International Conference on Control, Engineering & Information Technology (CEIT14)

Proceedings - Copyright IPCO-2014_ISSN 2356-5608

Development of analytical approach for linear

switched reluctance motor and its validation by
two dimensional FEA
MAHMOUD Imed and REHAOULIA Habib
High School of Sciences and Technology of Tunis (ESSTT),
5 av. Taha Hussein BP 56 1008 Tunis, Tunisia University
Laboratory of Signal, Image and Energy Control (SIME)
E-mail: [email protected]
E-mail: [email protected]
Abstract This paper gives an overview of modelling technique to describe the
nonlinear behaviour of saturated Linear Switched Reluctance Machine (LSRM).
The approach is based on the inductance versus position and current. Taking into
account the non-linearity of the magnetic circuit, models are expressed by either
Fourier series or polynomials where the only first three components are considered. The results of these analytical approaches are compared with those obtained
using finite element methods (FEM).
Keywords linear switched reluctance motor, analytical model, flux linkage,
magnetic saturation.
1.

Introduction

The Linear Switched Reluctance Motor (LSRM) is always operated in the magnetically saturated mode to maximize the energy transfer. The magnetic flux
linked by a single phase must be known to develop a sophisticated controller. The
inherent magnetic nonlinearity of the LSRM must be taken into account by appropriate modelling of the machine characteristics, [1-6-8].
In a LSRM, the phase inductances and flux linkages vary with rotor position due
to stator and rotor saliencies. The phase inductances and flux linkages at any rotor
position also vary with the instantaneous phase currents because of magnetic saturation. However, these variations can be modelled analytically using the data obtained through FEM or through experiments. These analytical expressions are

137

138

used to represent the LSRM dynamics and hence the machine performance can be
obtained, [11-23-24-26-27-28-29].
Obviously, the high degree of nonlinearity makes very difficult the modelling of
the flux linkage or the phase inductance. Many researchers have addressed the
problem of calculating the inductance or flux linkage from rotor position and
phase current analytically with various degrees of accuracy, [10-12-13-14-16].
In order to determine a refined model which describes the behavior of a saturated
reluctant structure, there are basically two ways to represent the static LSRM
characteristics. The first way is to plot the phase flux linkage variations with rotor
position and phase current. The second way is to plot the phase inductance variation with rotor position at different phase currents. These static characteristics are
highly nonlinear. Figure 1 shows a classification of the different LSRM modelling techniques, [2-3-7-9-19-20-35-39-40].

Fig.1. Classification of LSRM modelling.

Specifications of the designed prototype of the LSRM are shown in table 1.

139

Fig. 2. Main dimensions of the conceived actuator.

Table 1. Motor mechanical and electrical parameters
Number of modules

Tooth width (b)

3mm

Slot width (a)

3mm

Tooth pitch ( )

6mm

Phase separation (c)

1.5mm

Mover length

135 mm

stator length (L)

40.5 mm

Air gap width ( )

0.1mm

Step size

1.5 mm

Number of turns per

phase

2.

520

Inductance -based model of LSRM

In a LSR machine, the reluctance of the magnetic path in a given phase changes
with rotor movement. The reluctance is maximum when the stator and rotor poles
are unaligned and minimum when the poles are aligned. This variation in reluc-

140

tance reflects in the self-inductance of the respective stator phase. When the stator
and rotor poles are aligned, the self-inductance of the stator phase will be maximum and when the poles are unaligned, the self-inductance of the phase will be
minimum. The phase inductance in a LSRM is a periodic function of the rotor position. At any given rotor position, the phase inductance also varies with the instantaneous phase current because of magnetic saturation. Therefore, in the inductance-based model, the position dependency of the phase inductance is represented
by a limited number of Fourier series terms and the nonlinear variation of the inductance with current is expressed by means of polynomial functions, [23-25-3334]:
m

L( x, i ) = Lk (i ) cos kN r x

(1)

k =0

With

i , x et m are respectively the phase current, the position of the mover and

the number of terms in the Fourier series.

The accuracy and stability of numerical simulations are the main challenges which
should be met. To simplify the expression (1) only the first three terms of the Fourier series are considered. This leads to simplify the expression of the inductance
given by the equation (2), [4-20]:

L(x, ij ) = L0 (ij ) + L1(ij )cos(Nr (x ( j 1)

2
))
NNr

(2)

2
+ L2 (ij )cos(2Nr (x ( j 1)
))
NNr
With

L( x , i j ) and N are respectively the inductance associate to the phase j in

the position

x of the mover for the current i j and the number of phase.

To determine the three coefficients L0 , L1 et L2 , we use the inductance at three

positions: aligned position Lc ( i j ) , unaligned position Lop ( i j ) and midway position between the above two positions Li ( i j ) . Note that

Lop (i j ) can be treated as

141

a constant but,

Lc (i j ) and Li (i j ) are functions of the phase current i j and can be

approximated by the polynomials, [15-21-22-23-38]:

p

Lc (i j ) = an i n j

(3)

n =0
p

Li (i j ) = bn i n j

(4)

n =0

p is the order of the polynomials and an , bn are the coefficients. In our

research, p = 6 is chosen after we compare the fitting results of different p values, ( p = 3 , p = 4 , p = 5 and p = 6 have been tried and compared). Result the
Where

inductance of aligned position Lc ( i j ) and midway position

Li (i j ) are approxi-

mated respectively by the equation (5) and (6):

Figure 3 shows the evolution of the inductance versus current in the aligned and
the midway postions.
0.13

0.095

0.09

0.11

Inductance (H)

I n d u ctan ce (H )

0.12

0.1
0.09

0.085

0.08

0.075

0.08

FEM
Curve-fitting

FEM
Curve-fitting
0.2

0.4

0.6

Current (A)

0.8

0.2

0.4

0.6

Current (A)

a)

b)

Fig. 3. Evolution of the inductance vs current

in aligned (a) and midway positions (b)

0.8

142
6
5
4
3
2
Lc (i) = ai
1 + a2i + a3i + a4i + a5i + a6i + a7

a1 = -0.4883 a2 = 1.356
a3 = -1.153

a4 = 0.1993

(5)

a5 = 0.06603 a6 = -0.02222
a7 = 0.1253
4
3
2
Li (i) = bi1 6 + b2i5 + bi
3 + b4i + bi
5 + b6i + b7

b1= -0.3227 b2 = 1.186

b3 = -1.609 b4 = 0.9716

(6)

b5 = -0.2766 b6 = 0.03345
b7 = 0.09355
Consequently, the three coefficients for the Fourier series can be computed as:

L0 =

1 1

Lc + Lop ) + Li
(

2 2

(7)

L1 =

1
( Lc Lop )
2

(8)

L2 =

1 1

Lc + Lop ) Li
(

2 2

(9)

The stator phase inductance at the aligned position varies considerably with the
stator phase current because of the magnetic saturation. The unaligned inductance
does not vary much mainly because of the large reluctance that characterizes huge
air gap in the flux path. It can be observed that the inductance characteristics versus current with three positions obtained by the proposed model closely match
those obtained by finite element methods, figure 4.

143

0.13
0.12

Inductance (H)

0.11

Model
FEM

Aligned position

0.1
0.09
0.08

Unaligned position

Miday position

0.07
0.06
0.05
0

0.2

0.4

0.6

0.8

1.2

Current (A)

Fig. 4. Extreme left phase: Comparison of inductance versus current with three
positions__Model, *FEM.
0.12
0.11

Model
FEM

Inductance (H)

0.1
0.09
0.08
0.07
0.06
0.05
0

0.5

1.5

2.5

Displacement (mm)
Fig.5. Extreme left phase: Comparison of inductance versus position with different
currents__Model, *FEM.

144

A similar comparison is made for the characteristic of inductances versus positions for different values of currents, to show the effectiveness of the proposed
model. Figure 5 shows a good agreement between the model and the finite element method MEF.
Multiplying the expression of inductance by the current ( i ), it gives the expression of linkage flux:

( i, x ) = iL ( i, x )

(10)

Figure 6 gives the comparison of linkage flux produced by the left extreme phase
versus current for different positions. It can be observed that the linkage flux versus current for different position characteristics which are obtained by the proposed model closely match those obtained by finite element methods. These results prove the effectiveness of the proposed model.
0.09
0.08

Flux linkage (Wb)

0.07
0.06
0.05
0.04
0.03

Model
* FEM

0.02
0.01
0
0

0.2

0.4

0.6

0.8

1.2

Current (A)

Fig. 6. Extreme left phase: Comparison of linkage flux versus current with
different positions __Model, *FEM.

The total electromagnetic force is given by the following expression, [30-31-3236-37]:

145

F = F j ( i, x )

(11)

j =1

Where

N is the number of phase, Fj the force of phase j and i j the phase cur-

rent. Consequently, the force

Fj can be described by the following equation:

L ( x, i j ) i j di j
Wc , j

F j ( i, x ) =
= 0
x
x

(12)

L( x , i j ) is the inductance associate to the phase j in the position x of mover for

the current

ij .

If the current is applied to a given phase, with stator and rotor teeth unaligned, the
rotor will be attracted toward the balance position where the flux is maximum
(aligned position) developing a force generally expressed by:

Fj (i, x) =

1 L j ( x) 2
ij
2 x
[i ]=cte

(13)

12

Thrust force (N)

10

4
Model
FEM

0
0

0.5

1.5

Displacement (m)

2.5

3
-3

x 10

Fig. 7. Extreme left phase: Comparison of the thrust force as function of mover position
__Model, *FEM.

146

Figure 7 shows a perfect coincidence between the force characteristic determined

by the proposed model and that taken via the finite element method (MEF). This
represents a proof of the effectiveness of the proposed model.
-3

x 10

4
3
2

Linear model
Proposed model

1
0
0

0.5

1.5

2.5

3.5

Time (s)

a)
0.06
0.05
0.04
0.03
Speed (rad/s)

Position (m)

0.02
0.01
0
-0.01
-0.02
Linear model
Proposed model

-0.03
-0.04
0

0.5

1.5

2
Time (s)

b)

2.5

3.5

147

15

Thrust force (N)

10
5
0
-5
-10

Linear model
Proposed model

-15
0

0.5

1.5

2
Time (s)

2.5

3.5

c)
Fig. 8. Evolution of the position, speed and thrust force on four steps.
Figure 8, respectively illustrates the dynamic behavior of the machine for the
model without saturation and with saturation (refined model) starting from the
speed and of the angular position. The superposition of these results shows that
the dynamics of the evolution of the position is respected. A static error due to
friction affects the positions of balance: [5-41].
The refined model of the LSRM is characterized by a strongly oscillatory translation compared to the linear model. These oscillations are disturb the precision of
the position and the constancy speed often required by many industrial applications and especially in the medical fields. This problem often leads to losses of
synchronism, [11-17-18-26-28]
3. Conclusion
It is essential to have an accurate model of a Linear Switched Reluctance Motor
that describes its static characteristics. It has been shown in this paper that there
are different ways of modelling static characteristics of an LSRM. Developed
analytical models consider the variation of the phase inductance with rotor posi-

148

tion accounting for magnetic saturation. Results are compared to those obtained
via the 2D-FEM. The comparison shows a reasonable agreement, proving the validity of the proposed approaches.
4. References
[1] A .O. Khalil, Modeling and analysis of four quadrant sensorless control of a
switched reluctance machine over the entire speed range, The Graduate Faculty of
the University of Akron, PHD thesis, August, 2005.
[2] A.C. Pop, Finite Element Based Multiphysics Optimal Design of Switched Reluctance Motors Used in Electric Vehicles Propulsion, Journal of Computer Science and Control Systems, Volume 5, Number 1, May 2012.
[3] A. E. Y. Umenei, Models for extrapolation of magnetization data on magnetic
cores to high field, accepted for publication in IEEE Transactions on Magnetics,
2011.
[4] B. Fahimi, A New Approach to Model Switched Reluctance Motor Drive Application to Dynamic Performance Prediction, Control and Design, conference
IEEE 1 998, 0-7803-4489-8/98.
[5] C. Jae-Hak, Approximate optimization for minimum torque ripple of three
phase switched reluctance motor using response surface modeling, International
Journal of Applied Electromagnetics and Mechanics, 10.3233/JAE-2012-1548.
[6] D.A. Andrade, Characterization of Switched Reluctance Machines Using Fourier series Approach
, -7803-7116-X/01/$10.00 (C) 2001 IEEE.
[7] D. Susitra, Non-linear flux linkage modeling of switched reluctance machine
using MVNLR and ANFIS, International Journal of Applied Electromagnetics and
Mechanics, 10.3233/IFS-130766.
[8] E. Pdurariu, Switched Reluctance Motor Analytical Models, Comparative
Analysis, the International Conference on Optimization of Electrical and Electronic Equipment, OPTIM 2010, 978-1-4244-7020-4/10/$26.00 '2010 IEEE.
[9] E. Eyhab, Improving the energy conversion process in the switched reluctance
motors, International Journal of Applied Electromagnetics and Mechanics,
10.3233/JAE-121619.
[10] F.R. Salmasi and al, Virtual Auto-Tuning Position and Torque Sensors for
Switched Reluctance Motor Drives, IEEE Transactions on Industry Applications,
pp. 1355-1361, 2004.

149

[11] G. Singh, and al, Modelling and dynamic simulation of multiple-stack variable-reluctance step motors
, Comput and Elect Engng. Vol .1.pp.481-500 1974.
[12] H. Jin, Modelling of switched reluctance motor using Fourier series for performance analysis, Journal of Applied Physics, volume 93, number 10, 15 May
2003.
[13] H.P. Chi and al, Simplified flux-linkage model for switched-reluctance motors, IEE Proc. -Electr. Power Appl., Vol. 152, No. 3, pp. 577-583, May 2005.
[14] H.P. Chi, Flux-linkage based models for switched-reluctance motors, PHD
Philosophy, December 2005, National Cheng Kung University.
[15] H. Gao, Inductance Model-Based Sensorless Control of the Switched Reluctance Motor Drive at Low Speed, IEEE transactions on power electronics, vol. 19,
no. 6, November 2004.
[16] H. Torkaman, and E. Afjei, Comprehensive detection of eccentricity fault in
switched reluctance machines using high frequency pulse injection," IEEE Transactions on Power Electronics, Vol. 28, No. 3, 1382{1390, 2013.
[17] H.Torkaman, E. Afjei, and P. Yadegari, Static, dynamic, and mixed eccentricity faults diagnosis in switched reluctance motors using transient fnite element
method and experiments," IEEE Transactions on Magnetics, Vol. 48, No. 8,
2254{2264, 2012.
[18] I. Bentia, On the control of a rotary-linear switched reluctance motor, in
Proc. of the 5th International Symposium on Computational Intelligence and Intelligent Informatics (ISCIII '2011), Floriana (Malta), 2011, pp. 41-46.
[19] J. Nicholas, Modeling of a Saturated Switched Reluctance Motor Using an
Operating Point Analysis and the Unsaturated Torque Equation, IEEE transactions on industry applications, vol. 36, no. 3, May/June 2000.
[20] J. Hur, Modeling of switched reluctance motor using Fourier series for performance analysis, Journal of Applied Physics, 2003.
[21] J. MARIDOR, Design, Optimization, and Sensorless Control of a Linear Actuator, PHD thesis, cole polytechnique fdrale de lausanne, 2011
[22] J. Lee, Strurctural design optimization of electrical motors to improve torque
performance, PHD thesis, University of Michigan 2010.
[23] J.Y. Lee, Calculation of inductances in permanent magnet type transverse
flux linear motor, International Journal of Applied Electromagnetics and Mechanics, Volume 20, Numbers 3-4/2004.

150

[24] K. I. Hwu , Applying Powersys and Simulink to Modeling Switched Reluctance Motor, Tamkang Journal of Science and Engineering, Vol. 12, No. 4, pp.
429_438 (2009).
[25] K. Ha, Position estimation in switched reluctance motor drives using the first
switching harmonics of phase voltage and current, PHD thesis, Virginia Polytechnic Institute and State University, 2008.
[26] L. Lawler, A simulation model for the four phase switched-reluctance motor,
A Thesis, The University of Tennessee, Knoxville, May 2003.
[27] L. Szab, Dual motion switched reluctance motor for advanced industrial
applications, Proc. of OPTIM '2012 (in press).
[28] M. Feyrouz, Contribution lEtude dune Gnratrice Rluctance Variable, Universit Mentouri de Constantine, magistre en Electrotechnique, 2009
[29] M. Ruba, Novel Modular Fault Tolerant Switched Reluctance Machine for
Reliable Factory Automation Systems, Proc. of the 2010 IEEE International Conference on Automation, Quality and Testing, Robotics (AQTR '2010) THETA 17,
Cluj (Romania), 2010, Tome III, pp. 47-52.
[30] P. Dan-Cristian, Analytical Analysis of the Tubular Transverse Flux Reluctance Motor, Journal of Computer Science and Control Systems, Volume 5, Number 1, May 2012.
[31] R. M. Schupbach and al, Modeling Switched Reluctance Motors Under MultiPhase Excitation, University of Arkansas Department of Electrical Engineering
3217 Bell Engineering Center Fayetteville, AR 72701 1201 (501)575-3005,
Second Annual Report - Part 1 (AR2.1).
[32] S. Chevailler, Comparative study and selection criteria of linear motors,
PHD no 3569, Ecole Polytechnique Fdrale de Lausanne la facult Sciences et
Techniques de lIngnieur, Laboratoire d'actionneur intgrs.
[33] S. Misawa and al, A Rotor Position Estimation Using Fourier Series of Phase
Inductance for Switched Reluctance Motor, SPEEDAM 2010 International Symposium on Power Electronics, Electrical Drives, Automation and Motion.
[34] S. Song and al, A Comparative Study on Modeling Methods for Switched Reluctance Machines, Computer and Information Science Vol. 3, No. 2; May 2010
[35] S.W. Zhao, A self-Tuning Regulator for High-Precision Position Control of
linear Switched Reluctance Motor, IEEE Transactions On Industrial Electronics,
Vol 54, No.5, Otober 2007.

151

[36] T. Govindaraj, Finite Element Analysis of the Direct Drive PMLOM, 2012
Thangavel, licensee InTech, http://dx.doi.org/10.5772/46169
[37] V. IOAN-ADRIAN, Analytical flux linkage model of switched reluctance motor, Rev. Roum. Sci. Techn. lectrotechn. et nerg., 54, 2, p. 139146, Bucarest,
2009
[38] W. Lu, Modelling and control of a switched reluctance machine for electromechanical brake systems, PHD thesis, Ohio State University 2005.
[39] W. Shuyun ,Modeling and simulation of a piezodisc generator under central
load, International Journal of Applied Electromagnetics and Mechanics,
10.3233/JAE-121617.
[40] Z. Haijuin, Static characteristic and vibration dynamic reponse analysis of
switched reluctance motor system, International conference on mechatronics and
automation, proceeding of IEEE, August 2009.
[41] Z. Jingjun, Analysis of skewed angle effects of rotor
in switched reluctance motor on electromagnetic parameters using FEM, International Journal of Applied Electromagnetics and Mechanics, 10.3233/JAE-20121581.