The generalized mathematical model of heat

conduction in a complex multi-layered area

Cite as: AIP Conference Proceedings 1895, 090004 (2017); https://doi.org/10.1063/1.5007404
Published Online: 12 October 2017

A. Zaika, O. Hrytsiuk, E. Kobilskaya, et al.

ARTICLES YOU MAY BE INTERESTED IN

Mathematical model of heat transfer in an electric machine

AIP Conference Proceedings 2025, 080006 (2018); https://doi.org/10.1063/1.5064926

Mathematical modeling and numerical simulation of an energy generating hollow cylinder

AIP Conference Proceedings 2336, 030005 (2021); https://doi.org/10.1063/5.0045768

Mathematical models of the temperature field in the filament-extruder system during 3D

printing
AIP Conference Proceedings 2302, 100008 (2020); https://doi.org/10.1063/5.0033797

AIP Conference Proceedings 1895, 090004 (2017); https://doi.org/10.1063/1.5007404 1895, 090004

© 2017 Author(s).
The Generalized Mathematical Model of Heat Conduction in
a Complex Multi-layered Area
A. Zaika1,a), O. Hrytsiuk1,b),E. Kobilskaya1,c) and V. Lyashenko1,d)
1
Kremenchuk Mykhailo Ostrohradskyi National University, 20 Pershotravnev str., 39600 Kremenchuk, Ukraine

a) anton_zaika@rambler.ru
b) myboxua13@gmail.com
c) Corresponding author: kobilskaya1983@gmail.com
d) viklyash2903@gmail.com

Abstract. A mathematical model of thermal process in an electrical machine was built as an example, presented as a
three-layer cylinder where internal heat sources operate in one of the layers and heat is submitted to the other two by
means of heat conduction. A method of solving the boundary-value problems for heat conduction equation in a complex
area – a multi-layered cylinder with internal heat sources operating in one part of the layers and external ones in another
part, is proposed. A method of problem solution in conditions of uncertainty of one of the boundary condition at the
layers interface with conductive heat exchange between the layers is reviewed. The principle of method lies in the
averaging of temperature distributions radially in the internal layers. As a result of transformations at the layers interface
a boundary condition of the impedance-type conjugation appears. The analytical and numeric-analytical solutions of
simplified problems were obtained.

INTRODUCTION
In the mathematical modeling problems of thermal processes a range of challenges related to nonlinear nature of
the problems, complexity of investigation areas and boundary conditions at their interfaces, inhomogeneous medium
limiting the investigation areas, and tridimensionality of the problems appears in the mathematical modeling
problems of thermal processes appear. For linear and some of the nonlinear problems with linear boundary
conditions in canonical areas analytical methods are applied (Fourier method, integral transformation method,
thermal potential method, conformal representation method, different methods of problem linearization).
Opportunities of these methods require of the researchers to simplify output problems of heat exchange to the level
where problem solution becomes practically enforceable.
An important advantage of analytical solutions of heat exchange problems is a possibility to review their
dependence on parameters in an explicit form. Analytical solutions help to sort out the determining criteria among
all their diversity. It facilitates the physical experiment set up greatly. The accurate analytical solutions often play
the role of standard solutions and serve as a test task for checking the adequacy of more complex nonlinear models.
In comparison with the above mentioned analytical methods numerical methods gave the opportunity to get a full
quantity description of heat exchange in the complex form constructions and to use complex nonlinear mathematical
models. Nonlinear mathematical models describe the investigated thermal processes more accurately in a wide
geometrical and physical parameters change area [1, 2].
Numerical methods with use of contemporary electronic computer allow mathematical prediction of thermal
processes in quite complicated physical objects, particularly in modern electrical machines.
At the close of the XXth century the rapid growth of electrical machines building started due to the development
of small electrical machines for space branch, plane construction and electrical cars. Magnetic core and winding are
the active parts of any electrical machine. All the rest of the parts are constructive, providing necessary toughness,

Application of Mathematics in Technical and Natural Sciences

AIP Conf. Proc. 1895, 090004-1–090004-9; https://doi.org/10.1063/1.5007404
Published by AIP Publishing. 978-0-7354-1579-9/$30.00

090004-1
durability, slewing capacity, and cooling. An important role in their reliable work is played by the range of operating
temperatures of electrical machine.
Not withstanding the diversity of the types of electrical machines they have a lot of common structural elements:
wire with low electrical resistance, isolation material with high dielectric properties, magnetic conductors with
anisotropic properties in orthogonal directions. All this allows us building mathematical models of electromagnetic
and thermal processes in electrical machine. A method of temperature field investigation in a three-layer cylinder
was reviewed by us through the example of mathematical model of thermal process in the simplest electrical
machine.
Determination of three-dimensional temperature distributions generated by electromagnetic field activity in
electrical machines is an extremely complicated problem of mathematical modeling and mathematical physics [3, 4,
5, 6, 7, 8]. Electromagnetic fields are described by models containing Maxwell equations system, equation of charge
continuity and Ohm’s law [9, 10, 11, 12]
)& ))&
))& 4S )& 1 w E )& 1 w H
rot H J , rot E ,
N N wt N wt
)& ))& wUe )& )& )&
divE 4SUe , divH 0,  divJ 0, J V e E.
wt
))& )& )&
Here H , E , J , Ue , V e –magnetic fieldstrength, electric field strength, electric current density, electric charge density,
electrical conductivity. Presence of the fields and currents in an electrical machine stimulates the electrocaloric
effect in the subareas where electrical current flows. So an energy equation must be added to the Maxwell equation
in the mathematical model allowing determining the temperature distributions in the windings and between them
because the model will not be consistent without it. Furthermore, electrical conductivity V e of electrical machine
windings depends on the temperature. To determine the internal energy in electrical machine the thermal
conductivity equation is involved in the following form
we
Un div  O T  gradT   W  P, t , T  ,
wt
where O , U n are the thermal conductivity coefficient and winding materials density, respectively, e ³ cQ dT –
:
specific internal energy, cQ – generalized thermal capacity of the environment, W  P, t , T  – heat sources density in
)&
the electrical machine windings. The displacement current w E is neglected in many Maxwell equations
wt
)&
))& 4S )& 1 w E
rot H J
N N wt
)& ))&
in comparison with conduction current J generated by the electrical current density, H – magnetic field intensity in
)&
electrical machine windings, E – electric field intensity, N –speed of light. The assumption of that kind is appropriate
)&
because the displacement current w E in electrical machines makes much less influence on the temperature
wt
)&
distribution in comparison with conduction current J . Also Peltier heat appearing in the moving contacts is often
neglected.
After acceptance the abovementioned assumptions the Maxwell equations system can be written in the quasi-
steady-state approximation in the area :t in the following non-dimensional form
))&
))& )& )& wH ))& )& )&
rot H 4SV e E , rot E  , divH 0, J V e E W  P, t  .
wt
)&
Neglecting the displacement current w E allows simplifying the mathematical model greatly by changing the
wt
equation system and turning it from the hyperbolic into parabolic. Model is significantly simplified by decreasing
the number of boundary conditions.
Alternating electric current flowing through the electrical machine windings releases the energy in case of direct
current and in case of backward current as well. Furthermore, electrical energy of alternating current causes the

090004-2
change ofinternal energy of the conductors. Due to energy of alternating magnetic field and magnetic reversal, the
coil cores are heated by the whirling currents. Electric and magnetic energy cause the constructive elements
temperature rise. It is also called active energy, and the active heat sources density can be determined in this case by
the formula
U0 L(1  E T)
W J uU u R / Z , R , Z R 2  X L2 ,
S
where S, U0 , X L , U, Z, R are wire section area, specific resistance of winding wire, windings inductance, voltage in
windings, operating impedance, active impedance. If the electrical machine power factor cos M is known, active
J u U cos M
heat sources density is determined by the formula W .
s
Mathematical models that permit to investigate electromagnetic and thermal fields in electrical machine can be
designated as models of processes occurring in a complex area.On the basis of the foregoing in the simplest case the
heat exchange process in electrical machine can be presented in the form of temperature field in a three-layered
cylinder, where the internal heat sources operate in one of the layer (winding) and external heat sources operate in
two other cylinder layers.The simpler problem of temperature field investigation in a two-layered cylinder was
reviewed in works [5, 6].

Purpose of the paper

Purpose of the paper is to construct, using the simplest electrical machine as an example, a mathematical model
of temperature field in a multi-layered cylinder where internal heat sources operate in one part of the layers and
external sources operate in another part. It is also necessary to develop a method for determining the temperature
field in a complex multilayer area.
A schematic illustration of a three-layered cylinder is shown in Figure 1.

FIGURE 1. A three-layered cylinder with different thermal-physical characteristics of the layers

STUDY RESOURCES AND RESULTS

Let us consider the mathematical model of temperature field in a three-layered cylinder where internal heat
sources work in one of the layers and heat is transferred by means of thermal conductivity to two other layers having
different thermal-physical properties. Internal heat sources constitute Joule heat released at the time of electric
current flow through one of the layers. As an example let us consider the temperature field of an electrical machine
consisting of a two-layered rotor without heat sources and a stator with the windings of the copper wire or aluminum
inserts where electric current flows and generates Joule heat Figure 1.
Determination of the temperature distribution T (r , z , t ) in a three-layered cylinder where internal heat sources
operate in an external layer and heat is transferred from the external layer to the internal ones by means of thermal

090004-3
conductivity has a form of the following boundary problem on conjunction in the area
: u t : ^0  r  r0 , 0  z  l , 0  t  t0 `

­ I 2 U0 (1  E T )
1 w wTi w 2T wT ° , r1 d r  r2 (1)
Oi (r )  Oi 2i  ci Ui i ® S2
r wr wr wz wt °
¯0, 0  r  r0 , r0  r  r1 , r ! r2

Ti (r , z , 0) T0 , (2)

Ti (r , 0, t ) T0 , Ti (r , l , t ) T0 , (3)

wT1 wT3
0, O3 ª D (T3  Tc )  HV (T34  Tc4 ) º , (4)
wr wr ¬ ¼
r 0 r r2

T (ri 0 , z , t ) T1 (ri  0 , z , t ), i 1...3, (5)

wTi wTi 1 (6)

Oi Oi 1 , i 1...3,
wr r ri  0 wr r ri  0

where Oi , ci , Ui , i 1...3, Tc are respective thermal-physical characteristics and parameters of the parts of an
electrical machine,ambient temperature D , I , U0 , H , V , E are heat transfer coefficient, current strength, resistivity
of the windings wire, emissivity factor, Stefan-Boltzmann constant and temperature coefficient of
resistance,respectively. Here S is a total section area of the wire in the stator windings, particularly S3 S r22  r12  
layer surface area S3 .
The problem in an arrangement (1)-(6) is non-linear, so the analytical solution of the problem of that kind does
not exist. Let us simplify the problem according to the following algorithm to find the analytical solution.
It is enough to know the homogenized temperature distribution in the third and second layers to investigate the
temperature distribution in the inner cylinder. So to determine the transmission rate of heat flowing through the
inner cylinder surface, let us multiply the equation (1) by rdr and integrate over the layer thickness within the limits
from r1 to r2 .
r r2 r2 r r2
wT3 w2 w 2 I 2 U0 (7)
O3 r
wr
 O3
wz 2
³ T3rdr  c3 U3 wt r³
T3 rdr 
S3 2
³ (1  E T3 )rdr ,
r r1 r1 1 r1
r r2
wT3 wT3 (r1 , z , t )
O3 r
wr

r2 D 3 (T3  Tc )  HV (T34  Tc4 )  O3 r1  wr
.
r r1
After the transformations and using the ratio [13]
r
2 2
S3 r³
u3 ( z , t ) T3 (r , z , t )rd r, (8)
1
Eq. (7) can be displayed in the form of
S3 wT3 (r1 , z , t ) O3 S3 w 2u3c3 U3 S3wu3 I 2 U0 E I 2 U0 § r22 r12 · (9)
O3 r1  u  ¨  ¸
S32 ¨© 2 2 ¸¹
3
2 wr 2 wz 2 2wt 2 S3
 r2 ª¬D (u3  Tc )  HV (u34  Tc4 ) º¼ .

090004-4
 The proportion (9) will be considered as boundary condition of impedance type at the interface of the outer and
inner cylinders. With the following taken as u ( z , t ) | T3 (r1 , z , t ) | T2 (r1 , z , t ) , the boundary condition at the interface
of the second and the third layers of the cylinder with taking into account the direction of the heat flow, the
condition (9) has a form of
wT (r , z , t )
O2 2 1
O2 w 2T2

c3 U3O2 wT2 I 2 U0 EO2
 T2 

I 2 U0 O2 r22  r12
wr r1 wz 2
O3 r1 wt O3 r1S32 O3 S33 r1
(10)

O2 r2 2 ª
 D (T2  Tc )  HV (T2 4  Tc4 ) º¼ .
r1S3O3 ¬

After the determination of the boundary condition the temperature distribution in the second layer of the cylinder
reduces to the following problem with a condition that includes a tangent derivative alongside with a normal
derivative by the direction of
1 w wT2 w 2T wT (11)
O2 (r )  O2 22  c2 U 2 2 0,
r wr wr wz wt

r0  r  r1 , 0  z  l, t!0
T2 (r , z , 0) T0 , (12)

T2 (r , 0, t ) T0 , T2 (r , l , t ) Tl , (13)

(14)
wT (r , z , t )
O2 2 1
O2 w 2T2

c3 U3O2 wT2 I 2 U0 EO2
 T2 

I 2 U0 O2 r22  r12
wr r1 wz 2 O3 r1 wt O3 r1S32 O3 S33 r1
O2 r2 2 ª
 D (T2  Tc )  HV (T2 4  Tc4 ) º¼ ,
r1S3O3 ¬
T2  r1 0 , z , t 
wT  r , z , t 
O2 O1 1 10 . (15)
wr wr
The second boundary condition in this problem arrangement (11)-(15) is indefinite. The principle of method of
temperature distributions determination in a multi-layered cylinder in the conditions of uncertainty lies in the
stepwise averaging of the temperature fields along the radius and in the following determination of the temperature
field in the first layer 0  r  r0 . After finishing the averaging the backward process of temperature distributions
determination in the inner layers starts r0  r  r1 , r1  r  r2 , rN 1  r  rN .
If it is possible to neglect the temperature distribution along the radius in the second layer of the cylinder, too,
then, after averaging by analogy with (8) we get the boundary condition at the interface of the first and second
cylinders
wT (r , z , t )
O2 2 0
O2 w 2u2

c3 U3O2 wu2 I 2 U0 EO2
 u2 

I 2 U0 O2 r22  r12
wr r0 wz 2
O3 r0 wt O3 r0 S32 O3 S33 r0
2 (16)
O2 r2 2 ª 4 4 º O2 w u2 c U wu
 ¬D (u 2  Tc )  HV (u 2  T )
c ¼   2 2 2 0.
r0 S3O3 r0 wz 2
r0 wt

In this way the temperature distribution in the inner cylinder (layerS1) is determined from the solution of the
following problem in the area :1 u t ^0  r  r0 , 0  z  l , t ! 0`

090004-5
 1 w wT1 w 2T wT (17)
O1 (r )  O1 21  c1 U1 1 0,
r wr wr wz wt
T1 (r , z , 0) T0 , (18)
T1 (r , 0, t ) T0 , T1 (r , l , t ) Tl , (19)

O1
wT1 (r0 , z , t ) 2O1 w 2 u1 O1c3 U3 O1c2 U 2 wu1 I 2 U0 EO1
 (  )  u 
I 2 U0 O1 r22  r12

 
1
wr r0 wz 2 O3 r0 O2 r0 wt O3 r0 S32 O3 S33 r0
(20)
O1r2 2 ª
 D (u1  Tc )  HV (u14  Tc4 ) º¼ 0,
r0 S3O3 ¬

wT1  0, z , t  (21)
0,
wr
r
2 0 (22)
S1 ³0
u1 ( z , t ) T2 (r , z , t )rd r, S1 S r0 2 ,

r r0 r0 r
wT1 w2 w 0 (23)
O1r
wr
 O1
wz 2
³ T1rdr  c1U1 wt ³0
T1rdr 0,
r 0 0

O1r
wT1
r r0
2O1
O1c3 U3 O1c2 U 2 wu1 I 2 U0 EO1
w 2u1
 ( 
I 2 U0 O1 r22  r12 O r 2 
 1 2 ª¬D (u1  Tc )  HV (u14  Tc4 ) º¼ .
 (24)
) u 1
wr
r 0 wz 2
O3 O 2 wt O3 S3 2
O3 S33
S3O3
By using (24) in (23) and taking into account (22) we get
w 2 u1 c3 U3 c2 U 2 c1 U1 wu1
(   ) (
I 2 U0 E r2 2D
 )u1 
I 2 U0 r22  r12

 
wz 2 3O3 3O2 3O1 wt 3O3 S32 3S3O3 3O3 S33 (25)
r2 2D r 2HV Tc4 r 2HV 4
Tc  2
( ) 2 u1 0
3S3O3 3S3O3 3S3O3
and come to the following boundary problem for differential equation of second order
w2v wv (26)
 a2  Av  Cv 4 B, 0  z  l, t!0
wz 2
wt
v( z , 0) T0 , (27)
v(0, t ) T0 , v(l , t ) Tl , (28)
where
I 2 U0 E r2 2D c3 U3 c2 U 2 c1 U1 r2 2HV
A (  ), a2 (   ), C  ,
3O3 S3 2
3S3O3 3O3 3O2 3O1 3S3O3

B (

I 2 U0 r22  r12   r 2D T
2

r2 2HV Tc4
).
c
3O3 S33 3S3O3 3S3O3
Suppose that H 0 the problem (26)-(28) can be solved by analytical means.
After replacing the variables
A
v e P z  Ot u, P 0 O 2
, f (z, t) Be  Ot ,
a
Eq. (26) can be reduced to the form of
ut a 2 u zz  f  z, t  , t ! 0, 0  z  l (29)
With nonhomogeneous boundary and initial conditions
u  z , 0  M( z ) T0 e Pz , (30)

090004-6
 u  0, t  J1 (t ) T0 eOt , u l, t  J 2 (t ) Tl ePl Ot .
(31)
We seek the solution of the problem (29)-(31) as a sum of two functions
u  z,t  v  z,t   w  z,t  , (32)
where v  z,t  is a new required function and w  z,t  is an arbitrary function satisfying the nonhomogeneous
boundary conditions
z
w(z, t) > J 2 (t)  J1 (t)@  J1 (t). (33)
l
Function v  z,t  should satisfy the nonhomogeneous equation
vt  a 2 vzz f (z, t ), (34)
where
f ( z, t ) f  z , t   wt  a 2 wzz
and the additional initial and boundary conditions are
v(z, 0) M (z)  w(z, 0),
(35)
v(0, t ) J 1 (t )  w(0, t ),
v(l , t ) J 2 (t )  w(l , t ).
After plugging (33) in (35) the conditions become nonhomogeneous and the problem (34)-(35) is reduced to the
problem
ut a 2 u zz  f  z, t  , t ! 0, 0  z  l
u  z , 0  0,
u  0, t  u  l , t  0.
The solution of this problem has a form of
t l ª f 
n 2 S2 2
a (t W)
º
« 2 nS nS »
u  z, t  ³ ³ ¦ e [ f  [, W  d [d W
2
l sin z sin
«l n 1 l l »
0 0
«¬ »¼
t l
(36)
³ ³ G  z, [, t  W  f  [, W  d [d W.
00
Problem (26)-(28) of determining the temperature distribution in the inner cylinder is completely solved. We
build the finite-difference scheme of Crank-Nicolson [1, 2]. In the area :u t ^0 d z d l , t ! 0` we apply the
uniform grid by the length (coordinate z ) of the cylinder Z h ^ zm mh, m 0,1...M ` ; time grid (coordinate t )
Z 't ^t j j 't , j `
0,1... j0 with intervals: h l / M , 't t0 / j0 .
The problem solution in the area : is sought as a solution of a system of nonlinear algebraic equations for the
grid function u zm ,t j marked as uij .
uij 1  uij uij11  2uij 1  uij11  uij1  2uij  uij1 (37)
a2  Auij 1  C (uij 1 ) 4  B.
't 2h 2

Let us review and solve the linearized problem. The temperature values matrix obtained is used for the solution
of the problem (13)-(16) as a boundary condition.
For this purpose the finite-difference scheme of Douglas-Gann is built for the solution by alternating directions
method [1,2,3,4] in the area : u t {0 d r d r0 , 0 d z d l , t ! 0} with intervals h1 r0 / N , h2 l / M , 't t0 / j0
that has a form of

090004-7
 j 1/2 j 1/2 j 1/2 j 1/2 j 1/2
unj,m1/2  unj, m O2 un 1,m  un,m O2 un 1, m  2un,m  un 1,m
 
't / 2 rc2 U 2 2h1 c2 U 2 h12
j j j
O2 un, m 1  2un, m  un, m 1
 ,
c2 U 2 h22
j 1/2 j 1/2 j 1/2 j 1/2 j 1/2
unj,m1  unj,m1/2 O2 un 1,m  un,m O un 1, m  2un,m  un 1,m
 2 
't / 2 rc2 U 2 2h1 c2 U 2 h12
j 1 j 1 j 1
O2 un, m 1  2un, m  un, m 1

c2 U 2 h22
at the boundaries of the area
u2,j m1  4u1,j m1  3u0,j m1
0, n 0
2h1
for n N we substitute the value of the temperature distribution matrix at the boundary of outer cylinder.
Figure 2 shows the solution of problem (34)-(35), that is, the distribution of temperature in the inner layer of the
cylinder along z-coordinate.

m2
FIGURE 2. Solutions of the problem (34)-(35): curve 1-with D 1 , curve2- D 10 , curve3- D 100 K

CONCLUSIONS
The Generalized mathematical model of heat conduction in a complex multi-layered area is constructed in the
work. The method of solving the boundary problems for the heat conductivity equation in a complex area – multi-
layered cylinder where internal heat sources operate in one part of the layers and external sources operate in another
part in the conditions of uncertainty of one of the boundary conditions at the layers interface and with conductive
heat exchange between the layers, is proposed. The principle of method lies in the averaging the temperature
distributions by the radius in the internal layers if the boundary condition by the radius in one of the layers is not
determined. The temperature distribution averaging process along the radius decreases its dimensionality if it is
allowed by the physical model of the problem but as a result of the transformations the boundary condition of
impedance type appears at the interface of the layers. By the stepwise averaging by the radius and using the
conjunction condition at the interface of the layers the temperature distribution in the last layer is determined. The
next step of determining the temperature distributions in the multi-layered cylinder is the backward solution of the
problems from the last layer to the first. The analytical solution of the simplified problem for the three-layered
cylinder where internal heat sources operate in one of the layers and heat is transferred to other two layers by means
of heat conductivity is obtained. The analytical solution of simplified problem is obtained, the numerical

090004-8
experiments are conducted and the approximate solution of the nonlinear problem is obtained, the temperature
distributions are graphed.

REFERENCES
1. A.A. Samarsky and P.N. Vabishchevich, Computational Heat Transfer (Editorial URSS, Moscow, 2003).
2. R.D. Richtmyer, Difference Methods for Initial Value Problems (Interscience, New York, 1957), pp. 251–
321.
3. V. Lyashenko and E. Kobilskaya, “Contact of boundary-value problems and nonlocal problems in
mathematical models of heat transfer,” in AMiTaNS’15, AIP Conference Proceedings 1684, edited by M.D.
Todorov (American Institute of Physics, Melville, NY, 2014), paper 080009, 10p.
4. A. P. Slesarenko, O.P. Demyanchenko, V. P. Lyashenko, and E. B. Kobilskaya (2015) Visnyk of Kherson
National Technical University 3(54), 467–471.
5. V. P. Lyashenko and N.G. Kirilaha (2002) Visnyk of Zaporizhzhya National University 3, 64–69.
6. V. P. Lyashenko and T. A. Hryhorova (2010) Visnyk of V.N.Karazin Kharkiv National University 890, 47–
52.
7. O.A. Troitsky, V.I. Stashenko, V.G. Ryzhkov, V.P. Lyashenko, and E.B. Kobilskaya (2011) Problems of
Atomic Science and Technology 4, 111–117.
8. V. Lyashenko and T. Hryhorova, “Generalized mathematical model of thermal diffusion in powder
metallurgy,” in AMiTaNS’14, AIP Conference Proceedings 1629, edited by M.D. Todorov (American
Institute of Physics, Melville, NY, 2014), pp. 85–93.
9. R.P. Feynman, R.B. Leighton, and M. Sands, The Feynman Lectures on Physics, Vol. II: Mainly
Electromagnetism and Matter (New Millennium Edition, New York, 2010), pp. 324–410.
10. M.P. Galanin and Yu.V. Popov, Quasistationary Electromagnetic Fields in Inhomogeneous Media (Nauka,
Moscow, 1995). [in Russian]
11. I.E. Tamm, Fundamentals of the Theory of Electricity (Nauka, Moscow, 1966). [in Russian]
12. L.D. Landau and E.M. Lifshitz, Electrodynamics of Continuous Media (Nauka, Moscow, 1982).
13. A.V. Lykov, Theory of Heat Conductivity (Vyshcha Shkola, Moscow, 1967), pp. 31–32. [in Russian]

090004-9