6956 IEEE TRANSACTIONS ON INDUSTRIAL ELECTRONICS, VOL. 61, NO.
12, DECEMBER 2014
Armature Reaction Field and Inductance of Coreless
Moving-Coil Tubular Linear Machine
Liang Yan, Senior Member, IEEE, Lei Zhang, Zongxia Jiao, Hongjie Hu,
Chin-Yin Chen, and I-Ming Chen, Fellow, IEEE
Abstract—Analysis of armature reaction field and inductance
is extremely important for design and control implementation
of electromagnetic machines. So far, most studies have focused
on magnetic field generated by permanent-magnet (PM) poles,
whereas less work has been done on armature reaction field. This
paper proposes a novel analytical modeling method to predict the
armature reaction field of a coreless PM tubular linear machine
with dual Halbach array. Unlike conventional modeling approach,
the proposed method formulates the armature reaction field for
electromagnetic machines with finite length, so that the analytical
modeling precision can be improved. In addition, winding induc-
tance is also analytically formulated to facilitate dynamic motion
control based on the reaction field solutions. Numerical result is
subsequently obtained with finite-element method and employed
to validate the derived analytical models. A research prototype
with dual Halbach array and single phase input is developed.
Experiments are conducted on the reaction field and inductance
to further verify the obtained mathematical models.
Index Terms—Armature reaction, inductance, linear motor,
magnetic field, permanent magnet (PM).
N OMENCLATURE
φ Magnetic flux, Wb.
F Magnetomotive force (MMF), A.
l Half of the length of motor, m.
L, M Self-inductance and mutual inductance, H.
R Magnetic reluctance, Ω.
W Magnetic energy, J.
A Magnetic vector potential, Wb/m.
B Flux density vector, T.
Manuscript received October 23, 2013; revised January 24, 2014; accepted
February 17, 2014. Date of publication May 1, 2014; date of current version
September 12, 2014. This work was supported in part by the National Natural
Science Foundation of China (NSFC) under Grant 51175012, in part by the
National Key Basic Research Program of China (973 Program) under Grant
2014CB046406, in part by the NSFC under Grant 51235002, in part by the
Program for New Century Excellent Talents in University of China under Grant
NCET-12-0032, and in part by the Fundamental Research Funds for the Central
Universities.
L. Yan is with the School of Automation Science and Electrical Engineer-
ing, Beihang University, Beijing 100191, China, and also with the Research
Institute of Beihang University, Shenzhen 518000, China (e-mail: Lyan1991@
gmail.com).
L. Zhang, Z. Jiao, and H. Hu are with the School of Automation Science
and Electrical Engineering, Beihang University, Beijing 100191, China (e-mail:
zhanglei6988@126.com; zxjiao@buaa.edu.cn; hj@buaa.edu.cn).
C.-Y. Chen is with the Institute of Advanced Manufacturing Technology,
Ningbo Institute of Material Technology and Engineering, Ningbo 315201
China (e-mail: j0107@ms3.hinet.net).
I.-M. Chen is with the School of Mechanical and Aerospace Engineer-
ing, Nanyang Technological University, Singapore 639798 (e-mail: michen@
ntu.edu.sg).
Color versions of one or more of the figures in this paper are available online
at http://ieeexplore.ieee.org.
Digital Object Identifier 10.1109/TIE.2014.2321336
0278-0046 © 2014 IEEE. Personal use is permitted, but republication/redistribution requires IEEE permission.
See http://www.ieee.org/publications_standards/publications/rights/index.html for more information.
H Magnetic field intensity vector, A/m.
J Current density vector, A/m2 .
τp , τ r Pole pitch and axial width of radial magnets, m.
Br , Bz Radial and axial components of flux density, T.
I Current per turn, A.
k, Nk , p Number of positive coils, coil turns, and pole pairs.
ra , rb Outer and inner radii of internal permanent magnet
(PM), m.
ri , ro Inner and outer radii of internal PM, m.
I. I NTRODUCTION
D UE TO high power density, high efficiency, and compact
structure, PM linear machines (PMLMs) have wide appli-
cations in transportation [1], [2], military [3], [4], and manufac-
turing [5], [6]. Systematic investigation on air-core PMLMs has
been conducted by Boldea [7], Kremers et al. [8], Gysen et al.
[9], and Wang et al. [10], [11]. Impressive work has been done
on magnetic levitation systems and multi-degree-of-freedom
actuators. Various approaches have been proposed to analyze
magnetic field distribution of PM arrays [7], [12], [13]. An air-
core tubular PMLM with new topology of magnet array, i.e.,
dual Halbach array, has been proposed to increase the force
density and reduce the system vibration and noises [14], [15].
The magnetic flux density distribution of this electromagnetic
machine caused by PM poles alone has been studied in detail
[15]. However, armature is another important component that
influences the system output significantly. When the machine is
energized, the current in windings shifts and distorts the main
flux, which is known as armature reaction [16]. The magnetic
field on load is therefore a synthesis of PM field and armature
reaction field. Although armature reaction field is relatively
low, it may cause instability and tracking error, particularly
for high-precision systems [17]–[19]. Therefore, it is important
to analyze the armature reaction field precisely and, thus, to
improve the control performance of PMLMs.
Analytical modeling based on harmonic expansion is gen-
erally an efficient method to study armature reaction [20]. It is
more precise in the description of field distribution than conven-
tional permeance models [21]–[23] and current sheet models
[24]–[26]. It is also time saving and effective in design op-
timization of electromagnetic machines compared with finite-
element method (FEM) [27]–[29]. Researchers have obtained
analytical models of armature reaction fields on the premise of
infinite length of electric motors. For example, Wang et al. pre-
sented accurate analytical expression of armature reaction field
for a tubular PMLM with infinite length in [30]. Amara et al.
derived the armature reaction field of a tubular PMLM by
YAN et al.: ARMATURE REACTION FIELD AND INDUCTANCE OF MOVING-COIL TUBULAR LINEAR MACHINE
Fig. 1. Structure of a tubular PMLM with dual Halbach array.
assuming infinite length of an electric linear machine [31], [32].
The assumption of infinite length of linear motor apparently
simplifies the formulation procedure. However, it may also
reduce the modeling precision and unavoidably influences sub-
sequent system design and motion control implementation of
electric machines.
Therefore, this paper proposes an analytical modeling
method to describe magnetic field distribution of armature
reaction in moving-coil PMTLMs by taking finite motor length
into consideration. This method absorbs the merits of analytical
model based on harmonic expansion and permeance model.
The established model of armature reaction field is validated by
FEM solutions and utilized to formulate winding inductance.
It can benefit dynamic analysis and control of linear machines
[31], [33]. A research prototype with dual Halbach array and
single phase current is developed, and experiments are con-
ducted on both armature reaction field and inductance to verify
the developed analytical models. Along with the open-circuit
analytical model, the proposed armature reaction field model
can be employed for prediction of on-load field distribution of
linear machines [34].
II. M ACHINE S TRUCTURE
The structure of a tubular PMLM with dual Halbach array is
shown in Fig. 1. The major components include dual Halbach
array, coreless winding, and inner and outer back irons. The
radial magnets in different layers are magnetized in the same
direction and the axial ones in the opposite direction. This mag-
net arrangement helps to increase the radial component of flux
density in the air gap, whereas it reduces the axial component
significantly. It indicates that the dual Halbach array may offer
us two advantages, i.e., the axial force can be improved from
the increased radial flux, and the radial force disturbance and
vibration can be weakened from the decreased axial flux. A
coreless winding is selected for the design to have a near-linear
relationship between the current input and the force output,
which may benefit real-time motion control of the system.
III. F ORMULATION OF A RMATURE R EACTION F IELD
A. Assumptions
• The magnetic permeability of back irons is infinitely large
with respect to free space.
6957
Fig. 2. Five-region division of the machine space.
Fig. 3. Cylindrical coordinate systems CS1 and CS2.
• Laminated irons are used, and thus, eddy currents in stator
and mover are ignorable.
• The relative permeability of rare-earth PM is equal to that
of free space.
• All materials are isotropic in terms of magnetic property.
B. Governing Equations
Based on Maxwell equation and under Coulomb gauge, the
governing equation is derived as
∇2 A = −μ0 μr J. (1)
The machine space under study is separated into two types.
Type 1 is the central part without hatched lines, as shown
in Fig. 2, with the mover included; and type 2 is the part
filled with horizontal hatched lines. Because only magnetic flux
in the winding region affects the output performance of the
linear machine, this section mainly focuses on the study of the
central part. The hatched part will be employed for the study of
inductance computation in subsequent section. The central part
can be divided into two groups according to the current values.
Specifically, Group 1 includes two layers of PMs and air gaps
(Regions 1 and 3), and Group 2 is the current region (Region 2)
with a single phase current input. The two groups are governed
by Laplace’s and Poisson’s equations as follows:
∇2 Ai = 0, i = {1, 3} ∇2 A2 = −μ0 J. (2)
Due to the axial symmetric structure of the PMTLM, the
magnetic vector potential A only has θ component. Therefore,
in the cylindrical coordinate system CS1 shown in Fig. 3, A =
(0, 0, Aθ ). Since the machine is symmetric about the z-axis,
6958 IEEE TRANSACTIONS ON INDUSTRIAL ELECTRONICS, VOL. 61, NO. 12, DECEMBER 2014

(3). By taking variable separation, we assume that Aiθ (r, z) =

R(r)Z(z). Laplace’s equation is simplified to
1 ∂Z 2 (z)
= k2
z(z) ∂z 2
∂R2 (r) ∂R(r)
r2 +r + (k 2 r2 − 1)R(r) = 0
∂r2 ∂r
where k 2 is an eigenvalue. The general solution is
 
1
Aiθ = pi r + qi (si + ti z)
r
∞
+ {[ain J1 (kn r) + bin Y1 (kn r)]
n=1
× [cin ekn z + din e−kn z ]
+ [ein I1 (mn r) + fin K1 (mn r)]
× [gin cos(mn z) + hin sin(mn z)]} (6)

where J1 (kr) and Y1 (kr) are Bessel functions of the first and
second kind, respectively; and I1 (mr) and K1 (mr) are modi-
Fig. 4. Winding configuration and current density distribution. fied Bessel functions of the first and second kind, respectively.
As the axial flux density is periodic, the exponential terms are
Aθ is only a function of r and z. The governing equation in
the cylindrical coordinate system CS1 is cin = 0 din = 0.
 
∂ 2 Aiθ ∂ 1 ∂ By taking the curl of magnetic vector potential A, the radial
+ (rAiθ ) = 0 and axial flux densities are obtained as
∂z 2 ∂r r ∂r
   
∂ 2 A2θ ∂ 1 ∂ 1
+ (rA ) = −μ0 J. (3) Bir = pi r + qi ti
∂z 2 ∂r r ∂r
2θ r
∞
+ mn [ein I1 (mn r)+fin K1 (mn r)] sin(mn z  )
C. Current Source System n=1
Biz = 2(si + ti )pi
The coreless PMTLM adopts single phase winding for short ∞
stroke, ease of control, and low cost [35]. The same approach + mn [ein I0 (mn r)−fin K0 (mn r)] cos(mn z  ) . (7)
could be extended to electromagnetic machines with multiple n=1
phase windings. The current density distribution, accompany- 2) Magnetic Field in Current Region: Magnetic field dis-
ing with the winding configuration, is shown in Fig. 4. The tribution in the current region is the solution to the Poisson’s
harmonic expansion of the current density J is equation of (3). The solution to Poisson’s equation is the su-
∞     perposition of the homogeneous solution to Laplace’s equation
4J (2n−1)π (2n−1)π 
Js (z  ) = sin cos z and a particular solution to the Poisson’s equation. Substituting
n=1
(2n−1)π 2 τp (4) into the right side of Poisson’s equation and using variable
∞
   separation again yield
(2n − 1)π 
= Jn cos z (4)
τp Z(z) = cos(mn z  )
n=1
∂2R ∂R  
(mn r)2 + (mn r) − 1 + (mn r)2
where z  represents the axial position in the coordinate system ∂(mn r) 2 ∂(mn r)
CS2, as shown in Fig. 3. Its origin is at the geometry center of 2(mn r)2
the moving winding. The transformation between CS2 and the = Pn
π
static coordinate system CS1 fixed at the motor center is
where mn and Pn are calculated with
 
r =r z = z + zr (5) (2n − 1)π πμ0 Jn
mn = Pn = − .
τp 2m2n
where zr represents the mover position in CS1.
A particular solution to Poisson’s equation employing Struve
function [36] is given as
D. Analytical Solutions of Magnetic Field
∞

1) Magnetic Field of Noncurrent Region: Magnetic field of A∗2θ = L1 (mn r)Pn cos (mn z  ) . (8)
noncurrent region is the solution to the Laplace’s equation in n=1
YAN et al.: ARMATURE REACTION FIELD AND INDUCTANCE OF MOVING-COIL TUBULAR LINEAR MACHINE 6959

Therefore, the magnetic flux density in the current region is

 
1
B2r = − p2 r + q2 t2
r
∞
+ mn [e2n I1 (mn r) + f2n K1 (mn r)
n=1
+ L1 (mn r)Pn ] sin (mn z  )
B2z = 2(s2 + t2 )p2
∞
+ mn [e2n I0 (mn r) − f2n K0 (mn r)
n=1
+ L0 (mn r)Pn ] cos(mn z  ). (9)
3) Boundary Conditions: Boundary conditions are required
to determine the unique solution of the magnetic field. Sub-
jecting to Gauss’s law and Ampere’s law, magnetic field obeys
certain rules along the boundary of two different materials.
The component of magnetic flux density perpendicular to the
boundary is always continuous, i.e., B1n = B2n . The tangential
component of magnetic field is discontinuous by the amount
of surface current at the boundary, i.e., H1t − H2t = K. In
particular, in this paper, boundary conditions are
B1z |r=ri = 0 B3z |r=ro = 0 Fig. 5. Modeling process of EMC. (a) Number of coils. (b) Magnetic flow
distribution. (c) EMC.
B1r |r=ra = B2r |r=ra H1z |r=ra = H2z |r=ra
∞

B2r |r=rb = B3r |r=rb H2z |r=rb = H3z |r=rb . (10)
B2z = mn [e2n I0 (mn r) − f2n K0 (mn r)
Through solving the boundary condition equations, the follow- n=1
ing constraints are obtained: + L0 (mn r)Pn ] cos (mn z  ) . (13)
p1 = p 2 = p 3 = 0 To get the value of Q without complex computation, we utilize
q1 t1 = q2 t2 = q3 t3 = −Q equivalent magnetic circuit (EMC), instead of analytical mod-
UX = Y (11) eling based on harmonic expansion, to achieve a close solution.
Since the waveform of the flux density is already determined in
where U , X, and Y are as defined in the equations shown at the the analytical expression, Q is tolerable to estimate, and it will
bottom of the page. In coefficient matrix U not decrease much the precision of the analytical model.
r1 = mn ri r2 = mn ra r3 = mn rb r4 = mn ro . 4) Coefficients Determination: The coil numbers, magnetic
flow distribution, and EMC are shown in Fig. 5(a)–(c), respec-
Therefore, the analytical expressions of flux density in Regions tively. For a machine with single phase winding consisting of
i (i = 1, 3) and 2 can be simplified to (2k + 1) coils, the MMF generated by the nth coil is
∞
Q  Fn = (−1)n N I, n = −k, −(k − 1), . . . , k − 1, k. (14)
Bir = + mn [ein I1 (mn r) + fin K1 (mn r)] sin (mn z  )
r n=1
∞ Magnetic reluctance in the circuit is classified into three types, i.e.,

Biz = mn [ein I0 (mn r)−fin K0 (mn r)]cos (mn z  ) (12) g
Rn =
n=1 2πμ0 r̄τp
∞ g
Q  R+ =
B2r = + mn [e2n I1 (mn r) + f2n K1 (mn r) 2πμ0 r̄(l − zr − kτp )
r n=1 g
R− = (15)
+ L1 (mn r)Pn ] sin (mn z  ) 2πμ0 r̄(l + zr − kτp )

⎡ ⎤
I0 (r1 ) − K0 (r1 ) 0 0 0 0
⎢ I1 (r2 ) K1 (r2 ) −I1 (r2 ) −K1 (r2 ) 0 0 ⎥
⎢ ⎥
⎢ I0 (r2 ) −K0 (r2 ) −I0 (r2 ) K0 (ra ) 0 0 ⎥
U=⎢ ⎥
⎢ 0 0 I1 (r3 ) K1 (r3 ) −I1 (r3 ) −K1 (r3 ) ⎥
⎣ ⎦
0 0 I0 (r3 ) −K0 (r3 ) −I0 (r3 ) K0 (r3 )
0 0 0 0 I0 (r4 ) −K0 (r4 )
T
X = [ a1n b1n a2n b2n a3n b3n ]
Y = [ 0 Pn L1 (la ) Pn L0 (la ) −Pn L1 (lb ) −Pn L0 (lb ) 0 ]T
6960 IEEE TRANSACTIONS ON INDUSTRIAL ELECTRONICS, VOL. 61, NO. 12, DECEMBER 2014

where r̄ is the equivalent radii for gap reluctance, which is TABLE I

S TRUCTURE PARAMETERS IN FEM C OMPUTATION
defined as
ro − ri
r̄ = .
ln(ro /ri )
zt rb
The magnetic circuit equations are given as  2 2
   
+ B2r + B2z r dr dz
φ− R− − φ−(k−1) R−(k−1) = F−k z b ra ⎤
zt ro
φn Rn − φn+1 Rn+1 = Fn  
+ 2
B3r 2
+ B3z r dr dz  ⎦ (20)
φ k Rk − φ + R+ = Fk
k z b rb
φ− + φ+ + φn = 0. (16)
where the bound for z  is
n=−(k−1)

zb = −(kτp + τp /2) zt = kτp + τp /2. (21)

By solving the preceding equations, the main magnetic flux
between coil 1 and coil 0 (φ1 ) and that between coil 0 and coil For Regions 4 and 5 (type 2), the EMC method can provide a
−1 (φ0 ) is related with field model accurate enough for inductance computation. The
energy in type 2 is
φ1 l + (−1)k zr ⎡ l−z r ⎤
=− . (17) r o zb ro
φ0 l − (−1)k zr π ⎣
Wb = B+ 2  
r dr dz  + B−2 r dr dz  ⎦
Meanwhile, φ0 and φ1 can be deduced from the analytical μ0
z t ri −l−zr ri
expression of flux density, i.e., (9), as  2 
Ro − Ri 2 2
φ+ φ2−
= + . (22)
z  =3τ
p /4 z  =−τ
p /4 8πμ0 r̄2 l − zr − zt l + zr + zb
φ1 = 2πrB2r dz  φ0 = 2πrB2r dz  (18)
The magnetic energy resulting from the inductance matrix of
z  =τp /4 z  =−3τp /4 the five-coil winding is

where r is an arbitrary number from ra to rb . The range of z  in 1 T

W = [L] [Nk I][L] (23)
(18) is obtained from the magnetic flow distribution in Fig. 5(b). 2
By substituting (18) into (17), Q is obtained as where

∞ ⎡ ⎤
L11 M12 M13 M14 M15
2(−1)k r̄ yn
⎢ M21 L22 M23 M24 M25 ⎥
Q= n=1
zr ⎢ ⎥
lτp [L] = ⎢ M31 M32 L33 M34 M35 ⎥
⎣ ⎦
M41 M42 M43 L44 M45
where M51 M52 M53 M54 L55
[Nk I] = [ Nk I1 Nk I2 Nk I 3 Nk I 4 Nk I 5 ] T .
yn = e2n I1 (mn r̄) + f2n K1 (mn r̄) + L1 (mn r̄)Pn .
Assume that the five coils in the winding are completely equiv-
Therefore, all coefficients in analytical models are determined. alent in the inductive characteristics, i.e.,
Lii = Ljj , i, j = 1, 2, . . . , 5
IV. W INDING I NDUCTANCE
Mij = Mji = Lii , j = i (24)
Winding inductance influences system dynamics of electro-
magnetic machines significantly. Because of the employment the self-inductance of a single coil is given. When referring to
of single phase winding, only self-inductance exists. Magnetic the principle that the winding inductance is proportional to the
energy utilized to analyze the winding inductance is square of turns, the self-inductance of the single phase winding
with five coils is
B•H 1
W = dv = B2 dv. (19) Lw = 52 Lii . (25)
2 2μ0
V V

For Regions 1–3 (type 1), flux densities have been formulated, V. VALIDATION BY FEM
and the sum of energies in these areas is Although FEM is time consuming, it can achieve high-
precision results, taking nonlinear characteristics and flux leak-
Wa = W1 ⎡
+ W2 + W3 age into consideration. Therefore, FEM is utilized to validate

zt
ra
π ⎣  2 2
    the established analytical models, including magnetic field
= B1r + B1z r dr dz model and inductance model. The structure parameters of a
μ0
z b ri five-coil PMTLM in FEM computation is given in Table I.
YAN et al.: ARMATURE REACTION FIELD AND INDUCTANCE OF MOVING-COIL TUBULAR LINEAR MACHINE
Fig. 6. Armature reaction field variation versus z  validated by FEM. (a) Flux
density at r = (ri + ra )/2. (b) Flux density at r = (ra + rb )/2. (c) Flux
density at r = (rb + rs )/2.
A. Armature Reaction Field Validation
1) Flux Density Versus z  : Fig. 6(a)–(c) presents the flux
density variation, including radial and axial components, ver-
sus axial distance z  at r = (ri + ra )/2, r = (ra + rb )/2, and
6961
Fig. 7. Armature reaction field variation versus r by FEM.
r = (rb + rs )/2, respectively. Because the armature reaction
field is influenced by the mover position zr , the flux density
variation for different zr is also shown in the figure. The solid
curve is obtained at zr = 0. It indicates that the winding locates
at the center of the machine. The dotted curve is given at
zr = 18 mm, a pole pitch distance. It can be found that the
mover position zr only affects the distribution of the radial flux
component. It is mainly because magnetic reluctance change
caused by mover translation is hardly related to the axial
component, as indicated in Fig. 5. The analytical solutions fit
well with the FEM results.
2) Flux Density Versus r: Fig. 7 presents the magnetic
flux density variation versus radial distance at z  = τp /4 for
different mover position zr = 0 and zr = τp . It also shows that
the mover translation hardly affects the axial component of the
magnetic flux density. The radial component decreases for large
value versus radial distance r because of the increased cross
section for constant magnetic flux. The analytical solutions
are consistent with the FEM results, which indicates that the
analytical field model is of high precision comparable with
numerical computation.
B. Inductance Validation
The relationship of self-inductance Lw versus mover position
zr is shown in Fig. 8. It is shown that the difference between
the analytical solutions and numerical results is less than 3%,
which implies that the inductance model can be employed for
system dynamic analysis and control implementation. It also,
in turn, verifies the high precision of the flux field model.
The difference is mainly caused by the EMC method in the
computation of magnetic energy.
C. Comparison With Finite-Length Model
Magnetic flux density components of armature reaction ob-
tained by the developed mathematical model are compared
with conventional analytical model for infinite machine length.
The result is presented in Fig. 9. It shows that the developed
6962
Fig. 8. Inductance variation versus mover position zr by FEM.
Fig. 9. Comparison of magnetic field components of analytical models with
FEM results at r = (ri + ra )/2. (a) Radial component comparison. (b) Axial
component comparison.
magnetic field model fits with the numerical computation well,
which, in turn, indicates that the proposed method helps to
improve mathematical modeling of magnetic field distribution.
IEEE TRANSACTIONS ON INDUSTRIAL ELECTRONICS, VOL. 61, NO. 12, DECEMBER 2014
Fig. 10. Experimental study on armature reaction field. (a) Experimental
apparatus. (b) Research prototype.
TABLE II
S TRUCTURE PARAMETERS IN E XPERIMENTS
VI. VALIDATION BY E XPERIMENTS
A. Armature Reaction Field Measurement and Comparison
Experiments are conducted on magnetic field of an air-core
moving-coil PMTLM with dual Halbach array, as shown in
Fig. 10. The coil number of the single phase winding can be
either three or five, depending on the connection of the external
circuit. The main structure parameters of the linear machine are
listed in Table II. The current input per turn is 2A. The com-
parisons between the analytical results, FEM computations, and
experimental data are shown in Figs. 11 and 12 for different coil
numbers. The difference between the experimental results and
analytical solutions may mainly be caused by the measurement
error and the manufacturing error.
B. Experiments on Inductance and Comparison
The experiments on inductance are conducted by removing
resistance from the impedance as the capacitive reactance is
YAN et al.: ARMATURE REACTION FIELD AND INDUCTANCE OF MOVING-COIL TUBULAR LINEAR MACHINE
Fig. 11. Armature field versus z  for three-coil winding.
Fig. 12. Armature field versus z  for five-coil winding.
Fig. 13. Schematic of inductance measurement.
ignorable in the winding. Therefore, two steps are carried out to
measure the winding inductance. Specifically, Step 1 is to cal-
culate the resistance of winding with a multimeter, and Step 2
is to measure the impedance by inputting alternating current
with a frequency of 50 Hz. The schematic is shown in Fig. 13.
The experimental devices are shown in Fig. 14. Comparison
of analytical results, FEM computations, and experimental data
is presented in Fig. 15. It is found that the three sets of data
fit with each other quite well. The difference between the
analytical and measurement results is mainly caused by the
measurement error and the fabrication error of windings that
6963
Fig. 14. Inductance measurement on the research prototype.
Fig. 15. Inductance variation versus mover position zr validated by
experiments.
may influence the parameters in the models. However, because
the self-inductance of a coil is related with its neighboring
field, the self-inductances of the five coils are not exactly the
same. Meanwhile, the inductive coupling coefficient between
two coils is less than 1, and thus, the mutual inductance is
always less than the self-inductance. Therefore, (25), based on
the assumption, leads to a relatively large difference between
measurement data and analytical solutions.
Experiments are also conducted to test the linearity of the
electromagnetic motor and, thus, to further validate the estab-
lished mathematical models. Fig. 16 gives experimental results
of maximum force output versus current input. The linear
property can benefit the motion control implementation. The
thrust per volume and the thrust per copper losses of different
magnet topologies in given volume are investigated with FEM.
The result shows that the dual Halbach array helps to increase
force density of the system compared with single layer of
magnet array. More study on the force output of the linear motor
will be conducted at the next stage.
6964 IEEE TRANSACTIONS ON INDUSTRIAL ELECTRONICS, VOL. 61, NO. 12, DECEMBER 2014
Fig. 16. Experimental tests on the linear relationship of force output versus
current input.
VII. C ONCLUSION
The analysis of armature reaction field is significant for
high-precision motion control of electromagnetic machines. To
improve the modeling precision, this paper has proposed one
novel method to formulate the armature reaction field of a
coreless moving-coil PMTLM by taking finite motor length into
consideration. Winding inductance is subsequently analytically
derived. The modeling approach has the merits of analytical
technique based on harmonic expansion and EMC method.
Numerical computations are utilized to verify the analytical
models. The comparison shows that the mathematical mod-
els fit with numerical results closely. Furthermore, a research
prototype of the linear machine is developed. Experiments are
conducted to validate the mathematical models of armature
reaction field and inductance. It indicates that the established
models fit with the experimental results well. The study in this
paper can help design optimization and motion control of linear
machines. The proposed modeling method is effective and
could be implemented into analysis of other electromagnetic
machines.
R EFERENCES
[1] R. W. Cao, M. Cheng, C. C. Mi, and W. Hua, “Influence of leading design
parameters on the force performance of a complementary and modular lin-
ear flux-switching permanent-magnet motor,” IEEE Trans. Ind. Electron.,
vol. 61, no. 5, pp. 2165–2175, May 2014.
[2] K. J. Suzuki, Y. J. Kim, and H. Dohmeki, “Driving method of permanent-
magnet linear synchronous motor with the stationary discontinuous arma-
ture for long-distance transportation system,” IEEE Trans. Ind. Electron.,
vol. 59, no. 5, pp. 2227–2235, May 2012.
[3] M. Zhao, J.-M. Zou, Y.-X. Xu, J.-X. Zou, and Q. Wang, “The thrust
characteristic investigation of transverse flux tubular linear machine for
electromagnetic launcher,” IEEE Trans. Plasma Sci., vol. 39, no. 3,
pp. 925–930, Mar. 2011.
[4] L. Huang, H. T. Yu, M. Q. Hu, and H. X. Liu, “Study on a long pri-
mary flux-switching permanent magnet linear motor for electromagnetic
launch systems,” IEEE Trans. Plasma Sci., vol. 41, no. 5, pp. 1138–1144,
May 2013.
[5] A. Z. Shukor and Y. Fujimoto, “Direct-drive position control of a spiral
motor as a monoarticular actuator,” IEEE Trans. Ind. Electron., vol. 61,
no. 2, pp. 1063–1071, Feb. 2014.
[6] P. A. Commins, J. W. Moscrop, and C. D. Cook, “Novel tooth design for a
tubular linear motor for machine tool axis,” in Proc. IEEE ICM, Istanbul,
Turkey, Apr. 2011, pp. 660–665.
[7] I. Boldea, Linear Electric Machines, Drives, MAGLEVs Handbook.
Boca Raton, FL, USA: CRC Press, 2013.
[8] M. J. Kremers, J. H. Paulides, J. G. Janssen, and E. A. Lomonova, “Design
considerations for coreless linear actuators,” IEEE Trans. Magn., vol. 49,
no. 5, pp. 2271–2274, May 2013.
[9] B. J. Gysen, K. Meessen, J. H. Paulides, and E. Lomonova, “General for-
mulation of the electromagnetic field distribution in machines and devices
using Fourier analysis,” IEEE Trans. Magn., vol. 46, no. 1, pp. 39–52,
Jan. 2010.
[10] J. Wang and D. Howe, “Tubular modular permanent-magnet machines
equipped with quasi-Halbach magnetized magnets—Part I: Magnetic field
distribution, EMF, thrust force,” IEEE Trans. Magn., vol. 41, no. 9,
pp. 2470–2478, Sep. 2005.
[11] J. Wang, D. Howe, and G. W. Jewell, “Fringing in tubular permanent-
magnet machines: Part I. Magnetic field distribution, flux linkage,
thrust force,” IEEE Trans. Magn., vol. 39, no. 6, pp. 3507–3516,
Nov. 2003.
[12] J. de Boeij, E. Lomonova, and A. Vandenput, “Modeling ironless
permanent-magnet planar actuator structures,” IEEE Trans. Magn.,
vol. 42, no. 8, pp. 2009–2016, Aug. 2006.
[13] L. Yan et al., “Flux field formulation and back-iron analysis of tubular
linear machines,” IEEE Trans. Magn., vol. 48, no. 10, pp. 2617–2626,
May 2012.
[14] L. Yan, J. Hu, Z. X. Jiao, I.-M. Chen, and C. K. Lim, “Magnetic field
modeling of linear machines with double-layered Halbach arrays,” in
Proc. FPM, Beijing, China, Aug. 2011, pp. 1–6.
[15] L. Yan et al., “Magnetic field of tubular linear machines with dual Halbach
array,” Prog. Elect. Res., vol. 136, pp. 283–299, 2013.
[16] S.-M. Jang and S.-S. Jeong, “Armature reaction effect and inductance of
moving coil linear oscillatory actuator with unbalanced magnetic circuit,”
IEEE Trans. Magn., vol. 37, no. 4, pp. 2847–2850, Jul. 2001.
[17] G. G. Rigatos, “Adaptive fuzzy control of dc motors using state and
output feedback,” Elect. Power Syst. Res., vol. 79, no. 11, pp. 1579–1592,
Nov. 2009.
[18] J. X. Shen and K. J. Tseng, “Analyses and compensation of rotor position
detection error in sensorless PM brushless dc motor drives,” IEEE Trans.
Energy Convers., vol. 18, no. 1, pp. 87–93, Mar. 2003.
[19] J.-H. Horng, “Neural adaptive tracking control of a dc motor,” Inf. Sci.,
vol. 118, no. 1–4, pp. 1–13, Sep. 1999.
[20] R. P. Praveen et al., “A novel slotless Halbach-array permanent-magnet
brushless dc motor for spacecraft applications,” IEEE Trans. Ind.
Electron., vol. 59, no. 9, pp. 3553–3560, Sep. 2012.
[21] H. Vansompel, P. Sergeant, and L. Dupre, “A multilayer 2-dc2-d cou-
pled model for eddy current calculation in the rotor of an axial-flux PM
machine,” IEEE Trans. Energy Convers., vol. 27, no. 3, pp. 784–791,
Sep. 2012.
[22] Y. Fujimoto, T. Kominami, and H. Hamada, “Development and analy-
sis of a high thrust force direct-drive linear actuator,” IEEE Trans. Ind.
Electron., vol. 56, no. 5, pp. 1383–1392, May 2009.
[23] J. Asama, Y. Hamasaki, T. Oiwa, and A. Chiba, “Proposal and analysis
of a novel single-drive bearingless motor,” IEEE Trans. Ind. Electron.,
vol. 60, no. 1, pp. 129–138, Jan. 2013.
[24] J. L. He, E. Levi, Z. Zabar, L. Birenbaum, and Y. Naot, “Analysis of
induction-type coilgun performance based on cylindrical current sheet
model,” IEEE Trans. Magn., vol. 27, no. 1, pp. 579–584, Jan. 1991.
[25] Z. Q. Zhu and D. Howe, “Instantaneous magnetic field distribution in
permanent magnet brushless dc motors. IV. magnetic field on load,” IEEE
Trans. Magn., vol. 29, no. 1, pp. 152–158, Jan. 1993.
[26] T. F. Chan, L. L. Lai, and S. Xie, “Field computation for an axial
flux permanent-magnet synchronous generator,” IEEE Trans. Energy
Convers., vol. 24, no. 1, pp. 1–11, Mar. 2009.
[27] N. Bianchi, S. Bolognani, D. D. Corte, and F. Tonel, “Tubular linear
permanent magnet motors: An overall comparison,” IEEE Trans. Ind.
Appl., vol. 39, no. 2, pp. 466–475, Mar./Apr. 2003.
[28] Z. X. Lou, K. X. Yu, L. H. Wang, Z.-A. Ren, and C. Y. Ye, “Two-reaction
theory of homopolar inductor alternator,” IEEE Trans. Energy Convers.,
vol. 25, no. 3, pp. 677–679, Sep. 2010.
[29] H. Torkaman, E. Afjei, and M. S. Toulabi, “New double-layer-per-phase
isolated switched reluctance motor: Concept, numerical analysis, experi-
mental confirmation,” IEEE Trans. Ind. Electron., vol. 59, no. 2, pp. 830–
838, Feb. 2012.
[30] J. Wang, G. W. Jewell, and D. Howe, “A general framework for the
analysis and design of tubular linear permanent magnet machines,” IEEE
Trans. Magn., vol. 35, no. 3, pp. 1986–2000, May 1999.
YAN et al.: ARMATURE REACTION FIELD AND INDUCTANCE OF MOVING-COIL TUBULAR LINEAR MACHINE
[31] Y. Amara, G. Barakat, and P. Reghem, “Armature reaction magnetic field
of tubular linear surface-inset permanent-magnet machines,” IEEE Trans.
Magn., vol. 47, no. 4, pp. 805–811, Apr. 2011.
[32] Y. Amara and G. Barakat, “Analytical modeling of magnetic field in sur-
face mounted permanent-magnet tubular linear machines,” IEEE Trans.
Magn., vol. 46, no. 11, pp. 3870–3884, Nov. 2010.
[33] F. Cupertino, P. Giangrande, G. Pellegrino, and L. Salvatore, “End ef-
fects in linear tubular motors and compensated position sensorless control
based on pulsating voltage injection,” IEEE Trans. Ind. Electron., vol. 58,
no. 2, pp. 494–502, Feb. 2011.
[34] A. Bellara, H. Tiegna, Y. Amara, and G. Barakat, “On load analytical mod-
elling of the magnetic field for axial flux surface-inset permanent magnet
machines with semi-closed slots,” in Proc. ICEM, Marseille, France,
Sep. 2012, no. 1, pp. 2852–2858.
[35] Y. T. Chen, C. L. Chiu, Y. R. Jhang, Z. H. Tang, and R. H. Liang, “A
driver for the single-phase brushless dc fan motor with hybrid winding
structure,” IEEE Trans. Ind. Electron., vol. 60, no. 10, pp. 4369–4375,
Oct. 2013.
[36] Struve functions, Feb. 2014. [Online]. Available: http://www.maplesoft.
com/support/help/Maple/view.aspx?path=StruveH
Liang Yan (M’07–SM’12) received the Bachelor’s
degree from North China Institute of Technology,
Shanxi, China, in 1995, the Master’s degree from
Beijing Institute of Technology, Beijing, China, in
1998, and the Ph.D. degree from Nanyang Techno-
logical University (NTU), Singapore, in 2006.
From 1998 to 2002, he was with Beijing Institute
of Technology. From 2006 to 2009, he was with
NTU. He is currently a Professor with Beihang
University, Beijing. His research interests include
robotics, actuators and sensors, and navigation
systems.
Prof. Yan served as a Program and Publication Chairman for more than ten
IEEE/ASME conferences.
Lei Zhang received the Bachelor’s degree from
North China Electric Power University, Beijing,
China, and the Master’s degree from Beihang Uni-
versity (BUAA), Beijing, in 2014.
She is currently with the School of Automation
Science and Electrical Engineering, Beihang Univer-
sity. Her research interests include electromagnetic
actuators, sensors, and electrical circuit design.
Ms. Zhang was the recipient of the National Grad-
uate Scholarship in 2013, the Graduate Innovation
Fund in 2012, and the Excellent Master Thesis of
Beihang University in 2014.
6965
Zongxia Jiao received the B.S. and Ph.D. degrees
from Zhejiang University, Zhejiang, China, in 1985
and 1991, respectively.
From 1991 to 1993, he was a Postdoctoral Fellow
with Beihang University (BUAA), Beijing, China,
where he has been a Professor since 1994 and is
currently the Dean of the School of Automation
Science and Electrical Engineering. His research
interests include actuators, sensors, fluid power, and
transmission.
Prof. Jiao was the recipient of the Changjiang
Scholar Professor in 2006 and the Distinguished Young Scholar of China
in 2008.
Hongjie Hu received the Bachelor’s degree
from Harbin University of Science and Technology,
Harbin, China, in 1986, the Master’s degree from
Harbin Institute of Technology, Harbin, in 1993, and
the Ph.D. degree from Beihang University, Beijing,
China, in 2002.
He was a Posdoctoral Fellow and a Lecturer with
Beihang University, where he is currently an Asso-
ciate Professor. His research interests include elec-
trical machine design, mechanism design, control
theory, and engineering.
Chin-Yin Chen received the Ph.D. degree from Na-
tional Sun Yat-sen University, Kaohsiung, Taiwan,
in 2008.
He is currently a Researcher with the Institute
of Advanced Manufacturing Technology, Ningbo
Institute of Material Technology and Engineering,
Ningbo, China. His research interests include actua-
tor design, mechatronics, integrated structure/control
design, and robotics.
I-Ming Chen (M’95–SM’06–F’12) received the
B.S. degree from National Taiwan University, Taipei,
Taiwan, in 1986, and the M.S. and Ph.D. de-
grees from the California Institute of Technology,
Pasadena, CA, USA, in 1989 and 1994, respectively.
Since 1995, he has been with the School of Me-
chanical and Aerospace Engineering, Nanyang Tech-
nological University, Singapore, where he is also
the Director of the Intelligent Systems Centre. His
research interests include wearable sensors, human–
robot interaction, reconfigurable automation, parallel
kinematics machines, and smart material-based actuators.