ANALYTICAL, NUMERICAL AND EXPERIMENTAL VALIDATION OF COIL VOLTAGE IN INDUCTION MELTING PROCESS

International Journal of Electromagnetics ( IJEL ), Vol 1, No 1, August 2016
21
ANALYTICAL, NUMERICAL AND
EXPERIMENTAL VALIDATION OF COIL
VOLTAGE IN INDUCTION MELTING PROCESS
B Patidar, M.T.Saify, M.M.Hussain, S.K.Jha, and A.P.Tiwari
Bhabha Atomic Research Centre, Mumbai
ABSTRACT
This paper presents, mathematical model of induction heating process by using analytical and numerical
methods. In analytical method, series equivalent circuit (SEC) is used to represent induction coil and work
piece. Induction coil and workpiece parameters (resistance and reactance) are calculated by standard
formulas along with Nagaoka correction factors and Bessel functions. In Numerical method, magnetic
vector potential formulation is done and finite element method (FEM) is used to solve the field equations.
Analytically and numerically computed parameters such as equivalent coil resistance, reactance, coil
voltage, work piece power are compared and found that they are in good agreement. Analytically and
numerically obtained coil voltages at different frequencies are validated by experimental results. This
mathematical model is useful for coil design and optimization of induction heating process.
KEYWORDS: -
Electromagnetic, FEM, induction coil parameter, induction heating
1. INTRODUCTION
Induction heating is widely utilized in different industrial applications such as melting, forging,
welding, hardening etc [1, 2]. It is based on electromagnetic induction phenomenon, which is
very complex and results completely depend upon various design and process parameters [3].
Often, many experiments need to be conducted to optimize the melting process, which is
expensive and time consuming. Mathematical modeling and simulation of induction melting
process not only helps to avoid tedious experiments, it also helps to understand the process in
depth.
Most of literature on induction heating system design and optimization are based on
mathematical modeling and simulation [4]-[8]. Only few articles [9][10][11] are available, in
which analytical methods are used to explain the induction heating process. In this paper,
Analytical and numerical methods are used to calculate the coil voltages at different frequencies
and validated it with experimental results.
In Analytical method, Nagaoka short coil correction factor is used to account the end effect and
flux density outside the coil. Nagaoka modified correction factor is included in formulation to
account presence of work piece inside the coil [6]. Lead resistance and inductance are significant,
therefore they are also included in computation. Work piece resistance and power are calculated
by using Bessel function, because it gives more accurate values compared to generalized formulas

22
[6]. Coil equivalent inductance is used to select proper values of capacitors for coil tuning with
power source [1]. Analytical methods are generally applicable to simple geometries.
In Numerical method, magnetic vector potential formulation is used for computation of different
design parameters. Magnetic vector potential formulation requires less computation compared to
magnetic field formulation. Finite element method (FEM) based multiphysics software is used to
solve the field equations and compute various parameters such as coil impedance, voltage and
work piece power at different frequencies. Numerical methods can be applied to complicated
geometry, but it requires detailed understanding of coding language. Numerical methods are
iteration based and require lot of computation resources and time to solve the problem.
This paper is organized as follows: After introduction in section I, section II includes analytical
formulation which is used to calculate various parameters such as coil equivalent resistance,
reactance, coil voltage, work piece power, and coil efficiency. Mathematical model, numerical
method, simulation procedure and experimental set up details are also explained in section II. In
section III, all results computed from analytical and numerical method are displayed, analyzed
and validated with experimental results.
2. METHODS
2.1 Analytical formulation
2-D Axisymmetric geometry of induction melting process is shown in figure1 (a). Here, graphite
crucible is placed inside the multi turn copper induction coil. Graphite is used as a susceptor and
it heats up the charge by radiation and conduction. In induction melting process, Induction coil
and workpiece can be represented by using series equivalent circuit (SEC) as shown in figure
1(b).
Dc
dc
Sc
Dw
lc
lw
Refractory
Copper
Induction coil
Graphite
Crucible
Charge
z
r
1 (a)

23
R'w
Xg
Xc X'w
VL
R
IC
L
R
IC
L
R
IC
L Rc
XL
1(b)
Figure 1(a) 2-D geometry of induction coil and workpiece. 1(b). Series equivalent circuit of Coil
and workpiece (Graphite crucible)
In Series equivalent circuit, VL is the coil voltage (V), Ic is the coil current (A), Rc is the coil
resistance (Ω), Xc is the coil reactance (Ω), Rw is the work piece resistance (Ω), Xw is the work
piece reactance (Ω), XL is the lead reactance (Ω), RL is the lead resistance (Ω), R’
W is the work
piece resistance refereed to coil side (Ω), Xg is air gap reactance (Ω), X’W is work piece reactance
refereed to coil side (Ω).
Coil equivalent impedance (Zeq) and coil voltage (VL) are expressed by eq (1.1) and eq (1.2)
respectively,
= ( + + ′ ) + ( + + ′ + ) (1.1)
= (1.2)
Coil and workpiece parameters are calculated [9], [11] by using following formulas,
Coil resistance =
( )
(2.1)
Space factor ! = (2.2)
Coil skin depth " = #$# %
(2.3)
In these equations, Nc is the no of turns, lc is the coil height (m), ' is the coil diameter(m), ρc is
the copper resistivity (Ω-m) at 393 0
K, µ0 is the permeability of free space i.e. 4πX10 (-7)
, µr is the
relative permeability (µr=1 for copper, graphite),f is the power source frequency(Hz).
Coil reactance for (tc>2δc) Xc = Rc (3)
In this equation, tc is coil tube thickness (m)
Air gap reactance = (
∗ #$ %( * +)
(4.1)

24
,
∗
= , -1 − 0 +* +
( )
1 2 + 0 +* +
( )
1 (4.2)
, =
3
[3
$.67$ (8 9: )
;
]
(4.3)
In these equations, Dw is the workpiece diameter (m), δw is the workpiece skin depth (m), ,
∗
is the
modified Nagaoka correction factor, , is the short coil correction factor.
Workpiece resistance and reactance are calculated by using workpiece active and reactive power
respectively.
Workpiece active and reactive power is calculated by,
= = (>?@A@!)B - +
2 C D (5)
E = (>?@A@!)B - +
2C F (6)
Here, lH= Workpiece length (m)
p, q= Workpiece power loss constant, this parameters are find out using equations given in [2],
H= Magnetic field intensity (A/m)
Magnetic field intensity is calculated by,
C = (
∗ I
(7)
Lead resistance is calculated by standard ohms law considering the skin effect and proximity
effect. Lead reactance is calculated by using parallel transmission line formula.
2.2 Mathematical modeling
Induction heating is multiphysics process i.e combination of Electromagnetic field and heat
transfer. Electromagnetic field produced by induction coil is represented by maxwell equations
[4]-[8], [15], [16],[17] as shown below,
∇ × C
L
LM = NO
LM + PQ
LLLM (8.1)
∇ × O
LM = −
RS
LM
RT
(8.2)
∇. U
LM = 0 (8.3)
Constitutive equations,
U
LM = @A@!C
L
LM (8.4)

25
W
M = NO
LM (8.5)
In these equations, H is the Magnetic field strength (A/m), E is the Electric field strength (V/m),
σ is the Electrical conductivity (S/m), js
is the source current density (A/m2
), B is the Magnetic
flux density (T). From eq (8.3), magnetic vector potential is defined as,
U
LLLM= X × Y
M (9)
eq (8.4), eq (9) put into eq (8.1) gives,
X × X × Y
M = @A@!ZNO
LM + PQ
LLLM[ (10)
From eq (8.2), eq (9) and by applying vector identity, eq (10) can be written as,
∇ Y = @A@!(N
R
M
RT
− PQ
LLLM) (11.1)
Induction coil carries sinusoidal current, Hence eq(11.1) can be written as,
∇ Y = @A@!(N]Y
LLLLLLM − PQ
LLLM) (11.2)
For 2-D axisymmetric geometry shown in figure 2(a) and using cylindrical coordinates, eq (11.2)
can be written as,
In induction coil
R ^
R!
+
3
!
R^
R!
+
R ^
R_
−
^
!
= `]@A@!NYa − @A@!PQ
LLLM (12.1)
In Workpiece
R ^
R!
+
3
!
R^
R!
+
R ^
R_
−
^
!
= `]@A@!NYa (12.2)
In vacuum/air
R ^
R!
+
3
!
R^
R!
+
R ^
R_
−
^
!
= 0 (12.3)
In these equations, θ, z, r are the Cylindrical coordinates, Aθ is the θ Component of the magnetic
vector potential (V.sec/m), ω is the Angular frequency of coil current (rad/sec).
2.3 Numerical analysis
Finite Element Method (FEM) is used to solve the eq (12.1), (12.2) & (12.3) in geometry given in
figure 2(a). In this method, Solution domain is discretized into number of discrete sub domain or
elements. After discretization, integral formulation is obtained by using Galerkin technique,
which is known as weak form. After that, algebraic equations for each element are obtained in
matrix form. Assembling of all element equations gives the global matrices [18] as shown below
[b]cYd + [C]cYd = cWd (13)
In this equation, [K] is the stiffness matrix, [H] is the Mass matrix,

26
{A} is the Magnetic vector Potential, {J} is the Electrical current density.
Eq (13) is solved by using suitable boundary conditions and forcing function. Once, magnetic
vector potential is determined, then other quantities like, coil resistance, reactance, voltage, and
work piece power are computed.
2.4 Simulation
Simulation is done in three steps i.e. preprocessing, processing & post processing [1], [19], .In
preprocessing, 2-D axisymmetric geometry shown in figure 2(a) is used for simulation. Materials
and their properties given in table-I and table –II are assigned to different domains. Boundary
conditions are specified as per table -III. Current is used as forcing function. Triangular elements
are used for domain discretization. Mesh is refined on the surface of induction coil and workpiece
as shown in the figure 2(b) [19]. Induction coil carries sinusoidal current, therefore frequency
domain study is done for this analysis [16]. In processing step, magnetic vector potential is
computed in different domains as shown in figure 2(c). In post processing step, other process
related quantities like Magnetic flux density, Coil resistance, reactance, voltage and workpiece
power are computed.
Table-I. Workpiece dimension and properties
Table-II Induction coil dimensions and properties
Table-III Boundary condition and forcing function
Sr. No. Work Piece Description
1. Material Graphite
2. Outer Diameter 160 mm
3. Inner Diameter 117.5 mm
4. Height 180 mm
5. Electrical Conductivity at 2730
K 1.42653X105
S/m
6. Electric permittivity 1
7. Magnetic permeability 1
Sr.No. Induction coil Description
1. Material Copper
2. Inside diameter 215mm
3. Outside diameter 247mm
4. Height 200mm
5. Coil tube diameter 16mm
6. Coil tube thickness 2mm
7. No. of turn 9
8. Electrical conductivity at 2730
K 5.998X107
S/m
9. Electric permittivity 1
10. Magnetic permeability 1
Sr.No. Boundary condition Description
1. Outer boundary A=0
2. Asymmetry axis eY
ef
= 0

27
Table-IV Lead cable specification
2 (c) 2 (d)
Figure 2(a). Axisymmetric geometry, 2(b) Meshing, 2(c). Magnetic vector potential at 6.55
kHz, 2(d) Magnetic flux density at 6.55 kHz
Sr no Lead Cable Description
1 Material Copper
2 Length 3.61 m
3 Outer diameter 0.025m
4 Thickness 0.001 m
5 Insulation gap 0.015m

28
2.5 Experiment
Experiments were carried out on vacuum induction melting furance. Schematic diagram of
furnace is shown in the figure (3). 40 kW, 5 kHz, 400 V Power source was connected to the
induction coil. During experiment, various parameters such as Coil current, coil voltage were
measured and recorded at different frequency as shown in figure 4. High frequency current was
measured by rogowaski coil [20]. During experiment, Vacuum below 100 µbar and cooling water
pressure more than 2 kg/cm2
were maintained.
Water coolant
Vacuum
Copper
Induction
coil
Charge
Graphite
crucible
Induction
melting
power
source
3-Phase
mains
power
supply
Melting
vessel
Figure (3). Schematic diagram of Uranium melting
3. RESULT AND ANALYSIS
Experimental results are shown in the figure (4). From figure (4), it is observed that resonance
frequency of series RLC circuit is approximately (6 kHz). As power source output frequency
moves towards resonance frequency, induction coil voltage and current increase.
0 6.8 7.2 7.6 8.0 8.4
0
100
200
300
400
Voltage
Coil
voltage(V)
Frequency(kHz)
0
200
300
400
500
600
700
800
Current
Coil
Current(A)
Figure (4). Coil voltage and coil current during experiment

29
01 6.8 7.2 7.6 8.0 8.4
0.0
0.70
0.72
0.74
0.76
0.78
0.80
0.82 Copper
Graphite
Skin
Depth(mm)
Copper
Skin
Depth(mm)
Frequency(kHz)
0
14.5
15.0
15.5
16.0
16.5
W
orkpiece
Figure (5). Skin depth at different frequency
In figure (5), it is observed that skin depth in graphite is 20 times more than copper due to high
resistivity of graphite. Figure 6(a) and 6 (b), show the comparison of analytically and numerically
calculated equivalent coil resistance and reactance. The difference between analytical and
numerical resistance and reactance values are less than 7 % and 4 % respectively in specified
frequency range.
0 6.8 7.2 7.6 8.0 8.4
0.00
0.054
0.056
0.058
0.060
0.062
0.064
0.066
Equivalent
coil
resistance(Ohm)
Frequency(kHz)
Analytical
Numerical
6 (a)

30
0 6.8 7.2 7.6 8.0 8.4
0.00
0.05
0.34
0.36
0.38
0.40
0.42
0.44
0.46
Coil
equivalent
reactance(Ohm)
Frequency(kHz)
Analytical
Numerical
6 (b)
Figure 6(a) coil and workpiece equivalent resistance, 6(b) Coil and workpiece equivalent
reactance
From figure7 (a) and (b), it is observed that analytical and numerical calculated coil voltage and
workpiece power are in good agreement and difference is less than 4% for coil voltage and less
than 2 % for workpiece power in specified frequency range.
0 6.8 7.2 7.6 8.0 8.4
0
80
120
160
200
240
280
Coil
voltage(V)
Frequency(Khz)
Analytical
Numerical
7 (a)

31
0 6.8 7.2 7.6 8.0 8.4
0
5000
10000
15000
20000
25000
30000
Workpiece
power
(W)
Frequency(kHz)
Numerical
Analytical
7 (b)
Figure (7) (a) Coil voltage, (7) (b) workpiece power calculated by analytical and numerical
methods
Figure (8) shows the workpiece and coil power calculated by eq (7.1) and (7.2) at different
frequency. These data are used to calculate the coil efficiency at different frequencies and found
that coil efficiency is ~81%.
0 6.8 7.2 7.6 8.0 8.4
0
5000
10000
15000
20000
25000
30000
Coil
Power(W)
W
orkpiece
Workpiece
Power(W)
Frequency(kHz)
0
2000
4000
6000
Coilpower
Figure (8) Coil and work piece power at different frequency

32
0 6.8 7.2 7.6 8.0 8.4
0
100
150
200
250
300
350
400
Coil
Voltage(V)
Frequency(kHz)
Experiment
Numerial
Analytical
Figure (9) Analytically, numerically and experimentally obtained coil voltage
After validation of analytical and numerical model of induction heating process, experiment was
conducted to validate the coil voltage calculated by analytical and numerical method. Figure (9)
shows the comparison of analytical, numerical and experimental coil voltage. In this comparison,
lead impedance is included during analytical and numerical computation, because during
experiment, coil voltages are measured at power source output terminal. From figure (9), it is
observed that results are in good agreement.
4. CONCLUSION
Mathematical modeling and simulation of electromagnetic induction phenomenon in induction
heating process were carried out and results were validated by analytically obtained results.
Experiments were conducted to validate the coil voltage obtained by analytical and numerical
methods at different frequencies. Analytical and numerically calculated workpiece power at
different frequencies is very much close to each other and that is the main concern of induction
heating system designer. This model can be applied to optimize the coil parameters to get the
desired results in induction heating of different kind of materials and avoid the tedious
experiments. This model helps to understand the dynamics of induction heating process in depth.
As these analyses are restricted only to electromagnetic part of induction heating, this can be
further coupled to the power source for transient analysis and heat transfer physics to find out
temperature profile in workpiece.
5. REFERENCES
[1] Valery Rudnev, Don loveless, Raymond Cook, Micah Black, “Handbook of Induction heating”,
INDUCTOHEAT,Inc., Madison Heights,Michigan,U.S.A.
[2] E.J Davies and P.G. Simpson, Induction Heating Handbook. McGraw Hill, 1979.
[3] Phuong-Xuan Dang, “Improving the energy efficiency by process parameter optimization approach:
a case study of induction heating. International journal of renewable energy and environmental
engineering”, ISSN2348-0157, Vol.01, No.01, October 2013.

33
[4] C Chabodez, S Clain, R.Glardon, D Mari, J.Rappaz, M. Swierkosz, “Numerical modeling in
induction heating for axisymmetric geometries”, IEEE transactions on Magnetics.Vol33, No.1
January 1997, P 739-745.
[5] Jiin-yung Jang, Yu-We Chiu, “Numerical and experimental thermal analysis for a metallic hollow
cylinder subjected to step wise electromagnetic induction heating”, applied thermal engineering
27(2007), 1883-1894.
[6] K. sageghipour, J.A Dopkin,K.Li,” A computer aided finite element/experiment analysis of
induction heating process of steel, computer in industries,28(1996) 195-205.
[7] Q.-S.Chen, P. Gao, W.R.Hu, “Effects of induction heating on temperature distribution and growth
rate in large size Sic growth system, journal of crystal growth 266(2004) 320-326.
[8] Mohammad Hosin, Tavakoli, Hosin Karbasachi, Feridoun Samavat, “Influence of workpiece height
on induction heating process”, Mathematical and computer modeling, Elsevier journals.2011.
[9] Mark William Kennedy, Shahid Akhtar, Jon Arne Bakken,, Ragnhild E.Aune, “Review of Classical
Design Methods as applied to aluminum billet heating with induction coils”, EPD congress, San
Diego, California, February 27-March 3(2011),702-722.
[10] Lisiate Takau, Pat Bodger, “Low frequency modeling of induction heater using series equivalent
circuit, transformer equivalent circuit, and finite element analysis”, Australasian universities power
engineering conference, AUPEC 2013.Hobart,TAS, Australia,29 September-3 October 02013, p-1-6.
[11] Mark William Kennedy, Shahid Akhtar, Jon Arne Bakken,, Ragnhild E.Aune, “ Analytical and FEM
modeling of aluminum billet induction heating with experimental verification”, TMS( The mineral,
Metals & Materials Society),2012
[12] J Acero, J.M. Burdio, L.A. Barragan, J.I.Artigas, R. Aloso, “An electromagnetic based model for
calculating the efficiency in domestic induction heating appliances,” Power electronic specialist
conference 2006, PESC06, 37th
IEEE, P, 1-6.
[13] Milton Abramowitz, Irene A Stegun, “Handbook of mathematical functions, national bureau of
standard,” Applied Mathematics Series.55, December 1972.
[14] C.L. Wadhwa, “ Electrical power systems,” New age international publishers, New Delhi,2014
[15] Andrzej Krawczyk, John A. Tegopoulos, “Numerical modeling of eddy current,” Oxford science
publications, P-17.
[16] K.Djellabi, Meh Latreche, “Induction heating process design using Comsol multiphysics software
version 4.2a,” International journal of electrical, computer, electronics and communication
engineering, vol8 no 1 2014
[17] Sun Yefei, Sun Jing, Niu Dongjie, “Numerical simulation of induction heating of steel bar with
nonlinear material properties,” Proceeding of IEEE conference on automation and logistic Shenyang,
China August 2009.
[18] Huang, Hou-cheng, “Finite element analysis of heat transfer: theory and software,” Springer verlo,
London, 1994.
[19] Ion Carstea, Daniela Carstea, Alexandru Adrian Carstea, “A domain decomposition approach for
coupled field in induction heating device,” 6th
WEEAS international conference on system science
and simulation in engineering, Venice, Italy, November 21-23,2007,P63-70
[20] D.A.Ward And J.La.T.Exon, “Using Rogowski coils for transient current measurements”,
Engineering Science and Education Journal, June 1993.

ANALYTICAL, NUMERICAL AND EXPERIMENTAL VALIDATION OF COIL VOLTAGE IN INDUCTION MELTING PROCESS

More Related Content

Similar to ANALYTICAL, NUMERICAL AND EXPERIMENTAL VALIDATION OF COIL VOLTAGE IN INDUCTION MELTING PROCESS (20)

Recently uploaded (20)

ANALYTICAL, NUMERICAL AND EXPERIMENTAL VALIDATION OF COIL VOLTAGE IN INDUCTION MELTING PROCESS