Analysis on Electromagnetic Field of an Axial-

Flux Solid Aluminum Rotor Induction Motor

Qin Wei 1, Fan Yu 1, Zhu Xi 1, Li Shuo 1
1
Department of Electrical Engineering’ Beijing Jiaotong University
E-mail: [email protected]

Abstract —The Axial-Flux Solid Aluminum Rotor Induction

Motor (AFAIM), which is Electro-dynamic System [1-2], is
presented in this paper. The system creates electromagnetic field
by 3-phase alternating current in primary winds, induct eddy
field in secondary and create lift and thrust force. This paper
establishes a 2-D sub-loop electro -magnetic model, studies the
dynamic characteristic of the maglev, gets lift and thrust force
analytical solution. Finally, the paper analyses the model using
FEM software and gets the distribution of magnetic field. The
conclusion provides important method and theory basis for the
characteristic and magnetic-field of axial-flux solid rotor
induction motor. Fig.1 Rotation field electro-dynamic levitation device

I. INTRODUCTION
In an axial-flux motor, the air-gap flux flows in the axial
direction of the motor and the active current carrying
conductors are radially positioned [3] show as Fig.1and
Figure2. One of the principal features of this motor is the
relatively small inertia that makes it a suitable choice for servo
and high-speed applications. Axial-flux machine with a solid
rotor can be adopted into the construction of various devices
such as: electro-magnetic levitation device, and electric Figure3 laminated stator core with out winding
vehicles.
The electromagnetic part of Electro-dynamic System II. MOTOR STRUCTURE AND MODELING ANALYSIS
analytical equation is established and the solution is given.
The results of the distribution of magnetic field and the A. Machine construction
electromagnetic levitation force are shown. Comparison of
FEA (Finite Element Analysis) [4] software and analytical The Axial-Flux Solid Aluminum Rotor Induction Motor
solution reveal that two methods are consistent, and the prototype is shown in Fig. 1. The machine is constructed of
analytical model is effective and efficient. the secondary and the primary as shown in Fig. 2, 3-phase
windings are placed in the slotted stator core to produce the
time varying magnetic field and the nonmagnetic conducting
secondary are mounted on the non-rotating shaft. The machine
runs in an axial motion
The AFAIM offers some advantages over the conventional
Axial-Flux motor. It is considered due to the fact that lift force
can be generated by AFAIM. The repulsive force is
sufficiently large to suggest the possibility of its being used
for suspending a high-speed train. On the other hand, the
NSLIM also offers a lower cost besides it is easy to
manufacture.

B. AFAIM modeling
AFAIM is an axial magnetic field motor, the primary coil
placed along the radial, the magnetic circuit, the equivalent air
Fig.1Rotation field electro-dynamic levitation device prototype
gap length and the Polar distance are changing with increasing
This work was supported in part by the National Natural Science radius. The traditional research methods that the average
Foundation of China under grant No. 51077003, partly by National Natural
Science Foundation of China under grant No. 50807007, and partly
Doctoral Fund of Ministry of Education under grant No. 20090009110025,
China. The statements made herein are solely the responsibility of the
authors.
motor diameter method does not take into the change of Polar C. Field analysis
pitch, there is a big limitation. The magnetic field produced by the primary current of
Electromagnetic model of this device is established in this the machine can be described by the scalar magnetic
paper by using 2-D sub-loop electromagnetic model. The disc
potential Az :
primary will be divided along the radial into the rings. It is
complex to analysis magnetic field by using the ring model. ∂ 2 A2 z ∂ 2 A2 z
We splice each ring model along the radial direction as Figure ∇ 2 A2 z = + =0
4, and unfold it, which is similar with a Long and narrow ∂x 2 ∂y 2
rectangular structure linear motor as Figure 5. A conducting in region 2 (air space), 0 < y < g (1)
plate of thickness d and the primary as Figure 5 divides space
into four regions: region 1 the primary itself, region 2 (above ∂ 2 A3 z ∂ 2 A3 z
∇ 2 A3 z = + = i μ0 sσω A3 z
the conducting plate and below the primary) is the air-gap I, ∂x 2 ∂y 2
region3 the conducting plate itself, and region4 (below the
in region 3 (secondary), − d < y < 0 (2)
plate) is the air-gap II contains no magnetic sources or
conducting material. The radial distribution of air-gap 2 ∂ 2 A4 z ∂ 2 A4 z
magnetic field is converted to horizontal distribution. Main ∇ A4 z = + =0
parameters of each rectangular structure are connected to the
∂x 2 ∂y 2
pole pitch of each ring. in region 4 (air space), y < − d (3)
The geometry used to model the motor fields is shown in Where Az 2 Az 3 and Az 4 are related to the x and y components
Figure5.Let the upper surface of the conducting plate coincide
with the x -y plane (y = 0); the primary, which is located in a of B by:
plane parallel to the surface of the conducting plate, at y =g, B = ∇× A
moves in the x direction. ∂A ∂A
To complete the problem definition and to get to a unique By = − z , Bx = 2 z , E3 z = − jω A3 z
solution, the following boundary conditions are fixed: First, ∂x ∂y
the primary windings are modeled as two imaginary sets of For the machines, the general solutions of (1)–(63) are:
distributed phase windings laid on the primary iron surface, A2 z = (c1e py + c2 e − py )ei (ωt − kx ) (4)
the traveling wave excitation surface currents at the primary.
In the flux linkage calculation, the slots effect is usually A3 z = (c3e β y + c4 e− β y )ei (ωt − kx ) (5)
incorporated by applying the Carter coefficient. Second, the A4 z = (c5e py + c6 e− py )ei (ωt − kx ) (6)
machine is assumed to be infinitively long and has a periodic
construction. Third, the permeability of the stator core is The boundary conditions to be satisfied are:
assumed to be infinite whereas the relative permeability of the ∂A2 z
winding, air gap and permanent magnet as well as the shaft is Bx 2( y = g ) = = μ 0 J1 (7)
∂y ( y = g )
assumed to be one [4]-[5].
j1 = 2mW1kw1 I1 / pτ ⋅ e j (ωt − kx ) = J1e j (ωt − kx ) (8)
.

Where: m is the number of primary coil phase; W1 is the c

oil number; k w1 is the coefficient of the primary coil; I1 is th
e primary current， ω = 2π f ， k = π / τ 。 τ is the pole
pitch; p is the is the number of pole pairs; f is the fre
quency of the primary current.

H x 2( y =0) = H x 3( y =0) , A2 z ( y =0) = A3 z ( y =0) (9)

Figure 4 Sub-loop model of disc primary
H x 3( y =− d ) = H x 4( y =− d ) , A3 z ( y =− d ) = A4 z ( y =− d ) (10)
y
M o v in g − d ire c tio n
v
A4 z ( y =−∞ ) = 0 (11)
(1)
( jz = Jei(ωt−kx) )
The expression of H for the three regions are given by
1 ∂A2 z
g (2)
H x2 = (12)
x
μ0 ∂y
d (3)
1 ∂A3 z
(4)
H x3 = (13)
μ0 ∂y
Figure5 2-D Long and narrow rectangular linear motor
1 ∂A4 z
H x4 = (14)
μ0 ∂y
Application of the boundary conditions (7)–(11) to the above
(12)–(14) gives
K s [Δ 4 ( p + β ) − Δ 3 ( p − β )]
c1 = (15)
p Δ1Δ 4 − Δ 2 Δ3
K [Δ ( p − β ) − Δ 3 ( p + β )]
c2 = s 4 (16)
p Δ1Δ 4 − Δ 2 Δ 3
2Ks Δ4
c3 = (17)
Δ1Δ 4 − Δ 2 Δ3
−2 K s Δ 3 Fig.7 Air-gap flux density along circumference
c4 = (18)
Δ1Δ 4 − Δ 2 Δ 3
2 K s (Δ 4 − Δ 3 e 2 β d ) III. RESULTS AND DISCUSSION
c5 = , c6 = 0 (19)
Δ1Δ 4 − Δ 2 Δ 3 Fig.8 shows the analytical results of the lift force
Where: characteristic for the AFAIM under different levitated heights
with the different input current frequencies. The curve
Δ1 = ( p + β )e pg − ( p − β )e − pg indicates that the lift force increases as the air-gap decreases
Δ 2 = ( p − β )e pg − ( p + β )e − pg ， and the larger input current frequencies will produce a larger
lift force.
Δ 3 = ( p − β )e − β g ， Δ 4 = ( p + β )e β g 5

Therefore, the lift force and electromagnetic torque can be 4.5

written as: 4
τ
L
Flift = ∑ ∫
3.5

Re( By 3n ⋅ By∗3n − Bx 3n ⋅ Bx∗3n )dx (20)

4 μ 3
Lift force(N)

n =1,2,3... 0 0 air-gap=10mm
2.5 air-gap=5mm
τ
L
air-gap=1mm

∑ 2μ0 ∫0
2

Tdrag = 2π rn Re( Bx 3n ⋅ By∗3n )dx (21) 1.5

n =1,2,3...
1

Where: L is the width of each rectangular; τ is the pole 0.5

pitch. rn is the radius of the nth ring. 0

0 50 100 150 200 250 300
the primary current frequence(Hz)

D. FEA Program Fig.8 Curve between lift force and frequency

In this section, the computer-simulation model constructed
by the finite-element method (FEM) has been developed for 18

software ANSOFT to verify the analytical results. By using Analytical

16
the FEM simulation model with different operation Experiment
3-D-FEA
parameters, characteristics of the magnetic flux shown in 14

Fig.7, the lift force, and electromagnetic torque for the 12

proposed AFAIM can be obtained.

Lift force(N)

10
The finite element code comprises of two steps. The first
step is for automatic mesh generation, which requires only 8

geometrical data as input. This step generates modal co- 6

ordinates, element details and their connectivity. The second 4

step takes the input from the mesh generation program besides 2
material properties, boundary conditions [6]. A typical finite
element mesh is shown in Fig. 6. 0
1 2 3 4 5 6 7 8
the primary currten(A)

Fig.9 Curve between lift force and primary phase current

Fig. 9 shows the comparison among experimental,
analytical, and FEA results for lift force with different-current
in the secondary region. The analytical predictions, the FEA
analyzes and the experimental are shown in good agreement,
while those have an error corresponding to approximately 8%,
due to the end effect. The curve indicates that the larger input
current will produce a larger lift force.
Figure6 the mesh density of one pole of the stator
 All in all, from the analytical-equation derivation and
computer-simulation, it can be concluded that the AFAIM has
4.5 similar characteristics of the induction motor [7-9].
3-D-FEA
4 Analytical
Experment

CONCLUSION
3.5
IV.
3
Analytical models, formulated in Cartesian coordinates, have
lift force(N)

2.5
been developed for predicting the field distribution in Axial-
2
Flux Solid Aluminum Rotor Induction Motor. Since the
1.5 developed models are analytically based and take significantly
1 less computing time than the finite-element method, they are
0.5
eminently suitable for use during the initial design of Axial-
Flux Solid Aluminum Rotor Induction Motor. Predicted
0
0 0.005 0.01 0.015 0.02 0.025 0.03
air-gap(cm)
0.035 0.04 0.045 0.05 results from the analytical models have been compared with
finite-element predictions, and excellent agreement has been
Fig.10 Curve between lift force and air-gap length
achieved. It has been shown that the lift force and
Fig. 10 shows the comparison among experimental,
electromagnetic torque varies significantly with the pole
analytical, and FEA results for lift force with air-gap in the
number, and the primary current and the air gap. It has also
secondary region. The analytical predictions, the FEA
been shown that if the air gap is relatively large, the lift force
analyzes and the experimental are shown in good agreement.
and electromagnetic torque can be enhanced by employing
The curve indicates that the lift force increases as the air-gap
primary current.
decreases.
0.18
Analytical
0.16 3-D-FEA
Experiment
0.14 REFERENCES
0.12
[1] Lu Guang Yan., “Development and Application of the Maglev
Transportation System,” I.E.E.E, Transactions on Applied
torque(N.m)

0.1
Superconductivity, Vol. 18, No. 2, pp. 92-99, June 2002.
0.08
[2] Lu Guang Yan, “Suggestion for Selection of Maglev Option for Beijing-
0.06 Shanghai High-Speed Line”, I.E.E.E, Transactions on Applied
0.04
Superconductivity, Vol. 14, No. 2, pp. 936-939, June 2004.
[3] M. Mirzayee, M. Mirsalim and S. E.Abdollahi “Analytical Modeling of
0.02
a Disk Induction Motor Using Maple Software”, XII International
0
0 0.005 0.01 0.015 0.02 0.025 0.03 0.035 0.04 0.045 0.05
Symposium on Theoretical Electrical Engineering, CDproceeding,
air-gap(m) Warsaw, Poland, 2003.
Fig.11 Curve between torque and air-gap length [4] O. A. Mohammed, D. A. Lowther, M.Lean, and B. Alhalabi, “On the
Creation of a Generalized Design Optimization Environment for
Fig. 11 shows the comparison among experimental, Electromagnetic Devices,” IEEE Transactions on Magnetics, Vol. 37,
analytical, and FEA results for electromagnetic torque with the No. 5, pp. 3562-3565, September 2001.
different air-gap. The curve indicates that the electromagnetic [5] Wijono., H. Arof, H.W. Ping., “Analysis of magnetic field distribution
torque increases as the air-gap decreases. of a cylindrical discrete Halbach permanent magnet linear generator,”,
IET Electric Power Applications, Vol. 4, No. 8, pp. 629-636,
Novermber2009.
0.2 [6] M.Sc. M. Valtonen, D.Sc. A. Parviainen and Prof. J. Pyrhönen
0.18
Analytical
3-D-FEA “Electromagnetic Field Analysis of 3D Structure of Axial-Flux Solid-
0.16
Experiment
Rotor Induction Motor” SPEEDAM 2006 International Symposium on
0.14
Power Electronics, Electrical Drives, Automation and Motion
[7] Yaow-Ming Chen, Shu-Yuan Fan and Wie-Shin Lu., “Electro-magnetic
0.12
Force Analysis of the Magnetically Levitated Motor with Two
torque(N.m)

0.1
Directions of Movement”, I.E.E.E, Transactions on Industry
0.08
Applications, Vol. 42, No. 1, pp. 31-41, January2008.
0.06 [8] LV Gang., SUN Shouguang., MA Yunshuang., “Efficiency optimal
0.04 control of linear induction motor for urban rail transit”, Electrical
0.02 Machine and Control, Vol. 13, No. 4, pp. 490-495, July2009.
0
[9] Jonathan Bird, Thomas A. Lipo. , “Calculating the Force created by an
Electro-dynamic Wheel Using a 2-D Steady-State Finite-Element
0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1
s

Fig.12 Curve between torque and slip speed Method,” I.E.E.E, Transactions on Magnetics, Vol. 44, No. 3, pp. 365-
371, March2008.
Fig. 12 Fig. 9 shows the comparison among experimental,
analytical, and FEA results for electromagnetic torque with the
different slip speed. The analytical predictions, the FEA
analyzes and the experimental are shown in good agreement.
Curve in Fig.12 shows that the electromagnetic torque
increases as the slip speed increases and the larger slip-speed
will produce a larger electromagnetic torque before the torque
is saturated.