Professional Documents
Culture Documents
2
March-April 2011
Manuscript received and revised March 2011, accepted April 2011 Copyright © 2011 Praise Worthy Prize S.r.l. - All rights reserved
760
N. A. Rahim, W. P. Hew, A. Mahmoudi
Copyright © 2011 Praise Worthy Prize S.r.l. - All rights reserved International Review of Electrical Engineering, Vol. 6, N. 2
761
N. A. Rahim, W. P. Hew, A. Mahmoudi
1 T e t i t 1 T
Kp dt fe t fi t dt (10)
T 0 E pk I pk T 0
Copyright © 2011 Praise Worthy Prize S.r.l. - All rights reserved International Review of Electrical Engineering, Vol. 6, N. 2
762
N. A. Rahim, W. P. Hew, A. Mahmoudi
m is rotor angular speed, Dtot and Ltot respectively are Axial length of stator core Lcs can be written as:
machine outer diameter total and machine length total
including stack outer diameter and end-winding Bg p Do 1
protrusion from radial and axial iron stacks. Lcs (21)
4 pBcs
The generalized sizing equation approach can easily
be applied to double-sided axial-flux permanent-magnet where Bcs is flux density in stator core, and p is ratio of
TORUS type motor. The outer surface diameter (Do) can average air-gap flux density to peak air-gap flux density.
be written as: Axial length of rotor Lr becomes:
Dtot Do 2Wcu (17) where Bcr is flux density in rotor disc core, and Bu is
attainable flux density on permanent-magnet surface.
TABLE III Permanent-magnet length Lpm can be calculated as:
TYPICAL PROTOTYPE WAVEFORMS
Model e(t) i(t) Ki Kp
r Bg
Lcr Kc g (24)
Sinusoidal 0.5Cos Kf
2 Br Bg
Kd
Sinusoidal 2 0.5
where µr is magnets recoil relative permeability, Br is
permanent-magnet material residual-flux density, Kd is
Rectangular 1 1
leakage flux factor, Kc is Carter factor, Kf =Bgpk/Bg is
peak value corrected factor of air-gap flux density in
radial direction of AFPM motor. These factors can be
Trapezoidal 1.134 0.777
obtained from FEM analysis [19].
In AFPM motors, air-gap flux density and diameter
ratio are the two important design parameters having
significant effect on motor characteristics. To optimize
Triangular 3 0.333 motor performance, diameter ratio and air-gap flux
density must be chosen carefully. The optimum design
should maximize power density while maintaining
where, Wcu is protrusion of end winding from iron stack, desired efficiency within design restrictions (Table II). In
in radial direction. For back-to-back wrapped winding, design studies, diameter ratio and air-gap flux density are
protrusions exist towards machine axis as well as design parameters. Fig. 2 shows power density variation
towards the outsides, and can be calculated as: as a function of air-gap flux density, and diameter ratio
of the AFPM motor.
2 ADave
Di Di2
K cu J s III. Simulation and
Wcu (18)
2 Finite Element Analysis
where, Dave is average diameter of the machine, Js is The design was simulated via Finite Element Method
current density, and Kcu is copper fill factor. Note that for Magnetics (FEMM) Software. The simulation model
slotted machines, depth of stator slot is Lss=Wcu. Axial reached the output (2.7 kW) targeted for the electric
length Le of machine is given by: motorcycle.
Copyright © 2011 Praise Worthy Prize S.r.l. - All rights reserved International Review of Electrical Engineering, Vol. 6, N. 2
763
N. A. Rahim, W. P. Hew, A. Mahmoudi
(a) AFPM motor model in 2D with Fine Meshing (b) Magnetic-flux density FEMM simulation
The relatively symmetrical distribution of the FEA was for overview of saturation levels in various
magnetic-flux density relative to radial symmetrical axis parts of the machine, for comparison of flux densities
of the magnets indicates currents negligible influence on obtained from FEM with sizing analysis.
resultant magnetic field. Maximum flux density is higher Table V tabulates results for the comparison, which
in the stator core than in the rotor because the stator core was done at no load and for various parts of the machine.
is laminated steel that saturates at much higher values. The no-load flux density plots show consistency with
Table IV shows the parameters and the optimized sizing analysis, maximum flux density of rotor and of
TORUS motor dimensions calculated via sizing stator almost equal. Maximum and average air-gap flux
equations. densities from FEM and from sizing analysis agree, too.
TABLE IV Fig. 4 compares calculated back-EMF against
MOTOR DIMENSIONS
electrical angle of the designed motor, from FEA and
Rotor Inner Diameter 130 mm
from the no-load experiment at 700 rpm. The
Rotor Outer Diameter 230 mm experiments peak back-EMF was 90.4V, slightly less
Number of Windings 48 than the 95V computed value, agreeing closely with the
Number of Turns 12 computed waveform.
Magnetic Pole 16 Fig. 5 compares experiment torque against electrical
angle variation, and predicted torque from FEA; both
Magnet Thickness 7 mm
closely agree. At 30A rated current and 700rpm rated
Magnet Arc 18o speed, the motor produces 37.4Nm maximum torque
Magnet Material Nd-Fe-B, N35 while the simulation showed 36.78Nm.
Back-Iron Thickness 12 mm
TABLE V
Rated Voltage Line-to-Line 64V FLUX DENSITY COMPARISON OF THE DESIGNED MOTOR
(rms) Rotor Air-gap stator
Rated Phase Current (rms) 30A Bcr Bmax Bave Bcs
Output Power 1.8 kW FEM 1.2 0.81 0.52 1.15
Sizing Eq. 1.1 0.8 0.5 1.1
Copyright © 2011 Praise Worthy Prize S.r.l. - All rights reserved International Review of Electrical Engineering, Vol. 6, N. 2
764
N. A. Rahim, W. P. Hew, A. Mahmoudi
IV.1. Manufacturing
Design challenge in manufacturing the AFPM motor
is maintaining air gap between stator and rotor. Magnetic
interaction between rotor magnet and stator back-iron is
quite large (752 N simulated value for this motor). The
air-gap needs to be the smallest possible; in the design,
1mm. Fig. 6 shows the active parts assembled and the
rotors fabricated surface-mounted permanent-magnet
mount. Fig. 8. Shaft with hole-through for phase-winding terminal outlet
Windings were professionally hand-made; see Fig. 7.
They were placed on flat-stator-core surface. To prevent IV.2. Driving System
the windings from missing its position and from
vibration during motor operation, a type of epoxy resin To rotate the motor, the stator windings should be
was applied, giving the windings characteristics such as energized in sequence. Knowledge of the rotors
stiffness in working temperature, original dimensions, positioning is important, to understand which winding is
and good thermal conductivity for heat-release. energized following the energizing sequence. Rotor
Copyright © 2011 Praise Worthy Prize S.r.l. - All rights reserved International Review of Electrical Engineering, Vol. 6, N. 2
765
N. A. Rahim, W. P. Hew, A. Mahmoudi
Hall
sensors
windings
Copyright © 2011 Praise Worthy Prize S.r.l. - All rights reserved International Review of Electrical Engineering, Vol. 6, N. 2
766
N. A. Rahim, W. P. Hew, A. Mahmoudi
fe
fm (25)
2P
V. Experiment Results
Fig. 13 shows the experiment test-bench set-up in
University of Malayas Department of Electrical Fig. 13. Motor experiment test-bench set-up with NI® data logger
Engineering, for performance test of the in-wheel motor.
A National Instrument Data Acquisition System with
LabVIEW interface was used to obtain test data and
plot performance curves. Motor torque and back-EMF
were the main performance parameters obtained. During
cruising-speed test, secondary measurements such as
temperature rise in the motors critical parts were also
recorded.
Figs. 14 and 15 show graphs of back-EMF and torque.
Back-EMF maximum output was 180V peak-to-peak,
and torque output at rated current (30A) was about 37.4
N.m. Back EMF was acquired by mechanically turning
the wheel at a particular speed, and then measuring
terminal voltage.
Fig. 14. Comparison of Back-EMF results obtained from experiment
Motor speed was captured on a tachometer, which and from FEM simulation
obtained the speed from the Hall-sensor pulse train.
Under such conditions, the machine then acted as
generator. At no-load condition, terminal voltage of the
machine equaled generated back-EMF. Motor torque was
measured on a load-cell force sensor, which was
mounted on a free-rolling shaft. Constant, controlled
current was injected into the motor from an inverter. The
wheel was loaded with roller brake. Torque could be
increased to maximum value quickly, and to twice the
rated value.
Input power (Pi) is the electric energy that runs the
motor. Mechanical power delivered by the motor (Pm) is
torque and speed, and overall motor efficiency is the
ratio of input power to mechanical power: Fig. 15. Comparison of torque results obtained from experiment and
from FEM simulation
Pi 3V ph I ph (26)
Pm m (27)
Pm
(28)
Pi
Copyright © 2011 Praise Worthy Prize S.r.l. - All rights reserved International Review of Electrical Engineering, Vol. 6, N. 2
767
N. A. Rahim, W. P. Hew, A. Mahmoudi
Copyright © 2011 Praise Worthy Prize S.r.l. - All rights reserved International Review of Electrical Engineering, Vol. 6, N. 2
768
N. A. Rahim, W. P. Hew, A. Mahmoudi
Copyright © 2011 Praise Worthy Prize S.r.l. - All rights reserved International Review of Electrical Engineering, Vol. 6, N. 2
769