The current issue and full text archive of this journal is available on Emerald Insight at:

www.emeraldinsight.com/0332-1649.htm

Mathematical modeling of Induction

surface
induction surface hardening hardening
Jerzy Barglik
Department of Industrial Information Technology, 1403
Faculty of Material Sciences and Metallurgy,
Silesian University of Technology, Katowice, Poland Received 6 September 2015
Revised 12 January 2016
Accepted 26 January 2016
Abstract
Purpose – As far as the author knows the modeling of induction surface hardening is still a challenge.
The purpose of this paper is to present both mathematical models of continuous and simultaneous
hardening processes and exemplary results of computations and measurements. The upper critical
temperature Ac3 is determined from the Time Temperature Austenization diagram for investigated steel.
Design/methodology/approach – Computation of coupled electromagnetic, thermal and hardness
fields is based on the finite element methods, while the hardness distribution is determined by means of
experimental dependence derived from the continuous cooling temperature diagram for investigated steel.
Findings – The presented results may be used as a theoretical background for design of inductor-
sprayer systems in continual and simultaneous arrangements and a proper selection of their
electromagnetic and thermal parameters.
Research limitations/implications – The both models reached a quite good accuracy validated by
the experiments. Next work in the field should be aimed at further improvement of numerical models in
order to shorten the computation time.
Practical implications – The results may be used for designing induction hardening systems and
proper selection of field current and cooling parameters.
Originality/value – Complete mathematical and numerical models for continuous and simultaneous
surface induction hardening including dual frequency induction heating of gear wheels. Experimental
validation of achieved results. Taking into account dependence of the upper critical temperature Ac3 on
speed of heating.
Keywords Mathematical modelling, 3D FEM, Induction heating
Paper type Research paper

Introduction
Induction surface hardening is a kind of heat treatment where a hardened body is
heated by induction and then immediately cooled mostly by spraying its surface or
sometimes by merging the body into a container with a quenchant (Rudnev and Totten,
2014). The process consists of two consecutive stages: induction heating and intensive
cooling with a short austenization period between them. Let us consider a time
evolution of temperature in different zones of the hardened body (Figure 1).
Curves 1-4 in Figure 1 represent zones with different requirement of hardness and
microstructure of the material. In order to obtain the completely hardened material two
conditions should be satisfied. The first of them is connected with the uniform austenite
microstructure before cooling. The first condition is satisfied at the surface zone 1 where
the hardening temperature Th is bigger than the upper critical temperature Ac3:
T h ¼ Tjt¼th þ te 4 Ac3 (1) COMPEL: The International
Journal for Computation and
Mathematics in Electrical and
Electronic Engineering
Vol. 35 No. 4, 2016
pp. 1403-1417
Financial support from the National Centre for Research and Development (project PBS2/A5/ © Emerald Group Publishing Limited
0332-1649
41/2014) is kindly acknowledged. DOI 10.1108/COMPEL-09-2015-0323
COMPEL T
Th
35,4 A c3 1

A c1 2

3
1404
4
I II III
Ms
Ms f
th te tc

t
Notes: 1 – surface zone; 2 – transition zone;
3 – tempering zone; 4 – internal zone; Ac3,
Ac1 – upper and lower critical temperature,
Figure 1. respectively; Ms – martensite start temperature;
Time evolution of Msf – martensite finish temperature;
temperature in four Th – hardening temperature; th, te, tc – time of
zones of the
hardened body heating I, austenization II and quenching III,
respectively

The second condition is connected with parameters of cooling. Time evolution of

temperature during cooling (continuous cooling temperature (CCT) diagram) for
investigated steel is presented in Figure 2.
The final temperature after cooling should be smaller than the martensite finish
temperature Msf:
Tjt¼th þ te þ tc o M sf (2)
The hardness depends on velocity of cooling. In general the shorter time of cooling tc
means more harder microstructure. So for the surface zone 1 intensively cooled the
complete hardening is noticed characterized by high hardness (for the case shown

1,000
T °C A
900
Ac3
800
Ac1
F
700 P
600
500
400
B
300 Ms
200 Ms f
M
100 HV = 750 735 700 650 620 492 289 253
Figure 2. t c (s)
0
Continuous cooling 1 10 102 103 104 105
temperature diagram
for investigated steel Notes: A – austenite; M – martensite; B – Bainite; F – ferrite;
P – pearlite; HV – hardness in Vickers degrees
in Figure 2 it means the hardness of 700-750 Vickers degrees and acicular Induction
martensite microstructure). surface
For the transition zone the temperature T reaches lower values:
hardening
Ac3 4Tjt¼th þ te 4 Ac1 (3)
The austenite begins to form at the lower critical temperature Ac1, but the forming is
not completed. As a result the incomplete hardening characterized by lower hardness 1405
and microstructure containing the martensite with addition of bainite, ferrite and
pearlite is obtained. The shape and thickness of the hardened layer is controlled by the
basic process parameters like: field current frequency, time and power of heating, kind
of quenchant, cooling intensity, etc. For the zone 3 the temperature before cooling does
not reach the lower critical temperature:
Tjt¼th þ te p Ac1 (4)
As a result, the microstructure containing the tempered martensite is obtained. For the
internal zone 4 the temperature is distinctly lower than Ac1 and the microstructure does
not change. The material in zone 4 remains soft. However, it is necessary to notice
that induction heating is characterized by big heating rates reaching a range of about
1,000°C/s or sometimes even more. For such a rapid induction heating an increase of
both critical temperatures Ac1, Ac3 is noticed (Barglik, 2012). In order to determine these
values the Time Temperature Austenitization diagram should be determined by
measurements. Especially important is knowledge about the dependence between the
upper critical temperature Ac3 and velocity of heating. Example of such a dependence
for a carbon steel is shown in Figure 3.
For the slow heating (vh ¼ 0.01°C/s) the upper critical temperature Ac3 ¼ 835°C.
If the velocity of heating is higher the upper critical temperature increases. For
vh ¼ 350°C/s, Ac3 ¼ 1,035°C. In comparison with the critical temperature for a slow
conventional hardening it means the increase of about 200 K. So in order to obtain
requested austenite microstructure after fast induction heating the modified upper
critical temperature Ac3m should be applied and it is distinctly bigger than for the
conventional hardening:
Ac3m ¼ Ac3 þ DT ðvh Þ; DT ¼ 502501C (5)

1,100
Ac3, °C
1,050
1,035
1,000

950
350°C/s.

900

850

800 Figure 3.
750 Dependence of the
upper critical
700 temperature on
0.01 0.1 1 10 100 1,000 speed of heating
Speed of heating, °C/s
COMPEL In order to secure the uniform austenite microstructure after induction heating and time of
35,4 temperature equalization the hardening temperature Th should be taken bigger than Ac3.

Analyzed examples
As examples two kinds of the induction surface hardening technologies are analyzed:
1406 (1) continuous induction hardening of a bending brake made of steel C50E; and
(2) simultaneous induction hardening of gear wheel made of steel CrMo5.
Scheme of induction system for the first example is shown in Figure 4.
During the process the element sufficiently long in the direction of axis z is heated
and then cooled by the moving inductor-sprayer system. The line between points B and
C (Figure 4) should be hardened, however, a material near point D can remain soft.
Depth of the hardened layer between points B and C Δg ¼ 2.5 × 10−3 m.
The second process is realized as a consecutive induction heating by a ring inductor
and then immediate cooling by spraying (Figure 5). The inductor is equipped with the
flux concentrator (3) (Nemkov, 2013). In general cooling could be realized as well by
merging the gear into a container with a quenchant. Requirements for a hardness
distribution are discussed in (Barglik et al., 2014).

y 2 1
A A’
B B’
C C’
D

Figure 4. x
Cross-section of
induction heating 3
system for the
bending brake Notes: 1 – element; 2 – inductor-
sprayer system; 3 – symmetry axes

3
2
1

Figure 5.
Cross-section of
simultaneous
induction heating
system for gear
wheels Notes: 1 – gear wheel; 2 – inductor-
sprayer system; 3 – flux concentrator
Mathematical models Induction
Mathematical models for both tasks are given by differential equations describing surface
distribution of non-stationary and non-linear coupled electromagnetic and temperature
fields. The transient electromagnetic field is described by the equation for the magnetic
hardening
vector potential A (Lupi et al., 2015):
 
1 @A
curl curl A þ g þ v  curl A ¼ J z (6) 1407
m @t
where μ is the magnetic permeability; γ, the electric conductivity; v, the velocity of
movement; and Jz, the field current density.
But solving the transient problem is complicated because of a huge disproportion
between the period of the field current (a range of 10−4 s or less) and the time of heating
th (seconds). In order to shorten the calculation time to a reasonable value the
assumption that the magnetic permeability of any ferromagnetic part of the domain
may be considered as constant at individual cells of the system. Based upon this
simplification the problem may be analyzed as harmonic and the Equation (6) is
transformed into a form for the phasor of the magnetic vector potential A:


curl curl A þ jomgA þ mg v  curl A ¼ mJ z (7)

where j is the imaginary unit; and ω, the angular frequency.

The third left side term of (7) could be neglected for a slow moving inductor-sprayer
system (Doležel et al., 2003). Taking into consideration this simplification the phasor of
eddy current density Jind is expressed by:
J ind ¼ jogA (8)

The total volumetric power density pV released in the hardened body consists of two
components: the Joule losses pJ and the hysteresis losses pH:
2
J ind
pV ¼ pJ þ pH ; pJ ¼ ; pH ¼ pH ðjB j; f Þ (9)
g
where B is the magnetic flux density; and f the frequency.
For both electromagnetic problems the hysteresis losses are small in comparison
with the Joule losses. They could be neglected.
Temperature calculations are provided by means of the Kirchhoff-Fourier equation
supplemented by internal heat source represented by the volumetric Joule losses:
@T
divðlgradT Þrcp ðvgradT Þrcp ¼ pJ (10)
@t
where λ is the thermal conductivity; ρ, the density; and cp, the specific heat at a constant
pressure.
The boundary condition along the surface is given by the convection and radiation
for heating and natural cooling represented at the right side of (11) by the generalized
heat transfer coefficient αg (Barglik et al., 2008; Holman, 2009):
@T
l ¼ ag ðT S T a Þ (11)
@n
COMPEL and by the convection only for cooling:
35,4 @T

l ¼ ac T S T q (12)
@n
where n is the unit outward normal; αc, αg, the coefficients of convection and
generalized heat transfer, respectively; Ts, the temperature of body surface; Ta, the
1408 ambient temperature; and Tq, temperature of quenchant.
The mathematical model of induction surface hardening becomes more simple if the
electromagnetic field is formulated as 2D task and described by the tangent component
of the magnetic vector potential At only. However, both processes presented in the
paper have to be analyzed as 3D problem.
The first process has to take into consideration several different arrangements with
time dependent boundary conditions. For the second process: simultaneous induction
hardening of gear wheel by means of cylindrical inductor the 3D analysis are provided
for one-half of the tooth.

Results and their discussion

The first task is solved for the following parameters:
(1) Hardening body: bending brake:
• material: carbon steel C50E (its chemical composition in Table I);
• critical temperatures Ms ¼ 325°C, Ac1 ¼ 730°C, Ac3 ¼ 780°C, Ac3m ¼ 880°C;
• hardening temperature Th ¼ 930°C;
• dependence of hardness on velocity of cooling derived from the
CCT diagram;
• dependence of material properties on temperature are taken into account;
• hardness requirements 600-720 HV; and
• length, width and height, respectively, l ¼ 0.118 m, b ¼ 0.095 m, h ¼ 0.0125 m.
(2) Inductor and its supply parameters:
• copper conductor of the rectangular cross-section 3 mm × 9 mm;
• field current density Jz ¼ 1.05 × 107 A/m2;
• its frequency f ¼ 10 kHz, 20 kHz; and
• velocity of inductor-sprayer system movement v ¼ 2 × 10−3 m/s.
(3) Heat transfer and cooling parameters:
• quenchant: water in spray;
• temperature of quenchant Tq ¼ 10°C;

Table I. C Mn Si P S Cr Ni
Chemical
composition of C50E 0.47-0.55 0.6-0.9 0.4 0.035 0.035 0.4 0.4
steel C50 E Note: Chemical composition in percent mass
 • generalized heat transfer coefficient for heating αg ¼ 20 W/(m2 K); Induction
• convection heat transfer coefficient for cooling αc ¼ 500 W/(m K); and 2
surface
• ambient temperature Ta ¼ 30°C. hardening
The numerical algorithm is presented in Figure 6.
Calculations start with the completion of input data partly based on measurements
of material properties. At the beginning the stationary induction heating of the front 1409
part of the bending brake is provided. Computation of the quasi-stationary
electromagnetic field for the starting position of the inductor-sprayer system is
provided. Based upon known value of the volumetric Joule losses the non-stationary
temperature field is calculated. After time t1 ¼ 15 s the inductor-sprayer system starts
to move. Continual movement of inductor-sprayer system is replaced by a number n of
discrete shifts. It means a different configuration of the inductor-sprayer system in each
position and time dependent boundary conditions of the temperature field. The time
step Δt ¼ 0.5 s corresponds to the shift Δl ¼ 10−3 m. For each cycle the dependence of
material properties on temperature is taken into consideration. Results of temperature
distribution is saved in order to use them for further hardness determination. Then the
inductor-sprayer system is shifted to the next position and the computations are
repeated. When the final position of inductor-sprayer system is reached the cooling of
last part of the bending brake is calculated (t2 ¼ 5 s) and then the hardness distribution
is determined. The calculations are provided by the FEM based single-owned software
prepared in the framework of the Polish-Czech project (Barglik, 2012). The mesh for
electromagnetic field has about 250,000 elements. The mesh for temperature field

measurements
INPUT DATA

START

ELECTROMAGNETIC FIELD (⎥ B⎥, T ),  (T )

pJ, B
NEXT
POSITION cp(T ), (T ),
TEMPERATURE FIELD Radiation
c(T ), g(T )
procedures
T
t + Δt
DATABASE OF
TEMPERATURE EVOLUTION

T (t, x, y, z )

no FINAL n
POSITION

yes
Figure 6.
DETERMINATION CCT
Numerical algorithm
OF HARDNESS diagram for computation of
the continual
induction hardening
END
COMPEL contains about 30,000 elements. The time evolution of temperature during heating and
35,4 cooling for two points on a line parallel to the axis z passing through a center of BC
segment (see, Figure 4) is presented in Figure 7.
The temperature in this zone reaches its maximum of 1,000°C. There are some
temperature differences before points along this line in a range of not more than 40 K.
The lowest value of temperature is noticed in the front part of the body where induction
1410 heating starts. Such temperature evolution secures the complete hardening of the
material in the part of the body.
Figure 8 presents the time evolution of temperature in a lower part of the working
surface near point C (see, Figure 4).
Accordingly to requirements material in this zone of the bending brake should
remain soft. Average value of temperature after heating does not exceed 290°C,
however, there are bigger differences between points along this line being a range 70 K
with the maximum of 330°C for z ¼ −0.059 m. Such temperature evolution matches the
requirements and secures keeping the material soft.

1,000
900
800
700
600
T, °C

500
400

Figure 7. 300
Time evolution of 200
temperature on 100
points in the line
0
parallel to the axis z 0 10 20 30 40 50 60 70 80 90 100
passing through a t, s
center of the z=0m z = –0.03 m
segment BC
Source: See Figure 4

300

250

200
T, °C

150

100

Figure 8.
50
Time evolution of
temperature on
0
points in a line 0 10 20 30 40 50 60 70 80 90 100
parallel to the axis z t, s
passing through a
z=0m z = –0.03 m
point D
Source: See Figure 4
Calculated and measured hardness distribution along the working surface of the Induction
bending brake is presented in Figure 9. surface
There is a quite good accordance between measurements and computations with a
maximal inaccuracy of 28 HV. Also distribution of hardness along the working surface
hardening
match the requirements. At the upper corner of the hardened body the hardness
exceeds 420 HV. It means the incompletely hardened zone. At the central part of the
bending brake the hardness exceeds 600 HV. The material is fully hardened there. And 1411
finally at the lower part of the body near point D hardness is about 210 HV. It means
that material remains soft.
The second task of simultaneous hardening is solved for the following parameters of
the cylindrical coil-gear wheel system:
(1) Gear wheel and its material properties:
• material: steel 50CrMo4 with a chemical composition collected in Table II;
• critical temperatures Ms ¼ 300°C, Ac1 ¼ 720°C, Ac3 ¼ 800°C, Ac3m ¼ 850°C;
• hardening temperature Th ¼ 900°C;
• number of teeth n ¼ 41;
• diameter D ¼ 0.041 m;
• width b ¼ 0.0138 m;
• angle of pressure β ¼ 14.5°;
• dependence of hardness on velocity of cooling is derived from the CCT
diagram;

700

600

500
hardness, HV

400

300

200

100 Figure 9.
Hardness
0
0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1 distribution along
A B a /a0 C D the working surface
of the bending brake
calculations measurements

Table II.
Chemical
C Cr Mo Mn Si S P composition of
steel 50CrMo4
0.45-0.54 0.9-1.2 0.15-0.3 0.5-0.8 0.4 0.035 0.035 (percent)
 •
COMPEL dependence of material properties on temperature and magnetic permeability
35,4 on temperature and magnetic flux density are taken into account;
• hardness requirements 620-720 HV in thin contour zone; and
• depth of hardening zone g ¼ 2.6 × 10−4 m.
(2) Kind of induction heating:
1412 • single frequency; and
• dual frequency.
(3) Inductors and their supply parameters:
• ring copper inductor for medium frequency (MF) induction heating;
• MF field current IMF ¼ 1,275 A, 1,750 A, 3,660 A;
• its frequency f ¼ 10 kHz, 36 kHz;
• time of MF heating th ¼ 6 s, 10 s;
• average temperature after MF heating TMFfinal ¼ 500°C;
• time of technological break between MF and high-frequency (HF) heating
tb ¼ 0.5 s;
• ring copper inductor for HF induction heating;
• flux concentrator made of Fluxtrol 50;
• material properties and parameters of the concentrator (density ρ ¼ 6.1.103
kg/m3, relative magnetic permeability μr ¼ (63-120), electric conductivity
γ ¼ 0.2 S/m, maximal temperature Tmax ¼ 250-300°C;
• HF field current IHF ¼ 1,300 A;
• its frequency f ¼ 200 kHz; and
• time of HF heating th ¼ 1 s.
(4) Heat transfer and cooling parameters:
• quenchant: polymer solution Polyhartenol E8;
• generalized heat transfer coefficient for heating and natural cooling αg ¼ 15
W/(m2 K); and
• convection heat transfer coefficient for intensive cooling αc ¼ 5,000 W/(m2 K).
Numerical algorithm is presented in Figure 10.
Calculations start with the completion of input data and calculation of coupled
electromagnetic and temperature fields during induction heating I (Figure 10).
Dependences of all material properties on temperature as well as the dependence of the
magnetic permeability on temperature and on the module of the magnetic flux density
are respected. When the dual frequency induction hardening is applied the heating
cycle repeated two times: the first medium frequency step until the assumed
temperature TMFfinal is achieved and then after a short technological break necessary
for a removal of inductors the second HF heating step is provided until the hardening
temperature Th. Then the austenization (natural cooling) II and quenching (intensive
cooling) III are calculated. And finally the hardness distribution based upon the
 INPUT DATA
measurements Induction
surface
START hardening

ELECTROMAGNETIC FIELD (⎥ B⎥, T ), (T ) 1413

I
p v, B

cp(T ), (T ),
TEMPERATURE FIELD c(T ), g(T ) Radiation
procedures
T

TEMPERATURE FIELD II

c
TEMPERATURE FIELD III

DETERMINATION CCT
OF HARDNESS diagram Figure 10.
Numerical algorithm
for computation of
END the simultaneous
induction hardening
Notes: I – induction heating; II – austenitization; III – cooling

measured CCT diagram for investigated steel is determined. The calculations are
provided by the Flux 3D software with a number of numerical procedures elaborated at
the Silesian University of Technology. The number of the nodes for the electromagnetic
computations is over 90,000, for thermal computations about 7,000. The computation
time of one example takes approximately seven hours (Smalcerz, 2015). In case of the
simultaneous induction surface of gear wheels there is a special requirements for
hardness distribution within the tooth body (Figure 11).
The tooth body and its root is divided into four zones with different hardness and
microstructure requirements (Spezzapria et al., 2012; Schlesselmann et al., 2013). The
thin contour zone 1 is fully hardened with the acicular martensitic microstructure. Its
hardness reaches values between 620 and 720 HV. The transition zone 2 represents a
part of incompletely hardened material with a hardness lower than that in zone 1 and
with a mixed microstructure containing not only martensite, but also ferrite and
pearlite. The hardness of zone 3 representing the core of the tooth should be below 498
HV with the microstructure of the tempered martensite. The hardness of the internal
zone 4 containing the unchanged material is below 280 HV.
Two different arrangements of the inductor-sprayer system corresponding to
different kind of induction heating are analyzed. The single frequency induction
hardening by a MF inductor and water spraying and the dual frequency induction
heating by the MF inductor, then by the HF inductor equipped with a flux concentrator
and then cooling by the polymer solution spraying. Due to the symmetry of the
COMPEL 1
35,4
2

4
1414

Figure 11.
Zones of the tooth Notes: 1 – contour zone;
body with different 2 – transition zone;
hardness
requirements 3 – tempering zone;
4 – internal zone

arrangement, it is enough to analyze just a quarter of the tooth. The temperature

distribution after the single frequency induction heating is presented in Figure 12.
The heating is realized by the field current IMF ¼ 2,660 A of frequency f ¼ 36 kHz
during time tMF ¼ 6 s with a break between heating and cooling te ¼ 0.1 s.
The average temperature in the contour zone Tav ¼ 925°C. So it is sufficiently bigger
than the modified upper critical temperature Ac3m for steel 50CrMo4. Measurements are
provided at the experimental stand located in the ELKON company. After hardening
the gear wheel is tempered in temperature Tt ¼ 160°C during 7,200 s. Displacement of
the points where both values are compared are shown in Figure 13. The calculated
hardness is compared with measurements (Table III).
The dual frequency induction surface hardening is provided for the same gear
wheel. First, the medium frequency heating is realized at field current of IMF ¼ 1,750 A
and frequency of f ¼ 10 kHz in time of tMF ¼ 10 s (Figure 14).

Isovalues : TCelsius
Max
957.775
944.312
930.849
917.386
903.923
890.460
876.997
863.534
850.071
836.608
823.145
809.682
796.219
Figure 12.
782.756
Temperature
769.294
distribution before
755.831
cooling for the
742.368
single frequency
induction hardening
Notes: IMF = 2,660 A; f = 36 kHz; tMF = 6 s
Average temperature after the medium frequency induction heating Tav ¼ 485°C. After Induction
a short technological break for removal of inductors (tb ¼ 0.5 s) the HF induction surface
heating starts (IHF ¼ 1,300 A, f ¼ 200 kHz, tHF ¼ 1 s). Temperature distribution at the
working surface of the tooth before cooling is presented in Figure 15.
hardening
Average temperature at the working surface of the tooth Tav ¼ 880°C.
Measurements are provided at the induction hardening stand equipped with MF and
HF generators located at the Silesian University of Technology (Figure 16). 1415

1 2 3 4 Figure 13.
Distribution of
points for
comparison of
measured
hardness and
the computations

Points 1 (HV) 2 (HV) 3 (HV) 4 (HV) Table III.

Hardness
Measurements 650 658 662 620 distribution for
Calculations 635 635 645 596 single frequency
Note: Hardness distribution for selected points of tooth body for single frequency induction hardening hardening

Isovalues : TCelsius
Max
492.994
489.618
486.242
482.866
479.490
476.114
472.739
469.363
465.987
462.611
459.235
455.859
452.484
449.108
445.732 Figure 14.
442.356 Temperature
438.980 distribution after
435.604 medium frequency
induction heating
Notes: IMF = 1,750 A, f = 10 kHz, tMF = 10 s
COMPEL Isovalues : TCelsius
Max
35,4 896.803
881.960
867.117
852.275
837.432

1416 822.589
807.746
792.904
778.061
763.218
748.375
733.532
718.690
703.847
689.004
674.161
Figure 15. 659.319
Temperature 644.476
distribution before Notes: IMF = 1,200 A; f = 10 kHz; tMF = 6 s; IHF = 1,300 A; f = 200 kHz;
cooling
tHF = 1s; te = 0.1s

1
2

3
4

Figure 16.
Laboratory stand for
dual frequency
induction hardening Notes: 1 – MF inductor; 2 – HF
inductor; 3 – gear wheel; 4 – sprayer

The hardness distribution in the same points as shown in Figure 13 is collected

in Table IV.
In both analyzed cases a correct hardness distribution along the working surface of
the gear wheel is obtained. Quite reasonable convergence between the computations
and experiments is obtained. In both cases the working surface is hardened, but the
expected contour shape of hardened layer is obtained while the dual frequency
induction surface hardening process is provided.

Points 1 (HV) 2 (HV) 3 (HV) 4 (HV)

Table IV.
Hardness Measurements 650 680 685 640
distribution for dual Calculations 690 700 710 660
frequency hardening Note: Hardness distribution for selected points of tooth body for the dual frequency induction hardening
Conclusion Induction
Mathematical modeling of surface induction hardening is still uneasy because surface
of necessity to solve coupled non-linear and non-stationary electromagnetic and
temperature fields mostly in 3D formulation. Non-linear dependencies of material
hardening
properties, critical temperatures and heat transfer parameters are respected. Majority
of these material parameters and their temperature dependencies are measured. Some
specific features of the processes and their technical background were discussed. The 1417
reasonable, acceptable from technical point of view accordance between calculations
and measurements of hardness is achieved. The main problem seems still to be a big
time of computations. The analysis of one 3D task takes a couple of hours. Next
research in the domain should be concentrated on more efficient numerical algorithms
making possible to shorten effectively the computation time.

References
Barglik, J. (2012), “Induction hardening of steel tubes by means of internal inductor”, Journal of
Iron and Steel Research International, Vol. 19 Nos S1-S2, pp. 722-725.
Barglik, J., Czerwiński, M., Hering, M. and Wesołowski, M. (2008), Radiation in Modelling of
Induction Heating Systems, IOS Press, Amsterdam, pp. 202-211.
Barglik, J., Smalcerz, A., Przyłucki, R. and Doležel, I. (2014), “3D modeling of induction hardening
of gear wheels”, Journal of Computational and Applied Mathematics, Vol. 270, pp. 231-240.
Doležel, I, Barglik, J., Sajdak, C., Škopek, M. and Ulrych, B. (2003), “Modelling of induction heating and
consequent hardening of long prismatic bodies”, COMPEL: The International Journal for
Computation and Mathematics in Electrical and Electronic Engineering, Vol. 22 No. 1, pp. 79-87.
Holman, P. (2009), Heat Transfer, McGraw Higher Education, New York, NY.
Lupi, S., Forzan, M. and Alifierov, A. (2015), Induction and Direct Resistance Heating. Theory and
Numerical Modelling, Springer International Publishing, Berlin.
Nemkov, V. (2013), “Magnetic flux control in induction installations”, Proceedings of the
International Symposium Heating by Electromagnetic Sources, Padua, pp. 117-126.
Rudnev, V. and Totten, G. (2014), Induction Heating and Heat Treatment, ASM International,
Materials Park, OH.
Schlesselmann, D., Nikanorov, A., Nacke, B., Galunin, S. and Schon, Z. (2013), “Numerical
calculation and comparison of temperature profile and martensite microstructures in
induction surface hardening processes”, Proceedings of the International Symposium
Heating by Electromagnetic Sources, Padua, pp. 127-134.
Smalcerz, A. (2015), “The use of multifrequency induction heating for temperature distribution
control”, Archives of Metallurgy and Materials, Vol. 60 No. 2, pp. 721-725.
Spezzapria, M., Dughiero, F., Forzan, M. and Candeo, A. (2012), “Multiphysics FEM simulation of
contour induction hardening process on aeronautic gears”, Journal of Iron and Steel
Research International, Vol. 19 No. S1, pp. 95-98.

Corresponding author
Jerzy Barglik can be contacted at: jerzy.barglik@polsl.pl

For instructions on how to order reprints of this article, please visit our website:
www.emeraldgrouppublishing.com/licensing/reprints.htm
Or contact us for further details: permissions@emeraldinsight.com