a rectangular steel area in a structure made of copper (Fig. 2 b), a redistribution
of the generated heat is observed in its lower part (Fig. 4). The change in the temperature field
reflects the difference in the thermal conductivity coefficients of copper and steel.

t =6 sec t =120 sec t =300 sec

Fig.4. Temperature distribution in a non-homogeneous material
Published under an exclusive license by open access journals under Volume: 3 Issue: 10 in Oct-2023
Copyright (c) 2023 Author (s). This is an open-access article distributed under the terms of Creative Commons
Attribution License (CC BY).To view a copy of this license, visit https://creativecommons.org/licenses/by/4.0/

39
 International Journal of Discoveries and
Innovations in Applied Sciences
| e-ISSN: 2792-3983 | www.openaccessjournals.eu | Volume: 3 Issue: 10
For a specific point (2.5 cm, 11.5 cm) located on the left side of the cavity, table 1 presents
temperature values that indicate the influence of the non-homogeneity of the structure, namely a
decrease in temperature in the vicinity of sources.
Table 1. Temperature at point (2.5 cm, 11.5 cm)
Plate material t=6 sec t=120 sec t=300 sec
Homogeneous 0.0358 17.6320 30.3977
Non-homogeneous 0.0016 10.5583 23.5279
Conclusion
Thus, the presence of sources increases the temperature field in the vicinity of sources, and the
presence of non-homogeneity changes the overall temperature background of the structure.
References
1. Samarskii A. A., Vabishchevich P. N. Computational heat transfer. - M: Editorial, URSS, 2003.
- 784 p.
2. Tsaplin A.I. Modeling of thermophysical processes and objects in metallurgy: Teaching aid -
Perm: Publishing House of Perm State Technical University, 2011. - 299 p.
3. Zenkevich O. Finite element method in technology. – M.: Mir, 1975. - 452 p.
4. Segerlind L. Application of the finite element method. – M.: Mir, 1979. – 392 p.
5. Zhukov N. P., Mainikova N. F., Nikulin S. S. Antonov O. A. Solving heat conduction problems
using the finite element method. -Tambov, FSBEI HPE "TSTU", 2014. - 80 p.
6. Kuznetsov G.V., Sheremet M.A. Difference methods for solving heat conduction problems.
Tomsk, TPU, 2007. - 172 p.
7. Ivannikov L. M. Solution of a two-dimensional heat conduction problem by the finite element
method in MATHCAD. Khabarovsk, Pacific. National University, 2011. - 28 p.
8. Andreev A. Elasticity and thermoelasticity of layered composite shells. Mathematical model
and some aspects of numerical analysis. Saarbrucken, Deutschland: Palmarium Academic
Publishing, 2013. - 100 p.
9. Akhmat Ikramov, Askhad Polatov, Sukhbatulla Pulatov, Sardor Jumaniyozov. Computer
simulation of two-dimensional nonstationary problems of heat conduction for composite
materials using the FEM. AIP Conference Proceedings 2637, 040006 (2022);
https://doi.org/10.1063/5.0118684 Published Online: 20 October 2022.
10. Polatov A.M., Ikramov A.M., Jumaniyozov S.P., Sapaev Sh.O. Computer simulation of two-
dimensional unsteady-state heat conduction problems for inhomogeneous bodies by the FEM.
Modern Problems of Applied Mathematics and Information Technology (MPAMIT 2021). AIP
Conference Proceedings 2781, 020019 (2023); https://doi.org/10.1063/5.0144813 Published by
AIP Publishing.
11. Ikramov A.M., Jumaniyozov S.P., Sapaev Sh.O. Computer simulation of two-dimensional
stationary heat conduction problems taking into account point heat sources using the FEM.
Problems of Computational and Applied Mathematics, 2021, 3 (33). pp. 44 – 53.
12. Polatov A, M., Ikramov A. M., Jumaniyozov S. P., Sapaev Sh. O. Computer modeling of two-
dimensional non-stationary heat conduction problems for nonhomogeneous bodies using the
FEM. Problems of Computational and Applied Mathematics, 2022, 2 (39). pp. 61 – 71.

Published under an exclusive license by open access journals under Volume: 3 Issue: 10 in Oct-2023
Copyright (c) 2023 Author (s). This is an open-access article distributed under the terms of Creative Commons
Attribution License (CC BY).To view a copy of this license, visit https://creativecommons.org/licenses/by/4.0/

40