Article

Comparative Study of Permanent Magnet, Conventional, and

Advanced Induction Machines for Traction Applications
Tayfun Gundogdu 1,2 , Zi-Qiang Zhu 2, * and Ching Chuen Chan 3

1 Department of Electrical and Electronic Engineering, Hakkari University, Zeynel Bey Campus, Keklikpinar,
Pinarlar Cd., 30000 Hakkari, Turkey; tayfungundogdu@hakkari.edu.tr
2 Department of Electronic and Electrical Engineering, University of Sheffield, Mappin Street,
Sheffield S1 3JD, UK
3 Department of Electronic and Electrical Engineering, The University of Hong Kong, Hong Kong, China;
ccchan@eee.hku.hk
* Correspondence: z.q.zhu@sheffield.ac.uk

Abstract: This paper investigates and compares the torque-generating capabilities and electromag-
netic performance of advanced non-overlapping winding induction machines (AIM), conventional
induction machines (CIM), and interior-permanent magnet (IPM) machines for electric vehicle (EV)
applications. All investigated machines are designed based on the specifications of the Toyota Prius
2010 IPM machine. The steady-state and flux-weakening performance characteristics are calculated
by employing the 2D finite element method and MatLab, and the obtained results are quantitatively
compared. Furthermore, the torque-generating capabilities of three machines are investigated for
different electric loadings, and the machine having the highest torque-generating capability is deter-
mined as AIM. Moreover, the major parameters affecting the torque-generating capability, such as
magnetic saturation and magnet demagnetization, are examined in depth.

Citation: Gundogdu, T.; Zhu, Z.-Q.; Keywords: electric vehicles; induction machine; interior permanent magnet machine; non-overlapping
Chan, C.C. Comparative Study of winding; saliency; torque capability
Permanent Magnet, Conventional,
and Advanced Induction Machines
for Traction Applications. World
Electr. Veh. J. 2022, 13, 137. 1. Introduction
https://doi.org/10.3390/
Internal combustion engine vehicles are responsible for 21% of worldwide anthro-
wevj13080137
pogenic CO2 emissions [1]. Making the transition of the global vehicle fleet to zero-emission
Academic Editor: Hui Yang vehicle technology is critical for decarbonizing road transportation and fulfilling the envi-
ronmental and climate targets. Therefore, worldwide electric vehicle (EV) sales, including
Received: 22 June 2022
passenger cars, light trucks, and light commercial vehicles, reached 6.75 million units in
Accepted: 26 July 2022
Published: 28 July 2022
2021, corresponding to a 108% increase over 2020 [2]. It is very critical to choose the right
electrical machine topology for EV applications in order to maximize efficiency, transient
Publisher’s Note: MDPI stays neutral electromagnetic performance characteristics, flux-weakening capability, and cost. The
with regard to jurisdictional claims in worldwide five best-selling models in 2021 [3] and their electrical machine technologies are
published maps and institutional affil-
listed in Table 1. In addition, permanent magnet synchronous machines (PMSMs), particu-
iations.
larly interior-permanent magnet (IPM) machines, are used in the world’s top commercial
EVs and hybrid electric vehicles (HEVs), including Toyota/Prius, Nissan/Leaf, BMW/i3,
and numerous other vehicles. Other cars, on the other hand, including the BMW/X5,
Copyright: © 2022 by the authors.
Renault/Kangoo, GM/EV1, Chrysler/Durango, and a few others, employ induction ma-
Licensee MDPI, Basel, Switzerland.
chines (IMs) [4–11]. Moreover, Tesla Motors Inc., one of the world’s leading plug-in EV
This article is an open access article manufacturers, utilizes both IM (front) and IPM (rear) machines in its best-selling mod-
distributed under the terms and els, as seen in Table 1 [3]. In addition, Audi also utilizes the same traction topology in
conditions of the Creative Commons e-Tron models.
Attribution (CC BY) license (https://
creativecommons.org/licenses/by/
4.0/).

World Electr. Veh. J. 2022, 13, 137. https://doi.org/10.3390/wevj13080137 https://www.mdpi.com/journal/wevj

World Electr. Veh. J. 2022, 13, 137 2 of 18

Table 1. Top selling EV models in 2021 and their machine technology [3].

Top Selling Models Machine Technology

Tesla/Model 3 IM + IPM
Wuling/Hongguang Mini EV PMSM
Tesla/Model Y IM + IPM
Volkswagen/ID.4 PMSM
BYD/Qin Plus PHEV PMSM

The common characteristics of electric machines designed for traction applications

can be listed as [12,13]: (a) high starting torque; (b) high torque and high-power densities;
(c) high efficiency across a broad speed range; (d) low torque ripple; (e) light weight; and
(f) low cost.
Although there are many favorable characteristics of squirrel-cage IMs, such as robust-
ness, relatively low-cost, good control dynamics, and mature manufacturing technology,
having conductor bars in the rotor is its major disadvantage because the solid conductor
bars of the IM cause additional joule losses in the rotor and hence result in a relatively lower
efficiency compared to PMSMs. PMSMs, on the other hand, possesses superior advantages,
such as high torque, high power, and relatively high efficiency. However, the high cost
of NdFeB magnets has a significant impact on their popularity [8]. Moreover, there are a
number of comparative studies on the performance comparisons of electrical machines
employed in the traction applications in literature [4–28]. In these studies, the performance
characteristics of IMs, PMSMs, switched and synchronous reluctance machines, PM as-
sisted reluctance and brushless DC machines have been compared quantitatively. The
squirrel-cage IMs have been noted as having established production technology and the
ability to offer the required driving characteristics. PMSMs, on the other hand, have been
noted as having better efficiency and torque density. Moreover, according to a compre-
hensive quantitative comparison study on electrical machines for traction applications,
IPM machines have a higher power factor and efficiency than IMs, whereas IMs provide
competitive performance characteristics at a lower cost and with better overloading ca-
pability. Wound field, reluctance, and variable flux synchronous machines have all been
demonstrated to be less appealing for traction applications because of their poorer torque
density and efficiency characteristics [20,22,27,28]. In addition, in order to improve the
torque characteristics of PMSMs some unconventional methods have been proposed and
compared with conventional electrical machines employed in traction applications [29,30].
Torque-generating capacities are investigated in this paper, in addition to previous
research comparing the performance of IMs and PMSMs in the literature [5–9,12–14,19].
Therefore, the major goal of this study is to determine which machine is capable of produc-
ing higher torque under overload operating conditions and examine the underlying causes.
In addition, this comparison incorporates an advanced induction machine (AIM) with
non-overlapping windings, which was recently developed for EV/HEV applications [31].
This paper focuses on the analysis and quantitative comparison of electromagnetic
performance and design characteristics of the IPM machine, CIM, and AIM operated under
the same conditions and the same slot/pole number combinations. The major steady-
state electromagnetic performance and flux-weakening characteristics are provided and
discussed, with a particular focus on torque-generating capabilities. Moreover, since one
of the AIM’s most significant advantages is its comparatively small total axial length, the
impact of stack length on steady-state performance is systematically investigated.

2. Research Method
2.1. Concept
This study presents FEA and electromagnetic performance comparisons of different
types of electrical machines, namely PMSM and IM, employed in traction applications.
Among the considered machines, the IPM and CIM are adopted by conventional integer
World Electr. Veh. J. 2022, 13, 137 3 of 18

slot distributed windings (ISDWs), while AIM is adopted by advanced non-overlapping

windings (ANWs).

2.2. Description of the Tool

An IPM machine, CIM, and AIM, having the same operating and design specifications
as the Toyota Prius 2010 IPM machine, are modelled and analyzed by 2D, time stepping
FEA. Accordingly, the flux-weakening performance characteristics are calculated by the
Matlab tool by using the flux matrices obtained from FEA calculations.

2.3. Analysis Scheme

Various analyses, including transient, steady state, electric loading, and flux-weakening,
have been conducted in this study. Transient and steady-state analyses have been con-
ducted to reveal the rated performance characteristics such as electromagnetic torque,
torque ripple, back-EMF and induced voltage waveforms, flux line and density distribu-
tions, and saturation factors. Electric loading analyses have been conducted to compare
the torque-generating capabilities of the aforementioned machines. The flux-weakening
characteristics are calculated to compare the torque and power versus speed curves and
efficiency maps.

2.4. Research Results

Since the vehicle acceleration is of great importance, the torque-generating capabilities
of different traction machines need to be revealed. In this study, apart from the comparison
of the torque-generating capabilities, flux-weakening characteristics, torque ripple levels,
and total active material costs have been compared quantitively. The results presented
in this study show that the overall flux-weakening characteristics of IMs are comparable
to those of IPM machines and because of the demagnetization of the magnets of IPM
machine during overloading operating, the IMs deliver higher torque during acceleration
or overloading modes.

3. Design Specifications
In order to evaluate the torque-generating capabilities, the performance characteristics
of the studied machines are examined utilizing time-stepping 2D finite element method
(FEM) for rated and various electric loading operations. In order to investigate the torque-
generating capabilities, the electric loading will be varied from one to five times the rated
current value. For a fair comparison and to achieve comparable results, the same operating
conditions and geometrical parameters as shown in Table 2 are utilized.
For a fair comparison, the Toyota Prius 2010 IPM machine is directly adapted by ap-
plying the optimal specifications [21,32], and the IMs are designed with the same geometric
dimensions and pole number. It is worth mentioning that Toyota Prius 2010 IPM machine
has been used since all specifications and data for the Prius 2010 are fully available. How-
ever, other models are still not available because of the confidential issues of the company.
On the other hand, some essential geometric and operating design specifications of electrical
machines for traction applications can be found in [21–26]. Furthermore, the IPM machine
and the CIM share the identical stator slot/pole number combination (48S/8P) and winding
layout with 5-slot pitch single-layer windings. In addition, a double-layer non-overlapping
2-slot pitch winding topology is adopted for AIM having a 24S/8P combination.
IMs are optimized by using multi-objective global optimization via genetic algorithm,
as presented in [33]. Two-dimensional cross-sectional views and the design specifications of
the machines are shown in Figure 1 and given in Table 2, respectively. As shown in Table 2,
all the machines have the same main operational and geometrical specifications. Moreover,
the same iron core material whose core loss coefficients, namely hysteresis k h , eddy current
k c , and excessive k e are given in the table, has been assigned as core material for stator and
rotor parts of the machines. A strong grade of PM, whose remanence, relative permeability
µr , and coercivity Hc are given in the table, is assigned for the IPM machine. Moreover,
World Electr. Veh. J. 2022, 13, 137 4 of 18

because of the better heat and resistance characteristics of the copper over aluminium,
copper has been assigned as material of the squirrel-cage [34].

Table 2. Specifications of Prius IPM machine and IMs.

Parameters IPM CIM AIM

S/(R or M) */P 48S/16M/8P 48S/52R/8P 24S/26R/8P
Voltage limit (Vrms) 650 V × 85% × 2/π
Rated current (Apeak) 250
Number of coils per phase 8 8 8
Number of turns per coil 11 8 11
Number of series turn per phase 88 64 88
Fundamental winding factor 0.966 0.966 0.866
Number of parallel brunch 1
Slot fill factor 0.6
Phase resistance at 21 ◦ C 0.077 0.05612 0.0577
Stack length (mm) 50.8
Stator parameters
Outer diameter (mm) 264
Inner diameter (mm) 161.9 185.85 184.8
Tooth width (mm) 7.55 8.45 11.52
Slot opening (mm) 1.88 1.88 5.8
Slot height (mm) 30.9 15.4 22/11
Air-gap length (mm) 0.73 0.4 0.4
Rotor parameters
Tooth width (mm) — 6.83 11.97
Slot opening (mm) — 1.88 5.6
Slot height (mm) — 14 20
Magnet dimensions 49.3 × 17.88 × 7.16 — —
Cage material — Copper Copper
Iron grade DW310-35
kh 179.038
k c Veh. J. 2022, 13, x FOR PEER REVIEW
World Electr. 0.375 5 of 20
ke 0.262
Magnet grade NdFeB (N35) — —
µr 1.05 — —
Br for the IPM machine. 1.1
Moreover, because of the better heat and resistance characteristics
Hc (kA/m) −805.4
of the copper over aluminium, —
copper has been assigned as material of the —squirrel-cage
*[34].
R: Number of rotor slots. M: Number of permanent magnets. S: Number of stator slots.

(a) (b) (c)

Figure 1. Two-dimensional
Two-dimensionalviews
viewsofof
thethe
compared machines:
compared (a) IPM
machines: (48S/16M/8P).
(a) IPM (b) CIM
(48S/16M/8P). (b) CIM
(48S/52R/8P). (c)AIM
(48S/52R/8P). (c) AIM(24S/26R/8P).
(24S/26R/8P).

4. Performance and Active Material Cost Comparison

The performance characteristics of the IPM machine and IMs are determined using a
dq-axis reference frame with the d-axis oriented with the rotor field [35]. The obtained
results are presented in the following subsections.

4.1. Winding Factor and MMF Harmonics

 (a) (b) (c)

Figure 1. Two-dimensional views of the compared machines: (a) IPM (48S/16M/8P). (b) CIM
World Electr. Veh. J. 2022, 13, 137 5 of 18
(48S/52R/8P). (c) AIM (24S/26R/8P).

4. Performance and Active Material Cost Comparison

4. Performance and Active Material Cost Comparison
The performance characteristics of the IPM machine and IMs are determined using a
The performance characteristics of the IPM machine and IMs are determined using
dq-axis reference frame with the d-axis oriented with the rotor field [35]. The obtained
a dq-axis reference frame with the d-axis oriented with the rotor field [35]. The obtained
results are presented in the following subsections.
results are presented in the following subsections.
4.1. Winding
4.1. Winding Factor
Factor and
and MMF
MMF Harmonics
Harmonics
The calculated
The calculated winding
winding factors
factors and
and MMF
MMF amplitudes
amplitudesfor for1-ampere-1-turn
1-ampere-1-turnfor for24S/8P
24S/8P
(double-layer) and
(double-layer) and 48S/8P
48S/8P(single-layer)
(single-layer)winding
winding topologies
topologies areare
illustrated
illustratedin Figure 2. As
in Figure 2.
seen in Figure 2a, the fundamental winding factor and MMF amplitude
As seen in Figure 2a, the fundamental winding factor and MMF amplitude of the 24S/8P of the 24S/8P to-
pology are
topology are10.35%
10.35%lower
lowerthan
thanthose
thoseofofthe
the48S/8P
48S/8Ptopology.
topology.Therefore,
Therefore, itit can
can bebe predicted
predicted
that in order to obtain the same torque, the higher number of turns
that in order to obtain the same torque, the higher number of turns per phase is per phase is required
required
for the 24S/8P topology under the operating conditions with the same excitation current.
for the 24S/8P topology under the operating conditions with the same excitation current.
As seen
As seen in
in Figure
Figure 2b,
2b, the
the THD
THDof ofthe
theMMF
MMFof ofthe
the24S/8P
24S/8P topology
topology is is 49.68%
49.68% higher
higher than
than
that of
that of the
the 48S/8P
48S/8P topology.
topology. Hence,
Hence, itit can
can be
be expected
expected that
that the
the rotor bar copper
rotor bar copper loss
loss of
of
24S/8P topology
24S/8P topology will
will be
be considerably
considerably higher
higher than
than that
that of
of the
the 48S/8P
48S/8Pcounterpart.
counterpart.

1 2
24S/8P 48S/8P MMF Amplitude (At) THD (%)
0.8 1.6
Winding Factor

24S/8P (AIM) 28.34

0.6 1.2
48S/8P (IPM, CIM) 14.26
0.4 0.8

0.2 0.4
0
0
0
4
8

100
12
16
20
24
28
32
36
40
44
48
52
56
60
64
68
72
76
80
84
88
92
96
100
0
4
8
12
16
20
24
28
32
36
40
44
48
52
56
60
64
68
72
76
80
84
88
92
96

Harmonic Order
Harmonic Order
(a) (b)
Figure 2. Comparison of winding factors and MMF harmonics of the considered machines. (a)
Figure
Winding Comparison
2. factors of winding
of harmonics. factors
(b) MMF and MMF
harmonics harmonics
spectra for 1-turn of
andthe considered machines.
1-ampere
(a) Winding factors of harmonics. (b) MMF harmonics spectra for 1-turn and 1-ampere.
4.2. Induced Voltage and Back-EMF
4.2. Induced Voltage and Back-EMF
World Electr. Veh. J. 2022, 13, x FOR PEER REVIEW
The line back-EMF waveform for IPM machine and induced voltage waveforms 6 offor
20
The line back-EMF waveform for IPM machine and induced voltage waveforms for
IMs, their harmonic spectra, and THD percentages are illustrated in Figure 3. As seen, the
IMs, their harmonic spectra, and THD percentages are illustrated in Figure 3. As seen, the
back-EMF waveform of the IPM machine is highly distorted. The possible reasons behind
back-EMF waveform of the IPM machine is highly distorted. The possible reasons behind
the distorted
the distorted back-EMF
back-EMF waveforms
waveforms can
can be
be the
the combined
combined effect
effect of
of stator
stator slotting
slotting and
and the
the
heavily saturated stator core. On the other hand, since the number of turns per phase
heavily saturated stator core. On the other hand, since the number of turns per phase and and
hence the
hence the ampere-turn
ampere-turn magnitude
magnitude at
at the
the same
same current
current is
is higher
higher for
for the
the IPM
IPM machine,
machine, its
its
back-EMF amplitude
back-EMF amplitude is
is higher
higher than
than the
the IMs
IMs induced
induced voltage
voltage values.
values.

450 400
THD (%)
300 IPM 350
300 IPM 11.75
Voltage (V)

CIM
Amplitude (V)

150 CIM 7.64

250
0 AIM AIM 8.87
200
-150 150
-300 100
50
-450
0
0 60 120 180 240 300 360 1 3 5 7 9 11 13 15 17 19 21 23 25
Rotor Position (Elec. Deg.) Harmonic Order
(a) (b)

Figure 3. Comparison
Figure 3. Comparison of
of back-EMF
back-EMF waveforms
waveforms and
and their
their harmonic
harmonic spectra.
spectra. (a)
(a) Line
Line Back-EMF
Back-EMF and
and
induced
induced voltage
voltage waveforms.
waveforms. (b)
(b) Harmonic
Harmonic spectra
spectra of
of the
the back-EMF
back-EMF and
and induced
induced voltage.
voltage.

4.3. Flux Density and Flux Line Distributions

Figure 4 shows the flux density and flux line distributions of the considered ma-
chines. As expected, the flux density levels of the stator and rotor tooth parts are the high-
est. Among the designed machines, the CIM’s averaged core saturation level is the lowest
whilst the averaged core saturation of the IPM machine is the highest.
 0 60 120 180 240 300 360 1 3 5 7 9 11 13 15 17 19 21 23 25
Rotor Position (Elec. Deg.) Harmonic Order
(a) (b)

Figure 3. Comparison of back-EMF waveforms and their harmonic spectra. (a) Line Back-EMF and
World Electr. Veh. J. 2022, 13, 137 6 of 18
induced voltage waveforms. (b) Harmonic spectra of the back-EMF and induced voltage.

4.3. Flux Density and Flux Line Distributions

4.3. Flux Density and Flux Line Distributions
Figure 4 shows the flux density and flux line distributions of the considered ma-
Figure 4 shows the flux density and flux line distributions of the considered machines.
chines. As expected,
As expected, the fluxthe flux density
density levels oflevels of the
the stator stator
and rotorand rotor
tooth tooth
parts are parts are the high-
the highest.
est. Among the designed machines, the CIM’s averaged core saturation
Among the designed machines, the CIM’s averaged core saturation level is the lowest level is the lowest
whilst
whilst the averaged
averagedcorecoresaturation
saturation of the
of the IPMIPM machine
machine is theishighest.
the highest.

(a) (b)

World Electr. Veh. J. 2022, 13, x FOR PEER REVIEW 7 of 20

(c) (d)

(e) (f)
Figure
Figure 4.
4. Flux
Flux density and flux
density and fluxline
linedistributions
distributionsofof the
the machines.
machines. (a)(a)
IPMIPM machine
machine fluxflux density. (b)
density.
IPM machine
(b) IPM machineflux line.
flux (c)(c)
line. CIMCIMflux
fluxdensity.
density. (d)
(d) CIM fluxline.
CIM flux line.(e)
(e)AIM
AIMflux
flux density.
density. (f) (f)
AIM AIM flux
line.
flux line.

4.4. Torque and Torque Ripple

4.4. Torque and Torque Ripple
The electromagnetic torque of the machines is calculated by using the expressions
givenThe electromagnetic
between (1) and (3). Astorque of the
expressed machines
in (1), is calculated
there are two by using contribut-
variable components the expressions
given
ing thebetween
torque of(1)PMand (3). AsThe
machines. expressed in (1),
first variable there
term aredepends
in (1) two variable components
on the rotor excitationcontrib-
uting the torque of PM machines. The first variable term in (1)
that is the PM flux ψPM which depends on the properties of the PM material includingdepends on thesize,
rotor exci-
tation
maximum that is the PM
energy flux 𝜓BHmax
product which depends
, etc. and on the
the second properties
term, known as of reluctance
the PM material
torque,includ-
ing size, maximum energy product 𝐵𝐻 , etc. and the second term, known as reluctance
depends on the saliency of the rotor which is determined by inductance components Lq
and L . m is the phase number and p is the pole pair number.
torque,d depends on the saliency of the rotor which is determined by inductance compo-
nents 𝐿 and 𝐿 . 𝑚 is the phase number and 𝑝 is the pole pair number.
𝑚
𝑇 = 𝑝 𝜓 𝐼 + 𝐿 −𝐿 𝐼 𝐼 (1)
2
The IPM machine’s electromagnetic torque can be determined by modifying (1) with
flux components as shown in (2). Furthermore, the electromagnetic torque of a squirrel-
World Electr. Veh. J. 2022, 13, 137 7 of 18

 
m 

Tem = p ψPM Iq + Lq − Ld Id Iq  (1)

2 | {z } | {z }
Excitation Reluctance

The IPM machine’s electromagnetic torque can be determined by modifying (1) with
flux components as shown in (2). Furthermore, the electromagnetic torque of a squirrel-cage
IM may be estimated using (3), which was developed for stator flux-oriented IM drives [36].
The superscript “es” in (3) indicates that the quantity is in the synchronous reference frame
oriented to the stator flux.
3

Tem_IPM = p ψd Iq − ψq Id (2)
2
3  
Tem_I M = p ψdes Iqes (3)
2
For IPM machines and IMs, the current angle providing the maximum torque in motor
operation mode has been determined to be 270◦ and 0◦ electrical degrees, respectively.
Figure 5 illustrates the calculated electromagnetic torque waveforms and their harmonic
spectra. As clearly seen in the figure, although all the machines have a similar average
torque, the torque ripple percentage of the AIM is the highest. It is almost 2 times and
2.3 times higher than the CIM and IPM machines, respectively. Therefore, it is clear that
special care should be taken during the design stage of the AIM. Consequently, an effective
torque ripple reduction method involving utilizing u-shaped bridges on the rotor slots
of the AIM is presented in [33]. As for IPM machines, numerous different ways exist to
minimize the torque ripple, such as rotor, flux barrier, and magnet shape optimization,
World Electr. Veh. J. 2022, 13, x FOR PEER REVIEW 8 of 20
rotor skewing, harmonic current injection, stator slot optimization, and proper slot/pole
number combination selection.

250 240 10
THD (%)
230 200 8 IPM 3.76
Amplitude (Nm)
Torque (Nm)

6 CIM 3.15
210 160
AIM 5.85
Ripple (%) 120 4
190 IPM 8.5
80 2
170 CIM 10.15
0
AIM 19.5 40
150 1 3 5 7 9 11 13 15 17 19 21 23 25
0.5 0.52 0.54 0.56 0.58 0.6 0.62 0.64 0
0 2 4 6 8 10 12 14 16 18 20 22 24
Time (s) Harmonic Order
(a) (b)

Figure 5. Comparison of the torque waveforms and their spectra. (a) Electromagnetic torque
Figure 5. Comparison of the torque waveforms and their spectra. (a) Electromagnetic torque wave-
waveforms. (b) Harmonic spectra of torque.
forms. (b) Harmonic spectra of torque.
4.5. Flux-Weakening Characteristics
4.5. Flux-Weakening Characteristics
The machines’ flux-weakening performances were computed using a combined numer-
The machines’ flux-weakening performances were computed using a combined nu-
ical
merical analytical
and and analyticalmethod
methoddescribed
describedinin[9,37].
[9,37].The
Theobtained torque-speedand
obtained torque-speed andpower-
power-speed
curves of the of
speed curves machines are illustrated
the machines are illustratedin in
Figure
Figure6.6.Although
Although IMs IMshave
haveremarkable
remarkable flux-
weakening
flux-weakeningperformance
performance in the constant
in the constanttorque
torque region, theirperformance
region, their performance in the
in the con-constant
power region,
stant power particularly
region, in the
particularly deep
in the deep flux-weakening
flux-weakening area, fallsfar
area, falls farshort
short of of that
that of of IPM
machines [9]. As
IPM machines [9].seen in Figure
As seen 6a, 6a,
in Figure all all
thethe
machines
machines have
havethe
thesame
samemaximum output torque.
maximum output
torque. Although
Although IMs’ characteristics
IMs’ characteristics are poorer
are poorer than than
IPMIPM machine
machine at high-speed
at high-speed region,the max-
region,
the maximum
imum torque/power-speed
torque/power-speed characteristics
characteristics of of
thetheIMs
IMsare
arenot
not poorer
poorerthan
thanIPM IPM ma-machine.
Aschine. As clearly
clearly seen in seen in Figure
Figure 6, in6,some
in some speed
speed regions,IMs
regions, IMsshow
show even
evenbetter
betterflux-weak-
flux-weakening
ening characteristics
characteristics than thethan themachine.
IPM IPM machine. Therefore,
Therefore, in terms
in terms of flux-weakening
of flux-weakening char-
characteristics,
acteristics, the IM can replace IPM with a slight sacrifice of torque in the very deep flux-
the IM can replace IPM with a slight sacrifice of torque in the very deep flux-weakening
weakening region. Moreover, it is also revealed that the AIM has poorer flux-weakening
region. Moreover, it is also revealed that the AIM has poorer flux-weakening characteristics
characteristics than the CIM. The main underlying cause is that due to its lower funda-
mental winding factor, the AIM requires more turns to retain the average torque.

250
Torque (IPM)
200 Torque (CIM)
Torque (AIM)
m)
 IPM machines [9]. As seen in Figure 6a, all the machines have the same maximum output
torque. Although IMs’ characteristics are poorer than IPM machine at high-speed region,
the maximum torque/power-speed characteristics of the IMs are not poorer than IPM ma-
chine. As clearly seen in Figure 6, in some speed regions, IMs show even better flux-weak-
World Electr. Veh. J. 2022, 13, 137
ening characteristics than the IPM machine. Therefore, in terms of flux-weakening 8 of 18
char-
acteristics, the IM can replace IPM with a slight sacrifice of torque in the very deep flux-
weakening region. Moreover, it is also revealed that the AIM has poorer flux-weakening
characteristics thanmain
than the CIM. The the underlying
CIM. The main
cause underlying
is that due tocause is that
its lower due to itswinding
fundamental lower funda-
mental winding factor, the AIM requires more turns to retain
factor, the AIM requires more turns to retain the average torque. the average torque.

250
Torque (IPM)
200 Torque (CIM)
Torque (Nm) Torque (AIM)
150

100

50

0
World Electr. Veh. J. 2022, 13, x FOR PEER REVIEW 9 of 20
0 2 4 6 8 10 12 14 16 18 20
Speed (krpm)
(a)
120
Power (IPM)
100 Power (CIM)
Power (AIM)
80
Power (kW)

60

40

20

0
0 2 4 6 8 10 12 14 16 18 20
Speed (krpm)
(b)
Figure
Figure6.6.Flux-weakening characteristics.
Flux-weakening characteristics. (a) Torque-speed
(a) Torque-speed characteristics.
characteristics. (b) Power-speed
(b) Power-speed characteristics. charac-
teristics.
4.6. Efficiency Maps
Efficiency
4.6. Efficiency maps of the IPM machine and IMs are calculated for the 250Apeak current
Maps
excitation operation considering the phase resistances calculated at 80 °C. In addition,
Efficiency
mechanical lossmaps of the
Pmech and IPM machine
additional losses Pand IMs are calculated for the 250Apeak current
add including the friction, wind, and stray load
excitation operation considering the phase resistances calculated
loss have been taken into account as expressed in (4) and (5) at 80 ℃. In addition, me-
[38], respectively.
chanical loss 𝑃 and additional losses 𝑃 including the friction, wind, and stray
Pmech = as
load loss have been taken into account f 0 + k m2 f 02 in (4) and (5) [38], respectively.
k m1expressed (4)

𝑃 = 𝑘 𝑓f 1 + 𝑘 𝑓In 2
1.5 
(4)
Padd = 0.01Pout1 (5)
fn INL
where k m1 and k m2 are the mechanical loss coefficients, f 0 , f 1 , and f n are the fundamental,
working, and rated frequencies, respectively. Pout 𝑓 1 is. the𝐼 working output power and INL
𝑃 = 0.01𝑃 (5)
and In are the no load and rated current amplitudes,𝑓 respectively.
𝐼
Calculated efficiency maps for the IPM machine, CIM, and AIM are illustrated in
Figure 7. As clearly seen in Figure 7, the maximum efficiency is achieved at a different
where 𝑘 and 𝑘 are the mechanical loss coefficients, 𝑓 , 𝑓 , and 𝑓 are the funda-
speed range for each machine. Moreover, the differences between the highest efficiencies
mental, working, and rated frequencies, respectively. 𝑃 is the working output power
are not significant. While the highest efficiency is achieved between 3–4 krpm for IPM, it is
and 𝐼 and
achieved 𝐼 are4.5–7.5
between the nokrpm
loadfor
and
CIMrated
andcurrent
7–8 krpm amplitudes,
for AIM. Asrespectively.
seen from Figure 7c,
Calculated efficiency maps for the IPM machine,
the AIM has the highest efficiency at the deep flux-weakening CIM, and The
region. AIM are illustrated
following are in
Figure
some of7.the
Asmost
clearly seenmain
notable in Figure 7, the maximum efficiency is achieved at a different
findings:
speed
• range
The IPMfor each machine.
machine Moreover,
and CIM shows the
similar differences in
characteristics between
terms ofthe highestlower
efficiency: efficiencies
are not significant.
efficiency at theWhile
lowestthe
andhighest
highestefficiency is achieved between 3–4 krpm for IPM, it
speed regions;
is achieved between 4.5–7.5 krpm for CIM and 7–8 krpm for AIM. As seen from Figure 7c,
the AIM has the highest efficiency at the deep flux-weakening region. The following are
some of the most notable main findings:
• The IPM machine and CIM shows similar characteristics in terms of efficiency: lower
efficiency at the lowest and highest speed regions;
World Electr. Veh. J. 2022, 13, 137 9 of 18

• The efficiency of AIM at a lower speed is lower than those of the IPM machine and
CIM.
World Electr. Veh. J. 2022, 13, x FOR PEER REVIEW However, its efficiency at a higher speed is higher than those of the IPM
10 of 20 machine
and CIM.

250 96

94

75
92
200
80
88 86 84
90

88
150
Torque (Nm)

92 90

86

84
100
93 94

82
95

80
50
78
96
95 94 93 92 91 90 88 86 76
82
75
0
0 2 4 6 8 10 12 14 16 18 20
Speed (krpm)
(a)

(b)

(c)

Figure 7. Comparison of the efficiency maps of the considered machines. (a) IPM machine. (b) CIM.
(c) AIM.
 The current angle trend increases until peak inverter current limit (250Apeak) and
then starts to decrease with the increasing current. However, for the IM, the slip, i.e., the
difference between the stator and rotor magnetic fields, always increases with increasing
World Electr. Veh. J. 2022, 13, x FOR PEER REVIEW 11 of 20
peak current. Moreover, the AIM’s overall slip, delivering the maximum torque, is lower
than that of the CIM. The electromagnetic torque capability comparison of the machines
World Electr. Veh. J. 2022, 13, 137 is shown in Figure 9. Note that it is assumed that the conductors and isolation materials 10 of 18
Figure 7. Comparison of the efficiency maps of the considered machines. (a) IPM machine. (b) CIM.
can operate safely under the excitation of twice the rated current operation, and the limi-
(c) AIM.
tation of the current density is infinite. That means the thermal issues are ignored. As
4.7. Torque
clearly seenProduction
in Figure 9,Capability
the increase in torque capability of IMs with peak current is much
4.7. Torque Production Capability
fasterTothan
be able to comparemachine.
that of an IPM Furthermore, capabilities,
the torque-generating it is obviousthe from Figure 9ofthat
influence the
electric
torqueTocapability
be able toofcompare
the IPM the torque-generating
machine is better for capabilities,
lower electric the influence
load operations ofthan
electric
the
loading on the torque is investigated. Figure 8 depicts the variation current angle of IPM
loading
rated on the(250A).
current torque is investigated. Figure 8 starts
depicts the variation current angle of IPM
and slip percentage ofOnce the electric
IMs that deliver loading
the maximumto become
torque. Thehigher
current than the
angle rated cur-
delivering
and
rent,slip percentage
the torquetorque of IMs
capability that deliver
of the IMs the maximum torque. The current
◦ angle
◦ delivering
the maximum for various peakbecomes
current better.
changes between 80 to 98 .
the maximum torque for various peak current changes between 80° to 98°.
The current angle trend increases until peak inverter
100
Current Angle (Deg.) 10 current limit (250Apeak) and
then starts to decrease with the increasing current. However, for the IM, the slip, i.e., the
80 between the stator and rotor magnetic fields,8always increases with increasing
difference
peak current. Moreover, the AIM’s overall slip, delivering the maximum torque, is lower
60 6

Slip (%)
than that of the CIM. The electromagnetic torque capability comparison of the machines
40 in Figure 9. Note that it is assumed
is shown 4
that the conductors
IPM (C. Angle) and isolation materials
can operate safely under the excitation of twice the rated current operation, and the limi-
CIM (Slip)
tation20of the current density is infinite. That means the 2 thermal issues are ignored. As
clearly seen in Figure 9, the increase in AIM (Slip)
torque capability of IMs with peak current is much
0 0
faster than50 that
100 of150an 200
IPM250machine.
300 350 Furthermore,
400 450 500 it is obvious from Figure 9 that the
torque capability of the IPM machine
Current (A) is better for lower electric load operations than the
rated current (250A). Once the electric loading starts to become higher than the rated cur-
Figure
rent, 8.
8. Variation
Figurethe of
of current
torque capability
Variation currentofangle
the(for
angle IMsIPM
(for machine)
becomes
IPM and
and slip
better.
machine) slip percentage
percentage (for(for IMs)
IMs) with
with respect
respect to
to
peak
peak current.
current.
100 10
600
(Nm)(Deg.)

The current angle trend increases until peak inverter current limit (250Apeak) and
80
then500
IPM CIM AIM8
starts to decrease with the increasing current. However, for the IM, the slip, i.e., the
difference between the stator and rotor magnetic fields, always increases with increasing
Angle

60 6

Slip (%)
400
peak current. Moreover, the AIM’s overall slip, delivering the maximum torque, is lower
Torque

40
than300
that of the CIM. The electromagnetic torque
IPM (C. 4 comparison of the machines is
capability
Angle)
Current

shown in Figure 9. Note that it is assumed that the conductors and isolation materials can
20 safely under the excitation ofCIM
200
operate
(Slip) 2
twice the rated current operation, and the limitation
100
of the AIM (Slip)
current density is infinite. That means the thermal issues are ignored. As clearly
0 0
seen in Figure 9, the increase in torque capability of IMs with peak current is much faster
0 50 100 150 200 250 300 350 400 450 500
than that of an IPM machine. Furthermore, it is obvious from Figure 9 that the torque
0 50 100 150Current 200 250(A) 300 350 400 450 500
capability of the IPM machine is better for lower electric load operations than the rated
current8.(250A).
Figure Once
Variation of current Current
the electric
angle loading (A)starts
(for IPM to become
machine) and slip higher than(for
percentage theIMs)
rated current,
with respectthe
to
peak current.
torque9.capability
Figure Comparisonof the IMs becomes
of torque production better.
capabilities.

600
The torque ripple percentages for various electric loadings are compared in Figure
IPM CIM AIM
500seen, for the IPM machine, the higher the electric loading, the higher the torque
10. As
Torque (Nm)

ripple. On the other hand, for CIM, the higher the electric loading, the lower the torque
400
ripple. For AIM, the torque ripple increases with the increasing electric loading until the
300
200
100
0
0 50 100 150 200 250 300 350 400 450 500
Current (A)
Figure 9. Comparison
Comparison of torque production capabilities.

The torque ripple

ripple percentages
percentagesfor
forvarious
variouselectric
electricloadings
loadingsareare
compared
compared in in
Figure 10.
Figure
As seen,
10. for the
As seen, for IPM machine,
the IPM the higher
machine, the electric
the higher loading,
the electric the higher
loading, the torque
the higher ripple.
the torque
On the On
ripple. other hand,
the otherfor CIM,for
hand, theCIM,
higher the
the electric
higher theloading,
electricthe lower the
loading, the torque ripple.
lower the For
torque
AIM, the torque ripple increases with the increasing electric loading until the rated
ripple. For AIM, the torque ripple increases with the increasing electric loading until the current
is reached. As previously explained, the electromagnetic torque of PM machines consists
of PM and reluctance torque components. As seen in (1), the PM torque component is
World Electr. Veh. J. 2022, 13, x FOR PEER REVIEW 12 of 20

World Electr. Veh. J. 2022, 13, 137 11 of 18

rated current is reached. As previously explained, the electromagnetic torque of PM ma-
chines consists of PM and reluctance torque components. As seen in (1), the PM torque
component is proportional to the stator current, but the reluctance torque component is
proportional to
proportional to the
the square
stator current, but the
of the stator reluctance
current. torque component
The reluctance is proportional
torque component to
dimin-
the square of the stator current. The reluctance torque component diminishes as the saliency
ishes as the saliency ratio drops with increased electric loading. However, when the elec-
ratio drops with increased electric loading. However, when the electric load increases,
tric load increases, the impact of PMs becomes less prominent (PMs are getting demag-
the impact of PMs becomes less prominent (PMs are getting demagnetized) as seen in
netized) as seen in Figure 11. Furthermore, the saturation caused by PMs may affect the
Figure 11. Furthermore, the saturation caused by PMs may affect the increase in torque
increase in torque of the IPM machine. In fact, the IPM machine has PMs with constant
of the IPM machine. In fact, the IPM machine has PMs with constant flux magnitude.
flux magnitude. Even if the induced voltage is increased in the stator and the flux pro-
Even if the induced voltage is increased in the stator and the flux produced by the stator
duced by the stator windings is increased with increasing excitation current, the flux pro-
windings is increased with increasing excitation current, the flux produced by the PMs
duced by the PMs cannot be increased.
cannot be increased.

25
IPM CIM AIM
Torque Ripple (%)

20

15

10

0
0 50 100 150 200 250 300 350 400 450 500
Current (A)
Figure
Figure 10.
10. Comparison
Comparison of
of torque
torque ripple
ripple percentage for different electric loadings.

Essentially, as clearly seen in Figure 11, the flux generated by PMs dramatically reduces
as the excitation current increases. This figure explains the underlying reason behind the
lower torque generation of the IPM machine under overloading operating conditions. Once
the IPM machine is loaded from 50% to 200%, the flux production capability of PMs reduces
substantially. The flux produced by the stator windings dominates the flux produced by
the PMs. In other words, PMs are temporarily demagnetized by the stator field. Therefore,
since the flux density of the rotor core is much higher than the PM’s flux density, quite a low
flux can be produced by the PMs. Since the excitation and reluctance torque components of
the IPM machine are decreased dramatically with increasing excitation current, it could not
generate torque as high as IM (see Figure 9). For the IPM machine, the total flux is limited
by PMs and saliency. However, for IM, with the increasing injected current, both the stator
and rotor circuit’s fluxes are increased since the rotor bar current will be increased by the
increased excitation current. Therefore, since the total flux is increased, the obtained torque
will also be increased. In theory, in comparison with PM machines, if the current density
and heating issues are ignored, there is no torque limitation for the IMs. In order to estimate
the saturation levels of the machines, the saturation factors for both of the machines have
been calculated by using (6) and the variation of the calculated saturation factors with
respect to electric loading is illustrated in Figure 12.

MMFS + MMFR H + HR
k sat = 1 + = 1+ S (6)
MMFg Hg

where MMFS , MMFR , and MMFg are the magneto-motive force of stator, rotor, and air-gap,
respectively, and HS , HR , and Hg represent the surface integrations of flux intensity of the
same regions. As seen in Figure 12, the saturation levels are similar, and they increase
dramatically with increasing electric loading.
World Electr. Veh. J. 2022, 13, x FOR PEER REVIEW 13 of 20
World Electr. Veh. J. 2022, 13, 137 12 of 18

World Electr. Veh. J. 2022, 13, x FOR PEER REVIEW 14 of 20

limitation for the IMs. In order to estimate the saturation levels of the machines, the satu-
ration factors for both of the machines have been calculated by using (6) and the variation
of the calculated saturation factors with respect to electric loading is illustrated in Figure
12.
𝑀𝑀𝐹 + 𝑀𝑀𝐹 𝐻 +𝐻
𝑘 =1+ =1+ (6)
𝑀𝑀𝐹 𝐻
where 𝑀𝑀𝐹 , 𝑀𝑀𝐹 , and 𝑀𝑀𝐹 are the magneto-motive force of stator, rotor, and air-
gap, respectively, and 𝐻 , 𝐻 , and 𝐻 represent the surface integrations of flux intensity
of the same regions. As seen in Figure 12, the saturation levels are similar, and they in-
crease
Figure 11.dramatically
Figure Flux
11. density
Flux with
vectors
density increasing
of of
vectors PMs electric
forfor
PMs various loading.
electric
various loadings.
electric loadings.

5
Essentially, as clearly seen in Figure 11, the flux generated by PMs dramatically re-
IPM current increases. This figure explains the underlying reason be-
duces as the excitation
4
hind the lower torqueCIM generation of the IPM machine under overloading operating con-
ditions.
3 Once the IPM machine is loaded from 50% to 200%, the flux production capability
AIM
ksat

of PMs reduces substantially. The flux produced by the stator windings dominates the
2
flux produced by the PMs. In other words, PMs are temporarily demagnetized by the
stator field. Therefore, since the flux density of the rotor core is much higher than the PM’s
1
flux density, quite a low flux can be produced by the PMs. Since the excitation and reluc-
tance torque
0 components of the IPM machine are decreased dramatically with increasing
excitation0 current,
50 100 150not200
it could generate
250 torque
300 350 as high
400 as450 IM (see
500 Figure 9). For the IPM
machine, the total flux is limitedCurrentby PMs(A) and saliency. However, for IM, with the increas-
ing injected current, both the stator and rotor circuit’s fluxes are increased since the rotor
Figure
barFigure 12.
12. Variation
current Variation
will of saturation
be increased factor
by the with respect
increased to electric
excitation loading.
current. Therefore, since the total
flux is increased, the obtained torque will also be increased. In theory, in comparison with
PM4.8.machines,
Influence of
if Stack Lenght density and heating issues are ignored, there is no torque
the current
In this section, in order to observe the performance of AIM designed with the same
total axial length of the IPM machine and CIM, AIM1 and AIM2 have been designed, re-
spectively. As presented previously, the AIM was designed with the same stack length as
the Toyota Prius IPM machine. However, as shown in Table 3, once the AIM is designed
with the same stack length as the Toyota Prius IPM machine, its total axial length is ap-
World Electr. Veh. J. 2022, 13, 137 13 of 18

4.8. Influence of Stack Lenght

In this section, in order to observe the performance of AIM designed with the same
total axial length of the IPM machine and CIM, AIM1 and AIM2 have been designed,
respectively. As presented previously, the AIM was designed with the same stack length as
the Toyota Prius IPM machine. However, as shown in Table 3, once the AIM is designed
with the same stack length as the Toyota Prius IPM machine, its total axial length is
approximately 21.5% shorter than that of the IPM machine. In the same manner, it is
25% shorter than the CIM. Therefore, it is intended to investigate the performance of the
proposed AIM, which is designed with increased stack lengths.

Table 3. Overall comparison of key results.

IPM CIM AIM AIM1 AIM2

ns 88 64 84 60 60
` (mm) 50.8 50.8 50.8 72.4 76
` a (mm) 97.5 101.1 76.66 97.5 101.1
R ph (Ω) 0.077 0.0561 0.0577 0.0408 0.0371
Tem (Nm) 222.38 220.75 219.15 218.38 220.45
∆T (%) 8.45 10.15 19.63 19.5 19.8
s (%) 0 3.33 4.4 3.2 3.33
Pout (kW) 34.93 33.52 32.909 33.2 33.48
η (%) 85.532 83.00 79.32 84.21 84.47
MT (kg) 22.7 25.22 24.279 31.02 32.7
Cost (£) 76.7 69.71 68.55 70.77 76.97
World Electr. Veh. J. 2022, 13, x FOR PEER (A/mm2 )
Js REVIEW 21.77 28.52 28.76 23.92 15 21.6
of 20
Jr (A/mm2 ) — 18.17 16.69 13.35 12.65

4.9.
4.9.Comparison
Comparisonof Copper Losses
of Copper Losses
Figure
Figure1313
shows
showsthethe
calculated stator
calculated in-slot
stator winding
in-slot and end-winding,
winding and rotorand
and end-winding, bar rotor bar
copper losses of the considered machines. As seen in Figure 13, almost the same stator slot
copper losses of the considered machines. As seen in Figure 13, almost the same stator slot
copper losses are obtained except for the AIM. As clearly seen from the stator end-wind-
copper losses are obtained except for the AIM. As clearly seen from the stator end-winding
ing copper loss comparison, while the IPM machine, CIM, and AIM have almost the same
copper loss comparison, while the IPM machine, CIM, and AIM have almost the same
value, the AIM1 and AIM2 have almost half the value due to the low number of turns per
value, the AIM1 and
phase requirement. NoteAIM2
that, have
since almost
the stackhalf the value
lengths of thedue
AIM1 to and
the low
AIM2number
have beenof turns per
phase requirement. Note that, since the stack lengths of the AIM1 and AIM2
increased, their stator slot copper losses have also been increased. Therefore, it can be de- have been
duced that due to increased stack length, the number of turns and hence the stator copper it can be
increased, their stator slot copper losses have also been increased. Therefore,
deduced
loss that due
is reduced. The to increased
rotor stackoflength,
copper loss the CIMthe number of turns
is approximately halfand hence
of the AIM’stherotor
stator copper
loss is losses
copper reduced. The of
because rotor copper losslow
the remarkably of the CIM is approximately
(approximately half ofwinding
half of the 24S/8P) the AIM’s rotor
MMF
copperharmonics of 48S/8P
losses because ofwith a 5 slot-pitchlow
the remarkably single-layer winding topology
(approximately (see
half of the Figure winding
24S/8P)
2).
MMF harmonics of 48S/8P with a 5 slot-pitch single-layer winding topology (see Figure 2).

IPM CIM AIM AIM1 AIM2

4
Copper Losses (kW)

0
Stator (slot) Stator (end w.) Rotor
Figure 13. Comparison of copper losses including stator in-slot winding, stator end-windings, and
Figure 13. Comparison of copper losses including stator in-slot winding, stator end-windings, and
rotor bars.
rotor bars.
4.10. Overall Comparison
The electromagnetic performance and design characteristics of the IPM machine,
CIM, AIM, and AIM1 and AIM2, which are designed with different stack lengths, are
listed in Table 3. As can be observed, to be able to sustain the average torque, the number
of turns of AIM1 and AIM2 are reduced while their stack lengths are increased. Therefore,
it can be directly predicted that if the AIM topology is designed to have an equal total
World Electr. Veh. J. 2022, 13, 137 14 of 18

4.10. Overall Comparison

The electromagnetic performance and design characteristics of the IPM machine, CIM,
AIM, and AIM1 and AIM2, which are designed with different stack lengths, are listed in
Table 3. As can be observed, to be able to sustain the average torque, the number of turns
of AIM1 and AIM2 are reduced while their stack lengths are increased. Therefore, it can be
directly predicted that if the AIM topology is designed to have an equal total axial length
with the IPM machine and CIM, the phase resistance would be reduced as shown in Table 3.
As seen, if the AIM was designed with the same stack length as the IPM machine, its total
axial length would be 21.5% and 25% shorter than the IPM machine and CIM, respectively,
without sacrificing torque. On the other hand, if the AIM is designed to have the same total
axial length as the IPM machine, the efficiency of the AIM can be increased by up to 6.5%
without sacrificing the torque and output power. However, when the AIM is configured
with same total axial length as that of the CIM, its efficiency is increased by 1.8% due to the
increased output power.
The masses and the active material costs of each machine are calculated using the
expression specified in (7) and (8) for IM and IPM machines, respectively,
  
bsw

MTot I M = DW330 ( AScore + A Rcore )` + 2Dcus Nc AScoil · ` + τs − + DcuR RA Bar ` + 2A Ring ` Ring (7)
2
  
bsw
MTotPM = DW330 ( AScore + A Rcore )` + 2Dcus Nc ASco0l · ` + τs − + DPM NPM A PM ` (8)
2
where AScore , A Rcore , AScoil , A Bar , A Ring , and A PM indicate surface areas of stator and rotor
cores, stator coils, rotor bars, rotor end rings, and magnets respectively. Nc , NPM , R, and
` Ring are the total coil number, total magnet number, the number of rotor slots, and the
axial length of ring, respectively. DW330 , DCuS , DCuR , and DPM indicate the mass densities
of core, the stator copper, rotor bar copper, and magnet materials, respectively. In addition,
` is the stack length, bsw is the stator slot width, and τs is the average coil pitch. The cost of
the raw materials is calculated using the most recent material price data from [39] and [40],
as shown below:
• Copper: 7.354 £/kg—mass density: 7400 (kg/m3 );
• Steel: 0.44 £/kg—mass density: 8933 (kg/m3 );
• NdFeB35: 42.28 £/kg—mass density: 7520 (kg/m3 ).
As shown in Table 3, raw material costs of all machines are similar. Therefore, consid-
ering the overall performance and flux-weakening characteristics, IMs can easily replace
the expensive IPM machines. It is also revealed that if the stack length is increased, it is
possible to reduce the stator and rotor current densities simultaneously.

5. Conclusions
In this study, the electromagnetic performance and design characteristics of the IPM
machine, CIM, and AIM are quantitatively compared with particular reference to the torque-
generation capability. The key findings of such a comparison have been summarized for
the rated current operation condition and variable electric loading operations as follows:
• The overall flux-weakening characteristic of IMs are comparable to that of IPM machines;
• The flux-weakening characteristic of AIM are poorer than that of CIM;
• The overall efficiency of the IPM machine is higher than the CIM, and the difference
between the maximum efficiency regions is 1.041% only;
• The efficiency of AIM is higher than CIM in deep flux-weakening regions;
• The torque ripple of the AIM is nearly 57% and 50% higher than that of the IPM
machine and CIM, respectively, in the constant torque region;
• By extending the stack length without surpassing the total axial length of an IPM
machine or CIM, it is feasible to significantly improve the output power and efficiency
of AIM;
World Electr. Veh. J. 2022, 13, 137 15 of 18

• It is also possible to reduce both the stator and rotor current densities simultaneously
by extending the stack length;
Moreover, other important key findings related to the torque production capability
and electric-loading operations can be summarized as follows.
The higher the electric loading level:
• The much faster the torque increases for IMs machines, the higher the torque levels
become for the electric loading levels higher than the rated current (250Apeak), whilst
it is vice-versa for the electric loading levels lower than the 250Apeak;
• The lower the torque ripple for the CIM, whilst it is higher for the IPM machine and
the AIM;
• The higher the slip percentage for IMs;
• The higher the risk of demagnetization for the IPM machine.
It has been shown that non-overlapping winding topology is a very effective method,
yielding several advantages, such as shorter axial length without compromising efficiency
and torque, and simplicity in manufacturing. It has also been shown that the main disad-
vantages of the AIM topology are higher torque ripple and higher rotor copper loss due to
very high amplitudes of MMF harmonics.
It has also been demonstrated that because of the AIM topology’s vast design choices,
smaller and relatively efficient IMs may be designed. If improved electromagnetic perfor-
mance and efficiency are more significant design goals, it is possible to fulfill these criteria
by extending the AIM’s stack length without exceeding the entire axial length of the related
CIM design.

Author Contributions: Conceptualization and methodology, T.G. and Z.-Q.Z.; software, T.G.; formal
analysis, T.G., Z.-Q.Z. and C.C.C.; investigation, T.G. and Z.-Q.Z.; resources, Z.-Q.Z.; data curation,
T.G.; writing—original draft preparation, T.G.; writing—review and editing, Z.-Q.Z. and C.C.C.; visu-
alization, T.G.; supervision, Z.-Q.Z. and C.C.C.; project administration, Z.-Q.Z.; funding acquisition,
Z.-Q.Z. All authors have read and agreed to the published version of the manuscript.
Funding: This research was partially funded by Valeo Powertrain Electric Systems, 94017 Créteil
CEDEX, France.
Institutional Review Board Statement: Not applicable.
Informed Consent Statement: Not applicable.
Data Availability Statement: Not applicable.
Conflicts of Interest: The authors declare no conflict of interest.

Abbreviations

2D Two-Dimensional
AIM Advanced Non-overlapping winding Induction Machine
ANW Advanced Non-overlapping Winding
CIM Conventional Induction Machine
CO2 Carbon Dioxide
DC Direct Current
EMF Electromotive Force
EV Electric Vehicle
HEV Hybrid Electric Vehicle
IM Induction Machine
IPM Interior-Permanent Magnet
ISDW Integer Slot Distributed Winding
M Magnet Number
MMF Magnetomotive Force
NdFeB Neodymium Iron Boron
World Electr. Veh. J. 2022, 13, 137 16 of 18

P Pole Number
PM Permanent Magnet
PMSM Permanent Magnet Synchronous Machine
R Rotor Slot Number
S Stator Slot Number
THD Total Harmonic Distortion

Nomenclature

AScoil Surface Area of Stator Slot

DCu R Mass Density of Rotor Bar Copper Material
DCuS Mass Density of Stator Copper Material
` Ring Ring Axial Length
`a Total Axial Length
A Bar Surface Area of Rotor Slot
A PM Surface Area of Magnet
A Rcore Surface Area of Rotor Core Lamination
A Ring Surface Area of Rotor Ring
AScore Surface Area of Stator Core Lamination
DPM Mass Density of Magnet
DW330 Mass Density of Core Material (W330)
HR Surface Integrations of Rotor Flux Intensity
HS Surface Integrations of Stator Flux Intensity
Hg Surface Integrations of Airgap Flux Intensity
INL No Load Current Amplitude
Id D-axis Current
In Rated Current Amplitude
Iq Q-axis Current
Iqes D-Axis Current in Synchronous Reference Frame Oriented to Stator Flux
Jr Rotor Bar Current Density
Js Stator Current Density
Ld D-axis Inductance
Lq Q-axis Inductance
MTot I M Total Weight of IM
MTotPM Total Weight of IPM Machine
MMFR Rotor Magnetomotive Force
MMFS Stator Magnetomotive Force
MMFg Airgap Magnetomotive Force
MT Total Weight
NPM Number of Magnets
Nc Number of Coil
Pout1 Working Output Power
Padd Additional Power Loss
Pmech Mechanical Power Loss
Pout Rated Output Power
R ph Phase Resistance
Tem Electromagnetic Torque
bsw Stator Slot Width
f0 Fundamental Frequency
f1 Working Frequency
fn Rated Frequency
k m1, 2 Mechanical Loss Coefficients
k sat Saturation Factor
ns Serial Turn Number
τs Average Coil Pitch
ψPM Magnet Flux
ψd D-axis Flux
World Electr. Veh. J. 2022, 13, 137 17 of 18

ψdes D-Axis Flux in Synchronous Reference Frame Oriented to The Stator Flux
ψq Q-axis Flux
∆T Torque Ripple
` Stack Length
P Pole Number
R Rotor Slot Number
m Number of Phases
s Slip

References
1. Miller, J.; Khan, T.; Yang, Z.; Sen, A.; Kohli, S. Decarbonizing Road Transport by 2050: Accelerating the Global Transition to Zero-Emission
Vehicles; Briefing Paper; ZEV Transition Council, International Council on Clean Transportation: San Francisco, CA, USA, 2021.
2. Irle, R. Global EV Sales for 2021. Available online: http://www.ev-volumes.com/ (accessed on 18 February 2022).
3. Pontes, J. World Plugin Vehicle Sales-Top Brands. Available online: https://cleantechnica.com/2022/01/31/tesla-1-in-world-ev-
sales-in-2021/ (accessed on 18 February 2022).
4. Zeraoulia, M.; Benbouzid, M.E.H.; Diallo, D. Electric motor drive selection issues for HEV propulsion systems: A comparative
study. IEEE Trans. Veh. Technol. 2006, 55, 1756–1764. [CrossRef]
5. Dorrell, D.G.; Knight, A.M.; Evans, L.; Popescu, M. Analysis and design techniques applied to hybrid vehicle drive
machines—Assessment of alternative IPM and induction motor topologies. IEEE Trans. Ind. Electron. 2012, 59, 3690–3699.
[CrossRef]
6. Goss, J.; Popescu, M.; Staton, D. A comparison of an interior permanent magnet and copper rotor induction motor in a hybrid
electric vehicle application. In Proceedings of the 2013 International Electric Machines & Drives Conference, Chicago, IL, USA,
12–15 May 2013; pp. 220–225. [CrossRef]
7. Boldea, I.; Tutelea, L.N.; Parsa, L.; Dorrell, D. Automotive electric propulsion systems with reduced or no permanent magnets:
An overview. IEEE Trans. Ind. Electron. 2014, 61, 5696–5711. [CrossRef]
8. Yang, Z.; Shang, F.; Brown, I.P.; Krishnamurthy, M. Comparative Study of Interior Permanent Magnet, Induction, and Switched
Reluctance Motor Drives for EV and HEV Applications. IEEE Trans. Transp. Electrif. 2015, 1, 245–254. [CrossRef]
9. Guan, Y.; Zhu, Z.Q.; Afinowi, I.A.A.; Mipo, J.C.; Farah, P. Comparison between induction machine and interior permanent
magnet machine for electric vehicle application. COMPEL Int. J. Comput. Math. Electr. Electron. Eng. 2016, 35, 572–585. [CrossRef]
10. Yang, Y.; Castano, S.M.; Yang, R.; Kasprzak, M.; Bilgin, B.; Sathyan, A.; Dadkhah, H.; Emadi, A. Design and comparison of interior
permanent magnet motor topologies for traction applications. IEEE Trans. Transp. Electrif. 2017, 3, 86–97. [CrossRef]
11. El-Refaie, A.M. Motors/generators for traction/propulsion applications: A review. IEEE Veh. Technol. Mag. 2013, 8, 90–99.
[CrossRef]
12. Zhu, Z.Q.; Howe, D. Electrical machines and drives for electric, hybrid, and fuel cell vehicles. Proc. IEEE 2007, 95, 746–765.
[CrossRef]
13. Zhu, Z.Q.; Chan, C.C. Electrical machine topologies and technologies for electric, hybrid, and fuel cell vehicles. In Proceedings of
the 2008 IEEE Vehicle Power and Propulsion Conference, Harbin, China, 3–5 September 2008; pp. 1–6. [CrossRef]
14. Jiang, Y.; Krishnamurthy, M. Performance evaluation of ac machines for propulsion in a range extended electric auto rickshaw. In
Proceedings of the 2012 IEEE Transportation Electrification Conference and Expo (ITEC), Dearborn, MI, USA, 17–20 June 2012;
pp. 1–6. [CrossRef]
15. Wu, S.; Tian, L.; Cui, S. A Comparative study of the interior permanent magnet electrical machine’s rotor configurations for
a single shaft hybrid electric bus. In Proceedings of the 2008 IEEE Vehicle Power and Propulsion Conference, Harbin, China,
3–5 September 2008; pp. 1–4. [CrossRef]
16. Wang, A.; Jia, Y.; Soong, W.L. Comparison of five topologies for an interior permanent-magnet machine for a hybrid electric
vehicle. IEEE Trans. Magn. 2011, 47, 3606–3609. [CrossRef]
17. Liu, X.; Chen, H.; Zhao, J.; Belahcen, A. Research on the performances and parameters of interior PMSM used for electric vehicles.
IEEE Trans. Ind. Electron. 2016, 63, 3533–3545. [CrossRef]
18. Yamazaki, K.; Kumagai, M. Torque analysis of interior permanent-magnet synchronous motors by considering cross-
magnetization: Variation in torque components with permanent-magnet configurations. IEEE Trans. Ind. Electron. 2014, 61,
3192–3201. [CrossRef]
19. Bucherl, D.; Nuscheler, R.; Meyer, W.; Herzog, H.-G. Comparison of electrical machine types in hybrid drive trains: Induction
machine vs. permanent magnet synchronous machine. In Proceedings of the 2008 18th International Conference on Electrical
Machines, Vilamoura, Portugal, 6–9 September 2008; pp. 1–6. [CrossRef]
20. Zhu, Z.Q.; Chu, W.Q.; Guan, Y. quantitative comparison of electromagnetic performance of electrical machines for HEVs/EVs.
CES Trans. Electr. Mach. Syst. 2017, 1, 37–47. [CrossRef]
21. Wang, S.; Zhu, Z.; Pride, A.; Shi, J.; Deodhar, R.; Umemura, C. Comparison of different winding configurations for dual
three-phase interior pm machines in electric vehicles. World Electr. Veh. J. 2022, 13, 51. [CrossRef]
22. Aiso, K.; Akatsu, K. Performance comparison of high-speed motors for electric vehicle. World Electr. Veh. J. 2022, 13, 57. [CrossRef]
World Electr. Veh. J. 2022, 13, 137 18 of 18

23. Dmitrievskii, V.; Prakht, V.; Kazakbaev, V.; Anuchin, A. Comparison of interior permanent magnet and synchronous homopolar
motors for a mining dump truck traction drive operated in wide constant power speed range. Mathematics 2022, 10, 1581.
[CrossRef]
24. Orecchini, F.; Santiangeli, A.; Zuccari, F.; Alessandrini, A.; Cignini, F.; Ortenzi, F. Real drive truth test of the toyota yaris hybrid
2020 and energy analysis comparison with the 2017 model. Energies 2021, 14, 8032. [CrossRef]
25. Gronwald, P.-O.; Kern, T.A. Experimental validation and parameter study of a 2d geometry-based, flexible designed thermal
motor model for different cooled traction motor drives. World Electr. Veh. J. 2021, 12, 76. [CrossRef]
26. Vosswinkel, M.; Lohner, A.; Platte, V.; Hirche, T. Design, Production, and verification of a switched-reluctance wheel hub drive
train for battery electric vehicles. World Electr. Veh. J. 2019, 10, 82. [CrossRef]
27. Agamloh, E.; von Jouanne, A.; Yokochi, A. An overview of electric machine trends in modern electric vehicles. Machines 2020,
8, 20. [CrossRef]
28. He, T.; Zhu, Z.; Eastham, F.; Wang, Y.; Bin, H.; Wu, D.; Gong, L.; Chen, J. Permanent magnet machines for high-speed applications.
World Electr. Veh. J. 2022, 13, 18. [CrossRef]
29. Yun, D.; Kim, N.; Hyun, D.; Baek, J. Torque improvement of six-phase permanent-magnet synchronous machine drive with
fifth-harmonic current injection for electric vehicles. Energies 2022, 15, 3122. [CrossRef]
30. Li, C.; Guo, X.; Fu, J.; Fu, W.; Liu, Y.; Chen, H.; Wang, R.; Li, Z. Design and analysis of a novel double-stator double-rotor motor
drive system for in-wheel direct drive of electric vehicles. Machines 2022, 10, 27. [CrossRef]
31. Gundogdu, T.; Zhu, Z.-Q.; Mipo, J.-C. Design and analysis of advanced nonoverlapping winding induction machines for EV/HEV
applications. Energies 2021, 14, 6849. [CrossRef]
32. Olszewski, M. Evaluation of the 2010 Toyota Prius Hybrid Synergy Drive System; Oak Ridge National Laboratory: Oak Ridge, TN,
USA; United States Department of Energy: Washington, DC, USA, 2011.
33. Gundogdu, T.; Zhu, Z.Q.; Mipo, J.C. Optimization and improvement of advanced nonoverlapping induction machines for
EVs/HEVs. IEEE Access 2022, 10, 13329–13353. [CrossRef]
34. Kirtley, J.L.; Cowie, J.G.; Brush, E.F.; Peters, D.T.; Kimmich, R. Improving induction motor efficiency with die-cast copper rotor
cages. In Proceedings of the 2007 IEEE Power Engineering Society General Meeting, Tampa, FL, USA, 24–28 June 2007; pp. 1–6.
[CrossRef]
35. Vas, P. Vector Control of AC Machines; Clarendon Press: Oxford, UK; Springer: Berlin/Heidelberg, Germany, 1990; pp. 124–130.
[CrossRef]
36. Xu, X.; De Doncker, R.; Novotny, D.W. A stator flux oriented induction machine drive. In Proceedings of the PESC’88 Record.,
19th Annual IEEE Power Electronics Specialists Conference, Kyoto, Japan, 11–14 April 1988; Volume 2, pp. 870–876. [CrossRef]
37. Shin, M.-H.; Hyun, D.-S.; Cho, S.-B. Maximum torque control of stator-flux-oriented induction machine drive in the field-
weakening region. IEEE Trans. Ind. Appl. 2002, 38, 117–122. [CrossRef]
38. Pyrhönen, J.; Jokinen, T.; Hrabovcová, V. Design of Rotating Electrical Machines, 2nd ed.; John Wiley & Sons, Ltd.: West Sussex, UK,
2014; pp. 523–534; ISBN 978-1-118-58157-5.
39. Material Market Priced for 8 August 2017. Available online: https://www.lme.com/ (accessed on 22 February 2022).
40. NdFeB Sintered Rough 35H EXW Magnet. Available online: https://ise-metal-quotes.com/ (accessed on 22 March 2022).