Ahmad S.

Al-Adsani
Omid Beik

Multiphase
Hybrid
Electric
Machines
Applications for Electrified Powertrains
Multiphase Hybrid Electric Machines
Ahmad S. Al-Adsani • Omid Beik

Multiphase Hybrid Electric

Machines
Applications for Electrified Powertrains
Ahmad S. Al-Adsani Omid Beik
Department of Electrical Engineering Department of Electrical and Computer
Technology, Public Authority for Applied Engineering
Education and Training (PAAET) McMaster University
College of Technological Studies (CTS) Hamilton, ON, Canada
Kuwait City, Kuwait

ISBN 978-3-030-80434-3 ISBN 978-3-030-80435-0 (eBook)

https://doi.org/10.1007/978-3-030-80435-0

© Springer Nature Switzerland AG 2022

This work is subject to copyright. All rights are reserved by the Publisher, whether the whole or part of the
material is concerned, specifically the rights of translation, reprinting, reuse of illustrations, recitation,
broadcasting, reproduction on microfilms or in any other physical way, and transmission or information
storage and retrieval, electronic adaptation, computer software, or by similar or dissimilar methodology
now known or hereafter developed.
The use of general descriptive names, registered names, trademarks, service marks, etc. in this publication
does not imply, even in the absence of a specific statement, that such names are exempt from the relevant
protective laws and regulations and therefore free for general use.
The publisher, the authors, and the editors are safe to assume that the advice and information in this
book are believed to be true and accurate at the date of publication. Neither the publisher nor the authors or
the editors give a warranty, expressed or implied, with respect to the material contained herein or for any
errors or omissions that may have been made. The publisher remains neutral with regard to jurisdictional
claims in published maps and institutional affiliations.

This Springer imprint is published by the registered company Springer Nature Switzerland AG
The registered company address is: Gewerbestrasse 11, 6330 Cham, Switzerland
Preface

Development of road vehicles through electrified powertrains without compromising

their power capability, efficiency, performance, reliability, safety, driving range, and
cost has been the center of attention in academia and industry. This book is a user-
friendly reference and attractive subject for researchers and undergraduate and
graduate engineers who are interested in multiphase, permanent magnet and hybrid
machine topologies with a specific application in electric and hybrid electric
vehicles.
This book begins with an overview and principals of classical electric machine
operation, magnetic modeling, and characteristics of materials. Various classical
electric machine topologies, including brushed DC, and different AC machines are
discussed in Chap. 1.
Chapter 2 discusses hybrid electric machine concept together with a review of
different hybrid machine topologies, indicating their operational philosophy, advan-
tages, and disadvantages.
Chapter 3 presents a hybrid permanent magnet (HPM) machine topology that is
selected and analyzed in terms of its geometry, excitation field technique, back-
EMF, and developed electromagnetic torque for both no-load and on-load opera-
tions. The HPM topology consists of two different synchronous machines, a perma-
nent magnet (PM) and a wound field (WF) machine that are coupled on the same
rotor shaft, rotate with the same speed, and share the same multiphase stator.
Chapter 4 discusses an overview of multiphase electric machines. AC and
rectified DC output voltage waveforms of three- and nine-phase systems with
associated power electronics are presented. A comparison of three- and two nine-
phase machine winding, rectification characteristics, and losses for both HPM and
PM machine topologies is presented in Chap. 4.
Chapter 5 presents an overview of electric vehicles (EVs) and hybrid electric
vehicles (HEVs), their powertrains, and on-board energy sources. Different battery
technologies are discussed, and in the case of HEVs the feasibility of disconnecting
the internal combustion engine (ICE) from the electric drivetrain is studied. Driving
range, fuel economy, and emissions are evaluated over different driving cycles, and
at different vehicle powertrain hybridization ratios (HR) in Chap. 5.
v
vi Preface

In Chap. 6, a dynamic model of vehicle powertrain that includes a HPM generator

integrated into an ICE in an SHEV while considering a load demand is presented.
The ICE/HPM generator output power control scheme is modeled while maintaining
ICE efficiency within its optimal region. Several operating scenarios for the HPM
generator excitation scheme are assessed, and the HPM generator is characterized
utilizing a 32-phase brushless excitation scheme. In addition, different cases, such as
normal, boost, and buck functionality of HPM machine operation, are analyzed, and
a choice of the most appropriate operation mode has been selected to regulate the
total back-EMF via a WF excitation current control.
Dr. Al-Adsani wishes to express his sincere gratitude toward his wife, and the
authors extend special thanks to Dr. Nigel Schofield at the University of Hudders-
field for his valuable inputs and to the team at Springer for their care during the book
production.

Kuwait City, Kuwait Ahmad S. Al-Adsani

Toronto, ON, Canada Omid Beik
Contents

1 General Electric Machine Theory . . . . . . . . . . . . . . . . . . . . . . . . . . . 1

1.1 Magnetic Circuit Principles . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1
1.1.1 Magnetic Field Distribution and Flux Density . . . . . . . . . 1
1.1.2 Ferromagnetic Materials and Magnetization Curves . . . . . 3
1.2 Electric Machine Fundamentals . . . . . . . . . . . . . . . . . . . . . . . . . . 5
1.3 Overview of Classical Electric Machine Topologies . . . . . . . . . . . 6
1.3.1 Brushed DC and AC Machines . . . . . . . . . . . . . . . . . . . . 7
1.3.2 Brushless AC Machines . . . . . . . . . . . . . . . . . . . . . . . . . 10
1.3.3 Switch Reluctance Machines . . . . . . . . . . . . . . . . . . . . . 11
1.4 WF and PM Synchronous Machine Excitation Fields . . . . . . . . . . 13
1.4.1 Magnetic Flux Path Representation of WF
Synchronous Machines . . . . . . . . . . . . . . . . . . . . . . . . . . 13
1.4.2 Magnetic Flux Path Representation of PM
Synchronous Machines . . . . . . . . . . . . . . . . . . . . . . . . . . 14
2 Hybrid Electric Machine Concept . . . . . . . . . . . . . . . . . . . . . . . . . . . 17
2.1 History . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 17
2.2 Hybrid Electric Machine Classification . . . . . . . . . . . . . . . . . . . . . 17
2.3 Different Hybrid Machine Topologies . . . . . . . . . . . . . . . . . . . . . 18
2.3.1 PM Synchronous Machine with Claw Pole
Field Excitation (PSCPF) . . . . . . . . . . . . . . . . . . . . . . . . 18
2.3.2 Toroidal-Stator Transverse-Flux Machine (TSTFM) . . . . . 21
2.3.3 Hybrid Excitation Synchronous Machine (HESM) . . . . . . 23
2.3.4 Synchronous Permanent Magnet Hybrid AC
Machine (SynPM) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 26
2.3.5 Consequent Pole Permanent Magnet Hybrid
Excitation Machine (CPPM) . . . . . . . . . . . . . . . . . . . . . . 27
2.3.6 Field Controlled Torus-NS (FCT-NS) Machine . . . . . . . . 29
2.3.7 Dual-Rotor Machine . . . . . . . . . . . . . . . . . . . . . . . . . . . 32
2.3.8 Imbricated Hybrid Excitation Machine (IHEM) . . . . . . . . 33

vii
viii Contents

2.3.9 Series Double Excited Synchronous Machine

(SDESM) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 37
2.3.10 Switch Reluctance Machine with Stator Field
Assistance . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 38
2.3.11 Dual-Stator Hybrid Excited Synchronous
Wind Generator (DSHESG) . . . . . . . . . . . . . . . . . . . . . . 39
2.4 Summary of Surveyed Literature on HPM Machines . . . . . . . . . . . 43
3 Hybrid Permanent Magnet Machine Design . . . . . . . . . . . . . . . . . . . 45
3.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 45
3.2 Machine Volume Envelope Consideration . . . . . . . . . . . . . . . . . . 45
3.2.1 PM Machine Dimensions . . . . . . . . . . . . . . . . . . . . . . . . 47
3.2.2 PM Machine Stator Winding Layout . . . . . . . . . . . . . . . . 47
3.2.3 Stator Winding Fill Factor and Resistance . . . . . . . . . . . . 50
3.2.4 Finite Element Method Program . . . . . . . . . . . . . . . . . . . 52
3.2.5 Machine Back-EMF Prediction . . . . . . . . . . . . . . . . . . . . 55
3.2.6 PM Machine Analysis Via EMC Model . . . . . . . . . . . . . 56
3.3 WF Machine . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 61
3.3.1 WF Rotor Design . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 62
3.3.2 WF to PM Split Ratio . . . . . . . . . . . . . . . . . . . . . . . . . . 64
3.3.3 Comparative Analysis of WF Rotor Designs . . . . . . . . . . 65
3.4 HPM Machine Parameters . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 69
3.4.1 Torque Prediction and Saturation . . . . . . . . . . . . . . . . . . 69
3.4.2 Synchronous Inductance and Winding Resistance . . . . . . 73
3.5 HPM Machine Final Design Model Analysis . . . . . . . . . . . . . . . . 75
3.5.1 Rotor PM Demagnetization . . . . . . . . . . . . . . . . . . . . . . 75
3.5.2 Core Loss Prediction . . . . . . . . . . . . . . . . . . . . . . . . . . . 77
3.6 HPM Machine Thermal Model . . . . . . . . . . . . . . . . . . . . . . . . . . 81
3.6.1 General Principle of the Lumped Parameter Method . . . . . 81
3.6.2 Conduction Heat Transfer . . . . . . . . . . . . . . . . . . . . . . . . 82
3.6.3 Convection Heat Transfer . . . . . . . . . . . . . . . . . . . . . . . . 84
3.6.4 Radiation Heat Transfer . . . . . . . . . . . . . . . . . . . . . . . . . 86
3.6.5 HPM Machine Thermal Model . . . . . . . . . . . . . . . . . . . . 86
3.7 Comparison Between PM and Four HPM Machine
Topologies . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 90
3.8 Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 94
4 Multiphase HPM Generator Systems . . . . . . . . . . . . . . . . . . . . . . . . 95
4.1 Overview on Multiphase Machines . . . . . . . . . . . . . . . . . . . . . . . 95
4.1.1 Multiphase Windings Principles . . . . . . . . . . . . . . . . . . . 96
4.1.2 Rectified Voltage due to Three- and Nine-Phase
HPM Generator Systems . . . . . . . . . . . . . . . . . . . . . . . . 96
4.2 Nine-Phase HPM Generator Parameters . . . . . . . . . . . . . . . . . . . . 101
4.2.1 Nine-Phase Winding Layout and Back-EMF . . . . . . . . . . 101
4.2.2 Back-EMF and Torque Waveform Harmonics
Prediction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 103
Contents ix

4.2.3 Synchronous Inductance Prediction . . . . . . . . . . . . . . . . . 107

4.2.4 Construction of HPM Machines Prototype . . . . . . . . . . . . 111
4.2.5 Resistance and Inductance Measurements . . . . . . . . . . . . 114
4.3 Analysis Models . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 115
4.3.1 General dq Mathematical Model of HPM Generator . . . . . 115
4.3.2 Simulation Model . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 118
4.4 Three- and Nine-Phase HPM Generator System Studies . . . . . . . . 119
4.4.1 Impact on Synchronous Inductance and Rectifier . . . . . . . 121
4.4.2 System Sensitivity to Generator Synchronous
Inductance . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 126
4.4.3 DC-Link Voltage Quality . . . . . . . . . . . . . . . . . . . . . . . . 130
4.5 Loss Audit of Generator Systems . . . . . . . . . . . . . . . . . . . . . . . . . 134
4.5.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 134
4.5.2 Core Loss Prediction . . . . . . . . . . . . . . . . . . . . . . . . . . . 136
4.5.3 Passive and Active Converter Loss for HPM
and PM Generator Systems . . . . . . . . . . . . . . . . . . . . . . . 136
4.6 Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 141
5 Electric and Hybrid Electric Powertrains . . . . . . . . . . . . . . . . . . . . . 143
5.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 143
5.2 Overview of EVs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 143
5.2.1 EV Powertrain Configuration . . . . . . . . . . . . . . . . . . . . . 145
5.2.2 Battery Technology . . . . . . . . . . . . . . . . . . . . . . . . . . . . 146
5.3 Overview of HEVs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 146
5.3.1 HEV Powertrain Configurations . . . . . . . . . . . . . . . . . . . 148
5.4 Vehicle Driving Cycles . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 151
5.5 Series Hybrid Electric Vehicle (SHEV) . . . . . . . . . . . . . . . . . . . . 151
5.5.1 ZEBRA Battery . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 154
5.5.2 Internal Combustion Engine (ICE) . . . . . . . . . . . . . . . . . 156
5.5.3 Engine-Mounted Multiphase HPM Generator . . . . . . . . . . 159
5.6 Electric Vehicle Range Extender . . . . . . . . . . . . . . . . . . . . . . . . . 159
5.6.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 159
5.6.2 Literature Review of EV Range Extender Studies . . . . . . . 160
5.7 ICE/HPM Generator Range Extender in SHEVs . . . . . . . . . . . . . . 162
5.7.1 Vehicle Traction Machine Torque . . . . . . . . . . . . . . . . . . 163
5.7.2 Hybridization Ratio . . . . . . . . . . . . . . . . . . . . . . . . . . . . 165
5.7.3 Range Extender Sizing in SHEV Powertrain . . . . . . . . . . 165
5.7.4 Study Cases . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 169
5.8 Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 170
6 Operation and Characterization of Multiphase HPM
Generator in SHEV Powertrain . . . . . . . . . . . . . . . . . . . . . . . . . . . . 171
6.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 171
6.2 HPM Machine Back-EMF Control Strategy . . . . . . . . . . . . . . . . . 173
6.2.1 Control Strategy Analysis . . . . . . . . . . . . . . . . . . . . . . . . 174
6.2.2 DC-link Design Options . . . . . . . . . . . . . . . . . . . . . . . . . 178
x Contents

6.3 HPM Machine Output Power Control . . . . . . . . . . . . . . . . . . . . . 181

6.3.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 181
6.3.2 HPM Generator Operating Scenarios . . . . . . . . . . . . . . . . 184
6.3.3 Energy Loss Prediction for Two Driving Cycles . . . . . . . . 187
6.3.4 Solving Final Choice with Full Simulation Model . . . . . . 189
6.3.5 Thermal Analysis Results of the Investigated
HPM Machine . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 191
6.4 HPM Machine Characterization Using Brushless Excitor . . . . . . . . 194
6.4.1 32-Phase Brushless Excitation Scheme . . . . . . . . . . . . . . 197
6.4.2 Performance Curves . . . . . . . . . . . . . . . . . . . . . . . . . . . . 200
6.4.3 Efficiency Mapping . . . . . . . . . . . . . . . . . . . . . . . . . . . . 204
6.5 Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 206

References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 207
Index . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 221
Chapter 1
General Electric Machine Theory

1.1 Magnetic Circuit Principles

In electric machines, a magnetism phenomenon is utilized to build an electromotive

force (EMF) to supply an electric load, as in generators, or to drive a mechanical load
as in motors. In electric devices, four basic principles can describe how magnetic
field is used [1]:
(i) In a current-carrying conductor, a magnetic field is produced around that
conductor.
(ii) In transformer action, a time changing magnetic field induces a voltage in a coil
if it passes on it.
(iii) In generator action, a moving conductor in the presence of a magnetic field
induces a voltage and hence current flows through that conductor.
(iv) In motor action, a current-carrying conductor in the presence of a magnetic field
has an electromotive force induced on it.

1.1.1 Magnetic Field Distribution and Flux Density

The magnetic field that is created by current-carrying conductors based on Ampere’s

right-hand rule, as in Fig. 1.1a, shows a right hand with the thumb pointing in the
direction of current flow, while the magnetic field is rotating in the direction of the
other fingers. Notice in Fig 1.1b that the symbol ⨂ denotes a cross-sectional view of
the conductor carrying the current into the paper, while the symbol ⦿ denotes the
current flow out of the paper. A magnetic field intensity (H ) is characterized as an
effort a current is putting into establishing a magnetic field. The field intensity due to
excitation DC current (I), which passes in a coil with (N ) turns through magnetic
circuit path length (Lc), is calculated in (1.1) [2]. The strength of a magnetic field flux
density (B) is governed by H and core material, as in (1.2). μ represents the magnetic

© Springer Nature Switzerland AG 2022 1

A. S. Al-Adsani, O. Beik, Multiphase Hybrid Electric Machines,
https://doi.org/10.1007/978-3-030-80435-0_1
2 1 General Electric Machine Theory

Fig. 1.1 Magnetic field

direction due to a current-
carrying conductor coil (a)
Current-carrying conductor
coil and right-hand rule (b)
Magnetic flux direction for
in and out of the page
current directions

permeability of the material. A convenient way of representing the magnetizability

of materials is by comparing material permeability to the permeability of free space
(μo ¼ 4π  107 H/m), which is referred to as relative permeability (μr). Note in the
magnetic circuit that the flux density is governed by the flux (φ) and the cross-
sectional area (A) of the medium that it is passing through, as in (1.4).

NI
H¼ ð1:1Þ
Lc
B ¼ μH ð1:2Þ
μ
μr ¼ ð1:3Þ
μo
φ
B¼ ð1:4Þ
A

By considering the simple core shape in Fig. 1.1a, the magnetic path in the core
has a uniform shape, which has a reluctance value that depends on the path length,
core permeability, and cross-sectional area as in (1.5). The reluctance in the magnetic
circuit is like the resistance in the electric circuit, where one governs the flow of flux
and the flow of current, respectively, as in Fig. 1.2. In the magnetic circuit, the coil
has NI quantity that is called magnetomotive force (MMF). However, for permanent
magnet (PM) materials, the MMF is calculated as in (1.7), where Hc is the PM core
field intensity and Lc is the PM core length (thickness).
1.1 Magnetic Circuit Principles 3

Fig. 1.2 Electric and its magnetic circuit analogy (a) Electric circuit (b) Magnetic circuit

L
R ¼ ð1:5Þ
μA
MMF ¼ NI ð1:6Þ
MMF ¼ H c Lc ð1:7Þ

1.1.2 Ferromagnetic Materials and Magnetization Curves

When the magnetic circuit is divided into sections of materials that easily allow flux
line path to be formed, this is called ferromagnetic materials (FERMMs). In
FERMMs, there are atoms, and each atom has its own magnetic moment direction,
which is separated by a domain wall in each crystal boundary, as in Fig. 1.3a. These
magnetic moments tend to align in the same direction over domains containing many
atoms when they are subjected to a magnetic field intensity [2]. As H increases
further and further, more domain directions will align until all the domains are in the
same direction and when the material is magnetized to the maximum extent (satu-
ration region), as shown in Fig. 1.3a. Here, if the majority of the domains are in the
same directions after the applied field is removed, the material is said to be perma-
nently magnetized. Another important phenomenon that occurs in FERMMs is
called hysteresis. Hysteresis is described by referring to a typical B–H curve in
Fig. 1.3a. When a current flows through a coil warped around a ferromagnetic core,
MMF will then be created. As the MMF increases, so does H until the core saturates,
which is presented by point o to point a in Fig. 1.3a. Now, if the current decreased to
zero, the MMF and hence field intensity will go to zero. However, flux density will
not go to zero, which is presented by point a to point b in Fig. 1.3a. Here, the core
remains magnetized, even though the applied current and field intensity have gone to
zero. The magnetism that remains in the core is called residual magnetism, and this
effect creates a permanent magnet. If the applied current is reversed and slowly
increases in the negative direction, the flux density will be driven to zero, as
presented by point b to c in Fig. 1.3a. The negative field intensity needed to drive
B to zero is called coercive force, as presented by point c in Fig. 1.3a. As the current
4 1 General Electric Machine Theory

Fig. 1.3 General hysteresis loops for FERMMs (a) B versus H hysteresis loop showing path major
points (b) Soft ferromagnetic materials (c) Hard ferromagnetic materials

is made more negative, the core will eventually saturate and the flux density will
have a polarity opposite to that in the original case, as presented by point d in
Fig. 1.3a. Finally, if the current is reduced to zero and then made positive again, the
curve will join up with the original curve passing through points e and f [2], where
this closed loop joining points a, b, c, d, e, and f is called a hysteresis loop.
Thus, a PM material is typically a metal alloy, which after being subjected to field
intensity retains a substantial residual flux density (Br). In order to reduce the flux
density to zero, an H field direction opposite in sense to the original magnetizing
field must be applied. This impressed field magnitude must have a value (Hc) known
as the coercive force.
Here, materials magnetism can be categorized as permanent or temporary based
on their ability to be magnetized and hold their magnetism or their magnetism
1.2 Electric Machine Fundamentals 5

Fig. 1.4 Different cuts of PM materials available in the market [3] (a) Ferrite (b) Neodymium–
Iron–Boron (NdFeB) (c) Samarium–Cobalt (SmCo) (d) Aluminum–Nickel–Cobalt (AlNiCo)
(Magnetic materials: Goudsmit Magnetics, the Netherlands)

Fig. 1.5 Hard and soft

permanent magnet
demagnetization curves

vanishes as the DC supply source of the conductor that carries the current is turned
off [1]. Magnetic materials are relatively easy to magnetize since their relative
permeability values are high. FERMMs are classified as soft, in which the most
common magnetic materials include steels, iron, nickel, aluminum, cobalt, and rare-
earth elements. Fig. 1.3b illustrates the expected hysteresis loop behavior for the soft
FERMMs. Hard FERMMs, which have the expected hysteresis loop as in Fig. 1.3c,
comprise the permanent magnet materials such as alnico, the alloys of cobalt with
rare-earth elements such as samarium, copper–nickel alloys, chromium steels, and
other metal alloys. Fig. 1.4 shows different permanent magnet materials with special
cuts found in the market. In Fig. 1.5, the B–H demagnetization curves for several soft
and hard magnetic materials with different grades are illustrated. The PM material
grade is a number, which is specified after material type, to show different curves for
the same material based on cost, magnetic performance, and operational temperature
resistance.

1.2 Electric Machine Fundamentals

Electric machines are considered electromechanical power converters, such that they
convert mechanical power into electrical, as in generators, and convert electrical into
mechanical energy, as in motors. For generators, a source of mechanical power is
6 1 General Electric Machine Theory

Fig. 1.6 Fundamental

concepts associated with
electrical/mechanical
systems

required to rotate the machine shaft (prime mover), via applied torque (T ), at a fixed
or variable speed (ω) to develop an electromotive force (voltage difference) at the
machine terminals (v) and vice versa for motor action as illustrated in Fig. 1.6.
Hence, the electric field is considered the coupled medium between generation and
motor output quantities. The electric machines differ in their internal magnetic
source type, construction, and operation. In this chapter, the source of the magnetic
field in general electric machines and their stator and rotor geometry are discussed.

1.3 Overview of Classical Electric Machine Topologies

Electric machines consist of two major parts: stator and rotor. Stator is the stationary
part that does not move during operation, while the rotor is free to move and it can be
the inner or outer part of the machine. Both stator and rotor parts are made of
FERMMs that are discussed in the previous section. The stator accommodates the
alternating current (AC) conductors in slots that are cut on the inner periphery and in
some machines topologies in the outer periphery of the rotor structure. The coupling
between stator and rotor fields can be increased by selecting a low reluctance
material, which increases the flux density through the machine’s active parts. The
classification of various electrical machine topologies focuses on the machines with
or without commutators together with synchronous and asynchronous AC machines
types, as shown in Fig. 1.7. The utilization of these machine topologies in the
industry is subjected to meet the designer target in terms of efficiency, power
density, and cost while their usage ranges from light- to heavy-duty loads.
1.3 Overview of Classical Electric Machine Topologies 7

Fig. 1.7 Classical electric machine common classifications

1.3.1 Brushed DC and AC Machines

Direct current (DC) machines have essential features that made them continually find
application because of the relative simplicity and flexibility of their drive systems
compared with AC machines. In brushed DC machine topology, as in Fig. 1.8,
having a higher number of stator salient poles causes core saturation; hence, two,
four, and six poles are common. Their field winding is placed on the salient stator,
and the armature winding is placed on the round rotor. Through the field winding, a
DC current is applied to produce the flux, which presents the major component in the
general induced voltage formula. Such that, the induced voltage (e) in a conductor of
length (l) moving with linear velocity (v) in a non-time-varying magnetic flux
density is given in (1.8) [1, 4]. A unidirectional terminal voltage can be applied
through a brush and mechanical commutator assembly. For a single coil in DC
machines, a commutator action is to provide a full-wave rectification, and by
8 1 General Electric Machine Theory

Fig. 1.8 Radial view for two and four salient pole brushed DC machines topology. (a) Two salient
pole (b) Four salient pole

assuming sinusoidal flux distribution, the voltage waveform between brushes can be
transformed to a DC or average voltage (Ea) value between brushes as in (1.9) [4],
where ω represents the machine rotational speed.

e ¼ Blv ð1:8Þ
Z π
1 2
Ea ¼ ωNφ sin ωtd ðωt Þ ¼ ωNφ ð1:9Þ
π 0 π

DC machine working principle lays on the current flow through a coil within a
magnetic field, and then a magnetic force is produced to generate a torque that rotates
the rotor through four field excitation design topologies to display a wide variety of
volt-ampere or speed-torque characteristics for both dynamic and steady-state oper-
ation [4]. In DC generators, the field excitation topologies are called (i) separately
excited, (ii) shunt, (iii) series, (iv) cumulative compound (adds shunt and series
effect), and (v) differential compounded (subtract shunt and series effect) generator
[4]. Generally, these DC generator schemes are compared by their terminal voltage
regulation. Unlike DC motors, which are compared based on their speed regulation
capability. DC motors are driven from DC power supply. Unless otherwise specified,
the input voltage to a DC motor is assumed to be constant because that assumption
simplifies the analysis of motor comparison. Also, DC motors have five field
excitation topologies: (i) separately excited, (ii) shunt, (iii) series, (iv) compound,
and (v) permanent magnet [1].
However, brushed AC machines, as in Fig. 1.9, differ from DC machines in their
armature winding location. Their armature windings are almost always located on
the stator, while their field windings are located on the rotor. Generally, there are two
magnetic fields presented: magnetic field from rotor circuit DC current excitation
and another magnetic field from stator circuit. The interaction of these two magnetic
fields produces a torque in the machine, just like two PMs near each other that will
experience a torque that causes them to line up. The rotating magnetic field from the
1.3 Overview of Classical Electric Machine Topologies 9

Fig. 1.9 Radial view for two and four salient pole brushed three-phase AC machine topology. (a)
Two pole (b) Four pole

rotor field windings of an AC machine induces a three-phase set of AC voltages,

which are shifted by 120 electrical, into the stator armature windings calculated as
in (1.10). Conversely, a three-phase set of currents in the stator armature windings
produces a rotating magnetic field, which interacts with the rotor magnetic field,
producing torque in the machine [1]. Hence, the relationship between electrical angle
(θe) and the mechanical angle (θm) for AC machines with a number of poles (P) is
given in (1.11). Similarly, the relationship between electrical frequency ( fe) and the
mechanical frequency ( fm) of magnetic field rotation is given in (1.12). Note that it is
also possible to relate the electrical frequency in hertz to the resulting mechanical
speed (nm) of the magnetic fields in revolutions per minute (RPM) as in (1.13).

ea ðtÞ ¼ ωNφ sinðωtÞ

eb ðtÞ ¼ ωNφ sinðωt  120o Þ ð1:10Þ
ec ðtÞ ¼ ωNφ sinðωt  240o Þ

P
θ e ¼ θm ð1:11Þ
2

P
fe ¼ f ð1:12Þ
2 m
10 1 General Electric Machine Theory

nm P
fe ¼ ð1:13Þ
120

There are two rotor types, salient and nonsalient (round), in wound field
(WF) synchronous generators. The rotors are subjected to changing magnetic fields,
and it is constructed of thin laminations to reduce eddy current losses. Rotor DC field
winding can be supplied by DC source through slip rings and brushes as in Fig. 1.9,
or it can be through a special DC source mounted directly on the shaft of the
synchronous generator. Slip rings and brushes are applied for small synchronous
machines because no other methods are cost-effective [5]. On the other hand, large
generators and motors and brushless exciters are used to supply the DC field current
to the machine. A brushless exciter is a small AC generator with its field circuit
mounted on the stator and its armature circuit on the rotor [4]. By controlling the
small DC field current of the exciter generator, the rotor DC field winding of the
main WF synchronous generator is regulated.

1.3.2 Brushless AC Machines

An induction machine (IMs) is one in which AC current is supplied to the stator

directly and to the rotor by induction or transformer action. As in the synchronous
machine, the stator winding is like the synchronous generator discussed in the
previous section. When excited from a balanced three-phase source, it produces a
magnetic field in the air-gap rotating at synchronous speed as determined by the
numbers of poles and the applied stator voltage frequency. In IM topology, there are
two rotor types: squirrel-cage and wound rotor [4]. In this section, only the squirrel-
cage rotor is considered, as shown in Fig. 1.10. Compared with the wound rotor type,

Fig. 1.10 Radial view for brushless three-phase squirrel-cage IM machine topology
1.3 Overview of Classical Electric Machine Topologies 11

Fig. 1.11 Radial view for round and salient four-pole brushless three-phase PM machine topology.
(a) Salient PM rotor (b) Nonsalient PM rotor

the squirrel-cage rotor winding does not require slip rings and brushes; however, it
consists of conducting bars embedded in slots in the rotor iron core and short-
circuited at each end by conducting end rings. The squirrel-cage motor is substan-
tially a constant speed motor having a few percent drops in speed (slip) from no load
to full load. Different classes of squirrel-cage machines are presented in the literature
based on the effective resistance of the rotor-cage circuit [4]. Such that, the effect of
using these rotor-cage classes dictates machine torque-speed characteristics. Hence,
the extreme simplicity and raggedness of the squirrel-cage construction are excep-
tional advantages of this type of IM.
As for the brushless machine types, permanent magnet AC machines or brushless
PM machines are occasionally built to operate as synchronous machines with
rotating field winding replaced by a PM. Fig. 1.11 illustrates the brushless three-
phase PM synchronous machine having either salient or nonsalient PM rotor type.
The flux paths due to a four-pole PM AC machine that links stator phase coils with
rotor magnetic field are shown in Fig. 1.11a. Knowing that, if a constant torque is
exerted on the shaft to run the machine at a constant speed, this provides generator
action. On the other hand, if the three-phase winding is excited using a semicon-
ductor control switching pattern, then the machine is operating as a motor.

1.3.3 Switch Reluctance Machines

In terms of electric machine construction, switch reluctance machines (SRMs) are

considered a simple and rigid machine type. Their excitation winding is placed in the
salient or nonsalient stator only, where they always have salient magnetic rotor
shape. They operate using generated flux linkage due to stator applied current; its
path between stator and rotor tries to generate maximum torque through a tendency
12 1 General Electric Machine Theory

Fig. 1.12 Radial view for brushless SRM topology showing two stator poles to rotor poles ratio.
(a) 4/2 SRM (b) 6/4 SRM

to align rotor with the stator-produced flux linkage [4], as shown in Fig. 1.12. For the
control, the rotor position sensing is required in order to properly energize the stator
phase windings to produce torque. The SRM needs to be designed such that the
stator winding inductance varies with rotor position, while the stator core of SRM
requires high permeability magnetic material. The torque characteristics of SRM are
governed by the saliency of stator and rotor, which enhances the difference between
maximum and minimum inductances [4].
In SRMs, the torque is proportional to the magnitude of the phase current and
does not depend on its direction. Hence, unidirectional current can be used to supply
the stator winding through solid-state switches. Therefore, for motor drive, only half
of the solid-state switches are required to energize the stator phase through a single
current direction, which reduces the control electronics by half compared with the
other machine drive system [4], such as in brushless PM machines. The zero torque
position in the SRM cannot be presented if the ratio between the stator poles (SP) to
rotor poles is not an integer. For instance, SP/P for 6/4 SRM is 1.5, and hence there
will not be a simultaneous alignment of stator phase inductance.
However, in some instances, a SRM with an integer pole ratio is desirable; in this
case, the elimination of zero torque is attained by constructing the machine with an
asymmetric rotor [4]. Therefore, the rotor pole width is made wider than that of the
stator. In general, when a given phase is excited, the torque is such that the rotor is
pulled to the nearest position of maximum flux linkage. As excitation is removed
from this phase and the next phase is excited, the rotor is then pulled to a new
maximum flux-linkage position [4].
1.4 WF and PM Synchronous Machine Excitation Fields 13

1.4 WF and PM Synchronous Machine Excitation Fields

As discussed in the previous sections, the wound field (WF) and PM rotor types of
AC synchronous machines provide rotating magnetic fields that produce the three-
phase set of voltages in the stator coils as given in (1.10). The excitation field in the
WF rotor type is supplied by the DC voltage source through slip rings and brushes,
as in Fig. 1.9. While the PM rotor type does not need that, it instead requires spatial
arranging of soft or hard PM material, which can be accommodated on the rotor core
in common ways known as surface-mounted magnets, inset magnets, buried mag-
nets with radial magnetization, and buried magnets with circumferential magnetiza-
tion [5]. Note that, in this book, surface-mounted magnet rotor type is chosen for the
synchronous PM machine topology, as in Fig. 1.11.

1.4.1 Magnetic Flux Path Representation of WF

Synchronous Machines

By Ampere’s law, the current in a coil of wire wrapped around a ferromagnetic

material core produces a magnetic flux in the core. In a magnetic path representation,
the reluctance is the counterpart of electrical circuit resistance, and its unit is ampere-
turns per weber (A•t/Wb), while the MMF in magnetic path representation is
analogous to EMF voltage in an electrical circuit and its unit is ampere-turns (A•t).
The magnetic path representation translates the magnetic field behavior within
electric machine parts to a simplified manner, which otherwise is complex to analyze
for machine design process, as will be seen in detail in Chap. 3.
Assume that a concentrated stator winding is employed when the number of stator
slots is equal to the number of rotor poles. By considering a sectional view of a WF
synchronous machine, here the flux does not behave in a simple manner since there
are different ferromagnetic rotor and stator materials in addition to air-gap and
different cross-sectional flux path areas. Therefore, different cross-section path
reluctances are calculated using (1.5). Machine active parts are the parts through
which magnetic flux is passing causing generation of EMF voltage and electromag-
netic torque. There are nine different reluctances and one rotor coil MMF, which is
calculated as in (1.6), in the considered machine section. The flux-linkage path
reluctances are represented by left side stator yoke (R sy1), right side stator yoke
(R sy2 ), stator tooth (R st ), stator tooth tip (R stt ), air-gap (R g ), rotor tooth tip (R rtt ),
rotor tooth (R rt ), left rotor yoke (R ry1 ), and right rotor yoke (R ry2 ), as illustrated in
Fig. 1.13. Note, air-gap reluctance is very large compared with the other core
sections’ reluctances due to very low air permeability value. Hence, Kirchhoff’s
voltage law (KVL) can then be used to calculate the magnetic flux linkage.
14 1 General Electric Machine Theory

Fig. 1.13 Magnetic flux

path representation of a WF
synchronous machine
section

Fig. 1.14 Magnetic flux

path representation of a PM
synchronous machine
section
1.4 WF and PM Synchronous Machine Excitation Fields 15

1.4.2 Magnetic Flux Path Representation of PM

Synchronous Machines

Given the same concentrated winding assumptions for the PM machine topology as
in WF machines, different cross-section path reluctances are calculated using (1.5).
There are eight different reluctances and one rotor PM MMF, which is calculated as
in (1.7), in the considered machine section. For the stator and air-gap of the PM
machine section, the flux-linkage path reluctances are similar to those found in the
WF machine case, while the PM rotor reluctances are represented by rotor PM (R m),
left rotor yoke (R ry1 ), and right rotor yoke (R ry2 ), as illustrated in Fig. 1.14. Again,
KVL can be used to calculate the magnetic flux linkage that will be shown in detail in
Chap. 3.
Chapter 2
Hybrid Electric Machine Concept

2.1 History

As the demand for less expensive and more efficient electrified powertrain grows,
the need for optimized electric machines becomes more apparent. An interesting
electric machine topology that leads to simplified powertrains is hybrid excitation
electric machines. In hybrid excitation machines, there exist two magnetic fields.
This provides a flexible field control capability with an acceptable power density and
without the need for an expensive power converter control system.
Different methods of hybrid excitation field regulation topologies, including a
PM combined with a WF excitation, have been considered in the literature [6–15,
16–39]. By combining PM and WF excitation, here referred to as hybrid PM (HPM)
machine, the advantages of both PM and WF synchronous machines are utilized.
The HPMs can be classified based on their magnetic excitation field paths (series or
parallel) and based on their place in the machine stationary, rotary or both parts, as in
Fig. 2.1.

2.2 Hybrid Electric Machine Classification

For HPM machine topologies, there are at least two excitation field sources that
provide the net machine excitation. In general, a PM source provides the main
excitation, and a wound field component acts to regulate the machine flux distribu-
tion either by boosting or by weakening the PM field depending on the direction of
the wound field DC excitation current. The DC field winding may be placed on the
rotor part of the machine as the PMs [22, 23, 33, 39], which necessitates slip rings
and brushes or an exciter, or on the stator [17–22, 24, 26, 28, 31, 37–39].

© Springer Nature Switzerland AG 2022 17

A. S. Al-Adsani, O. Beik, Multiphase Hybrid Electric Machines,
https://doi.org/10.1007/978-3-030-80435-0_2
18 2 Hybrid Electric Machine Concept

Fig. 2.1 General

classification of HPM
machines

Fig. 2.2 Cross sections of the permanent magnet synchronous machine with claw pole field
excitation (PSCPF) [16]

2.3 Different Hybrid Machine Topologies

For the HPM machines to operate as a variable voltage generator, the range of
air-gap flux density variation has to be designed to match the anticipated application
requirements. A number of HPM machine topologies have been reported in the
literature in recent years. The reported HPM machine topologies will be reviewed
next. An assessment will be made for each topology with a view of arriving at a
topology that will be studied in the following chapters.

2.3.1 PM Synchronous Machine with Claw Pole Field

Excitation (PSCPF)

PM synchronous machine with claw pole field excitation (PSCPF) is briefly

discussed by Zhao and Yan [16], where machine components and the associated
flux linkages are detailed; Fig. 2.2 shows the machine cross section. The PSCPF is
composed of two parts, one called the main part and the other the assistant part. Both
2.3 Different Hybrid Machine Topologies 19

Fig. 2.3 Flux path of the

(PSCPF) machine as
reported in [16]

Fig. 2.4 Simplified construction figure of HESG as reported in [17] (a) Axial section view (b)
Radial section view

parts of the machine share one common stator. Referring to Fig. 2.2, the assistant
part is composed of components 2–5; these represent the claw pole structure. The
field winding is placed on the stator; therefore, slip rings and brushes are not
required.
When current flows through the field winding (component 5), the magnetic path
of the DC flux is through the inner cylinder of component 3 (axial); the bottom of
component 3 (radial); the outer cylinder of component 3 (axial); the air-gap δ1
(radial); plane magnet pole (axial); the main air-gap δ (radial); stator iron core
(radial); air-gap δ (radial); claw pole magnet pole 2 (radial); magnetic shaft (axial);
air-gap δ2 (radial); and inner cylinder of component 3. The magnetic path of the PM
is through the claw pole magnet pole; air-gap δ (radial); stator iron core; air-gap δ
(radial); claw-plane magnetic pole; PM (N pole); and rotor iron core and PM
(S pole), as illustrated in Fig. 2.3.
Zhao and Yan also discussed an improved PSCPF machine, referred to as the
hybrid excitation synchronous generator (HESG), as illustrated in Fig. 2.4. It is
20 2 Hybrid Electric Machine Concept

Fig. 2.5 A new type hybrid excitation claw pole synchronous machine (HECPSG) components
[40]

Fig. 2.6 HECPSG machine

assembly [40]

basically a similar structure to that of the PSCPF, the dissimilarity being that the
latter has clapboard inserts that are made of nonmagnetic material. The clapboard
introduces an air-gap and thus reduces the coupling between the PM and wound field
excitation, making the two fields independent of each other. For both the PSCPF and
HESG designs, the PM and wound field excitations act independently; that is, they
are magnetically in parallel.
In 2007, Chao-hui et al. [18] presented a study of a new HPM machine based on
the HESG topology called the hybrid excitation claw pole synchronous generator
(HECPSG). The structure of the HECPSG is shown in Figs. 2.5 and 2.6.
The stator of the HECPSG consists of multiphase windings. The claw poles of the
rotor are magnetized by a cylindrical wound coil and a cylinder-shaped permanent
magnet, which is axially magnetized. The flux under one pole pair consists of two
2.3 Different Hybrid Machine Topologies 21

Table 2.1 HESPSG advantages and disadvantages

Field
Advantages Disadvantages path
(1) Good flux weakening [18] (1) Rotor structure is Parallel
(2) The structure of the claw pole is helpful to arrange more relatively complex*
magnet poles when the rotor diameter is relatively small
[18]
(3) Slip rings and brushes are not required [18]
Note: Deduced by the author (*)
Reported in the literature [18]

parts: one is produced by the permanent magnets and the other produced by the coil
exciting current [18]. The magnetic field from one claw pole passes through the
air-gap and stator core and back to another claw pole. No detailed discussion is given
for the interaction between the PM and winding fields, that is, potential for demag-
netization, heating effects, and reaction effects. Furthermore, the contribution from
each field source to the stator-induced back-EMF is not discussed. Table 2.1 sum-
marizes the advantages and disadvantages of the HECPSG topology.

2.3.2 Toroidal-Stator Transverse-Flux Machine (TSTFM)

Spooner et al. [19]. discussed hybrid excitation of AC and DC machines for rail
traction and engine-mounted generators. Transverse-flux AC synchronous machines
are excited by means of a simple DC coil mounted on the stator, as shown in
Fig. 2.7a. Consequently, they are naturally brushless, they are reported to have
low rotor losses (since the rotor has no permanent magnet poles), and they are
mechanically suited to very high speed. However, the authors do not consider high-
frequency losses that may occur in the solid rotor poles. The basic machine cross-
section schematic is illustrated in Fig. 2.7a, consisting of two stator sections joined
by a soft-magnetic outer casing and separated by the field coil. The rotor has two
similar sections, one in each stator section and mutually displaced in space, in this
case by 180 mechanical. Each rotor section has a salient structure, Fig. 2.7b. The
field coil DC current establishes a set of north poles on rotor Sect. 1 and a set of south
poles on rotor Sect. 2, as illustrated in Fig. 2.7b. Each stator coil encloses both stator
core sections and experiences alternate north and south rotor poles as the rotor turns.
The flux-linking of a stator coil is equivalent to that in a conventional radial field
machine design of half the total core length [19] since there are empty spaces
between the rotor soft-magnetic iron poles. A major problem for designers is the
provision of sufficient magnetic material to carry flux between the two rotor sections.
Furthermore, there is a substantial leakage flux when the stator sections are faced by
the large effective air-gap of the “empty” or high reluctance rotor sections.
Fixing magnets in the empty spaces of each rotor section, as shown in Fig. 2.7c,
provides a pole opposite to those established by the field winding and enhances the
22 2 Hybrid Electric Machine Concept

Fig. 2.7 Transverse-flux machine components as reported by Spooner et al. [19] (a) Machine cross
section (b) Machine soft-magnetic rotor (c) Machine rotor with saliency and permanent magnets

mechanical rotational symmetry (balance). The flux that passes through the machine
shaft due to the permanent magnets is subtracted from that due to the excitation field
current and so makes possible a greater flux-per-pole for each rotor section. The
required field current can thus be reduced from the design of Fig. 2.7b, and leakage
flux is also reduced [19]. Thus, transverse-flux machine arrangements appear to be
an attractive option for small- and medium-size generators [19].
Spooner et al. [40] presented a rotary toroidal version of the transverse-flux
hybrid excitation machine, based on the work of Evans and Eastham transverse-
flux AC machine topology. The machine construction is illustrated in Fig. 2.8,
showing a toroidal wound stator core of multiphase windings, DC field winding
located inside the toroidal core, and two rotating discs with alternate permanent
magnet and soft-magnetic poles. The flux-linkage paths throughout the machine
parts due to both the PM’s and stationary field coil are illustrated in Fig. 2.9. If the
two rotor poles are only provided by PMs, the flux path can be traced from one rotor
plate containing north pole magnets, crossing the air-gap into the toroidal stator, and
then traveling circumferentially across the second air-gap into the south magnet pole
on the opposite plate, through the plate into the shaft and back to the first plate to
2.3 Different Hybrid Machine Topologies 23

Fig. 2.8 Toroidal

transverse-flux machine
reported by Spooner et al.
[19]

close the loop at the north pole [20], as shown in Fig. 2.9a. A modification to the
design of Fig. 2.9a has soft-magnetic poles between the respective north and south
PM poles, as illustrated in Fig. 2.11b [19], resulting in additional flux paths. Thus,
flux from the north pole on the right-hand side plate crosses to the stator but then
comes back to the same rotor disc via the soft iron pole [20], as shown in Fig. 2.9b. In
this case, flux does not generally pass through the rotor shaft. However, during the
operation of the machine, flux travels through both paths, subject to reluctance
variation in the shaft. Finally, there is a third flux path due to the field excitation
coil that drives flux through the rotor shaft, rotor plate, iron poles, air-gap, stator, and
the second iron poles on the opposite disc [20], as illustrated in Fig 2.9b, c for both
strengthening and weakening modes, respectively. The toroidal transverse-flux
machine configurations are brushless machines generating an AC output that is
modified by the DC field winding excitation current [21]. For both transverse-flux
topologies illustrated in Figs. 2.7, 2.8 and 2.9, the main PM field and moderating
wound field are magnetically in parallel, their advantages and disadvantages being
noted in Table 2.2.

2.3.3 Hybrid Excitation Synchronous Machine (HESM)

Naoe and Fukami discussed the structure of a hybrid excitation synchronous

machine (HESM) [22]. The machine has a conventional AC stator and a two-part
rotor construction where each part is separated by an air-gap. One rotor part has PM
excitation and the other part wound field excitation. Each rotor part is independent of
the other and, in the case reported, is of radial field design. The HESM is illustrated
schematically in Fig. 2.10. The flux produced by the field winding is designed not to
pass through the PMs because of their large reluctance, thus keeping the field
24 2 Hybrid Electric Machine Concept

Fig. 2.9 Flux paths in the

toroidal transverse-flux
machine, as reported by
Spooner et al. [19]. (a) Flux
paths due to PMs alone;
without rotor iron poles (b)
Flux paths due to both PMs
and DC field excitation in
strengthening mode; with
rotor iron poles (c) Flux
paths due to both PMs and
DC field excitation in
weakening mode; with rotor
iron poles

winding MMF low [22]. Hence, the machine air-gap flux can be modified by the field
winding current direction and magnitude. The PM and rotor wound field excitation
sources are magnetically in parallel. Table 2.3 summarizes the advantages and
disadvantages of the HESM topology.
2.3 Different Hybrid Machine Topologies 25

Table 2.2 Advantages and disadvantages of toroidal-stator transverse-flux machine topologies

Field
Advantages Disadvantages path
(1) Control is relatively simple [19] (1) The magnetic path of the electrical Parallel
(2) The short axial length makes this excitation is relatively large, which neces-
machine suitable for directly mounting to sitates relatively high excitation MMF*
an engine shaft replacing, in part, the (2) Mechanics are complex*
flywheel [19] (3) The design magnetic field of the toroi-
(3) Slip rings and brushes are not required dal is restricted by the machine diameter
[19] [19]
Note: Deduced by the author (*)
Reported in the literature [19]

Fig. 2.10 Structure of the

HESM with a two-part rotor
[22]

Table 2.3 Advantages and disadvantages of the HESM

Field
Advantages Disadvantages path
(1) Machine is simple in structure (1) Slip rings and brushes exist, which Parallel
[22] increases complexity and maintenance costs*
(2) Short magnetic path [22]
(3) The air-gap flux can be easily
controlled by the field current [22]
Note: Deduced by the author (*)
Reported in the literature [22]

2.3.4 Synchronous Permanent Magnet Hybrid AC Machine

(SynPM)

The synchronous permanent magnet hybrid AC machine (SynPM) was presented by

Xiaogang and Lipo [23]. The machine is a combination of four PM poles and two
wound field excitation poles on the same rotor, as illustrated in Fig. 2.11. The PM
poles provide the major part of air-gap flux, while the wound field excitation poles
act as a flux regulator to adjust the air-gap flux distribution. By appropriate connec-
tion of the stator coils and rotor winding excitation, the net phase flux linkage and
hence back-EMF may be weakened or strengthened. Considering one of the stator
26 2 Hybrid Electric Machine Concept

Fig. 2.11 Cross section of the of SynPM machine reported by Xiaogang and Lipo, showing one
phase belt of the stator winding [40]

Fig. 2.12 Back-EMF of

one coil of the phase belt
winding [40]

phase belt coils, the coil back-EMFs for the three excitation modes are as shown in
Fig. 2.12, while Fig. 2.13 illustrates the corresponding open circuit flux lines due to
positive, zero, and negative DC field currents. A phase belt is formed by connecting
three coils of the same phase in series, as shown in Fig. 2.11; thus, the resulting
phase back-EMFs for the cases of positive, zero, and negative field winding current
are as shown in Fig. 2.14. Slip rings and brushes are required for this machine
topology. For the machine discussed, excitation produces around 67% of the total
air-gap flux [23]. The flow of the flux is radial for both PM and DC field windings,
which are magnetically acting in parallel. Table 2.4 summarizes the advantages and
disadvantages of the SynPM topology.
2.3 Different Hybrid Machine Topologies 27

Fig. 2.13 Flux lines of the six-pole SynPM machine presented by Xiaogang and Lipo [40]. (a) Full
positive field current (b) Zero field current (c) Full negative field current

2.3.5 Consequent Pole Permanent Magnet Hybrid Excitation

Machine (CPPM)

Tapia et al. discussed a consequent pole permanent magnet hybrid excitation

machine [24–25]. The machine combines fixed PM excitation with variable flux
via a field winding fixed in the stator. The machine is similar to the transverse-flux
machine reported by Spooner et al. [19]. However, Tapia et al. discussed a greater
number of design options and discussed the design in greater depth. The machine
consists of a rotor divided into two sections, each section having radially magnetized
surface-mounted permanent magnets interleaved with laminated iron poles, as
illustrated in Fig. 2.15a. The magnetization of each rotor section is shifted 1-pole-
pitch with respect to the other section.
The stator is composed of two laminated tooth sections inside a solid outer soft
magnet yoke. A conventional three-phase AC winding is located in slots around the
periphery of the inner stator diameter, and a circumferential field winding is placed
between the two stator sections, as illustrated in Fig. 2.15a. The field winding is
excited by DC current. For no field current, the machine excitation is due to the rotor
PMs alone and is essentially radial, each PM linking with a consequent soft iron pole
on the same machine half. When excited with positive current, the flux generated by
the field winding flows in a direction such that it adds to the PM flux and the flux
closes its path in the same half stator, as illustrated in Fig. 2.15b. If the field current is
negative, the direction of the air-gap flux is as shown in Fig. 2.15c. Figure 2.15d
shows further views of the CPPM components. The stator and rotor yokes provide a
low reluctance path for the axial flux, which is considered an important feature of the
machine operation. The current of the field winding is externally controlled in order
to provide variable excitation.
28 2 Hybrid Electric Machine Concept

Fig. 2.14 Example of

resultant coil and phase
back-EMF for different field
winding excitation
conditions [40]. (a) Back-
EMF of one circuit with full
positive excitation (b) Back-
EMF of one circuit with zero
excitation (c) Back-EMF of
one phase with full negative
excitation
2.3 Different Hybrid Machine Topologies 29

Table 2.4 Advantages and disadvantages of the SynPM

Field
Advantages Disadvantages path
(1) Machine is comparatively simple in (1) Slip rings and brushes exist [23] Parallel
structure* (2) The combination of four-pole or
(2) In addition, it is easy to fabricate short two-pole field flux in field weakening,
magnetic paths. A high power density is with the six-pole stator flux, will result in
suggested, but no data are quoted* a number of space and time-harmonic
components and undesirable torque pul-
sations and vibration [23]
(3) At high speed, when the field is
weakened, a high iron loss in the stator
might appear*
Note: Deduced by the author (*)
Reported in the literature [23]

Fig. 2.15 Consequent pole PM hybrid excitation machine (CPPM) [40]. (a) Magnetic structure of
the CPPM machine [25] (b) Magnetizing effect of the field flux (c) Demagnetizing effect of the field
flux (d) 3 kW CPPM prototype
30 2 Hybrid Electric Machine Concept

Fig. 2.16 Field controlled Torus-NS machine (FCT-NS) [40]. (a) Machine components (b) Stator
assembly (c) Rotor assembly

2.3.6 Field Controlled Torus-NS (FCT-NS) Machine

Aydin et al. discussed an axial flux machine designed to improve the flux weakening
operation of the previously reported axial flux, toroidal PM machines [26]. The
machine is essentially an axial field version of the CPPM and was referred to as the
field controlled Torus-NS (FCT-NS) machine. The machine construction consists of
two outer rotor discs carrying axially magnetized permanent magnets alternatively
placed with slotted magnetic iron pole pieces. There are two slotted stator cores, an
inner and outer core, realized by tape wound laminations inserted with multiphase
AC windings and a DC field winding between the stator inner and outer cores, as
illustrated schematically in Fig. 2.16. Variations on the FCT-NS design were
presented by Lipo and Aydin [27, 28].
Figure 2.17 shows the main flux direction of a two-pole portion of the FCT
machine at the average diameter [26] (a); rotor flux directions (b); air-gap flux
directions (c); and operating principle of the FCT machine (d) for zero (i), positive
(ii), and negative (iii) field current. Figure 2.17e shows the FCT stator and rotor
components. Figure 2.18 illustrates schematics of the single-rotor-single-stator
topology (a); the NN- and NS-type double-rotor-single-stator (b and c); double-
stator-single-rotor (d); and multistage (e) concepts.
Figure 2.19 illustrates the hardware of the NN-type FCAFPM machine as
reported in the literature. The CPPM and variants are all parallel permanent magnet
and wound field magnetic designs. Table 2.5 summarizes the advantages and
disadvantages of the CPPM and variants as reported in [24–29].

2.3.7 Dual-Rotor Machine

Amara et al. proposed a dual-rotor machine that is composed of two rotors placed
together (one wound and the other with PMs) inside the same stator assembly, as
shown in Fig. 2.20. The design employs juxtaposed magnetic circuits that, according
2.3 Different Hybrid Machine Topologies 31

Fig. 2.17 Field controlled Tours-NS type (FCT-NS) [40]. (a) Main flux direction of the FCT
machine [26] (b) Rotor flux directions [26] (c) Air-gap flux directions (d) Operating principle (e)
FCT rotor and stator components

to the authors, avoids the risk of PM demagnetization [30]. Flux weakening is

achieved via excitation of the wound rotor to create a flux opposite to that created
by the rotor PMs [30]. The design is similar in form to the HESM presented in Sect.
2.3.3 [22] but having slightly different rotor topologies.
32 2 Hybrid Electric Machine Concept

Fig. 2.18 Reported combinations of the FCAFPM machines [40]. (a) Single-rotor-single-stator (b)
NN-type double-rotor-single-stator (c) NS-type double-rotor-single-stator (d) Double-stator-single-
rotor (e) Multistage

Fig. 2.19 NN-type FCAFPM machine reported in [40]. (a) Stator view pre-impregnation (b)
Complete stator assembly (c) Rotor assembly

2.3.8 Imbricated Hybrid Excitation Machine (IHEM)

Amara et al. also proposed an imbricated hybrid excitation machine (IHEM), as

illustrated in Fig. 2.21. The rotor is composed of two magnetically isolated parts, one
containing the PM excitation, and the other is used to direct flux created by an
2.3 Different Hybrid Machine Topologies 33

Table 2.5 Advantages and disadvantages of CPPM and variants as reported in [24–29]
Field
Advantages Disadvantages path
(1) Control of the CPPM is convenient (1) Additional DC winding in the stator Parallel
(2) The air-gap flux can be controlled reduces the power density, such that the
without affecting the magnetization char- additional air-gap surface associated with
acteristics of the PMs this winding does not participate in the
(3) A wide range of air-gap flux control energy conversion process
can be obtained with a low DC excitation
field ampere-turn requirement
(4) Slip rings and brushes are not required

Fig. 2.20 Dual-rotor

machine [40]

Fig. 2.21 Imbricated hybrid excitation machine (IHEM) [40]. (a) Machine cross section (b) Rotor
structure

excitation coil that is located on either the rotor or the stator, the latter case avoiding
all sliding contacts. The stator is composed of two identical parts linked by a yoke, as
shown in Fig. 2.21a. The main goal of this design was to ensure that the flux created
by the excitation winding does not pass through the PM; hence, the possibility of
demagnetization is greatly reduced [30].
34 2 Hybrid Electric Machine Concept

Furthermore, Vido et al. proposed two improved versions of the IHEM [31], the
(i) homopolar and (ii) bipolar hybrid excitation synchronous machines, HHESM and
BHESM, respectively, as illustrated in Fig. 2.22. Cross-sectional schematics of both
prototypes are shown in Fig. 2.22. The rotors consist of three parts, one a solid core,
one part laminated core, and a set of permanent magnets. The schematics show an
axial cut of the stator and rotor for both prototypes, which are six-pole pairs. The two
machine rotors have the same dimensions. By comparing the two topologies, it can
be observed that the lateral permanent magnets are not present in the BHESM
prototype [31].
The various flux paths created by excitation coils, lateral PMs (side magnets), and
azimuth PMs may be divided into two categories: homopolar and bipolar flux paths.
The homopolar flux path represents a flow of flux through machine parts in axial and
radial directions. The bipolar design has flux paths in either radial or axial direction.
Therefore, the flux generated by the field DC coils has only one path, which is
homopolar in nature, as shown in Fig. 2.23a. Moreover, the homopolar path for the
lateral PMs can be observed in Fig. 2.23a. The flux generated by the PMs has two
distinct paths, one of which is bipolar, as shown in Fig. 2.23b, c, which creates north
and south poles under the active parts [32]. The flux path generated by the azimuth
PMs is primarily oriented perpendicular to the axial direction of the machine [32]. A
portion of the flux generated by the lateral PMs is oriented in the axial direction of
the machine via the rotor flux collector, as shown in Fig. 2.23c. In other words, the
fluxes created by either the PMs or the wound field excitation that exhibits a
homopolar path only give rise to one type of pole (either north or south), depending
on the direction in which the magnets are magnetized and the polarity of current in
the DC field coils [32]. Flux only passes once through the air-gap under the active
part, and then it returns first via the stator end shields and then via the rotor flux path,
as illustrated in Fig. 2.23d [32]. Figure 2.24a shows flux paths created by the DC
field coils for the BHESM design. This bipolar configuration passes through two
annular excitation coils. Each coil acts in one polarity of pole [31]. The flux created
by an excitation coil goes through active parts and an air-gap (homopolar path).
Figure 2.24b shows the bipolar flux path created by PMs, where this bipolar flux
passes through active parts and air-gap, creating north and south poles. Figure 2.24c
shows the PM leakage flux path, which is not through the active parts and hence does
not contribute to torque production [31]. Figure 2.25 shows homopolar flux paths
created by PMs, as reported in [31]. For homopolar hybrid excitation machines, the
total flux passing through the stator windings exhibits a DC component, while for
bipolar hybrid excitation machines, the total flux passing through the armature
windings does not have a DC component [31]. Thus, although air-gap flux control
is effective for both the HHESM and BHESM machines, the DC current excitation
efficiency is better for the HHESM because of the solid rotor core parts [31]. For the
HHESM operating with enhanced excitation flux, magnetic saturation occurs when
the magnetic pole in which the DC excitation is acting is saturated, even if the other
pole is still not saturated [32]. However, for the BHESM, magnetic saturation occurs
only when both magnetic poles are saturated, from which the authors conclude that
the BHESM has a wider excitation flux variation [32]. The efficiency of the hybrid
2.3 Different Hybrid Machine Topologies 35

Fig. 2.22 Homopolar and bipolar hybrid excitation synchronous machines [40]. (i) Schematic
(i) Schematic (ii) Prototype rotor details (ii) Prototype rotor details (iii) Prototype stator and rotor
(iii) Rotor laminations (a) First prototype machine (HHESM) (b) Second prototype machine
(BHESM)
36 2 Hybrid Electric Machine Concept

Fig. 2.23 HHESM various flux paths due to deferent excitations [40]. (a) Homopolar flux path due
to DC coils (b) Bipolar flux path (azimuthal magnets) (c) Bipolar flux path due to PMs (d)
Homopolar flux path due to PMs

excitation is better for the HHESM than it is for the BHESM design because of the
leakage flux path, as shown in Fig. 2.24c [31], which does not contribute to torque
production. Table 2.6 summarizes the advantages and disadvantages of the IHEM
topology.

2.3.9 Series Double Excited Synchronous Machine (SDESM)

Fodoren et al. present the series double excited synchronous machine (SDESM) that
has series excitation circuits [33, 34]. The parallel excitation circuit reported in some
of the previously presented topologies suffer from the drawback of construction
complexity [33]. The main advantage of the SDESM appears in applications where
2.3 Different Hybrid Machine Topologies 37

Fig. 2.24 BHESM various flux paths created by different excitations [40]. (a) Flux paths created
by DC field coils (b) Bipolar flux path created by a rotor PM (c) Leakage flux path created by PM

Fig. 2.25 Homopolar paths of fluxes created by PM [40]. (a) First homopolar path (b) Second
homopolar path

Table 2.6 IHEM advantages and disadvantages

Field
Advantages Disadvantages path
(1) The flux created by the excitation winding does (1) Machine is complex* Parallel
not pass through the PM [30] (2) Substantial amount of mag-
(2) The risk of magnet demagnetization is reduced netization leakage*
[30]
(3) Slip rings and brushes are not required [30]
Note: Deduced by the author (*)
Reported in the literature [30]

the electric drive operates under partial loads for most of the time [33]. Fodoren et al.
presented a design procedure, prototype, and test results for a SDESM design in [33].
The proposed SDESM design has the field excitation winding fixed on the rotor in
a series magnetic configuration with the surface PMs, as shown in Fig. 2.26. The
38 2 Hybrid Electric Machine Concept

Fig. 2.26 SDESM design presented by Fodoren et al. [40]. (a) SDESM cross section, basic
principle (b) SDESM cross section, actual design

Fig. 2.27 Hardware of SDESM [40]. (a) SDESM rotor (b) Test bench

stator is that of a commercial induction motor, while the rotor was constructed as
shown in Fig. 2.27a. The three-phase stator winding is single-layer, with three slots-
per-pole-per phase [33]. Table 2.7 summarizes the advantages and disadvantages of
the SDESM topology as reported in [33].
2.3 Different Hybrid Machine Topologies 39

Table 2.7 SDESM advantages and disadvantages [33]

Field
Advantages Disadvantages path
(1) Construction is simple (1) Rotor sliding contacts to rotor DC Series
(2) Good flux weakening with reduced excitation field
demagnetization risk
(3) Reduced iron losses in extended speed
operating region

Fig. 2.28 Cross section and actual switch reluctance motor with field assistance [40]. (a) Motor/
generator assembly (b) Cross section of the motor/generator (c) Motor/generator hardware

2.3.10 Switch Reluctance Machine with Stator Field

Assistance

Afjei et al. presented a new configuration of switch reluctance machine with stator
field assistance that represents a hybrid generator topology, albeit with no PMs, as
illustrated in Fig. 2.28 [36, 37]. This machine design was intended to be utilized in a
hybrid vehicle motor/generator unit. The proposed hybrid machine consists of two
40 2 Hybrid Electric Machine Concept

stator and two rotor sections placed on both sides of the field coil assembly, as
illustrated in Fig. 2.28b [36]. Here, the magnetic flux produced by the field coil
travels through the guide and shaft to the rotor poles and then to the stator poles,
finally closing through the motor housing [36].
A variant of the Afjei design that incorporates PMs was presented by Chau et al.
for small wind applications [38]. This hybrid design has a unique structure that,
which it is claimed, contributes to simplified mechanical manufacturing and mag-
netic fixing. The machine design is illustrated in Fig. 2.29 [38]. Due to the extra
air-bridge that is in parallel with each PM, an amplification of the effect of the flux in
the reinforcing mode is achieved, as illustrated in Fig. 2.29b, where the field winding
MMF is opposing the PM MMF. The PM flux leakage will increase, causing an
amplification of the effect of the flux weakening, as shown in Fig. 2.29b [38]. Thus,
as with a proper design of the air-bridge width, a wide flux regulating range can be
obtained by virtue of a small DC field excitation [38]. The PM and rotor wound field
excitation sources are magnetically in series. Table 2.8 summarizes the advantages
and disadvantages of the PM brushless hybrid generator topology.

2.3.11 Dual-Stator Hybrid Excited Synchronous Wind

Generator (DSHESG)

Liu et al. presented a novel dual-stator hybrid excited synchronous wind generator
(DSHESG), as illustrated in Fig. 2.30 [39]. The proposed generator stator is com-
posed of outer stator, inner stator, and field winding. The rotor consists of the PMs,
claw poles, rotor yoke, and cup rotor [39]. There are two independent parallel
magnetic circuits in the DSHESG due to PMs and the DC excitation coil. Here,
the series PM magnetic circuit consists of PMs, air-gap, cup rotor, and laminated
stator core. The rotor series magnetic circuit consists of claw poles, air-gap, lami-
nated core of the outer stator, and bracket of the field winding. It is claimed that this
topology overcomes some of the previous hybrid machine topologies’ weakness, for
example, the benefit of the two independent magnetic circuits, which reduces the
leakage flux problems, and there is a reduced risk of PM demagnetization. The PM
and rotor wound field excitation sources are magnetically in parallel. Table 2.9
summarizes the advantages and disadvantages of the DSHESG topology.

2.4 Summary of Surveyed Literature on HPM Machines

The review of publications has highlighted a number of notable issues. The

published HPM machine designs are, in general, complicated and have particular
weaknesses in their design, for example, excessive flux paths that lead to high
leakage (magnetomotive force loss) [21, 26, 33] and PM demagnetization
2.4 Summary of Surveyed Literature on HPM Machines 41

Fig. 2.29 Cross section,

magnetic field distributions,
and actual prototype PM
brushless hybrid generator
[40]. (a) Generator cross
section (b) Magnetic field
distributions under different
DC excitations (c) Prototype
generator
42 2 Hybrid Electric Machine Concept

Table 2.8 PM brushless hybrid generator advantages and disadvantages

Field
Advantages Disadvantages path
(1) Wide flux regulating capability with small DC field (1) There is a risk of magnet Parallel
excitation [38] demagnetization*
(2) The rotor has neither PMs nor field windings, which (2) There is a risk of stator
offers high mechanical integrity [38] teeth saturation*
(3) Slip rings and brushes are not required [38] (3) There is a substantial
amount of flux leakage*
Note: Deduced by the author (*)
Reported in the literature [38]

Fig. 2.30 Detailed structure and magnetic circuits of DSHESG [40]. (a) Structural 3D FEA model
(b) Structure cross section (c) Magnetic circuit of PM (d) Magnetic circuit of DC field winding (e)
Actual prototype
2.4 Summary of Surveyed Literature on HPM Machines 43

Table 2.9 DSHESG advantages and disadvantages

Advantages Disadvantages Field path
(1) Good field control capability [39] (1) Machine complex assembly* Parallel
(2) Reduced leakage flux problems [39] (2) Slip rings and brushes exist*
(3) Low risk of magnet demagnetization [39]
Note: Deduced by the author (*)
Reported in the literature [39]

[33]. Table 2.10 collects data for the different hybrid excitation machines reported in
the literature review. Note that Table 2.10 is not complete due to the lack of some
information in the published papers. The open-circuit back-EMF regulation capa-
bility that has been reported for some of the HPM machine designs is generally based
on experimental or finite-element analysis (FEA) results. The variation of open-
circuit back-EMF due to the machine DC field current of the reported HPM machine
topologies varies from 42 to 175% relative to the machine back-EMF with zero DC
field excitation. However, Table 2.10 does not give a clear picture regarding the best
machine topology for vehicle application; thus, in order to get an adequate and fair
comparison, the machine mass, volume, and thermal limits should be considered
along with their performance and back-EMF regulation capabilities. Furthermore,
the power rating of some of the actual HPM machine prototypes discussed in the
literature ranged from 0.65 to 10 kW, as detailed in Table 2.10. Some of the
publications highlighted the area of interest for HPM machine topologies, for
example, traction, wind power, and vehicle systems.
Table 2.10 Main particulars of HPM machines presented in research publications
44

No-load back-EMF
DC field DC field regulation capability
current range Ampere-turns Bucking Boosting Control Rating Area of
Topology (A) (A-t) mode (%) mode (%) range (%) (kW) interest References
HECPSG ( 20) to (20) – 33* 33* 66* 0.65 – Zhao [18]
TSTFM – – – – – 10 Fixed on an Spooner
engine shaft [19]
TSTFM (0) to (2) +3000 – 27+ – 5 – Novinschi
[20]
HESM – – – – – – – Naoe [22]
SynPM – – – – – – – Xiaogang
[23]
CPPM – 25+ 25+ 50+ 3 – Tapia [24]
300
– 44+ 54+ 98+ 3 – Tapia [25]
500
FCTPM – – – 87* 10 – Aydin
1200 [26, 29]
( 18) to (18) 0 18+ 24+ 42+ 1.34 – Aydin [28]
IHEM ( 0.7) to (1.8) – 50+ 125+ 175+ – Traction Amara
system [30]
HHESM ( 7) to (5) – 70+ 30+ 100+ 3 – Vido
BHESM [31, 32]
SDESM (0) to (24) – – 24.5+ – 5.5 – Fodoren
[33]
Switch reluctance mot. With field 0 to 1 – Unclear Unclear Unclear – Vehicles Afjei
assistance generator [36, 37]
Brushless PM hybrid generator – 300 53+ 86+ 138+ – Wind power Chau [38]
+600
DSHESG ( 0.8) to (0.8) – – – – – Wind power Liu [39]
2 Hybrid Electric Machine Concept

Note: Experimental (+) FEA (*)

Chapter 3
Hybrid Permanent Magnet Machine Design

3.1 Introduction

The multiplicity of machine designs reviewed in Chap. 2 creates a certain level of

difficulty when selecting the most appropriate HPM machine design topology since
each author declared some degree of novelty but did not reference their design
against any benchmark solution. It was anticipated from previous research that many
future HPM machine applications would be direct engine-mounted having a large
outer diameter to active axial length aspect ratio and possible facility for some
through-shaft element (for flywheel or multiple geared outputs) necessitating an
essentially “donut” shaped volume envelope constraint [41–43]. Consequently, an
available 3-kW, surface magnet mounted, brushless permanent magnet
(PM) machine having the above volumetric attributes was chosen as a benchmark
PM design from which to develop a suitable HPM generator in this book since the
average power demand of 3 kW is typical of small 1.0 to 1.2 ton urban vehicles. This
chapter discusses the design and analysis of the benchmark brushless PM machine
and testing thereof to validate the calculation models and procedures. The developed
tools were then used to analyze and design WF rotors within the same stator
constraints and with an adequate axial length split ratio of the brushless PM machine,
from which a preferred rotor design solution was chosen. Thus, a machine that
combined the brushless PM and chosen WF design features was proposed. This
HESM structure, as discussed in Chap. 2 (Sect. 2.3.3), was compared along with
three other HPM machine topologies (all designed around the benchmark brushless
PM machine) to attempt a comparative analysis of the four competing topologies.
Conclusions are presented, and the chosen HPM machine design solution is subse-
quently adopted for further study in Chaps. 4, 5 and 6.

© Springer Nature Switzerland AG 2022 45

A. S. Al-Adsani, O. Beik, Multiphase Hybrid Electric Machines,
https://doi.org/10.1007/978-3-030-80435-0_3
46 3 Hybrid Permanent Magnet Machine Design

3.2 Machine Volume Envelope Consideration

The proposed and investigated hybrid generator, in this book, is designed as a direct
engine-mounted machine having a large outer diameter to active axial length aspect
ratio similar to the engine-PM machine system layout shown in Fig. 3.1 [44]. The
radial flux hybrid machine topology consists of two different synchronous machines:
PM and WF machines coupled on the same rotor shaft. Based on the application
volume envelope constraint, the design ensures (i) similar stator winding configura-
tion and geometry for both PM and WF sections, (ii) similar air-gap flux-density
waveform for both sections of the machine, and (iii) HPM machine adequate rating
and axial length based on buck/boost capability of the WF rotor. Consequently,
using available laminations, a three-phase, 3 kW, surface magnet mounted, brushless
PM machine having the set volumetric attributes, as will be discussed in the next
subsections, is designed as a benchmark machine from which to develop a suitable
HPM generator rating to meet a 3–4 kW average power demand for small (1.0–1.2
ton) vehicles [45].

Fig. 3.1 Direct engine-mounted PM synchronous generator layout [44]

3.2 Machine Volume Envelope Consideration 47

3.2.1 PM Machine Dimensions

As stated earlier, direct engine coupling with a large outer diameter to active axial
length aspect ratio and possible facility for some through-shaft element (for flywheel
or multiple geared outputs) necessities a “donut” shaped machine volume envelope.
Consequently, an available 3-kW, surface magnet brushless permanent magnet
(PM) machine having the required volumetric attributes was chosen as a benchmark
PM machine design. The benchmark brushless PM machine is illustrated in Fig. 3.2,
showing main dimensions and a photograph of the machine assembly. Table 3.1 lists
the main specification of the benchmark PM machine.

3.2.2 PM Machine Stator Winding Layout

The stator winding details of the benchmark PM machine are given in Table 3.2. A
phase belt (i.e., number of slots per pole per phase) for the three-phase PM stator

Dso
Stator back-iron Winding
hsy

Wst hst Tooth

hstt Wstt τst hgPM

LPMa
LM
hry τPMrp
Rotor back-iron Magnet
DPMri
(a)

(b)

Fig. 3.2 Benchmark PM Machine (a) Main Dimensions (b) Machine Assembly
48 3 Hybrid Permanent Magnet Machine Design

Table 3.1 Main design dimensions of benchmark brushless PM machine

Stator outer diameter (Dso) 230 mm
Stator inner diameter (Dsi) 167.2 mm
Stator yoke thickness (length) (hsy) 13 mm
Stator tooth width (Wst) 9.2 mm
Stator tooth tip width (Wstt) 12.4 mm
Stator tooth thickness (hst) 16.5 mm
Stator tooth tip thickness (hstt) 1.9 mm
Spacing between stator teeth 2.42 mm
(at tooth tip) (τst)
No. of stator slots (S) 36 –
Half stator slot area (Ass) 60.6
1026 m2
Rotor outer diameter (DPMro) 165.6 mm
Rotor inner diameter (DPMri) 130 mm
Rotor yoke thickness (length) (hry) 15 mm
Magnet thickness (LM) 2.8 mm
Magnet material NdFeB 37 MGOe
Spacing between rotor poles (at pole tip) (τPMrp) 1.8 mm
Number of rotor poles (P) 32 –
Air-gap thickness (hgPM) 0.8 mm
Active axial length (LPMa) 25 mm

Table 3.2 Stator winding details of benchmark brushless PM machine

No. of phases (n) 3
No. stator slots/rotor poles (nsp) 36/32
Coils per pole per phase (ncpp) 0.375
Slot pitch (Sρ) 160 elec. ¼ 10 mech.
Slot span (Sγ) 1
Pole pitch (Pρ) 180 elec. ¼ 11.25 mech.
PM pole span (PPMγ ) 10.11 mech.
Coil pitch (Cρ) 180 elec.
Coil span (Cγ ) 1 slot
Conductor diameter (Dc) (mm) 0.56
No. of parallel conductors per single turn (nPc) 12
No. of stator winding layers (nl) 2
No. of turns per coil (nstc) 6
No. of series coils per phase (nsc) 12
No. of turns per phase (nst) 36

winding is connected as 3/8; hence, a concentric winding with 36 coils wound

around stator teeth is realized. Each phase has four sets of three concentrated coils
connected in series as illustrated in Fig. 3.3a. Note that the middle coil of the three
concentrated coils is wound opposite to the other two coils, which results in the
largest amplitude back-EMF in the three-phase winding. In Fig. 3.3b, the
3.2 Machine Volume Envelope Consideration 49

Fig. 3.3 Three-phase PM

machine stator lamination Phase A winding start
and phase (A) winding Phase A winding start
scheme (a) A phase winding
configuration (b) Phase
back-EMF configuration
vectors

Phase AA
Phase
winding end
winding end

(a)

E1 2
2
Stator slot
Stator slot numbers
3
3
numbers
4
10 11 11
12
12
13
19 20 20
21
3-phase winding 21
22
28 29
29
30
kw1= 0.9455 30
31

(b)

fundamental winding factor (kw1) is calculated using the methods described in [46],
which incorporates both the slot pitch and distribution factor in one equation.
Therefore, by numbering the stator slots using vector U, the phase back-EMF vectors
(Ei) and kw1 are calculated as follows in (1, 2, 3, 4) [46].
50 3 Hybrid Permanent Magnet Machine Design

! jπP
S jU ðiÞj
E i ¼ signðU ðiÞÞ:e ð3:1Þ

 
nl S=3 
1  X !
kw1 ¼  E ð3:2Þ
nl S=3  i¼1 i 

U ¼ ½1  2  2 3 3  4 10  11  11 12 12  13 19  20  20 21 21  22 28
29  29 30 30  31 ð3:3Þ

 36 
X  
2
3 !
 E i  ¼ e jπP
1  2e jπP πP πP πP πP πP
þ 2e2j S  e3j S þ e9j S  2e10j S þ 2e11j S  e12j S þ e18j S
πP πP

S S

 i¼1 
πP πP πP πP πP πP πP
2e19j S þ 2e20j S  e21j S þ e27j S  2e28j S þ 2e29j S  e30j S Þ ð3:4Þ

where P, S, and nl are the number of poles, slots, and stator layers. The three-phase
fundamental winding factor determines the machine back-EMF voltage waveform
peak. Note that, from machine stator winding design point of view, as the funda-
mental winding factor approaches 1, a better back-EMF voltage utilization is
achieved.

3.2.3 Stator Winding Fill Factor and Resistance

A mechanical or electrical slot fill factor is an important parameter in machine stator

winding design. The mechanical slot fill factor accounts for the conductor wire with
its insulation layer, where the electrical slot fill factor accounts for the conductor wire
without the insulation layer. Filling the slot between stator teeth with a desired
number of conductors and with an adequate conductor cross-sectional area, as in
Fig. 3.4, contributes to compact and optimized machine design through minimizing
the coil resistance, which in turn decreases the developed heat and core loss during
operation. The three-phase back-EMF voltage generation is governed by rotational
speed, flux, and number of turns, as in (1.10); hence, the required number of turns
can be calculated by specifying the desired speed and phase voltage, assuming a
constant flux lines passing in the stator teeth due to either rotor PM or wound field
DC excitation. Winding space in machine slot design goes through an iterative
process to find a slot area that would accommodate the required number of turns
with a certain conductor cross-section area that produce a minimum resistance.
However, for a given stator lamination, the mechanical or the electrical slot fill
factor is obtained by knowing the following:
3.2 Machine Volume Envelope Consideration 51

Fig. 3.4 Schematic for electrical fill factor determination of a stator slot

(i) Slot-liner (insulation paper) thickness.

(ii) Single conductor diameter with varnish coat (wire with insulation layer) (Dc).
(iii) Parallel or single conductor area with varnish coat (Ac).
(iv) Effective conductor diameter (wire without insulation layer) (Dec).
(v) Parallel or single effective conductor area (Aec).
(vi) Slot winding area (half of the stator slot area with slot-liner) (Ass).
A stranded wire is recommended for stator winding of small electric machines to
achieve the required Ac with bendable conductors, which introduces the number of
parallel conductors per turn (nPc) term, in the effective or noneffective conductor
area calculation, as in (3.5). By using machine AutoCAD drawing, the slot winding
area can simply be determined, and both the mechanical and electrical slot fill factors
(kmf, kef) are respectively calculated as in (3.6).

Ac ¼ 0:5πD2c nPc
ð3:5Þ
Aec ¼ 0:5πD2ec nPc

nstc Ac
k mf ¼
Ass
ð3:6Þ
nstc Aec
k ef ¼
Ass

Note, as the effective conductor diameter per single turn increases, the coil
resistance decreases. Using the effective stator slot fill factor, the phase winding
resistances (Rs) for the benchmark PM machine are calculated using (3.7) and (3.8),
52 3 Hybrid Permanent Magnet Machine Design

where τrewr is the approximated stator mean end winding radius and Lsc is the copper
length per stator coil.

Lsc ¼ nstc ð2L þ 2πτsewr Þ ð3:7Þ

nsc nstc Lsc

Rs ¼ ð3:8Þ
σkef Ass

3.2.4 Finite Element Method Program

There are a number of tools that have been developed over the years to investigate
and predict electric machine behavior in linear and nonlinear modes. Here, rotary
machine analysis is achieved by utilizing the finite-element numerical techniques
[47]. FEMM is a commercial software package that can solve several low-frequency
electrostatic and electromagnetic problems in a 2D plane. In this chapter, the various
machine geometries are analyzed using their equivalent 2D FE models. Before
embarking on actual machine hardware assembly and testing, FEMM is used to
analyze the proposed machine construction, changes to geometry, parameters,
injected currents, coil number of turns, and so on to obtain the adequate variables,
which satisfy the desired mode of operation.
In addition, while the selected HPM machine topology design meets certain
design criteria, such as low saturation and demagnetization risks, a quantitative
comparison of HPM machine topologies considered most likely to be applicable
for high-volume automotive manufacture is investigated within the same volumetric
and slot current density constraints of that of the reference PM machine design.
Magnetostatic problems are problems in which the fields are time-invariant. In
this case, the field intensity (H ), flux density (B), and current density (J ) must obey
(3.9) and (3.10) [42]:

J ¼Δ
H ð3:9Þ

Δ∙B ¼ 0 ð3:10Þ

subject to a constitutive relationship between B and H for each material [48]:

B ¼ μH ð3:11Þ

If the material is nonlinear (saturating iron or permanent magnet material), the

permeability μ is a function of B as in (3.12) [48]:
3.2 Machine Volume Envelope Consideration 53

B
μ¼ ð3:12Þ
H ðBÞ

The FEMM field solver converges such that (3.9), (3.10) and (3.11) are satisfied
via a magnetic vector potential approach. Flux density is written in terms of the
vector potential, Av, as in (3.13) [48]:

B ¼ Δ
AV ð3:13Þ

This definition of B always satisfies (3.10). Then, (3.9) can be rewritten as [48]
 
1
J ¼Δ
Δ
AV ð3:14Þ
μðBÞ

For a linear isotropic material and assuming the Coulomb gauge (Δ.Av ¼ 0),
(3.14) reduces to [48]:
 
1 2
J¼ Δ AV ð3:15Þ
μ

FEMM retains the form of (3.14) so that magnetostatic problems with a nonlinear
B–H relationship can be solved.
The advantage of using the vector potential formulation is that all the conditions
to be satisfied have been combined into a single equation. If Av is found, B and H can
then be deduced by differentiating Av [48]. Hence, the FEMM solver technique is
based on the division of the volume or domain in which this equation is valid into
smaller volumes or domains or so-called finite elements. Within each element, a
simple polynomial is used to approximate the solution via some iterative methods
such as conjugate-gradient and Newton–Raphson methods.
The procedure for numerical computation of magnetic field problems using
FEMM is divided into three steps: preprocessing, processing, and postprocessing.
In the preprocessing mode, the machine outline is drawn (in AutoCAD in this case)
and then uploaded into the FEMM solver. The machine material properties are
defined, boundary conditions assigned, and mesh generation achieved. In the
processing mode, the relevant Maxwell’s equations are used to solve the problem
and to obtain the field distribution in the analyzed domain of the electric machine. In
the postprocessing mode, calculations of characteristics, as well as parameters of the
analyzed electric machine, are obtained.
Figure 3.5 illustrates the benchmark brushless PM 2D machine model used in
FEMM, showing the peripheral boundary condition (a), the mesh detail (b), and
defined material (c). Material specification in FEMM is via blocks. Each block has a
defined name, a material assigned from the FEMM library, and an associated current
or circuit if relevant referring to Fig. 3.5c. The PM machine consists of five material
areas:
54 3 Hybrid Permanent Magnet Machine Design

Fig. 3.5 Benchmark brushless PM machine 2D FEA model (a) Boundary conditions (b) Mesh
detail (c) Material definitions

(i)Air: for all air-gaps inside the machine and around it.
(ii)US Steel Type 2-S, 0.018-inch thickness: for the stator part of the machine.
(iii)Carpenter electrical iron: for the rotor part of the machine.
(iv) NdFeB 37 MGOe: sintered neodymium–iron–boron permanent magnet mate-
rial used for the machine rotor permanent magnets.
(v) Copper: used for the stator phase windings.
Having set up the machine problem at the preprocessing stage, the FEMM solves
the magnetic field equations and produces results, for example, the distribution of
3.2 Machine Volume Envelope Consideration 55

magnetic flux in the machine. Postprocessing of the solved field solution can be
carried out to further analyze machine performance.
To consider machine rotation, it is necessary to produce multiple field solutions
from which the machine back-EMF can be determined. In FEMM, a Lua script code
can be written to perform multiple field solutions. An example of the Lua script tasks
are listed below:
(a) Start by opening the desired file in FEMM.
(b) Apply a “For-loop” function that is controlled by a specified rotor angular step.
(c) Solve and analyze the model.
(d) Save the analyzed data in a new file.
(e) Rotate the desired parts of the machine by angular step.
(f)Solve and analyze the new rotated model.
(g) Loop back to (d) or exit the for loop function when the desired final angle value
is reached.
(h) Stop the analysis.
Hence, it is relatively straightforward to obtain the desired data from the FEMM
solver for stationary and rotating magnetostatic fields and thus predict the electric
machine characteristic behavior for different operating conditions.

3.2.5 Machine Back-EMF Prediction

The benchmark PM machine back-EMF versus rotor position is obtained by manip-

ulating the FEMM postprocessing data that was collected for different rotor posi-
tions for machine no-load and varying load test cases. A Matlab script file was
written to collect and manipulate the FEMM data.
The proposed stator winding layout for a quadrant of the benchmark PM machine
is shown in Fig. 3.6. Here, flux φ passing through a single coil is predicted at discrete
rotor position angles over two electrical cycles, via the FE solver line integration
function based on (3.16) [49]. Referring to Fig. 3.6, Av1 and Av1’ are the magnetic
vector potentials of two coil sides. The previous step is applied for three consecutive
stator coils. Based on the coil number of turns and rotor speed (ω), the generated φ
and rotor position angle data arrays produced by FEA, as in Fig. 3.7a, are then
substituted into (3.16) and (3.17) to predict the back-EMF waveforms for each of the
three consecutive coils and single phase at no-load, as shown in Fig. 3.7.

φcoil ¼ ðAV1  AV10 ÞL ð3:16Þ

 
dφcoil
EMF ¼ ωnstc ð3:17Þ
dθr
56 3 Hybrid Permanent Magnet Machine Design

Fig. 3.6 One quarter of the benchmark PM machine stator winding layout

where dθr is the difference in rotor position angle and ω is the angular electrical
speed in rad/sec as calculated by (3.18):

ω ¼ 2π f e ð3:18Þ

where fe is the machine electrical frequency that is calculated by (3.19) and Ns is the
rotor mechanical speed in rpm.

PN s
fe ¼ ð3:19Þ
120

3.2.6 PM Machine Analysis Via EMC Model

Magnetic fields are present around current-carrying conductors and also exist around
magnetized objects such as permanent magnet material [50]. The magnetic circuit
quantities are represented by magnetic flux, reluctance, and magnetomotive force.
These magnetic quantities can be thought of by an analogy to electrical circuit
quantities, such that the electrical circuit analysis rules can be applicable to equiv-
alent magnetic circuits (EMCs) [51]. The equivalent magnetic circuit (EMC) repre-
sentation based on the lumped parameter method gives an approximate to the
machine field solution since it represents a lumped, and somewhat idealized, section
of the machine magnetic circuit. Nevertheless, the lumped parameters solution gives
fast field calculation that can consequently speed up iterative design procedures. In
this section, an EMC model is developed for one-quarter of the benchmark brushless
PM machine (due to symmetry) since the solution thereof can be compared with the
FEA model and measured data are discussed in the preceding section. Having gained
confidence in the PM machine EMC model, the model will then be developed to
3.2 Machine Volume Envelope Consideration 57

0.4
Coil 1 flux-linkage

Coils flux-linkage (mWb)

0.3 Coil 2 flux-linkage
Coil 3 flux-linkage
0.2
0.1
0
0 100 200 300 400 500 600 700 800
-0.1
-0.2
-0.3
-0.4
Rotor angle (elec. deg.)
(a)
10
8
Three coils back-EMF (V)

6
4
2
0
-2 0 100 200 300 400 500 600 700 800
-4
-6 (+) Coil 1 EMF
-8 (--) Coil 2 EMF
(+) Coil 3 EMF
-10
Rotor angle (elec. deg.)

(b)
Three coils and phase back-EMF (V)

100
80
60
40
20
0
-20 0 100 200 300 400 500 600 700 800
-40 (+) Coil 1 EMF
-60 (--) Coil 2 EMF
-80 (+) Coil 3 EMF
Phase EMF
-100
(c)

Fig. 3.7 Benchmark PM machine open-circuit back-EMF waveforms at 3000 rpm (a) Flux-linkage
waveforms for three consecutive stator coils (b) Back-EMF waveforms for three consecutive stator
coils (c) Back-EMFs per phase and for three consecutive stator coils

design a WF rotor within the benchmark PM machine stator. These two machines,
(i) surface magnet and (ii) wound field rotor, both having the same stator magnetic
and winding designs, will then be combined to form the prototype HPM machine
(HESM topology). The lumped parameter method has a number of limitations [51]:
(a) Only one value of relative permeability μ, from FEMM databases, is used for
each soft magnetic section, as in (3.20), (3.21) and (3.22) [51]:
58 3 Hybrid Permanent Magnet Machine Design

ΔB
μ¼ ¼ μ0 μr ð3:20Þ
ΔH

μi
μri ¼ ¼ 2141 ð3:21Þ
μ0

μs
μrs ¼ ¼ 1834 ð3:22Þ
μ0

where ΔH is the difference in magnetic field intensity, μri is the relative permeability
of iron, and μrs is the relative permeability of steel.
(b) Soft magnetic material saturation is not included, although it can be accounted
for by setting a maximum flux density for each section.
(c) Leakage reluctance is not included.
The considered flux density is normal to the surface flux density (Bn), which is
calculated by (3.23) [51]. The flux passing through each part of the magnetic circuit
is calculated by (3.24) [51] along with the magnetomotive force due to both the
permanent magnets and wound field poles (WF rotor case) by (3.25) and (3.26) [51]:

φ
Bn ¼ ð3:23Þ
Ac

MMF
φ¼ ð3:24Þ
R

MMF ¼ H c LM ð3:25Þ

MMF ¼ n f I f ð3:26Þ

where MMF is the magnetomotive force, R is the component reluctance, Hc is the

material coercivity, nf is the excitation field number of turns, and If is the DC
excitation field current. Furthermore, (3.26) represents the MMF of a wound coil,
thus, by using (3.27) [51], the reluctance of each part of the machine may be
calculated and the relevant EMC model is constructed.

h
R ¼ ð3:27Þ
μWL

where h is the material thickness, W is the width, and L is the axial length relevant to
the flux path. The lumped parameter model reluctance definition was given in
3.2 Machine Volume Envelope Consideration 59

Fig. 3.8 Lumped parameter model for benchmark PM machine. (a) Different sections for the EMC
model (b) Reluctances for one section enclosing a stator tooth (c) EMC representation

Chap. 1. Due to machine stator winding symmetry, calculation of the reluctance

paths of the benchmark PM machine EMC model, based on the lumped parameter
method of one-quarter of the machine cross section, is developed and illustrated in
Fig. 3.8. The respective machine sections forming the EMC network of MMFs and
reluctances are shown in Fig. 3.8a, flux path for a rotor pole and stator tooth is shown
in Fig. 3.8b, and a circuit representation of the EMC model is shown in Fig. 3.8c.
Assuming a clockwise flow of magnetic flux in each loop of the EMC model,
Fig. 3.7c, and applying KVL, a set of equations is derived as in (3.28) and (3.29).

φL ¼ R1 :MMF ð3:28Þ

60 3 Hybrid Permanent Magnet Machine Design

0.4
FEA Tooth No. 9
0.3
Magnet ic f lux (mW b)
Analytical using EMC
0.2
0.1
0 80
0 10 20 30 40 50 60 70
-0.1
-0.2
-0.3 Tooth No. 1

-0.4
Rotor angle (mech. deg.)

Fig. 3.9 Magnetic flux comparison for the PM machine sections shown in Fig. 3.8a

Table 3.3 Comparison of benchmark PM machine flux density

Flux density
Section No. (tooth Rotor angle (mech. (T) EMC and FEA flux density
body) deg.) EMC FEA difference (%)
1 0 1.27 1.34 5.4
2 10 1.29 1.22 5.5
3 20 0.96 0.93 4.0
4 30 0.57 0.56 0.5
5 40 0.18 0.19 6.1
6 50 0.18 0.19 5.6
7 60 0.60 0.57 5.8
8 70 1.16 1.22 5.4
9 80 1.29 1.34 4.1

X
φmb ¼ # φLi  " φLj ð3:29Þ
i, j

where φL is the vector of loop fluxes (n
1), R is the matrix of reluctances (n
n),
MMF is the vector of MMFs (n
1), φmb is the vector of resultant fluxes through
each circuit branch (m
1), and the loop magnetic flux #φLi and "φLj represent the
downward and upward magnetic flux directions, respectively, passing through two
neighboring loops, i and j.
Solving Eqs. (3.28) and (3.29), the flux flowing in each part of the benchmark PM
machine EMC model is computed and compared with that of FEA, as illustrated in
Fig. 3.9, for the section of the machine shown in Fig. 3.8a, b. The circled areas in
Fig. 3.9 show magnetic flux for stator teeth. Note that the graphs in Fig. 3.9 are for
one static solution at a fixed rotor angle; however, flux in each part of the machine
changes as the rotor rotates. Table 3.3 lists the flux density in the benchmark PM
machine tooth body in different sections as specified in Fig. 3.8a. The maximum
difference between the results obtained from EMC and FEA is 6.1%, and it occurs in
Sect. 5 where the flux leakage is highest for the presented static field solution. Note
3.3 WF Machine 61

Fig. 3.10 Comparison of benchmark PM machine predicted and measured results of the coil and
phase back-EMFs at 250 rpm

that the machine operates on the knee region of the steel B–H curve and the highest
flux density occurs in the tooth body that is 1.34 T. The saturation effect is small due
to machine geometry that results in low flux leakages and hence not considered in the
EMC modeling. To further confirm the validity of the FEA model, back-EMF
measurements are obtained from the laboratory prototype benchmark PM machine,
as in Fig. 3.2b, and compared with those of analytical results as illustrated in
Fig. 3.10. To form a stator winding for the benchmark PM machine, coils are
wound around one stator tooth as well as three consecutive teeth that form
one-quarter of a full phase winding. The analytical, FEA, and measured results
show a good agreement, hence confidence in the applied method, tools, and model
accuracy.

3.3 WF Machine

The benchmark PM machine is used as the base of the design and hence forms the
PM section of the HPM generator. To ensure the design is within the application
volume constraint, the same stator geometry and winding configuration are consid-
ered for both PM and WF sections. To design a WF rotor, the lumped parameter
model for the benchmark PM machine is modified to model flux paths on a wound
field rotor. The WF rotor is designed to have a similar pole number (3.32) and shape
as for the benchmark PM machine. However, there is a slight difference between
rotor pole spacing due to practical limitations on WF rotor coils placement. To
ensure similar stator flux levels, the WF rotor is designed to develop a close nominal
air-gap flux density and stator flux linkage at full WF excitation current as the PM
rotor does with an unexcited stator.
62 3 Hybrid Permanent Magnet Machine Design

3.3.1 WF Rotor Design

The same steps used to develop the PM section EMC model are applied to generate a
suitable WF rotor design. The equivalent magnetic circuit model of one of the WF
machine designs is illustrated in Fig. 3.11c. The flux flowing in each part of the WF
machine EMC model is computed analytically and compared with that of FEA, as
illustrated in Fig. 3.12a. Subsequently, for nf¼100 turns and If¼ 3 A, which gives a
rotor slot current density of 4.9 MA/m2, the EMC model is adjusted in an iterative
process to investigate the impact of key rotor magnetic circuit dimensions on overall
design, the driving criteria being to achieve the required air-gap flux density and

Fig. 3.11 Lumped parameter model for WF machine. (a) Different sections for the EMC model (b)
Reluctances for one section enclosing stator and rotor tooth (c) EMC representation
3.3 WF Machine 63

Fig. 3.12 WF machine stator teeth flux comparison via FEA and EMC along with stator tooth
iterative process using different air-gap thicknesses. (a) Magnetic flux comparison for sections
shown in Fig. 3.11a (b) Stator tooth flux for different air-gap thicknesses and flux density

stator flux linkage subject to minimizing rotor excitation MMF, saturation, slot
current density, and rotor copper losses. Figure 3.12b shows flux variation of WF
stator tooth for different air-gap thicknesses (δ) with corresponding air-gap normal
flux densities (Bn). For a WF air-gap thickness of δ ¼ 0.35 mm, the stator tooth peak
magnetic flux will be equitable to that of the benchmark PM machine; however, it
results in a flux density of 1.34 T, which is outside the defined boundary. Therefore,
the WF section air-gap is increased to 0.4 mm, resulting in a peak stator tooth flux
density of 1.195 T. Final WF rotor geometry is shown in Fig. 3.11a, b.
Hence, for HPM machine operation stability, the air-gap flux density waveforms
for the PM and WF parts are plotted via FEA for one quadrant, as previously shown
in Figs. 3.8a and 3.11a, to check their symmetry, as shown in Fig. 3.13. This
64 3 Hybrid Permanent Magnet Machine Design

Fig. 3.13 Air-gap flux density waveforms due to PM and WF machines via FEA

symmetry in the machine air-gap flux density waveforms will minimize the possi-
bility of adding more air-gap MMF harmonic effects to the HPM machine back-
EMF, which increases total harmonic distortion (THD) due to the addition of PM
and WF machine elements back-EMF waveforms. The impact of stator slot opening
with the rotor poles width at different locations is shown through the nonuniform
air-gap flux density waveforms. Note that it is recognized in the literature that
cogging torque harmonics can be reduced when nonoverlapping winding, minimal
slot opening, and the right combination of pole and slot numbers with adjusted rotor
pole width are achieved [52, 53].

3.3.2 WF to PM Split Ratio

In the previous section, the WF machine radial air-gap length is selected using the
EMC model to produce a phase flux linkage level as for the benchmark PM machine
axial length at a fixed rotor speed of 3000 rpm, the WF to PM generator split ratio
can be deduced to meet certain operation constraints. The operating philosophy in
SHEVs using an internal combustion engine/HPM generator scheme acts as an
axillary power system to provide average power for the vehicle powertrain where
the main battery system handles the transients power demand. By adopting this HPM
machine operating philosophy in SHEVs, the size of the engine and hence both fuel
consumption and emission are then reduced. The speed variations during transients
and dynamics are to be adjusted via the WF excitation of the HPM generator. Hence,
an expected vehicle DC-link voltage variation of 30% due to vehicle transients and
dynamics has been reported in the literature [46]. Therefore, the HPM generator-
rectified output voltage is required to follow these voltage variations. The second
3.3 WF Machine 65

Fig. 3.14 Magnetic flux due to axial length variation via the EMC model of the WF machine

iterative process is implemented here to come up with the adequate split ratio
between PM and WF machines’ axial lengths (WF/PM) based on the operation
philosophy of the final HPM generator design. The active length has a great
influence on the machine weight and output voltage; hence, it should be selected
carefully. Thus, by reducing the axial length of the WF machine element in several
steps and observing the stator teeth peak magnetic flux, the optimized WF machine
axial length based on operational philosophy is identified via the EMC model. The
second iterative process results via the EMC model are shown in Fig. 3.14, where the
adequate WF machine axial length is found to be 10 mm. Thus, this WF DC
excitation of the HPM generator final design will provide the required 30% buck/
boost back-EMF voltage capability in the SHEV powertrain. This HPM generator
will operate without demagnetizing the PM machine magnet during back-EMF buck
mode since the magnetic linkage between the HPM generator rotors is magnetically
decoupled.

3.3.3 Comparative Analysis of WF Rotor Designs

Having developed the lumped parameter model for the WF rotor and stator, the
major rotor magnetic dimensions were varied to investigate their impact on the
design, the driving criteria being to achieve the required air-gap flux density while
minimizing rotor excitation MMF and saturation. Several WF machine rotor designs
(with identical stators) were identified, the results for which are detailed in Table 3.4.
Here, the design that can produce maximum air-gap normal flux density (Bgn) with
the lowest DC excitation current, rotor slot current density (Jrs), field copper loss
(Pfc), and DC excitation field voltage (Vf) will present the best design choice to be
used in the final HPM machine design.
66 3 Hybrid Permanent Magnet Machine Design

Table 3.4 Dimension details for WF machine rotor designs

WF rotor design parameters 1 2 3 Units
Rotor outer diameter (DWFro) 166.75 166.75 166.75 mm
Rotor inner diameter (DWFri) 109 100 109 mm
Rotor tooth width (Wrt) 8.54 9.2 7.6 max. mm
11 min.
Rotor tooth tip width (Wrtt) 13.2 14.2 13 mm
Rotor tooth thickness (hrt) 27 17.5 16.5 mm
Rotor tooth tip thickness (hrtt) 1.3 2.2 0.8 mm
Rotor yoke thickness (hry) 5 15 11.5 mm
Spacing between rotor teeth 3.2 1.8 3.1 mm
(at tooth tip) (τrt)
Pole number (P) 32 32 32 –
Active axial length (LWFa) 10 10 10 mm
Half rotor slot area (Ars) 59.3 36.3 41.7
1026 m2
Air-gap thickness (hgWF) 0.45 0.45 0.45 mm

Note that in the FEA and lumped parameter models, the WF rotor designs consist
of three materials:
(a) Air: for all air-gaps inside and around the machine.
(b) US Steel Type 2-S, 0.018-inch thick laminations: for the stator and rotor parts of
the machine.
(c) Copper: for both field excitation winding circuits on the rotor and phase winding
circuits on the stator.
Figure 3.15 illustrates FEA results for the three WF rotor designs satisfying the
above design criteria.
To better inform the choice of WF rotor design, some analysis was carried out
utilizing (3.30) (3.31), (3.32), (3.33) and (3.34) and considering two test cases:
Case (1): Calculating the WF copper loss and required supply voltage Pfc and
Vf respectively, when the rotor slot current density Jrs is kept constant at 3.7
MA/m2. Here, field current If and rotor tooth coil turns nf are varied to choose the
most appropriate number for nf.
Case (2): Calculating the air-gap flux density Bgn, and Jrs, Pfc, and Vf when nfis kept
constant and If varies.
The analysis test cases will support the decision on the adequate excitation field
number of turns and the maximum allowable DC excitation current value for the
selected WF rotor design.

J rs Ars
nf ¼ ð3:30Þ
If
3.3 WF Machine 67

Fig. 3.15 FEA results for three WF designs considered for detailed study. (a) Design 1 (b) Design
2 (c) Design 3

Lfc ¼ n f ð2LEFa þ 2πτrewr Þ ð3:31Þ

Pn f Lfc
Rf ¼ ð3:32Þ
σkP Ars

Pfc ¼ I 2f R f ð3:33Þ
68 3 Hybrid Permanent Magnet Machine Design

Fig. 3.16 Case (1) results with a vertical line intersection at 35.9 V; 100 turns

Table 3.5 Case (1) results WF rotor design Bgn (T)

when Jrs constant at 3.7
1 0.374
MA/m2
2 0.153
3 0.183

Table 3.6 Summary of the case (2) results when nf equals 100 turns
WF rotor design If (A) Jrs (MA/m2) Pfc (W) Vf (V) Bgn(max) (T)
1 4.2 7.07 288.0 68.8 0.666
2 5 13.8 666.4 133 0.529
3 6 14.4 834.7 139 0.682

V f ¼ I fRf ð3:34Þ

where τrewr is the rotor mean end winding radius, Lfc is the copper length per
excitation field coil, kp is the slot fill factor, σ is the conductor conductivity, and Rf
is the wound rotor field coil resistance.
The results of test case (1) illustrate that Vf for all wound field rotor designs equals
35.9 V when nf equals 100 turns, but with different copper losses, as shown in
Fig. 3.16. Furthermore, Table 3.5 compares air-gap maximum flux density values
when Jrs equals 3.7 MA/m2. Thus, test case (1) shows that design 1 presents the
preferable rotor design geometry.
The results of test case (2) are summarized in Table 3.6 and Fig. 3.17 and
demonstrate that WF rotor design 1 is considered the best choice of the three designs.
The maximum air-gap flux density for all of the WF machine designs is around
0.67 T; however, the slot current density and copper losses at this air-gap flux density
are high for designs 2 and 3. Thus, WF rotor design 1 represents the best choice since
it can realize Bgn(max) with the lowest values of If, Jslot, Pfc, and Vf.
3.4 HPM Machine Parameters 69

Fig. 3.17 Air-gap flux versus slot current density for the three WF rotor designs

Thus, rotor design 1 was adopted as the most suitable solution and used in the
design of a WF machine having the same stator cross section as the benchmark
brushless PM design, but with a 10-mm as opposed to a 25-mm active axial length.
Applying full rotor field excitation, the open-circuit back-EMFs of this WF
machine design are illustrated in Fig. 3.18. Note that Fig. 3.7 is repeated so that
the back-EMFs due to the 25-mm PM machine and the 10-mm WF machine may be
compared. The choice of active axial lengths for the PM and WF rotor machines
simply comes from a consideration of the contribution of the WF machine back-
EMF to that of the PM machine back-EMF.

3.4 HPM Machine Parameters

The FEA is used again in this section to solve the machine electromagnetic field and
hence determine the machine parameters in the presence of complex magnetic circuit
geometry and nonlinear material properties.

3.4.1 Torque Prediction and Saturation

Electromagnetic torque developed in the machine stator-to-rotor air-gap can be

obtained for each rotor angular increment by using the FEMM solver Maxwell’s
stress tensor integral routine [48]. This method is a standard FEA technique based on
prescribing a force per unit area produced by the electromagnetic field on a defined
surface [48]. Torque is computed by integrating the stress tensor along a line running
through the center of the air-gap between the rotor and the stator, as illustrated in
70 3 Hybrid Permanent Magnet Machine Design

Fig. 3.18 WF machine open-circuit back-EMF waveforms at 3000 rpm. (a) Flux-linkage wave-
forms for three consecutive stator coils (b) Back-EMF waveforms for three consecutive stator coils
(c) Back-EMFs per phase and for three consecutive stator coils

Fig. 3.19 [47]. The same contour is used for the WF machine. For both PM and WF
stators, current densities are defined in the three-phase stator winding slots as per the
phasor relationship illustrated in Fig. 3.20. With the WF current set at a normal
maximum of 4.2 A, as will be clarified later, torque as a function of rotor position can
3.4 HPM Machine Parameters 71

Fig. 3.19 FEMM model of

PM machine section,
highlighting the
circumferential contour line
taken for Maxwell stress
integration

Fig. 3.20 Phasor

relationship of stator
currents

Fig. 3.21 Predicted torque versus rotor angle for the HPM machine sections
72 3 Hybrid Permanent Magnet Machine Design

Fig. 3.22 Magnetic saturation characteristics of the PM and WF HPM machine parts. (a) PM peak
torque characteristic (b) WF peak torque characteristic

be calculated, as illustrated in Fig. 3.21. The HPM electromagnetic capability is

determined by gradually increasing the three-phase stator current magnitudes until
the developed peak torque per ampere saturates, as illustrated in Fig. 3.22 for the
(a) PM and (b) WF machine sections. Based on the stator current density of both the
PM and WF machine parts, the peak power of each part can be calculated individ-
ually by (3.35), where ωs is the mechanical angular speed of the rotor in rad/sec. The
results show that the peak power that can be achieved from the PM part equals
5311 W and 2357 W for the WF part due to the relatively high rotor current density
(7.07 MA/m2). However, the peak torque and power capabilities will be duty rated
due to the machine internal losses, such as iron and copper losses. Thus, the machine
thermal power rating is less, as will be shown later.
3.4 HPM Machine Parameters 73

PRated ¼ T d ωs ð3:35Þ

3.4.2 Synchronous Inductance and Winding Resistance

The HPM generator self- and mutual-inductances are predicted using 2D FEA
models and an empirical factor applied to account for end-winding. To consider
the contribution from the self- and mutual-inductance components, the series coils
are reduced to a single effective coil, and the inductances are calculated for WF and
PM sections separately. For the three-phase machine configuration, the spatial
displacement between the single stator coils is 120 electrical, as illustrated in
Fig. 3.23. The machine magnetic field is then solved with and without the PM or
WF rotor excitation. The self-inductance of this single effective coil, which repre-
sents the equivalent phase coil, can be calculated by (3.36) [54]. For the case of the
PM part of the HPM machine, the rotor PM pieces are changed to air in the FEA
model, DC current is injected in the phase coil for which the self- and mutual-
inductances are calculated, and the other phase currents are set to zero. The FEA
results for the self-inductance is based on (3.36), while the mutual-inductance
between, for example, phases (A) and (B), is based on (3.37) [54]:

Fig. 3.23 Single coil

representation of three-
phase HPM machine
74 3 Hybrid Permanent Magnet Machine Design

Table 3.7 HPM generator inductances calculated from FEA with and without rotor
Mutual inductances between phases as a % of
LSelf
Item Stator LSelf (μH) MAB MAC
Three-phase PM 146.0 7.5 7.5
(123)a (9.1)a (9.1)a
WF 178.7 4.9 4.9
(49.3)a (9.1)a (9.1)a
WF rotor inductance (with stator) Mutual inductances between WF and stator phases as a % of
Lf
Item Lf (mH) MfA MfB MfC
Three-phase 68 2.7 0.56 3.4
a
Values without rotor

Table 3.8 HPM generator measured inductances and resistance

Inductance (μH) (with rotor)
End-
Stator Self-winding winding Resistance (mΩ)
Item section FEA Measured 2Lend Predicted Measured
Three- PM 146.0 175 29.0 40.84 50.00
phase WF 178.7 308 129 20.95 31.00
WF rotor (with stator) 68
103 91
103 23
103 13.06
103 11.35
103

λ e ne λ t
LSelf ¼ ¼ ð3:36Þ
I in I in

λel
M¼ ð3:37Þ
I in

where λe is the total phase flux linkage, ne is the number of turns per phase, λt is the
flux linkage per turn per phase, Iin is the injected phase current and λel is the flux-
linking with the coupled coil of interest, for example, the effect of phase A excitation
on phase B. A similar procedure can be followed to obtain the self- and mutual-
inductances for the other phases.
Table 3.7 lists predicted inductances via FEA while Table 3.8 compares the
predicted and measured inductances and stator per-phase and WF rotor resistances.
Note that the stator inductances are calculated for two cases: (i) with the presence of
the PM and WF rotor sections and (ii) without rotor sections. Since the inclusion of
the rotor reduces the reluctance of flux paths, the inductances are higher than the
inductances calculated without rotors as seen from Table 3.7. The 2D FEA is used to
calculate the machine inductances as it concludes more accurate results compared to
the EMC model due to the inclusion of flux leakages. Predictions of synchronous
inductance do not include end-winding effects due to the 2D FEA limitations. The
end-winding effect on inductance can be calculated from 3D FEA proposed in [55]
3.5 HPM Machine Final Design Model Analysis 75

or from a series of measurements made on machines of varying axial length. The

latter is used in this chapter where actual winding self-inductances of the assembled
HPM machine stators were measured during the machine assembly and included in
the self-inductance as follows:

Lself m ¼ 2Lend þ Lself FEA ð3:38Þ

where Lend is the end winding inductance and Lself  m and Lself  FEA are measured
and predicted self-inductances, respectively. Thus, as the machine axial length
increases, the impact of the end-winding inductance on total inductance reduces,
as seen in Table 3.8.

3.5 HPM Machine Final Design Model Analysis

The final design specification of the proposed three-phase HPM machine is illus-
trated in Table 3.9. The predicted HPM machine rotor PM demagnetization along
with core and copper losses will be analyzed based on this final design specification.

3.5.1 Rotor PM Demagnetization

The hard magnetic material used on the PM rotor is sintered neodymium–iron–boron

(NdFeB) that has a linear second-quadrant B–H characteristic at 20  C [51]. How-
ever, at higher temperatures, say 120  C, the second-quadrant characteristic becomes
nonlinear and there is potential for recoil working of the material. In terms of recoil
working, an important area to consider is demagnetization or partial demagnetization
due to the thermal operating environment of the permanent magnet. Figure 3.24
illustrates a permanent magnet material showing a typical open-circuit load line and
how the second-quadrant characteristic becomes nonlinear with elevated tempera-
ture. N35H is close to the design material specification used for the prototype PM
machine; thus, 120  C operation could be problematic. To check the potential for
rotor demagnetization, the FEMM post-processor file information was analyzed to
calculate the B–H working point of each element defined in the rotor FEA mech.
Figure 3.25 illustrates one of the rotor PMs with its internal mesh elements (tri-
angles). Each of these elements has a flux density value that varies with the element
and rotor position. The flux density value for each rotor PM element can be projected
on the PM material demagnetization curve to ensure that they fall within the desired
operating limits or not. In this analysis, each material specified in the 2D FEMM
machine model contains many numbers of elements. Each element is produced by
three nodes, as shown in Fig. 3.25b, and each node has an X, Y coordinate and vector
76 3 Hybrid Permanent Magnet Machine Design

Table 3.9 HPM generator final design specifications

HPM stator section
Item PM section WF section
Stack length (mm) 25 10
Clearance between PM and WF rotors (mm) 5
Yoke outer diameter (mm) 230.4
Yoke inner diameter (mm) 204.4
Yoke thickness (mm) 16.5
Stator tooth-body width (mm) 9.2
Stator tooth-body length (mm) 16.7
Stator tooth-tip width (mm) 12.4
Stator tooth-tip length (mm) 1.9
Stator slot opening (mm) 2.45
Stator slot area (mm2) 121.2
Number of (stator slots/rotor poles) (36/32)
Number of stator turns per coil 6
Number of rotor turns per coil 100
Air-gap thickness (mm) 0.8 0.4
Shaft outer diameter (mm) 130
Shaft outer diameter (mm) 40
Magnet thickness (mm) 2.8
Magnet length (mm) 14.4
Rotor yoke outer diameter (mm) 160 122
Rotor yoke inner diameter (mm) 130 109.2
Rotor tooth-body width (mm) – 8.54
Rotor tooth-body length (mm) 27
Rotor tooth-tip width (mm) 13.2
Rotor tooth-tip length (mm) 1.3
Rotor slot opening (mm) 3.2
Rotor slot area (mm2) 59.3

potential. Equations (3.39), (3.40), (3.41) and (3.42) [40] are solved using data from
the FEMM solver.
   
2Av ¼ X j Y k  X k Y j þ ðX k Y i  X i Y k Þ þ X i Y j  X j Y i ð3:39Þ

    
X k  X j Avi ðX i  X k ÞAvj X j  X i Avk
BX ¼ þ þ ð3:40Þ
2Av 2Av 2Av
3.5 HPM Machine Final Design Model Analysis 77

Fig. 3.24 Rotor permanent magnet material characteristics [56]

    
Y j  Y k Avi ðY k  Y i ÞAvj Y i  Y j Avk
BY ¼ þ þ ð3:41Þ
2Av 2Av 2Av

qffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
jBj ¼ B2X þ B2Y ð3:42Þ

Figure 3.25c illustrates the rotor permanent magnet operating points plotted on
the material second-quadrant B–H characteristic and for a typical full-load operating
point. Note that this analysis is further used to assess the HESM and DESM
topologies.

3.5.2 Core Loss Prediction

The prediction of iron losses in rotary machines has been discussed in many papers
[57–61]. The most commonly used iron loss formulas divide the losses into hyster-
esis, eddy current, and excess losses. Li et al. discuss the separation of iron losses
general equation into two equations to calculate iron losses [59], such that one of the
equations relates to the existence of alternating magnetization and the second one is
related to the rotary magnetization iron losses. In addition, Li et al. calculate iron
losses via four different methods showing that there were no significant disparities
between the analytically obtained iron losses [59]. Boglietti et al. presented the well-
known iron loss model; however, some of the new coefficients in their equation
78 3 Hybrid Permanent Magnet Machine Design

Fig. 3.25 PM pole in FEMM post-processing mode. (a) PM pole showing FEA mesh triangles (b)
Stylized close up of a PM pole illustrating element coordinates (c) Typical PM operating points at
full-load current

depend on the knowledge of the chemical and physical characteristics of the

considered magnetic material [59]. In this section, no separation of these iron loss
elements will be applied and iron loss will generally be predicted by polynomial
regression used to curve fit actual iron loss data of United Laminated steel material
(0.47 mm, 26 gauge) [62], obtained at several frequencies. The manufacturer’s
measured iron loss curves present the iron losses in watts per kilogram versus
magnetic flux density, as illustrated in Fig. 3.26. Therefore, the general equation
that represents those iron loss curves is to be obtained in order to predict the iron loss
curve at any operating frequency. Here, two main steps have been applied via
polynomial regression before solving the final iron loss equation, as in (3.43). Iron
loss due to current harmonics is not included in the following loss prediction steps.
3.5 HPM Machine Final Design Model Analysis 79

Fig. 3.26 Manufacturer’s

loss data for United
Laminated steel (0.47 mm)
[62]

Piron ðBÞ ¼ k e B3m þ kh B2m þ k a Bm þ kd ð3:43Þ

where the Piron is the iron loss, Bm is the peak flux density, and ke, kh, ka, and kd are
the polynomial coefficients (frequency-dependent terms).
Hence, the first step in this derivation was to get the approximate loss curve via
polynomial regression for each of the manufacturer’s iron loss curves as in (3.43).
The second step has been applied by using the approximate obtained curve equations
to plot the different ke coefficients via their respective frequencies. Here, again an
equation via polynomial regression is derived as in (3.44), where in this equation ke
is a function of frequency and ae2, ae1, and ae0 are the regression coefficients.

ke ð f Þ ¼ ae2 f 2 þ ae1 f þ ae0 ð3:44Þ

By the second step procedure, kh, ka, and kd are presented as in (3.45), (3.46) and
(3.47).

kh ð f Þ ¼ ah2 f 2 þ ah1 f þ ah0 ð3:45Þ

ka ð f Þ ¼ aa2 f 2 þ aa1 f þ aa0 ð3:46Þ

k d ð f Þ ¼ ad2 f 2 þ ad1 f þ ad0 ð3:47Þ

Table 3.10 presents the obtained regression coefficient values that are used to get
an approximate value for ke, kh, ka, and kd at different frequencies and then be used to
solve (3.43), as in Fig. 3.27, for different flux density peak values.
80 3 Hybrid Permanent Magnet Machine Design

Table 3.10 Coefficients of Regression coefficients

Eqs. (3.44), (3.45), (3.46) and
ae2 0.000100
(3.47)
ae1 0.004100
ae0 0.287300
ah2 0.000040
ah1 0.000200
ah0 0.874300
aa2 0.000010
aa1 0.025900
aa0 1.076700
ad2 0.000007
ad1 0.003000
ad0 0.148000

Fig. 3.27 Curve fitting for HPM machine core loss prediction

The last step of this HPM machine iron loss prediction is to find the mass of each
stator section, such as stator back iron, tooth, and tooth tip, as shown in Fig. 3.28,
since the iron loss curves are given in W/kg. Then, the approximated iron loss curves
are used to find the iron loss of each part depending on the peak value of its flux
density waveform. Finally, the loss from the different stator parts is summed for both
HPM machine parts (PM and WF stators) to get the total machine iron loss at the
specified rotor speed. Table 3.11 details the different stator section masses of the
HPM machine parts, their calculated iron loss, and the total mass and iron loss. Thus,
by (3.44), Fig. 3.27, and Table 3.11, the predicted HPM machine iron losses at
3 krpm and with no-load and rated WF excitation equal 411.8 W.
3.6 HPM Machine Thermal Model 81

Fig. 3.28 Stator sections

for iron loss calculations
with FEA results at no-load
condition (PM part)

Table 3.11 Different stator and rotor sections masses and core losses for the fundamental
harmonic only of three-phase PM and HPM machine at no-load
HPM machine
WF Total
PM stator WF stator WF rotor LWFa
LPMa (25 mm) LWFa (10 mm) LWFa (10 mm) (10 mm)
Machine Mass Core Mass Core Mass Core Core
section (kg) Losses (kg) Losses (kg) Losses Losses
(W) (W) (W) (W)
Back-iron 1.748 34.60 0.700 8.85 0.139 28.80 37.65
(19.8 W/ (12.6 W/ (207.2 /
kg) kg) kg)
Teeth 0.968 191.20 0.387 38.50 0.472 90.50 129.00
(197.5 W/ (99.4 W/ (192.2 /
kg) kg) kg)
Teeth-tip 0.148 13.80 0.059 2.96 0.055 2.39 5.35
(93.2 W/ (49.8 W/ (43.2 W/
kg) kg) kg)
Total 2.865 239.60 1.146 50.30 0.665 121.90 172.20

3.6 HPM Machine Thermal Model

3.6.1 General Principle of the Lumped Parameter Method

Heat dissipation in all energy generation and conversion machines arise as a

consequence of losses and need to be assessed since they may severely affect
machine integrity and lifetime. As NdFeB magnet materials are temperature
82 3 Hybrid Permanent Magnet Machine Design

sensitive, it is necessary to accurately predict temperature distribution to prevent

damage and demagnetization. Thus, the importance of thermal management within
machines has led to a wider range of work devoted to the development of electric
machine thermal models [63–71]. Therefore, detailed thermal modeling of the HPM
machine integral parts is required since the knowledge of thermal behavior in
different cases and conditions can prevent overheating and improve the utilization
of the machine. In this section, a general description of the thermal model for both
PM and WF machine sections will be developed via commercial software along with
thermal analysis results.
As with electromagnetic analysis, there are two main methods used for thermal
modeling: lumped parameter (analytical) and the finite-element method [63–
67]. Finite-element methods have some undesirable deficiencies due to the compu-
tation times, 2D cross-section simplification, and loss of accuracy [70]. Thus, a
commercial software package, the Motor-CAD [66] program, which is based on the
lumped parameter method, will be utilized in this thesis to obtain the equivalent
thermal model and predict the temperature on the machine sensitive parts, based on
natural convection cooling for both PM and WF machine sections.

3.6.2 Conduction Heat Transfer

The principle of the lumped parameter method consists of dividing the machine into
basic thermal elements that represent a combination of conduction, convection, and
radiation heat transfer processes. Heat is transferred in electric machines by means of
conduction in solid and lamination parts and by convection and radiation between
surfaces that are in contact with air. The determination of thermal constraints is an
essential prerequisite for estimating these temperature distributions in order to ensure
that constraints imposed, for example, by insulating and PM materials, are not

Fig. 3.29 Illustrative lumped parameter thermal networks for three heat flow directions
3.6 HPM Machine Thermal Model 83

violated. Here, the lumped parameter method uses T-equivalent circuit, as shown in
Fig. 3.29, to represent the relationship between the heat flow and the main compo-
nent temperature in the axial, radial, and circumferential directions, which have been
used for the thermal analyses of some electric machines [63, 64].
In each network, one of the terminals represents the main temperature θm of the
component at which any internal heat generation (u) or thermal storage capacitance
(C) is introduced. The other two terminals represent the appropriate surface temper-
ature of the component. For each network, there is a central node that gives the main
temperature of the component if there were no internal heat generation or storage
[63]. The values of the thermal resistances in each network are derived from the
independent solutions of the heat conduction equations in the axial, radial, and
circumferential directions. These thermal resistances are given in terms of the
dimensions of the general cylindrical shape and the axial, radial, and circumferential
thermal conductivities Ka, Kr, and Kc as shown in (3.48), (3.49), (3.50), (3.51),
(3.52), (3.53), (3.54), (3.55) and (3.56) [63, 64].

L
R1a ¼   ð3:48Þ
2πK a r 21  r 22

L
R2a ¼   ð3:49Þ
2πK a r 21  r 22

L
R3a ¼   ð3:50Þ
6πK a r 21  r 22

2  3
1 4 2r 22 ln rr12
R1r ¼ 1  2 5 ð3:51Þ
4πK r L r 1  r 22

2  3
2 r1
1 4 1 2r ln
  15
r2
R2r ¼  2 ð3:52Þ
4πK r L r 1  r 22

2  3
1 4 2   4r 21 r 22 ln rr12
R3r ¼ r1 þ r2 2
  2  5 ð3:53Þ
8πK r L r 1  r 22
84 3 Hybrid Permanent Magnet Machine Design

2π
R1c ¼  ð3:54Þ
r1
2LK c ln r2

2π
R2c ¼  ð3:55Þ
r1
2LK c ln r2

2π
R3c ¼  ð3:56Þ
6LK c ln rr12

Hence, the negative values of the interconnected thermal resistances R3a, R3r, and
R3c show that the main temperature θm, which is related to all three networks, has to
be lower than the temperature given by the central node. The total thermal capaci-
tance of the cylinder geometry is shown in (3.57) [63].
 
C ¼ ρcΡ π r 21  r 22 L ð3:57Þ

where ρ is the material density, cΡ is the material specific heat, r1 is the cylinder outer
radius, r2 is the cylinder inner radius, and L is the axial length.

3.6.3 Convection Heat Transfer

There are two types of convection heat transfer related to electric machines, natural
convection, and forced convection. When heat is removed by means of ventilation or
circulating liquid inside the machine, it is called forced convection heat transfer.
Natural convection means that neither an external blower nor any coolant liquids
exist [72]. Natural convection heat transfer is a primary function of the temperature
difference between the component and the internal or external air. These types of
convections occur via the air next to the heated body. For example, electric machine
convection heat transfer can occur in the air-gap, endcap air, and at its surface. The
thermal resistance Rc that can be utilized to model the convection heat transfer at the
frame, air-gap, and endcap air is calculated as in (3.58) [63–71].

ðT s  T a Þ 1
Rc ¼ ¼ ð3:58Þ
Q hAca

where Ta presents the ambient temperature, Ts is the emitting surface, Q is the

heat dissipation, Aca is the surface area in contact with the ambient, and h is the heat
transfer coefficient.
3.6 HPM Machine Thermal Model 85

The main difficulty in the previous equation lies with the calculation of the heat
transfer coefficient, which depends on many variables such as the temperature
difference between the heated component and air, the component geometry, and
the property of the component surface. However, accurate values for the previous
variables are very hard to obtain since electric machines are constructed in different
manners and shapes. Therefore, the previous obstacle is resolved by using a proven
empirical formulation (correlation), which is provided by Motor-CAD that gives a
good approximate value for the heat coefficient for any convection surface in the
machine [66]. Natural and forced convection from simple electric machine surfaces,
such as cylinders, flat plates, and more complex surface structures such as open and
closed channels of various shapes and sizes, can be found in [73]. Such correlations
are usually based on empirical dimensionless analysis. Some of the dimensionless
numbers like Nusselt (NNu), Reynolds (NRe), Grashof (NGr), Prandtl (NPr), Taylor
(NTa), and Rayleigh (NRa) are used as in (3.59) and (3.60) [66, 73].
For natural convection, the typical form of correlation is

N Nu ¼ aðN Gr N Pr Þb ð3:59Þ

and for forced convection, the typical form is

N Nu ¼ aðN Re Þb ðN Pr Þc ð3:60Þ

where a, b, and c are constants giving in the Motor-CAD correlation.

For various electric machine structures, the magnitude of NRe is used to judge if
there is laminar or turbulent flow in forced convection systems. Similarly, the NRa
number, which is equal to NGrNPr, is used in natural convection systems [66, 73]. In
addition, the NTa number can be used to judge if there is a laminar, vertex, or
turbulent flow in the air regions around the rotating rotor since the case is different
with these regions [63, 71, 72]. However, some adjustments have to be made to the
heat transfer coefficients, which are related to the air regions around the rotor, where
stator or rotor slot opening is not included in the NTa number. Empirical results
obtained by Gazley [63, 66] suggest that if the flow is laminar and the slots are in the
rotor side only, then there is a decrease by 10% in the heat transfer compared to the
smooth air-gap, and if the slots are in both sides, stator and rotor, then the decrease is
by 20%. However, if the flow is turbulent and the slots are in the stator side only,
then there is an increase by 10% in the heat transfer compared to the smooth air-gap,
and if the slots are in the rotor side only, then the increase is by 20% [66]. Turbulent
flows give enhanced heat transfer, and they add resistance to the heat flow path in a
forced convection system [66]. In addition, if there is a mixed heat transfer due to a
combination of both natural and forced convection, then (3.61) is required to
compensate for this heat flow mixture, in which the machine orientation determines
whether the sign between the natural and forced heat coefficient is positive or
negative [66, 72]:
86 3 Hybrid Permanent Magnet Machine Design

h3MIXED ¼ h3Forced  h3NATURAL ð3:61Þ

where a positive sign is used for assisting and transverse flow and a negative sign is
for the opposing flows.
Each surface requires one value of heat transfer coefficient for stationary case (hs)
and another heat transfer coefficient value for the rotating case (hr) [63]. The heat
transfer coefficient between the rotating body and air-gap can be defined in terms of
the dimensionless NNu number, air-gap length Lg, and the thermal conductivity of air
Kair, as in (3.62).

N Nu K air
h¼ ð3:62Þ
Lg

3.6.4 Radiation Heat Transfer

Radiation heat transfer can be described as the transfer of heat from a surface due to
energy transfer by electromagnetic waves [74]. Such that, (3.63) [74] describes the
radiation resistance of the lumped circuit:

1
RR ¼ ð3:63Þ
Aco hR

where Aco is the component surface area and hR is the heat transfer coefficient.
The heat transfer coefficient is calculated from [74].
 
σ s ε1 F 12 T 41  T 42
hR ¼ ð3:64Þ
T1  T2

where σ s is the Stefan–Boltzmann constant, ε1 is the emissivity of surface 1, F12 is

the view factor (how well surface 2 is viewed by surface 1), T1 is the absolute
temperature of surface 1, and T2 is the absolute temperature of surface 2.

3.6.5 HPM Machine Thermal Model

The Motor-CAD software is used to model and solve the machine equivalent thermal
network by inputting the desired machine geometric data using the graphical radial
and cross-section editor, along with material names and predicted machine losses.
The thermal model of the HPM machine, which does not account for the circum-
ferential heat flow, is dealt with as two individual thermal machine models, which
3.6 HPM Machine Thermal Model 87

Fig. 3.30 PM machine

steady-state thermal analysis
results via Motor-CAD (a)
Radial cross section (b)
Axial cross section

represent the PM and WF parts, due to the limitation imposed by the program. This
thermal program limitation has minor impact on the predicted temperatures of the
HPM machine parts due to the separation distance between the PM and WF sections
88 3 Hybrid Permanent Magnet Machine Design

Fig. 3.31 WF machine

steady-state thermal analysis
results via Motor-CAD. (a)
Radial cross section (b)
Axial cross section

of the machine. The thermal parameters such as conduction, convection, radiation,

and thermal resistances are calculated by the program, and hence the machine
thermal model is created.
For the housing outer cooling by natural convection case, the thermal analysis
results for both machines (PM and WF) at steady state are illustrated in Figs. 3.30
and 3.31, where the transient temperature curves are shown in Figs. 3.32 and 3.33.
The thermal analysis results were satisfactory for the PM machine; however, the WF
3.6 HPM Machine Thermal Model 89

Fig. 3.32 PM machine thermal transient analysis curves via Motor-CAD

Fig. 3.33 WF machine thermal transient analysis curves via Motor-CAD

machine thermal analysis showed an elevated temperature, especially in the rotor,

such as 189.7  C at the rotor winding and 194.2  C at the rotor tooth. These high
temperatures are due to the relatively high rotor copper losses caused by the rated
DC excitation current, which in this analysis is 4.2 A. However, the required DC
excitation current to regulate the selected HPM machine-rectified DC-link voltage
when the machine system is placed in a hybrid EV powertrain is actually discrete and
less than the rated value, as will be discussed in Chap. 5, and this in turn will
decrease the rotor copper losses. For example, if the rated excitation current
decreased by half the copper losses will be 75% less than the rated copper losses,
and an acceptable steady-state machine internal part temperatures will be gained.
Hence, the applied thermal analysis is a good starting point to predict the HPM
machine distributed temperatures before embarking with the machine actual testing
at different operating conditions.
Thus, different excitation current values are analyzed next to monitor the decrease
in temperature in the rotor tooth, which present the hottest rotor spot, while
90 3 Hybrid Permanent Magnet Machine Design

Fig. 3.34 WF machine rotor tooth thermal transient analysis curves for different excitation currents
via Motor-CAD

maintaining the machine parts core losses constant as for the worst case (rated
excitation current). Figure 3.34 shows the decrease in the WF rotor tooth temperature
due to the decrease in the excitation current, where the temperature decreased from
194 to 85  C for If equals 4.2 to 0.5 A, respectively. Note that the tooth temperature
did not decrease beyond 85  C due to the fixed core losses applied to the Motor-
CAD, which have been predicted previously at rated speed and excitation current
condition.

3.7 Comparison Between PM and Four HPM Machine

Topologies

The discussion of the HPM machine has so far been centered on the HESM
topology, that is, a two-part machine having separate PM and WF sections. The
electromagnetic and the thermal analysis tools discussed in the preceding sections
were used to assess four contending HPM machine topologies, as detailed in
Chap. 2, and compare them against the benchmark brushless PM machine previously
discussed. The selected topologies are illustrated in Fig. 3.35 along with the refer-
ence PM machine. The machines are compared for the same volume envelope and in
terms of their maximum phase back-EMF when the DC field excitation current
equals zero and when it is varied with both positive and negative flux paths, and
whether the permanent magnets are at a risk of demagnetization due to the total field
weakening by the wound field excitation, as summarized in Table 3.12. The volumes
and component masses for the different HPM machines are calculated and compared
in Table 3.13.
In all of the selected HPM machine topologies, the magnetic flux path is radial
except the CPPM design, which has both radial and axial flux paths when it is
operated with zero or negative excitation currents and a radial flux path when
3.7 Comparison Between PM and Four HPM Machine Topologies 91

Fig. 3.35 PM and several HPM machine models used in the analysis. (a) PM (b) HESM (c) DESM
(d) SynPM (e) CPPM

Table 3.12 Comparison of maximum back-EMF per phase for the selected HPM topologies and
the reference PM machine
Machine type
PM HESM DESM CPPM SynPM
EMFPk/phase (V) Excitation
If ¼ +4.2 A 60 70.21 30.23 (65.39) 54.53
If ¼ 4.2 A 30 54.50 21.99 32.24
If ¼ 0 A 70 45 64.55 26.10 (65.39) 44.53
% variation of EMFmax/phase 0 66.7 24.3 31.57 24.4
PM demagnetization No No Some No No
Rotor excitation No Yes Yes No Yes
Note: Rotor Jrs ¼ 7.07 MA/m2; nf ¼ 100 turns; speed ¼ 3 krpm; and machine axial length ¼ 0.04 m
92 3 Hybrid Permanent Magnet Machine Design

Table 3.13 Total volume and mass for the different HPM machines
Machine type
PM HESM DESM CPPM SynPM
Copper (kg) 0.652 0.967 1.406 0.917 0.841
Iron (kg) 10.89 9.923 9.579 10.85 10.58
PMs (kg) 0.382 0.239 0.382 0.167 0.312
Total mass (kg) 11.93 11.23 11.37 11.81 11.93
Total volume (1023 m3) 1.48 1.45 1.45 1.48 1.48

Fig. 3.36 Back-EMF waveform for SynPM and the PM machine. (a) SynPM back-EMF wave-
forms (b) PM back-EMF waveform

operating with positive excitation. The analysis results demonstrate that all of the
HPM topologies are capable of producing similar back-EMF waveforms that may be
varied, to a greater or lesser extent, by the secondary source of field excitation.
However, the topologies demonstrate different peak amplitudes depending on
the excitation current magnitude and polarity except for the SynPM. Excitation of
the SynPM field creates a modulated back-EMF waveform due to asymmetry in the
air-gap field distribution, as illustrated in Fig. 3.36a and presented in Table 3.12
where the data in parenthesis relates to the PM element of back-EMF and the lower
values to the peak of the modulated period. In the SynPM design, some of the rotor
poles consist of permanent magnets, B, C, and D in Fig. 3.35d, with an iron tooth
third pole, E in Fig. 3.35a. Therefore, the resultant SynPM back-EMF waveform is
modulated by the rotor excitation field. This feature is undesirable in the application
considered since it leads to a modulation of the machine back-EMF and hence the
connecting rectifier DC output.
This modulation feature only arises because of the stator winding and pole
number combination constraint adopted for the comparison study. For example, if
the stator had a conventional short-pitch winding and the rotor pole number was set
to a six-pole pair, the rotor field coil would buck or boost the PM-induced back-EMF
as opposed to modulating the back-EMF, as illustrated in Fig. 3.36a. However, this
3.7 Comparison Between PM and Four HPM Machine Topologies 93

Fig. 3.37 Comparison between PM flux density distribution points for the HESM and DESM
designs when If ¼ 4.2A (field weakening) and at full stator field current. (a) Typical HESM PM
operating points at full-load current (b) Typical DESM PM operating points at full-load current

change in pole number would require deviation from the benchmark PM volumetric
constraints. Furthermore, the rotor field excitation has a demagnetizing effect at high
buck operation, as with the DESM topology of Fig. 3.35c. Hence, the SynPM
topology was not considered for further study.
By comparing the four topologies in terms of their back-EMF production without
excitation current, the DESM shows the highest back-EMF value, but it gives the
lowest voltage regulation at 24%, with full excitation current. The HESM flux
linkage has short radial paths that minimize leakage flux and hence it gives a high
voltage regulation region with rotor excitation at 66%. In terms of rotor permanent
magnet demagnetization, the HESM appears to better withstand demagnetization
compared to the DESM, as illustrated in Fig. 3.37, showing the PM working points
for both the HESM (a) and DESM (b) superimposed on the PM B–H characteristic.
This feature is essential because the rotor excitation flux path does not pass through
the permanent magnet pieces.
The obvious drawback of the HESM and DESM designs is the DC rotor excita-
tion via brushes and slip rings. However, compact exciters would mitigate this
problem. The CPPM has no slip rings and brushes, but it has the lowest back-
EMF and highest leakage flux components. Thus, the HESM presents the most
attractive design among the analyzed topologies since it has the highest back-EMF
regulation capability, no PM demagnetization risk from the rotor excitation field,
low leakage flux, and low part (stator and WF rotor teeth) saturation risk. Moreover,
the HESM mass is similar to the other topologies.
94 3 Hybrid Permanent Magnet Machine Design

3.8 Conclusion

This chapter has discussed the analysis tools and models developed to assess HPM
machine topologies and design the WF rotor of a HESM topology that will form the
base machine in subsequent chapters. The FEA analysis was performed using
FEMM and used to numerically evaluate several electric machines reported in
Chap. 2 against a benchmark PM electromagnetic and volumetric design. Further-
more, several WF rotor designs have been analyzed, from which it was found that the
rotor slot area is a key constraining factor. Hence, WF design 1 represents the best
design geometry among the proposed designs.
The magnetic circuit representation of PM and WF test machines was obtained,
and the magnetic flux was calculated by using a lumped parameter equivalent
magnetic circuit model and checked by using 2D FEA. The analytical and numerical
results for both machines were successfully compared and showed a good correla-
tion. In addition, PM and WF test machines have been investigated for their back-
EMF waveforms (no-load case) and power rating capabilities (on-load case) for a
specified stator slot current density.
In addition, several HPM topologies have been analyzed for the same volume
envelope. The validity of the FEA results for the PM was confirmed by actual
machine measurements. Moreover, the Motor-CAD lumped parameter thermal
network software was used to predict the transient temperature rise in some of the
HPM machine parts, such as back-iron, stator windings, and WF rotor winding. The
predicted results illustrate the temperature rise in the important machine parts due to
the natural convection case. Thus, combining both PM and WF machines in one
machine housing, as with the HESM, will make the most of their respective
advantages.
Chapter 4
Multiphase HPM Generator Systems

4.1 Overview on Multiphase Machines

The concept of multiphase machines is not new, and the roots of multiphase machine
modeling can be traced back to 1959 [75] with many papers published to date [76–
137]. Levi et al. and Bojoi et al. have presented a comprehensive survey of existing
and future applications for multiphase machines [78–80], reviewing machines found
in various applications ranging from electric shipboard generation and ship propul-
sion, hybrid electric vehicles, wind generation and more electric aircraft, to high-
power industrial applications where, for a variety of reasons, a three-phase machine
does not satisfy the desired specification requirements, particularly with regard to
hardware implementation or fault criticality. Greater efficiency, higher torque den-
sity with reduced torque pulsations, lower per-phase power handling requirements,
increased fault tolerance, and improved noise characteristics are the major advan-
tages of multiphase machines over their three-phase counterparts, as reported in [81].
The modeling and optimization of multiphase machines in terms of their power
density and harmonic content is the subject of considerable interest, since the designs
are not standard “off-the-shelf” machines [82, 83]. Levi et al. discuss the degree of
freedom merits that are presented by the choice of phase numbers greater than three,
an example being multiphase power electronic converters that can be utilized to
control independent machines within a multi-motor drive system, or to enhance the
torque production above that of standard three-phase machines by injecting higher
stator current harmonics [78, 84]. In drive system applications, Singh et al. presented
a straightforward approach to developing an indirect field oriented control scheme
for a six-phase induction machine with an arbitrary displacement between the two
three-phase winding sets [77]. In addition, high level drive control via a space vector
pulse width modulation algorithm for a multilevel, multiphase, voltage source
converter was presented by Lopez et al. [81]. The losses and pulsating torque in
multiphase induction machines have been described by Williamson and Smith in
[82]. In addition, Brisset et al. have presented a comparative study of different

© Springer Nature Switzerland AG 2022 95

A. S. Al-Adsani, O. Beik, Multiphase Hybrid Electric Machines,
https://doi.org/10.1007/978-3-030-80435-0_4
96 4 Multiphase HPM Generator Systems

configurations of nine-phase, concentrated winding, axial flux, permanent magnet

synchronous generators for direct-drive wind turbines [83]. Moreover, Arwyn et al.
have investigated a high phase number flux-switching permanent-magnet brushless
machine via 2D FEA. They presented a comparison of the electromagnetic perfor-
mances of three-, four-, five-, and six-phase machine variants within the specific
context of an aerospace application [85]. Arwyn concluded that the high phase
number offered more torque density over conventional three-phase machines along
with lower sensitivity to the level of rotor saliency [85].

4.1.1 Multiphase Windings Principles

To realize near-sinusoidal back-EMF waveform for the AC machines, more slot per
pole per phase is required, and hence, as the phase number increases, it becomes
gradually difficult to achieve that. However, matching a quasi-rectangular back-
EMF waveform by stator current waveform provides enhanced torque production.
The multiphase machines are divided into symmetrical or asymmetrical machines
based on their electrical displacement between two adjacent phases. A symmetrical
multiphase machine is attained once the electrical angular displacement between any
two stator phases equal 2π/n; this can always be achieved if the phase number (n) is
odd prime number, such as 5, 7, 11, .... Where the number of phases is an even or an
odd and not a prime number, such as 4, 6, 8, 9, 10, ..., the multiphase machine phase
winding can be realized in a different way [78]. In this way, the machine windings
can be seen as the number of sets having a phases each. Such that, the number of
phases equals a ∙ k, where for multiphase (above three) machines a ¼ 3, and k ¼ 2,
3, 4, 5, . . ., and the electrical angular displacement between the first phases of two
consecutive sets is π/n electrical degree, leading to asymmetrical multiphase
machine [80]. In the literature, many multiphase machine designs and applications
have been reported, as illustrated in Tables 4.1 and 4.2. It is shown in the tables that
different electric machine types with more than three-phase winding schemes can be
utilized in several industrial applications.

4.1.2 Rectified Voltage due to Three- and Nine-Phase HPM

Generator Systems

An interesting feature of the design studied in Chap. 3 was the stator tooth and rotor
pole combination. The 36-slot, 32-pole combination was primarily a design feature
to help mitigate against cogging torque. However, the HPM-rectified output voltage
of the three-phase winding scheme must be conditioned to remove the harmonics
introduced by the passive rectification stage. Traditionally, this is facilitated via
electrolytic capacitors and line inductors placed on the DC side of the output of the
4.1 Overview on Multiphase Machines 97

Table 4.1 References of multiphase electric machine types and applications

Number of
Application Machine type phases References
Ship propulsion IM 15 [89]
Brushless PM 12 [86, 87]
Synchronous WF 9 [90]
Synchronous WF 6 [91]
Shipboard generation Brushless PM 12 [88]
Synchronous WF 6 [92]
Electric and hybrid electric Brushless hybrid PM 9 [94]
vehicles Brushed hybrid PM 9 [93, 95, 96]
IM 6 and 9 [76]
IM 6 [97]
Brushless PM 5 [98]
Brushless DC 5 [99–101]
Brushless DC 4 and 5 [102]
Wind generation Brushless PM 12 [114]
Brushed hybrid PM 9 [103–109,
115]
Brushless axial flux 9 [83, 111, 112]
PM
Brushless PM 6 and 9 [113]
Brushless PM 6 [110]
IM 6 [116]
More electric aircraft Switch reluctance 6 [119]
Switch reluctance 4 [118]
Brushless PM 3, 4, 5, and 6 [85]
Brushless PM 4 [117]
Brushless DC 5 [120, 121]

Table 4.2 References of general multiphase machines modeling and analysis

Machine type Number of phases References
IM 4, 5, 6, 9, and 15 [82]
IM 3, 4, 6, and 12 [128]
IM 6 [122, 123]
IM 5 [124, 125]
Brushed hybrid PM 9 [46]
Brushless PM 7 and 9 [136]
Brushless PM 7 [133, 134]
Brushless axial flux double rotor PM 7 [137]
Brushless PM 5 [131, 132]
Brushless PM 2, 3, and 5 [135]
Switch reluctance 5 [129, 130]
98 4 Multiphase HPM Generator Systems

Fig. 4.1 Three- and nine-phase HPM generator with power circuit converter. (a) HPM model. (b)
Three-phase. (c) Nine-phase

full-bridge rectifier (AC/DC). The move to higher phase numbers reduces the level
of ripple voltage on the generated DC output voltage and thus reduces the required
capacitor and/or inductor passive device rating, as will be demonstrated later. Thus,
for the chosen machine topology, an evaluation can be made and benefits can be
quantified. Figure 4.1 illustrates the HPM machine showing the wound field rotor
and PM rotor cross sections schematically (a), and the three-phase (b) and nine-
phase (c), equivalent circuits when connected to a simple full-bridge rectifier and
resistive load. In terms of the machine phase peak back-EMF voltage (EMFPk), the
4.1 Overview on Multiphase Machines 99

general form of the average DC-link voltage for loss-less full-bridge rectification is
given by [138]:
Z π=n  
4 2n π
V DClink ¼ EMF Pk cos ðωt Þd ðωt Þ ¼ EMF Pk sin ð4:1Þ
ð2π=nÞ 0 π n

which, for a three-phase system, simplifies to

pffiffiffi
3 3
V DClinkð3phaseÞ ¼ EMF Pk ¼ 1:654EMF Pk ð4:2Þ
π

and for the nine-phase system, (4.1) solves to

6:1564
V DClinkð9phaseÞ ¼ EMF Pk ¼ 1:96EMF Pk ð4:3Þ
π

This average DC output voltage is derived from the machine AC voltage wave-
form, as illustrated in Fig. 4.2 for the three- and nine-phase cases, from which the DC
voltage ripple can be defined:

V r ¼ V DClinkð max Þ  V DClinkð min Þ ð4:4Þ

From Fig. 4.2, it can be seen that the DC-voltage ripple reduces with increased
phase number and has a frequency of 2n times that of the machine fundamental
frequency [139]. Moreover, the move to a higher phase number distributes current
across a larger number of rectifier legs providing the designer with an opportunity to
use rectifiers having lower power loss (but similar peak current) distributed around
the periphery of the machine, leading to the potential for improved packaging and
cooling. With the advent of higher temperature rectifier devices, for example, silicon
carbide (SiC) [140], the rectifiers could potentially be mounted within the machine
housing.
However, at high operating frequency, the machine synchronous inductance, LS,
can have a significant effect on output power and voltage regulation that might lead
to an incorrect decision regarding the feasibility of utilizing the HPM generator with
a passive rectification stage. If a DC voltage is to be generated by a three-phase
machine with a sinusoidal back-EMF via a passive, full-bridge, three-phase rectifier,
then (4.2) is valid only for the no-load case. Under load conditions, there will be
voltage drops due to the diode reverse recovery time (trr) causing commutation
overlap, and due to machine synchronous inductance, LS. If a linear rise of DC
generated current (IDC-link) from zero to the rated value is assumed or a constant di/dt,
the influence of the machine operating frequency, synchronous inductance, and load
current in terms of DC-link output voltage drop can be calculated from [138]:
100 4 Multiphase HPM Generator Systems

Fig. 4.2 AC L-L voltage and rectified DC output for three- and nine-phase machines at no-load. (a)
Three-phase. (b) Nine-phase
4.2 Nine-Phase HPM Generator Parameters 101

V dr ¼ 2 n f e LS I DClink ð4:5Þ

Therefore, limitations are imposed on the DC output power and voltage regula-
tion for both three- and nine-phase configurations; hence, LS is evaluated for both
topologies.

4.2 Nine-Phase HPM Generator Parameters

Having studied the individual coil and phase back-EMFs, as in Chap. 3, the
possibility of connecting the stator as a nine-phase winding became apparent.
Importantly, the stator lamination, rotor magnetization, and other key magnetic
dimensions did not have to be changed; hence the relative merits of a three-phase
versus a nine-phase winding topology could be evaluated fairly. Thus, a general
investigation into multiphase machine topologies is undertaken via a direct three- to
nine-phase stator winding comparison study.

4.2.1 Nine-Phase Winding Layout and Back-EMF

As discussed in Chap. 3, the WF rotor of the HPM generator is fed from an external
DC source via slip rings and brushes. The WF rotor lamination has 32 slots, and the
WF DC winding is formed of 32 series coils woundaround the rotor teeth. Here, the
phase belt (i.e., number of slots per pole per phase) for the nine-phase HPM stator
winding is 1/8, and with subsequence phases being displaced by 40 electrical, hence
a concentric winding with 36 coils wound around stator teeth is realized. For the
three-phase winding, as in Chap. 3, each phase has four sets of three concentrated
coils connected in series while for the nine-phase winding, there are two design
variations where four coils are connected in series, as illustrated in Fig. 4.3. Note that
the middle coil of the three concentrated coils is wound opposite to the other two
coils, which results in the largest amplitude back-EMF in the proposed three-phase
winding. The fundamental winding factor (kw1) is calculated again using the
methods described in [46], which incorporates both the slot pitch and distribution
factor in one equation. Therefore, by numbering the stator slots using vector U, the
phase back-EMF vectors (Ei) and fundamental winding factor (kw1) are calculated
for both nine-phase winding designs using Eqs. (4.5, 4.6, 4.7, 4.8, 4.9, and 4.10).
 
nl S=9 
1  X !
kw1 ¼  E ð4:6Þ
nl S=9  i¼1 i 
102 4 Multiphase HPM Generator Systems

Fig. 4.3 Two nine-phase HPM machine stator lamination and phase (A) winding schemes. (a)
Nine-phase winding design 1. (b) Nine-phase winding design 2. (c) Design 1 (A) phase back-EMF
configuration vectors. (d) Design 2 (A) phase back-EMF configuration vectors

U design 1 ¼ ½1  2  2 3 19  20  20  21 ð4:7Þ

 36 
2 9   
X !
 Ei  ¼ e j S 1  e j S þ 2e2j S þ e18j S  2e19j S þ e20j S
πP πP πP πP πP πP
ð4:8Þ

 i¼1 

U design 2 ¼ ½1  2 10  11 19  20 28  29 ð4:9Þ

 36 
2 9   
X !
 E i  ¼ e jπP
1e jπP πP πP πP πP πP πP
þ e9j S  e10j S þ e18j S  e19j S þ e27j S  e28j S ð4:10Þ

S S

 i¼1 

The nine-phase winding design 2, Fig. 4.3b, results in 1.53% and 4.20% higher
fundamental winding factor and hence back-EMF as compared to design 1 and the
4.2 Nine-Phase HPM Generator Parameters 103

Fig. 4.4 Normalized back-EMF waveforms of the three- and nine-phase HPM machines

previous three-phase windings. Therefore, nine-phase winding design 2 is adopted

as the preferable HPM generator winding scheme. The higher phase voltage trans-
lates to a better power density given a fixed load current. In addition, going to a
higher number of phases, the rectified output voltage has lower ripple, better quality,
and hence lower requirements for filtering. To numerically evaluate the HPM
machine design performance, a reliable model is required. FEA is used to solve
the machine electromagnetic field and hence determine the machine terminal param-
eters in the presence of complex magnetic circuit geometry and nonlinear material
properties, as discussed in Chap. 3. These parameters are then used in a Matlab/
Simulink SimPower model to simulate the machine and nonlinear effects of the
diode rectification schemes.
Phase back-EMF waveforms for the coil distributions shown in Fig. 4.3 were
calculated for the three- and nine-phase designs, as illustrated in Fig. 4.4, showing
the back-EMF coefficient waveforms with normalized 1.0 p.u. magnitudes and
electrical rotor angle. The utility of Fig. 4.4 is simply to illustrate the waveform
shape since it is envisaged that the more trapezoidal EMFs could yield more torque
and hence power capability than a winding designed to realize a sinusoidal back-
EMF. The sinusoidal back-EMF is considered since such a machine leads to easier
analytic performance estimation.

4.2.2 Back-EMF and Torque Waveform Harmonics

Prediction

In this section, 2D FEA analyses are carried out on the back-EMF waveform at
no-load for three- and nine-phase generator winding schemes for the PM, WF, and
the HPM machine. For normalized back-EMF waveforms, a fast Fourier transform
(FFT) is applied on expected operating scenarios in HEV powertrain to investigate
104 4 Multiphase HPM Generator Systems

Fig. 4.5 Amplitudes of normalized back-EMF harmonic orders due to different WF excitation
modes. (a) Three-phase HPM machine. (b) Nine-phase HPM machine

Fig. 4.6 Predicted buck/boost capabilities for back-EMF and torque waveforms of the nine-phase
HPM machine. (a) back-EMF at no-load. (b) electromagnetic torque

the THD based on the fundamental, third, fifth, and seventh harmonic orders.
Figures 4.5a and b show harmonic spectrum for normalized back-EMF voltage
waveform at no-load for three WF excitation scenarios. These scenarios are charac-
terized as (i) boost (If ¼ +3A), (ii) buck (If ¼ 3A), and (iii) normal (If ¼ 0A)
operation modes, as shown for the nine-phase machine in Fig. 4.6a. The FFT
analyses have shown an improvement in the third and fifth harmonic amplitude at
buck mode for three- and nine-phase winding schemes compared to buck and normal
operation modes. In addition, the nine-phase winding scheme showed larger third
and fifth harmonic amplitude compared to the three-phase winding; hence, larger
back-EMF waveform harmonic distortion is expected in the nine-phase HPM
machine scheme.
4.2 Nine-Phase HPM Generator Parameters 105

Table 4.3 Predicted back-EMF THD for three- and nine-phase PM, WF, and HPM machines
EMF EMF1/ EMF3/ EMF5/ EMF7/ THD
Machine type (rms) V EMF (%) EMF1 (%) EMF1 (%) EMF1 (%) (%)
Three-phase PM 0.79 96.3 7.79 0.94 0.05 7.85
Three-phase WF 0.77 96.6 5.88 0.35 0.28 5.90
Three-phase 0.78 96.4 7.35 0.74 0.07 7.39
HPM (boost)
Three-phase 0.80 96.2 8.60 1.32 0.14 8.71
HPM (buck)
Nine-phase PM 0.84 95.5 12.8 2.35 0.07 13.0
Nine-phase WF 0.80 96.2 8.51 1.38 0.59 8.67
Nine-phase HPM 0.83 95.7 11.8 1.51 0.10 11.9
(boost)
Nine-phase HPM 0.85 95.1 14.5 3.88 0.32 15.0
(buck)

Further analysis of the normalized back-EMF waveforms is illustrated in

Table 4.3. The three-phase PM part of the HPM machine has a higher back-EMF
THD than the three-phase WF machine part. The three-phase HPM machine back-
EMF at buck mode showed a higher THD compared with all of the three-phase
machine scenarios, as shown in Table 4.3. Moreover, degradation in the THD
presented by the more trapezoidal back-EMF waveform in the nine-phase PM part
compared with the three-phase and the nine-phase WF part is noticeable. At normal
mode scenario, the THD is 12.98%, and at boost mode, the THD drops to 11.9%;
however, at buck mode, the THD increased to 15.01%. Thus, the nine-phase HPM
machine operating at buck mode scenario provides the highest THD for the normal-
ized back-EMF waveform among all of the scenarios.
Similarly, the electromagnetic machine torque waveform is analyzed via 2D FEA
at rated load for the three- and nine-phase winding schemes. Here, a stator slot
current density of 3.4 MA/m2 is chosen based on the allowable machine thermal
limit at rated load. The HPM machine with three WF excitation scenarios is applied
on the three- and nine-phase winding schemes. As shown in Fig. 4.6b, a 36% buck/
boost torque capability is achieved for the HPM machine.
To investigate the electromagnetic torque THD, an FFT is applied on the
expected HEV powertrain operating scenarios. The second, fifth, eighth, and so on
harmonics are the negative sequence harmonics that create a negative torque, where
the fundamental, fourth, seventh, tenth, and so on harmonics present a positive
torque sequence. Furthermore, the multiple of third harmonics, such as the third,
sixth, ninth, and so on contribute to the zero sequence, which results in a circulating
current in the machine and needs to be accounted for in unbalanced systems.
Figure 4.7 shows the harmonic spectrum for normalized electromagnetic torque at
rated load for three- and nine-phase HPM machines. The FFT shows that at normal
operation (zero WF excitation), some of the nine-phase harmonic amplitudes are less
than three-phase winding scheme, such as the second, fourth, and fifth harmonics.
Hence, the buck mode harmonics are found to be dominant in the three- and
106 4 Multiphase HPM Generator Systems

Fig. 4.7 Amplitudes of normalized electromagnetic torque harmonic orders due to different WF
excitation modes. (a) Three-phase HPM machine. (b) Nine-phase HPM machine

Table 4.4 Predicted electromagnetic torque THD for three- and nine-phase PM, WF, and HPM
machines
T (rms) T1/T T3/T1 T5/T1 T7/T1
Machine type Nm (%) (%) (%) (%) THD (%)
Three-phase PM 0.70 87.2 0.30 0.55 0.04 0.89
Three-phase WF 0.69 87.1 2.93 0.33 0.16 4.03
Three-phase HPM (boost) 0.70 87.2 0.98 0.33 0.11 1.24
Three-phase HPM (buck) 0.70 87.1 1.18 1.04 0.15 2.63
Nine-phase PM 0.70 96.8 0.31 0.12 0.19 0.51
Nine-phase WF 0.69 96.7 1.69 0.04 0.09 4.01
Nine-phase HPM (boost) 0.70 96.8 0.67 0.10 0.14 1.23
Nine-phase HPM (buck) 0.71 96.8 0.45 0.17 0.31 2.04

nine-phase HPM machines. It is also noticed that there is a small increase in the
second and eighth spatial harmonics in the nine-phase winding scheme in the buck
mode, where the other harmonics decrease compared to three-phase winding.
The FFT showed an improvement in the nine-phase HPM machine harmonic
contents over three-phase in boost mode except for the second harmonic, which is
responsible of the excitation of the second mechanical deformation (distorting) and
therefore it is significant [141]. Table 4.4 illustrates the predicted torque THD for the
PM, WF, and HPM machines.
4.2 Nine-Phase HPM Generator Parameters 107

4.2.3 Synchronous Inductance Prediction

The calculation of machine synchronous inductance (LS) based on stator winding

schemes has been discussed in the literature [142–147]. The most common method
for analytically predicting the self- and mutual-inductances based on magnetic flux-
linkage is presented by the winding function method [142]. This method uses a
Fourier series to predict a winding function for the machine phase winding config-
uration and the number of poles. This function is utilized in common flux-linkage,
self-, and mutual-inductance equations [144]. El-Refaie and Thomas have discussed
a closed-form technique for analyzing surface PM machines equipped with
fractional-slot concentrated windings as per the machine being studied. The tech-
nique takes advantage of some established analytical methods to analyze a wide
range of fractional-slot concentrated winding configurations [145]. Furthermore, the
impact of winding layer number on surface PM machine synchronous inductance
was investigated by El-Refaie and Thomas [146]. Hence, based on the comparison
between self- and mutual-inductances via analytic, FEA, and measured values, as
presented by El-Refaie et al. [147], it can be concluded that FEA prediction can give
satisfactory results if the end winding effect is properly taken into account. In this
book, the HPM machine self- and mutual-inductances for the PM and WF machines
are predicted with 2D FEA models, and an empirical factor is applied to account for
end-winding inductances. To consider the contribution from the self- and mutual-
inductance components, the series phase coils of both the three- and nine-phase
machines can be reduced to a single effective coil. The machine magnetic field is
then solved without the PM or WF rotor excitation. The self-inductance of this single
effective coil, which represents the equivalent phase coil, can be calculated by
(4.11). For the case of the PM part of the HPM machine, the rotor PM pieces are
changed to air in the FEA model, DC current is injected in the phase coil for which
the self- and mutual-inductances are calculated, and the other phase currents are set
to zero. The FEA results for the self-inductance is based on (4.11), while the mutual-
inductance between, for example, phases (A) and (B), is based on (4.12) [148]:

λ e ne λ t
LSelf ¼ ¼ ð4:11Þ
I in I in
λel
M¼ ð4:12Þ
I in

where λe is the total phase flux-linkage, ne is the number of turns per phase, λt is the
flux-linkage per turn per phase, Iin is the injected phase current, and λel is the flux-
linking with the coupled coil of interest, for example, the effect of phase A excitation
on phase B. A similar procedure can be followed to obtain the self- and mutual-
inductances for the other phases. Table 4.5 presents the self- and mutual-inductances
for both the PM and WF sections of the HPM machine stators, as predicted by FEA.
Note that calculations are presented with and without the respective rotors since
108 4 Multiphase HPM Generator Systems

Table 4.5 Comparison of three- and two nine-phase HPM generator inductances calculated from
FEA with and without rotor
Mutual inductances between phases as a % of LSelf
MAD MAE
Item Stator LSelf (μH) MAB MAC MAG MAF MAH MAI
Three- PM 146.0 7.5 7.5 Not applicable
phase (123)a (9.1)a (9.1)a
nstc ¼ 6 WF 178.7 4.9 4.9
(49.3)a (9.1)a (9.1)a
Nine- PM 190.5 6.0 5.8 5.9 31.4 5.8 6.0
phase (140.4)a (1.9)a (0.7)a (0.94)a (45)a (0.7)a (1.9)a
nstc ¼ 14 WF 272.7 10.9 10.0 10.3 18.5 10.0 10.9
(56.1)a (1.9)a (0.7)a (0.94)a (45)a (0.7)a (1.9)a
Nine- PM 317.7 6.0 5.8 5.9 31.4 5.8 6.0
phase (233)a (1.87)a (0.71)a (0.94)a (49.7)a (0.71)a (1.86)a
nstc ¼ 18 WF 331.0 10.9 10.0 10.3 18.9 10.0 10.9
(92.7)a (1.87)a (0.71)a (0.94)a (47.3)a (0.71)a (1.87)a
WF rotor induc- Mutual inductances between WF and stator phases as a % of Lf
tance (with stator)
Item Lf MfA MfB MfC MfD MfE MfH MfI
(mH) MfG MfF
Three- 68 2.67 0.56 3.39 Not applicable
phase
nstc ¼ 6
Nine- 68 2.02 1.48 0.29 0.89 1.89 0.29 1.48
phase
nstc ¼ 14
Nine- 68 2.60 1.90 0.38 1.14 2.40 0.38 1.90
phase
nstc ¼ 18
a
Values without rotor

early tests to confirm the FEA predictions were carried out on stator assemblies as
opposed to full machine constructions.
To consider the machine synchronous inductance, the voltage equation for one
phase of a three-phase machine winding, as derived from first principles, is given by
[149]:

dLSelf iA dM AB iB dM AC iC
vA ¼ RA iA þ þ þ þ EMF A ð4:13Þ
dt dt dt

If the machine self- and mutual-inductances are not a function of their respective
phase currents, or rotor position, which is the case for normal operation of the
machine studied, (4.13) simplifies to
4.2 Nine-Phase HPM Generator Parameters 109

diA di di
vA ¼ RA iA þ LSelf þ M AB B þ M AC C þ EMF A ð4:14Þ
dt dt dt

For balanced winding construction and phase current excitation, that is,

M AB ¼ M AC ¼ M CA ¼ M ð4:15Þ

and

iA þ iB þ iC ¼ 0
diA diB diC
þ þ ¼0
dt dt dt

thus,
 
diA diB diC
¼ þ ð4:16Þ
dt dt dt

Hence, (4.14) reduces to

diA
vA ¼ RA iA þ ðLSelf  MÞ þ EMF A
dt
ð4:17Þ
di
vA ¼ RA iA þ LS A þ EMF A
dt

where LS is usually termed the phase synchronous inductance. It can be shown that
for an ideal, balanced sinusoidal winding distribution excited with balanced sinu-
soidal phase currents, the phase synchronous inductance becomes [149]:

  
2π 1
M ¼ LSelf  LL cos ¼  LSelf  LL ð4:18Þ
3 2

where LL is the leakage inductance and hence, for classical AC (sinusoidal) machine
analysis,

3
LSð3phaseÞ ¼ L  LL ð4:19Þ
2 Self

However, for the machine of interest, the phase windings are formed from
concentrated coils; hence, the self- and mutual-inductances are calculated via FEA
to yield

LSð3phaseÞ ¼ 1:075LSelf ð4:20Þ

Thus, the mutual coupling and hence inductances have less significance for the
concentrated winding of interest. For the nine-phase machine configuration, the
110 4 Multiphase HPM Generator Systems

Fig. 4.8 Single coil representation of the three- and nine-phase HPM machines

spatial displacement between the single stator coils is 40 electrical, as illustrated in
Fig. 4.8, which also shows the three-phase coil schematic for completeness. For the
nine-phase machine, the method for calculating the synchronous inductance can be
as per the three-phase case, but extending the equations to account for the mutual
couplings of all coils. Thus, with the same assumptions and expanding (4.14), the
phase voltage equation is now

diA di di
vA ¼ RA iA þ LSelf þ M AB B þ M AC C
dt dt dt
diD diE diF
þ M AD þ M AE þ M AF ð4:21Þ
dt dt dt
di di di
þ M AG G þ M AH H þ M AI I þ EMF A
dt dt dt

where by symmetry MAB ¼ MAI, MAC ¼ MAH, MAD ¼ MAG, MAE ¼ MAF and (4.21)
can be rewritten as
 
diA diB diI
vA ¼ RA iA þ LSelf þ M AB þ
dt dt dt
   
diC diH diD diG
þ M AC þ þ M AD þ ð4:22Þ
dt dt dt dt
 
diE diF
þ M AE þ þ EMF A
dt dt
4.2 Nine-Phase HPM Generator Parameters 111

Again, it can be shown that for an ideal, balanced sinusoidal winding distribution
excited with balanced sinusoidal phase currents, the phase synchronous inductance
becomes

  
2π
M AB ¼ LSelf  LL cos ð4:23Þ
9
  
4π
M AC ¼ LSelf  LL cos ð4:24Þ
9
 
 6π
M AD ¼ LSelf  LL cos ð4:25Þ
9
  
8π
M AE ¼ LSelf  LL cos ð4:26Þ
9

hence, for classical AC (sinusoidal) machine analysis,

9
LSð9phaseÞ ¼ L  LL ð4:27Þ
2 Self

However, for the machine of interest, the phase windings are found from con-
centrated coils; hence, the self- and mutual-inductance values are again calculated
via FEA to yield

LSð9phaseÞ ¼ 1:67LSelf ð4:28Þ

As with the three-phase winding, the mutual couplings have less significance than
the phase self-inductance to the total synchronous inductance value.

4.2.4 Construction of HPM Machines Prototype

The construction process of the HPM machine prototype based on the final stator
windings layout, as in Figs. 3.3a and 4.3a, and design specifications as in Table 3.9,
is illustrated in Fig. 4.9. Once the two stators with different axial thickness are
soldered, as in Fig. 4.9a, they are wound in the three- and nine-phase configuration.
For each HPM winding configuration, two stators with 25 and 10 mm axial length
are wound as a three-phase configuration, and the other two stators are wound as a
14-turn nine-phase configuration. Based on the calculated lengths and parallel
conductors per-turn for the three- and nine-phase configuration, the windings of
the different phases are prepared for the hand winding, as shown in Fig. 4.9b. One of
the important steps that are required before winding is preparing the slot liners to
protect the stator winding from damaging the insulation layer. Figure 4.9c shows the
four wound stators of three- and nine-phase HPM machines.
112 4 Multiphase HPM Generator Systems

Fig. 4.9 HPM machine stator fabrication with both winding configurations. (a) Welded stator
pack. (b) Preparation of stator windings. (c) Four wound stators

The HPM machine rotor consists of three parts: shaft, PM rotor lamination, and
WF rotor lamination. The PM rotor back-iron was carefully prepared to ensure a
clean and smooth surface for the permanent magnets to be glued on to get a similar
PM rotor assembly as in Fig. 3.2b. As for the WF rotor lamination design is
highlighted in the dark dotted circle in Fig. 4.10a. The construction process is similar
to that of the stator. For the WF rotor, the DC field coils were professionally hand-
wound to ensure a high packing factor. Even so, 80 turns could be fitted in and not
4.2 Nine-Phase HPM Generator Parameters 113

Fig. 4.10 Assembly of 10-mm WF rotor of the HPM machine section. (a) WF rotor iron sheet. (b)
Finalized WF rotor

Fig. 4.11 HPM machine final prototype assembly. (a) Rotor with plain shaft. (b) Machine with
finalized rotors. (c) Machine viewed at no coupling side

100 turns giving a packing factor of 0.48. Figure 4.10b shows the finalized WF rotor
after the vacuum pressure impregnation (VPI) and backing process. Figure 4.11
shows the HPM machine rotor while it is placed inside the machine for the plain
shaft and overall rotor-designed hardware viewed at the WF rotor. Finally, two holes
114 4 Multiphase HPM Generator Systems

were drilled into the HPM machine housing to assure adequate access to the stator
phase windings of both stators along with the fixation of brushes and slip rings on the
machine shaft, as illustrated in Fig. 4.11c.

4.2.5 Resistance and Inductance Measurements

As mentioned previously, the predictions of synchronous inductance for the three-

and nine-phase stator winding configurations do not include winding end effects due
to the 2D FEA limitations. Hence, some consideration is given since the machine
axial length to diameter ratio is relatively small. Different analytical approximation
methods that predict end-winding inductance have been reported in [147] and [150–
153], for example. However, these techniques tend to be empirical and associated
with some uncertainty. The most appropriate method to determine end-winding
effects is 3D FEA [152] or from a series of measurements made on machines of
varying axial length. The latter was considered in this book, where actual winding
self-inductances of the assembled HPM machine stators were measured via an
Agilent 4284, a precision LCR meter having a measuring bandwidth of 20 Hz to
1 MHz. The measured inductance results can be characterized as in (4.29).

LSelf m ¼ 2Lend þ L ð4:29Þ

where LSelf-m is the measured self-inductance, Lend is the end-winding inductance,

and L is the self-inductance determined from 2D FEA. Note L equals LPMa for the
PM stator and LWFa for the WF stator. Thus, as the machine axial length increases,
the impact of the end-winding inductance on total inductance reduces, as shown in
Table 4.6. Furthermore, for the three-phase machine case, (4.20) has also been
verified by measuring the line-to-line inductance, which includes both self- and
mutual-line-inductance values, and dividing this value by 2, and then subtract the
self-inductance from this value in order to obtain the line mutual-inductance that will

Table 4.6 Comparison of three- and nine-phase HPM generator measured inductances and
resistance
Inductance (μH) (with rotor)
End-
Stator Self-winding winding Resistance (mΩ)
Item section FEA Measured 2Lend Predicted Measured
Three- PM 146.0 175 29.0 40.84 50.00
phase WF 178.7 308 129 20.95 31.00
nstc ¼ 6
Nine- PM 190.5 217 26.5 78.56 90.00
phase WF 272.7 481 208 55.05 70.00
nstc ¼ 14
WF rotor (with stator) 68  103 91  103 23  103 13.06  103 11.35  103
4.3 Analysis Models 115

Table 4.7 Mutual-inductance calculation based on line-to-line inductance measurements of the

three-phase machine
LSelf
PM part of the HPM machine 2(LSelf +M ) (measured) M M
(LPMa ¼ 0.025 m) (measured) (μH) (μH) (measured) (predicted)
Three-phase (nsct¼6) 372 175 0.063LSelf 0.075LSelf

be added to the self-inductance to get the line synchronous inductance as in

Table 4.7. The disparity between FEA and measured inductance values was
predicted, and it showed satisfactory results. Furthermore, sensitivity analysis has
been carried out, as in Sect. 4.4.2, to account for and predict the impact of different
LS values on the analyzed system outcomes.
Moreover, the stator winding resistance (Rs) was calculated for the three-phase
HPM machine stators in Chap. 3, as given in Table 3.8. In this section, the resistance
for the 14-turn nine-phase scheme is predicted and measured, as given in Table 4.6.
Again, the shown results illustrate the comparison between the three- and nine-phase
stator resistances and predicted and measured WF rotor resistance, along with the
difference between the predicted and measured data of the total stator winding
resistances, which is due to the mean end winding approximation and measurement
error.

4.3 Analysis Models

In this section, a generalized mathematical model is given to support the classical

sinusoidal waveform analysis only of the HPM generator along with simulation
dynamic models of three- and nine-phase HPM generator systems.

4.3.1 General dq Mathematical Model of HPM Generator

The HPM generator is operating at all times to supply the vehicle DC-link with
constant power at driving mode and charge the battery at vehicle stop mode. The
generalized mathematical model of the HPM generator, considered here, is presented
in the dq coordinates with the rotor excitation field on the d-axis. In the d coordinate
system, the steady-state voltage equation is expressed as in (4.30) [154]:

V d Rs ωLq Id 0
¼ þ ωλHPM ð4:30Þ
 Vq ωLd Rs Iq 1
λHPM¼ λPM  λWF ð4:31Þ
116 4 Multiphase HPM Generator Systems

Fig. 4.12 Equivalent

steady-state circuits of the
HPM machine in the dq
axes. (a) d-axis. (b) q-axis

λWF ¼ kex I f ð4:32Þ

where Id, Iq, Vd, Vq, Ld, and Lq are the d and q components of armature currents,
terminal voltages, and armature synchronous inductances, respectively, λHPM is
the amplitude of the total excitation flux-linkage in the reference d-axis, λPM is the
amplitude of the PM excitation flux-linkage in the reference d-axis, λWF is the
amplitude of the WF excitation flux-linkage in the reference d-axis, kex is the
wound field excitation constant, and Rs is the armature resistance. Figure 4.12
represents the equivalent circuits in dq axes of the HPM machine. Hence, by
introducing the steady-state phasor results, as in (4.33, 4.34, 4.35, 4.36, 4.37, and
4.38) [35], the phasor diagram of the HPM generator for two operating cases is
illustrated in Fig. 4.13.

V s ∠δ þ ðRs þ jX s ÞI s ∠δ ¼ EMF d ∠0 ð4:33Þ

λs ¼ λd þ jλq ð4:34Þ
4.3 Analysis Models 117

Fig. 4.13 HPM generator

phasor diagram in steady-
state region for three
operating cases. (a) Phasor
diagram with zero wound
field excitation mode from
no-load to rated resistive
load. (b) Phasor diagram in
the voltage regulation mode
at constant resistive load

λq ¼ L q I q ð4:35Þ

λd ¼ Ld I d þ λHPM ð4:36Þ


jX s I s ¼ jω Ld I d þ jLq I q ð4:37Þ
118 4 Multiphase HPM Generator Systems

EMF d ¼ jωλs ¼ jωλHPM ð4:38Þ

where Xs equals 2πfeLS.

In this book, the dynamics of the HPM generator variables are characterized
initially by these two operating cases. Figure. 4.13a presents the HPM generator
phasor diagram due to a gradual increase in a resistive load with zero wound field
excitation, as will be investigated in Sect. 4.4. For the induced back-EMF regulation
mode at constant resistive load, the HPM generator phasor diagram is illustrated in
Fig. 4.13b. δ in Eq. (4.33) and Fig. 4.13 presents the angle of the phase voltage Vs
and phase current Is with respect to the reference back-EMF vector.

4.3.2 Simulation Model

The HPM generator system’s dynamic model consists of two major sections. The
first section generates the machine back-EMF waveforms. Here, by using normal-
ized sinusoidal and trapezoidal back-EMF waveforms, obtained analytically and via
FEA data, the control over machine back-EMF frequency and amplitude for all of
the machine phases becomes relatively simple, as in (4.39) and (4.40), where ωnorm
is the normalized angular speed at 3 krpm.

EMF φ1 ¼ ωðλPM  λWF Þ cos ðωnorm t Þ ð4:39Þ

⋮ ⋮ ⋮ ⋮
2π
EMF φn ¼ ωðλPM  λWF Þcos ðωnorm t  Þ ð4:40Þ
n

The second section is presented by the line resistance, synchronous inductance,

and passive rectification stage that link machine terminal voltage with a resistive
load bank, which is modeled simply by utilizing SimPower built-in blocks with
certain modification to their internal values, where these predicted internal values of
the phase resistance, synchronous inductance, diode internal resistance, and diode
voltage drop along with an adequate analysis time step will account for more
accurate system behavior. The three- and nine-phase machine dynamic models are
based on the following assumptions.
• The three-phase stator winding is based on a benchmark assumption that utilizes a
distributed winding to produce a sinusoidal back-EMF, so the space harmonics of
the magnetic field of the stator are neglected and on the FEA trapezoidal back-
EMF waveforms based on the actual machine with concentrated windings.
• The nine-phase stator winding is based on the FEA trapezoidal back-EMF
waveform.
• The iron losses in the stator and rotor are neglected.
• Nonlinear effects of saturation and hysteresis are neglected.
4.4 Three- and Nine-Phase HPM Generator System Studies 119

Fig. 4.14 Three-phase HPM generator with power circuit converter dynamic model

• The stator and rotor winding resistances are not a function of frequency or
temperature.
Figures 4.14 and 4.15 illustrate the dynamic model of the three- and nine-phase
HPM generator systems that are adopted in this book to analyze the system out-
comes, such as DC-link ripple, voltage regulation, and average output power due to
different internal parameters and loading conditions, which are considered the main
issue of this chapter.

4.4 Three- and Nine-Phase HPM Generator System Studies

In this section, a comparison of three- and nine-phase HPM generators with different
back-EMF waveforms and stator number of turns for the same slot current density
and machine geometry will be presented along with the quality and regulation of
both the machine and DC-link voltage terminals. Note that the analysis undertaken
in this chapter is for the case of zero WF excitation. Simoes and Vieira present a
general form of output power calculation of any machine [99]:
Z T
n
Po ¼ EMF ðt Þiðt Þdt ¼ nK p EMF Pk I Pk ð4:41Þ
T 0

where EMF(t) and EMFPk are the instantaneous phase back-EMF and its peak value,
respectively, similarly i(t) and IPk are the phase current and its peak value, T is the
period of the back-EMF waveform, and Kp is the back-EMF waveform factor that
120 4 Multiphase HPM Generator Systems

Fig. 4.15 Nine-phase HPM generator with power circuit converter dynamic model

accounts for variation from a pure sinusoidal to one that is more trapezoidal.
Furthermore, (4.41) shows that a high number of phases will allow high power
density only if Kp factor is presented by the more trapezoidal back-EMF waveform.
This general equation is investigated by implying the above assumption for
different machine configurations without any rectification stage, as shown in
Fig. 4.16b. Then the passive rectification stage was added to the systems to predict
output power at rated conditions, as shown in Fig. 4.16c. Finally, the active
rectification stage, as shown in Fig. 4.16a, is considered.
For the active rectification stage, as shown in Fig. 4.16a, different techniques of
forced commutation are discussed in [138], and the authors’ view is that the most
suitable one for this application is the sinusoidal-pulse-width modulation (SPWM).
This AC–DC conversion technique overcomes the third-order harmonic problem,
which presents the lowest order harmonic in the phase angle, extinction angle, and
symmetrical angle control techniques. However, due to the high switching frequency
of the SPWM technique, the higher order harmonics would have higher amplitudes,
which could easily be filtered out [138]. The diode losses in the passive rectifier and
the switching losses in the active rectifier case are investigated in Sect. 4.5.2.
4.4 Three- and Nine-Phase HPM Generator System Studies 121

Fig. 4.16 Multiphase

generators without and with
power circuit converters. (a)
With active rectifier (b)
Without passive rectifier (c)
With passive rectifier

4.4.1 Impact on Synchronous Inductance and Rectifier

The three- and two sets of nine-phase HPM generator configurations with and
without passive power circuit rectifier are analyzed for sinusoidal and trapezoidal
back-EMF waveforms, for the same stator slot current density and with a machine
geometry that will give similar stator copper losses, based on (4.42, 4.43, 4.44, and
4.45) by assuming that LS ¼ LSelf for all of the considered stator winding configu-
rations. Psc is the stator copper losses, Jss is the stator current density, 3 and 9 denote
three- and nine-phase configurations.

Psc3 ¼ 3I 2rms3 Rs3 ð4:42Þ

ns3 I rms3
J ss3 ¼ ð4:43Þ
Ass

rffiffiffiffiffiffiffiffiffiffiffiffi
Psc3
I rms9 ¼ ð4:44Þ
9Rs9
122 4 Multiphase HPM Generator Systems

Table 4.8 Comparison of different HPM machine configurations at rated Psc when LS ¼ LSelf and
without applying excitation current to the wound field part
Without passive
rectifier With passive rectifier
Machine System load
Machine output (Po) output (Po) PDC-link
(average) (W) (average) (W) (average) (W)
Winding Phase Phase Case (a) Case (b) Case (c) Case (c)
configuration EMFrms Irms (LS ¼ 0) (LS ¼ LSelf) (LS ¼ LSelf) (LS ¼ LSelf)
(V) (A)
Three-phase 61.5 23.7 4388 3043 2826 2574
(sinusoidal)
(nsct ¼ 6)
Three-phase 69.6 23.7 4914 3773 3490 3229
(trapezoidal)
(nsct ¼ 6)
Nine-phase 58.2 9.73 5083 4037 3339 3077
(trapezoidal)
(nsct ¼ 14)
Nine-phase 74.8 7.87 5283 4085 3393 3140
(trapezoidal)
(nsct ¼ 18)

ns9 I rms9
J ss9 ¼ ð4:45Þ
Ass

The results illustrated in Table 4.8 show the predicted improvement by the
trapezoidal back-EMF waveform that is offered by the fractional slot HPM machine
stator winding configuration, over the three-phase machine with sinusoidal back-
EMF, for four cases. For Table 4.8, where the machine synchronous inductance is
1.0LSelf, both nine-phase schemes with a trapezoidal back-EMF give more machine
output power than the three-phase trapezoidal back-EMF waveform when the
passive rectification stage is not included in the system.
However, once the passive rectification stage is added to the system, the machine
terminal and DC-link average output power of the three-phase configuration with a
trapezoidal back-EMF waveform becomes slightly greater than the 18-turn nine-
phase machine configuration. Moreover, due to the passive diode switching func-
tion, the average output power drops by 7.1% and 7.5% for both three-phase
machine back-EMF waveforms (sinusoidal and trapezoidal), respectively, where
for the nine-phase machine configuration (14- and 18-turn), the drop is 17%. In
addition, the rectification stage efficiency for all machine configurations is around
92.5%. Thus, it is known that there is no benefit in power output to be gained by
increasing the machine phase number above three for sinusoidal phase back-EMF
waveform.
Figures 4.17 and 4.18 demonstrate different phase back-EMF, terminal voltage,
and current waveforms for the four analyzed HPM machine configurations without a
rectification stage, at rated conditions only and with a rectification stage for
4.4 Three- and Nine-Phase HPM Generator System Studies 123

Fig. 4.17 Phase voltages and current waveforms of the six-turn three-phase HPM generator system
at different loading conditions (LS ¼ LSelf). (a) Sinusoidal EMF3-phase without rectifier (b) Sinusoi-
dal EMF3-phase with rectifier (c) Trapezoidal EMF3-phase without rectifier (d) Trapezoidal EMF3-
phase with rectifier

increments from minimum to rated resistive load power demands. The generator
phase terminal voltage, as in Fig. 4.17b, shows clearly the three-phase rectifier
six-step switching states that are changed every sixth of a cycle (V levels are
VDC-link/3 and 2VDC-link/3), resulting in a nonsinusoidal generator phase current,
as in Fig. 4.17b.
For the nine-phase generator configuration, there are four possibilities of equiv-
alent load configurations that can be obtained if an active rectification stage is
utilized based on the generated IGBT’s gate signals [155]. However, for the passive
nine-phase rectification stage, which is utilized in this thesis, the 4–5 load configu-
ration is generated via the rectification stage (18 diodes) and a maximized phase
current is achieved [155]. Therefore, Figs. 4.18b and (d) show clearly the 4–5
configuration effect (nine-phase rectifier 18-step switching states that are changed
every 18th of a cycle) in the phase terminal voltage amplitude due to 4VDC-link/9
and 5VDC-link/9 levels. Furthermore, it is shown that as the phase current increases,
due to the load power demand, the clear effect of the machine LS on the output power
124 4 Multiphase HPM Generator Systems

Fig. 4.18 Phase voltages and current waveforms of the 14- and 18-turn nine-phase HPM generator
system at different loading conditions (LS ¼ LSelf) (a) Trapezoidal EMF9-phase (without rectifier and
nstc ¼ 14) (b) Trapezoidal EMF9-phase (with rectifier and nstc ¼ 14). (c) Trapezoidal EMF9-phase
(without rectifier and nstc ¼ 18) (d) Trapezoidal EMF9-phase (with rectifier and nstc ¼ 18)

rating becomes dominant, which imposes a limitation on the system rating for all of
the machine configurations.
Moreover, the detailed results for the case with passive rectifier are presented in
Table 4.9. Here, the nine-phase HPM machine configuration with two sets of tooth
number of turns, 14- and 18-turn, are investigated. To assure an adequate compar-
ison between the three- and nine-phase HPM machine systems at rated conditions, it
is desirable to have their DC-link output voltages at similar levels.
The results in Table 4.9 illustrate that the 14-turn nine-phase machine gives a
similar average DC-link voltage to the three-phase with sinusoidal back-EMF
waveform case, but with less ripple value, as in Fig. 4.19. For DC-link % ripple
improvement, both nine-phase configurations give similar improvements in % ripple
value. The increase in output power compared to the three-phase with sinusoidal
back-EMF is 19.5% and 21.9% for nine-phase machine configurations with 14- and
18-turn, respectively. However, for the three-phase machine configurations, the
trapezoidal back-EMF waveform has an output DC-link voltage ripple that is
4.4 Three- and Nine-Phase HPM Generator System Studies 125

Table 4.9 Comparison of three- and nine-phase HPM generator system data for LS ¼ LSelf
Three-phase Three-phase Nine-phase Nine-phase
(nstc¼6) (nstc¼6) (nstc¼14) (nstc¼18)
Units (sinusoidal) (trapezoidal) (trapezoidal) (trapezoidal)
Machine phase current A 23.7 23.7 9.735 7.87
(Irms)
Phase EMFPk V 86.9 89.9 69.9 89.9
Phase voltage (VPk) V 58.4 81.5 56.2 68.5
Peak VDC-link V 84.5 113.3 81.44 103.4
Average VDC-link V 80.8 102.2 80.6 102.2
DC-link voltage ripple V 11 14.37 1.94 2.4
(Vr)
% VDC-link p.u. 0.136 0.14 0.024 0.024
ripple/average
Ripple voltage kHz 4.8 4.8 14.4 14.4
(VDC-link) frequency ( fr)
Average electromag- W 2826 3490 3339 3393
netic power (Po)
Average DC-link power W 2575 3229 3077 3140
(PDC-link)
Ratio of DC-link power p.u. 1.00 1.254 1.195 1.219
nsc – 12 12 4 4
nPc – 12 12 5 4
Conductor length per m 0.6044 0.6044 1.403 1.8
tooth (PM stator)
Conductor length per m 0.4214 0.4214 0 .9834 1.2600
tooth (WF stator)
Rs Ω 0.081 0.081 0.160 0.245
LS based on mH 0.37 0.37 0.71 1.16
(LS ¼ LSelf)
Note: Excitation current ¼ 0 A; axial length ¼ 40 mm; mass ¼ 11.2 kg; volume ¼ 1.45  103 m3;
copper area per single conductor ¼ 2.46176  107 m2; WF rotor turns per tooth ¼ 100; total
conductor length of the WF rotor part ¼ 7.5 m; stator fe ¼ 800 Hz; Jss ¼ 2.4 MA/m2; Psc ¼ 136.5 W;
and rotor speed ¼ 3 krpm

worse than the three-phase machine with a sinusoidal back-EMF; however, it pro-
vides the highest DC-link average output power. Hence, the results suggest that to
achieve the most average DC-link output power, the three-phase configuration with
trapezoidal back-EMF presents the best choice among the other compared cases.
Furthermore, the nine-phase machine with 14-turn will be compared with the three-
phase machine configuration with sinusoidal back-EMF due to the similarity in
DC-link voltage level at rated conditions as shown in Fig. 4.20. The trapezoidal
back-EMF three-phase machine is to be compared with the 18-turn nine-phase
machine, as shown in Fig 4.20.
126 4 Multiphase HPM Generator Systems

Fig. 4.19 DC-link voltage

waveforms for the three- and
nine-phase HPM
configurations at rated
conditions (LS ¼ LSelf)

Fig. 4.20 Voltage regulation of HPM machine at different loads

4.4.2 System Sensitivity to Generator Synchronous

Inductance

In this section, two more comparisons via sensitivity analysis based on LS are carried
out. The first comparison is between sinusoidal back-EMF three-phase machine and
trapezoidal back-EMF nine-phase machine with 14-turn per tooth configuration. The
second comparison is between the trapezoidal back-EMF three-phase machine and
the 18-turn nine-phase machine configurations. The comparison is performed for the
same stator slot current density and machine geometry and for a machine synchro-
nous inductance value of LSelf, which presents the sensitivity analysis benchmark
based on the three-phase machine. In other words, the sensitivity analysis investi-
gates the impact of four LS9 values (0.5, 0.75, 1, and 1.25LSelf) with respect to LS3 of
4.4 Three- and Nine-Phase HPM Generator System Studies 127

Fig. 4.21 Output power ratio curves for three-phase machine with sinusoidal back-EMF versus
14-turn nine-phase machine

LSelf on the output power ratio, machine terminal voltage, and DC-link voltage of the
three- and nine-phase machine systems at different operating conditions.
For the first comparison, in both the three- and nine-phase systems, degradation in
the output power is expected at higher loads with higher values of machine self-
inductance due to the reactive voltage drop presented by this inductance. The
presented machine stator winding schemes offer lower machine self-inductance
value, which presents an advantage in this application, if they were to be compared
with over-lapping or single layer distributed stator winding schemes. Figure 4.21
shows the effect of different LS9 values with respect to one value of LS3, for different
load conditions, on the output power ratio between the sinusoidal back-EMF three-
and 14-turn trapezoidal back-EMF nine-phase configurations. The system power
ratio illustrates that the advantage of the 14-turn nine-phase configuration with
trapezoidal back-EMF over the three-phase with sinusoidal back-EMF waveform
is directly related to the values of both configurations’ synchronous inductances,
where the advantage offered by the nine-phase scheme may become a disadvantage
with higher values of LS9 in the proposed HPM machine system. Furthermore,
Table 4.10 illustrates the HPM machine terminal and DC-link voltage regulation
due to different LS values for the nine-phase configuration (14-turn). Here, better
voltage regulation capability in both machine terminal and DC-link voltage is gained
by the nine-phase machine with a trapezoidal back-EMF at rated conditions, as
shown in Table 4.10 and Fig. 4.22. Furthermore, the sensitivity analysis showed the
expected voltage regulation capability as the synchronous inductance increases or
decreases in the nine-phase machine with respect to a single value of the three-phase
machine synchronous inductance.
The second comparison, as shown in Fig. 4.23, presents the effect of different LS9
values with respect to one value of LS3, for different load conditions, on the output
power ratio between the trapezoidal back-EMF three-phase machine and the 18-turn
 128

Table 4.10 Voltage regulation of three- and nine-phase machine systems for the first comparison case
Machine terminal voltage regulation (%) DC-link voltage regulation (%)
Machine 0.5LSelf 0.75LSelf 1.0LSelf 1.25LSelf 0.5LSelf 0.75LSelf 1.0LSelf 1.25LSelf
Three-phase (sinusoidal) – – 56.5 – – – 77.0 –
(nsct¼6)
nine-phase (trapezoidal) 15.8 24.7 43.2 78.9 34.6 45.1 67.3 110
(nsct¼14)
4 Multiphase HPM Generator Systems
4.4 Three- and Nine-Phase HPM Generator System Studies 129

Fig. 4.22 Different terminal and DC-link voltage regulation curves of the three-phase with
sinusoidal back-EMF versus the 14-turn nine-phase HPM machines. (a) Terminal voltage regula-
tion (LS3 ¼ 1.0LSelf). (b) DC-link voltage regulation (LS3 ¼ 1.0LSelf)

trapezoidal back-EMF nine-phase machine configurations. The system power ratio

illustrates that the previous advantage, as illustrated in Fig. 4.21, offered by the
14-turn nine-phase machine with more trapezoidal back-EMF over the sinusoidal
back-EMF three-phase machine disappears and the power ratio degrades in this
second comparison. Furthermore, Table 4.11 illustrates the HPM machine terminal
and DC-link voltage regulation due to different LS values for the nine-phase config-
uration (18-turn). The voltage regulation capability in both machine terminal and
DC-link voltage is almost equal at rated conditions, as shown in Fig. 4.24. The
sensitivity analysis also showed the predicted voltage regulation capability as the
synchronous inductance increases or decreases in the nine-phase machine with
respect to a single value of the three-phase machine synchronous inductance.
130 4 Multiphase HPM Generator Systems

Fig. 4.23 Output power ratio curves for three-phase machine with trapezoidal back-EMF versus
18-turn nine-phase machine

4.4.3 DC-Link Voltage Quality

In many systems with high DC-link voltage ripple, a smoothing (filter) capacitor is
introduced to eliminate or minimize DC-link ripple, as illustrated in Fig. 4.25. The
filter capacitor is the most vulnerable and one of the most expensive components in
the drive system [156], where its life is affected by the ripple current and it has a
much shorter life compared to the other components. This poses a clear drawback if
it is to be used in a harsh environment or where maintenance access is denied or
difficult.
In Table 4.12, the voltage ripple equals 14.37 V and 2.4 V for the three- and nine-
phase (18-turn) machines with a trapezoidal back-EMF waveform, respectively. For
the three-phase HPM machine case, to reduce ripple from 14.37 to 2.4 V, the
smoothing capacitor (CS) would need to be 1.3498 mF using equation (4.46)
[157], where the IDC-link equals 31.1A and fr equals 4.8 kHz. However, this calcu-
lated value requires adjustment by an iterative process that has been applied via the
numerical model, which suggested a new value for CS to achieve this reduction in
ripple, where the new capacitor value based on the dynamic model prediction equals
(1.35/52.1) mF and the RMS current (ICrms) passing through it equals 1.362 A.

I DClink
C 3ϕ ¼ ð4:46Þ
Vγ f γ

Hence, an equal DC-link voltage ripple between three- and nine-phase machines
was achieved. A suitable commercial capacitor was found online [158]. Based on the
available capacitors and the SimPower model DC-link voltage levels, a 200-V rating
capacitor (twice the DC-link voltage) was initially suggested. Based on the electro-
lytic capacitor datasheet, the capacitor internal resistance (ESR) at 105  C, ripple
Table 4.11 Voltage regulation of the three- and nine-phase machine systems of the second comparison case
Machine terminal voltage regulation (%) DC-link voltage regulation (%)
Machine 0.5LSelf 0.75LSelf 1.0LSelf 1.25LSelf 0.5LSelf 0.75LSelf 1.0LSelf 1.25LSelf
Three-phase (trapezoidal) – – 43.2 – – – 64.4 –
(nsct¼6)
Nine-phase (trapezoidal) 14.5 23.8 44.6 87 30.0 40.3 62.3 108
(nsct¼18)
4.4 Three- and Nine-Phase HPM Generator System Studies
131
132 4 Multiphase HPM Generator Systems

Fig. 4.24 Different terminal and DC-link voltage regulation curves of three-phase with trapezoidal
back-EMF versus 18-turn nine-phase HPM machines. (a) Terminal voltage regulation (LS3 ¼
1.0LSelf). (b) DC-link voltage regulation (LS3 ¼ 1.0LSelf)

frequency, and RMS current were accounted for in the DC-link capacitor selection
process. To calculate the DC-link capacitor losses (PCap-loss) and rating (PCap-rating)
and check if they satisfy the design requirements, (4.47) and (4.48) were utilized.
 2
I Crms
PCaploss ¼ ESR ncp ð4:47Þ
ncp
4.4 Three- and Nine-Phase HPM Generator System Studies 133

Fig. 4.25 DC-link voltage

waveform with and without
smoothing capacitor effect
for the three-phase
trapezoidal back-EMF HPM
machine system at rated
conditions

Table 4.12 The encountered iterative process for selecting the DC-link capacitors
Selected voltage Capacitor ESR ICrmsM PCap-loss PCap-rating Satisfy
level (V) (μF) (Ω) (A) ncp (W) (W) requirements
68 1.46 0.71 1 2.707 0.736 No
33 3 0.42 2 2.781 1.058 No
200 22 4.45 0.31 3 2.750 1.283 No
15 6.13 0.26 4 2.841 1.658 No
10 9.75 0.17 6 3.013 1.690 No
68 1.05 0.91 1 1.947 0.869 No
350 33 2.25 0.56 2 2.086 1.411 No
22 3.22 0.44 3 1.990 1.870 No
15 7.45 0.24 6 2.302 2.575 Yes

PCaprating ¼ ðI CrmsM Þ2 ESR ncp ð4:48Þ

where ICrmsM is the datasheet RMS capacitor current and ncp is the number of
capacitors in parallel. Table 4.12 illustrates the results of the iterative process for
choosing the adequate number of DC-link capacitors, which are to be connected in
parallel and satisfy the design requirements. Thus, the results suggest that six
capacitors of 15μF designed to handle 350 V are to be connected in parallel in the
DC-link terminal of the three-phase system case to reduce the DC-link voltage
ripple.
However, more investigation on some of the alternatives to reducing this voltage
ripple by using an LC filter, with different DC-link series inductor (LDC-3) and
capacitor (CS-3) values, for the three-phase case, is carried out. By adding LDC in
the three-phase system simulation models, as shown in Fig. 4.26, the prediction of
this inductance in the three-phase case has been done, based on the best and optimum
combination of both CS-3 and LDC-3 values. Therefore, an investigation is applied on
134 4 Multiphase HPM Generator Systems

Fig. 4.26 SimPower models for three-phase HPM generator systems with DC-link inductance

Fig. 4.27 Impact of DC-link inductance on DC-link voltage ripple in three-phase HPM machine
system

the three-phase HPM generator dynamic model with four different capacitor values
and a range of inductance values, as illustrated in Fig. 4.27.
Due to the harsh environment at which CS is to be placed in, a decrease in the
capacitor requirements by substituting an adequate inductor to achieve the desired
quality of DC-link output will improve the powertrain system robustness in HEVs.
Note that the addition of this DC-link inductance improves the robustness of the
vehicle powertrain while degrading DC-link voltage regulation. However, this
degradation in voltage is not a significant drawback since it can be compensated
for, for example, by increasing the number of stator turns.

4.5 Loss Audit of Generator Systems

4.5.1 Introduction

The proposed nine-phase HPM generator system in this book is an alternative to the
three- or nine-phase PM generator with an active converter, which provides a
4.5 Loss Audit of Generator Systems 135

Table 4.13 Comparison of HPM rectifier system with PM active conversion

PM PM
HPM connected to connected to HPM connected to connected to
Item passive rectifier VSC passive rectifier VSC
Generator WF-PM PM WF-PM PM
topology
Stator winding Three-phase Three-phase Nine-phase Nine-phase
No. of IGBTs – 6 – 18
(bridge)
No. of diodes 6 6 18 18
(bridge)
Control scheme WF control using Vector WF control using Vector
voltage regulator control voltage regulator control
Control Simple Complex Simple Complex
complexity
Phase current – Current – Current
measurement sensors sensors
Rotor angle – Encoder/ – Encoder/
measurement resolver resolver
Slip rings and 2 – 2 –
brushes
Shaft speed Fixed Variable Fixed Variable
Active switching – 1.0 – 1.0
losses (p.u.)
PM axial length 0.75 1.0 0.75 1.0
(p.u.)
WF axial length 0.3 – 0.3 –
(p.u.)
Total machine 1.05 1.0 1.05 1.0
volume (p.u.)
Active converter – 1.0 – 1.0
volume (p.u.)
Passive rectifier 1.0 1.0 1.0 1.0
volume (p.u.)

simplified power conversion scheme, specifically by eliminating the active switching

devices and reducing the associated measurement devices (phase current and rotor
angle sensors) and control hardware. Table 4.13 shows a general comparison of the
HPM generators using a passive rectification system with PM generators using an
active voltage source converter (VSC) for both three- and nine-phase stator winding
configurations. Hence, this section investigates some of the undiscussed losses that
appear in the HPM generator system with passive rectification and the PM generator
system with active rectification. The losses of interest are mainly due to the stator
core and passive and active switches. An assessment of these losses is computed as a
feasibility study of using a HPM generator with a passive rectifier system as an
alternative to a PM generator with an active rectifier system.
136 4 Multiphase HPM Generator Systems

4.5.2 Core Loss Prediction

Assuming similar core loss for three- and nine-phase winding configurations, the
used core loss for HPM machines is conducted previously as in Chap. 3 for the
25-mm PM and 10-mm WF machines. For completeness, the same calculation
method was used to assess the core losses for a PM machine having 40 mm axial
length, as in Table 4.14. The HPM machine iron loss at 3 krpm with rated WF
excitation is 411.8 W. The same volume of PM machines with an axial length of
40 mm has predicted a loss of 383.4 W.

4.5.3 Passive and Active Converter Loss for HPM and PM

Generator Systems

For a HPM generator with passive rectification, a simple diode loss equation is used
in the SimPower dynamic model to calculate the rectification stage losses. The
average diode conduction loss across the switching period (Tsw ¼ 1/FS) is given
by (4.49) [159], where IDav is the average diode current, IDrms is the rms diode
current, uD0 is the maximum forward voltage drop, and rD is the diode internal
resistance.
Z T SW 
1
PCD ¼ uD0 :iD ðt Þ þ r D :i2D ðt Þ dt ¼ uD0 :I Dav þ r D :I 2Drms ð4:49Þ
T SW 0

The datasheet of a three-phase bridge rectifier 36MT series is used to get the
typical values for uD0 and rD. Hence, Table 4.15 illustrates the passive rectification
stage losses in three- and 18-turn nine-phase HPM generator configurations with
trapezoidal back-EMF at rated conditions.
Predicting the losses in the semiconductor active power switches, such as IGBTs,
is complicated if they are to be compared with the passive ones. Some papers have
discussed active semiconductor switch loss calculations [160–163] with minor
differences in their assumptions. The proposed IGBTs loss equations by Casanellas
[162] are used to predict semiconductor power losses when switching sinusoidal
currents. The loss mechanisms in an IGBT are
• Turn-on losses (PSW-on) where energy is required to activate the device; the rise
time (tr) for the device is related to the gate resistor (RG), which limits the gate
charge.
• Conduction losses (PIGBT); that is, when the device is conducting, there is a
voltage drop across the P–N junction (VCO) and a proportional relationship with
the collector current (Icc) and an ohmic loss component.
• Power losses in the reverse biased diode related to (VF) (PDIODE).
Table 4.14 Different stator and rotor section masses and core losses for the fundamental harmonic only of PM and HPM machine at no-load
PM machine HPM machine
4.5 Loss Audit of Generator Systems

WF total
Stator (LPMa ¼ 40 mm) PM stator (LPMa ¼ 25 mm) WF stator (LWFa ¼ 10 mm) WF rotor (LWFa ¼ 10 mm) (10 mm)
Machine Mass Core losses Mass Core losses Mass Core losses Mass Core losses
section (kg) (W) (kg) (W) (kg) (W) (kg) (W) Core losses (W)
Back-iron 2.797 55.4 1.748 34.6 0.7000 8.9 0.139 28.8 37.7
Teeth 1.549 305.9 0.968 191.2 0.3874 38.5 0.471 90.5 129.0
Teeth-tips 0.238 22.1 0.148 13.8 0.0594 3.0 0.055 2.4 5.4
Total 4.584 383.4 2.865 239.6 1.1458 50.3 0.665 121.9 172.2
137
138 4 Multiphase HPM Generator Systems

Table 4.15 Losses due to passive rectification stage in HPM generator system
Electrical power losses (W) Three-phase system Nine-phase system
Losses per diode 18.44 3.49
Smoothing capacitor losses 2.302 (six capacitors of 15 μF) –
Total rectification stage losses 110.64 62.82

• Turn-off losses (PSW-off) related to the fall time (tf) of the device and collector
current.
• Reverse recovery losses (PRR) where the diode must recover from a conducting to
blocking state—this is relational to the charge (QRR) and the time (trr).
While these characteristics of the IGBT can be used to model its losses, many
parameters change with junction temperature. The model developed by Casanellas
considers temperature and operationally varying characteristics and provides linear
approximations for use in modeling. The model produced by Casanellas has been
verified via calorimetric test and is considered to give accurate results to within
5–10% [163]. Power silicon losses with sinusoidal current control can thus be
calculated from the turn-on losses, which are estimated using (4.50) [162]:
2 
1 I
PSWon ¼ V DClink t rN F S cm ð4:50Þ
8 I cn

The conduction losses, including the third harmonic sine modulation that allows
full DC-link utilization, are estimated using (4.51) [162]:
 pffiffiffi pffiffiffi  
1 2 3 3 V  V co 2
PIGBT ¼ þ M index cos ðθÞ  M index cos ðθÞ : cen I cm
8 9π 45π I cn
þ PIGBT 3rdharmonic
 pffiffiffi 
1 3
¼ þ M index cos ðθÞ :V co I cm ð4:51Þ
2π 12

The diode conduction losses, including the third harmonic sine modulation that
allows full DC-link utilization, are estimated using (4.52) [162]:
4.5 Loss Audit of Generator Systems 139

Table 4.16 Parameter definitions and typical values for semiconductor loss calculations
Definition Parameter Units 1200 V
Rated collector–emitter forward voltage drop Vcen V 2.00
Rated collector–emitter forward voltage drop Vco V 1.00
Collector current rating Icn A 400
Collector current Icm A NA
Modulation index Mindex – 1
Switching frequency FS Hz 10000
Phase angle cos(θ) Rads NA
Reverse recovery charge Qrrn C 1.60 n
IGBT rated fall time at rated current tfn ns 350
Diode recovery fall time at rated current (Icn) trrn ns 250

 pffiffiffi pffiffiffi  
1 2 3 3 V  V co 2
PDIODE ¼  M index cos ðθÞ þ M index cos ðθÞ : cen I cm
8 9π 45π I cn
þ PIGBT 3rdharmonic
 pffiffiffi 
1 3
¼  M index cos ðθÞ :V co I cm
2π 12
ð4:52Þ

The turn-off losses are estimated using (4.53) [162]:

 
1 I
PSWoff ¼ V DClink I cm t fN F S þ cm ð4:53Þ
3π 24I cn

The diode reverse recovery losses are estimated using (4.54) [162]:
"  2 !  #
0:38 I cm I cm 0:8 I cm
Prr ¼ F s V DClink Qrrn 0:28 þ þ 0:015 þ I cm t rrn þ 0:05
π I cn I cn π I cn
ð4:54Þ

The parameters trn, Fs, Icm, Icn, Mindex, cos(θ), Vcen, Vco, and tfn and their typical
values, as obtained from the device datasheet (CM400DY-24NF), are defined in
Table 4.16. IGBT power loss per device is calculated at 10, 15, and 20 kHz
switching frequency and then multiplied by the total number of devices, as shown
in Table 4.17. Table 4.18 illustrates the comparison between passive and active
rectification stage losses for the three- and nine-phase systems. The results suggest
that all systems have similar efficiency. Table 4.18 predicts that the total system loss
in the three-phase case with active rectification is less than that with passive
rectification by 5%, where the total system loss for the nine-phase case with active
rectification is greater than that with passive rectification by 5.8%. Therefore, for a
 140

Table 4.17 Calculated loss due to active rectification stage in PM generator system for several switching frequencies
FS ¼ 10 kHz FS ¼ 15 kHz FS ¼ 20 kHz
Power loss Three-phase Nine-phase Three-phase Nine-phase Three-phase Nine-phase
Definition components system (W) system (W) system (W) system (W) system (W) system (W)
Turn-on loss PSW-on 0.0287 0.00316 0.04305 0.00474 0.0574 0.00633
IGBT conduction PIGBT 5.91 1.9 1.9 1.9 1.9 1.9
loss
Reverse biased PDIODE 1.99 0.65 0.65 0.65 0.65 0.65
diode loss
Turn-off loss PSW-off 0.92 0.3 1.38 0.45 1.84 0.6
Reverse recovery PRR 2.03 0.98 3.04 1.46 4.06 1.95
loss
Losses per IGBT IGPTloss 12.93 4.8 15.45 5.93 17.97 7.06
Total rectification RecStageloss 77.59 86.42 92.7 106.75 107.81 127.07
stage loss
4 Multiphase HPM Generator Systems
4.6 Conclusion 141

Table 4.18 Comparison between passive and active rectification stage losses of three- and nine-
phase machine systems
Three-phase HPM machine with passive Three-phase PM machine with active
Loss types rectification losses (W) rectification losses (W)
Copper 136.5 136.5
Core 411.8 383.39
Rectification 110.64 107.81 (FS ¼ 20 kHz)
DC-link 2.302 (six capacitors of 15 μF) –
capacitor
Total 661.24 627.7
Loss types Nine-phase HPM machine with pas- Nine-phase PM machine with active
sive rectification losses (W) rectification losses (W)
Copper 136.5 136.5
Core 411.8 383.39
Rectification 62.82 127.07 (FS ¼ 20 kHz)
DC-link – –
capacitor
Total 611.16 646.96

more reliable system to be implemented in the HEV powertrain, the nine-phase HPM
machine with a passive rectification stage is a good candidate for vehicle applica-
tions when compared with the other systems.

4.6 Conclusion

In this chapter, an overview of different multiphase machines and their applications

are discussed. Based on a similar stator slot current density and machine geometry, a
comparison between three- and nine-phase HPM machine configurations along with
their DC-link voltage quality and regulation is investigated. SimPower is used to
analyze and predict the HPM generator system’s behavior at both the machine and
DC-link terminals while the WF excitation is set to zero.
In Table 4.9, it is shown that the highest output power is achieved by the
trapezoidal six-turn three-phase HPM machine configuration. Both nine-phase con-
figurations, with 14- and 18-turn, with trapezoidal EMF waveform produce a very
low DC-link ripple value with less DC-link output power when compared with the
three-phase configuration with trapezoidal back-EMF, when LS ¼ LSelf for all of the
considered stator winding configurations. Furthermore, the impact of different
values of LS9 (with respect to a single value of LS3) has been investigated via
sensitivity analysis, and it showed the imposed limitation on the system output
DC-link voltage and power due to different values of LS9.
Passive and active rectification stage losses in the semiconductor power switches,
such as diodes in the passive and IGBTs in the active cases, for three- and nine-phase
HPM and PM machines are investigated. The overall predicted losses and efficiency
142 4 Multiphase HPM Generator Systems

of both systems, HPM with passive and PM with active switches, and different
winding configurations, have been compared to judge the feasibility of utilizing
HPM machine with a passive rectification stage as an alternative to a PM machine
with active rectification stage in a HEV powertrain system. Thus, in terms of system
robustness, reliability, and cost, the nine-phase HPM machine with a separate DC
excitation source seems to be a suitable configuration that can be considered a
replacement of the PM machine system in a HEV powertrain.
Chapter 5
Electric and Hybrid Electric Powertrains

5.1 Introduction

Worldwide concerns over increasing energy consumption and pollution due to

transportation systems are a primary motivation for alternatively fueled or more-
electric vehicles. Electric vehicles (EVs) are set to play an increasingly predominant
role in the automotive market since they address the energy and environmental
impact of an expanding road transport population by offering a more energy-efficient
and less-polluting powertrain alternative to conventional internal combustion engine
(ICE) vehicles. However, when compared with conventional gasoline or diesel-
fueled vehicles, all-electric vehicles are disadvantaged in driving range because of
the relatively low energy storage capacity of the onboard batteries, excessive battery
mass and refueling thereof, and recharging time, associated with existing electro-
chemical battery technologies. For example, a 2.5-ton ICE urban vehicle will have a
typical range of 400 km [95] compared with 126 km for the equivalent all-electric
vehicle [164]. On the other hand, hybrid electric vehicles (HEVs) address the vehicle
range issue while offering improved fuel economy and emission reduction when
compared with conventional ICE vehicles, but only when the vehicle has intermit-
tent or transient duty operation.

5.2 Overview of EVs

The driving range of electric vehicles depends mainly on the stored energy on an
onboard high-voltage battery system and vehicle gross weight. The driving distance
is also affected by driving behaviors, vehicle architecture, road conditions, type of
battery used, and vehicle age. Many companies now are producing all-electric
vehicles, an example of which is listed in Table 5.1 for the year 2020, including

© Springer Nature Switzerland AG 2022 143

A. S. Al-Adsani, O. Beik, Multiphase Hybrid Electric Machines,
https://doi.org/10.1007/978-3-030-80435-0_5
144 5 Electric and Hybrid Electric Powertrains

Table 5.1 Top EV models in the year 2020 [165]

Minimum (p.u.) Max. driving range Wheel-drive
Vehicle model cost (miles) architecture
Hyundai Kona 0.36 258 Front
Electric
Nissan Leaf 0.31 226 Front
Mini Cooper SE 0.29 110 Front
Electric
BMW i3 0.43 153 Rear
Kia Niro EV 0.38 239 Front
Chevrolet Bolt 0.35 259 Front
Tesla Model Y 0.48 216 Rear
Hyundai Ioniq Electric 0.32 170 Front
Jaguar I-Pace 0.67 234 All
Tesla Model X 0.77 351 Rear
Audi e-tron 0.75 204 All
Porsche Taycan 1.00 201 Rear
Tesla Model 3 0.37 322 Rear
Tesla Model S 0.72 402 Rear
Note: 1 p.u. cost equals 103,500 US dollars

Fig. 5.1 Several EV models (a) Nissan Leaf [166] (b) BMW i3 [167] (c) Chevrolet Bolt [168] (d)
Tesla Model S [169]
These vehicle photos are not sponsored, endorsed, or promoted by Nissan North America, Inc.,
BMW North America, LLC, General Motors Co. (GM), Tesla, Inc., or their affiliates
5.2 Overview of EVs 145

their expected driving range, wheel-drive architecture, and per-unit price. Figure 5.1
shows pictures of four selected vehicle models.

5.2.1 EV Powertrain Configuration

In this book, the powertrain presents the system that propels the vehicle, which is
significantly different between electric and conventional ICE-powered vehicles. For
example, it is estimated that a conventional ICE vehicle powertrain has 2000
components, where EVs have substantially fewer powertrain components
[170]. Hence, the EV powertrain system is expected to require less labor to produce

Fig. 5.2 EV powertrain components (a) Simplified powertrain configuration (b) Detailed
powertrain components [171]
146 5 Electric and Hybrid Electric Powertrains

compared with conventional gasoline or diesel engine vehicles [170]. Figure 5.2
presents a simplified and detailed EV powertrain component.

5.2.2 Battery Technology

Battery technology is a key element in EV and HEV powertrains. The specific

power, specific energy, pack size, charging time, cycle life, and cost per kW are
the major constraints that dictate the most adequate battery technology to be used
and developed for future EV and HEV powertrains. For more than 100 years, the
specific power capability of lead-acid battery technology has continually been
improved, but their specific energy has slightly increased, typically between
30 and 50 Wh/kg. This imposes a threshold on their continuous use in the automo-
tive industry. Moreover, the lead-acid battery, an example of which is shown in
Fig. 5.3a, has a short cycle life compared with competing technologies [172]. An
alternative battery technology is nickel–metal hydride (Ni-MH), as shown in
Fig. 5.3b, which has an energy density that falls between 60 and 120 Wh/kg. This
technology has been used in Toyota Prius; however, it is not the best choice for EVs.
A significant performance improvement can be gained by utilizing sodium–nickel
chloride (Na-Ni-Cl) battery technology, which is commonly referred to as ZEBRA
battery technology. It is considered safe and low-cost with a long cycle life since it
can be discharged completely without degrading its lifetime expectancy. The
ZEBRA battery shown in Fig. 5.3c has a specific energy that falls between 60 and
120 Wh/kg. On the other hand, the ZEBRA battery has great potential to be used in
association with other energy sources, such as ICE or supercapacitors in HEV
powertrains. The most promising battery technology is the lithium-ion (Li-ion),
where its energy density is between 100 and 265 Wh/kg. However, the Li-ion
battery, as shown in Fig. 5.3d, suffers from material availability and overcharging
safety issues and higher price compared with the previously mentioned battery
technologies. Table 5.2 illustrates a brief comparison in terms of specific energy
and power, energy and power density, cycle lifetime, and energy and power capacity
cost between different battery technologies.

5.3 Overview of HEVs

Driving range and battery charging time are the major EV barriers. However, HEVs
and plug-in HEVs alleviate some of the drawbacks experienced with EVs in terms of
range while providing an intermediate powertrain concept in the move from con-
ventional to an all-electric rational [177]. The major difference between HEVs and
plug-in HEVs is the onboard battery size with respect to the ICE size. The plug-in
HEV powertrains have a larger and more powerful battery pack compared with the
5.3 Overview of HEVs 147

Fig. 5.3 Different EV and HEV battery technology (a) Lead-acid battery [173] (b) Ni-MH battery
pack [174] (By Claus Ableiter—Own work, CC BY-SA 4.0) (c) ZEBRA battery [175] (Von Claus
Ableiter—Eigenes Werk, CC BY-SA 4.0) (d) Li-ion battery pack [176] (By Claus Ableiter—Own
work, CC BY-SA 4.0)

Table 5.2 Comparison of different battery technologies [172]

Specification Lead-acid NiMH ZEBRA LI-ion
Specific energy (Wh/kg) 30–50 60–120 100–120 100–265
Energy density (Wh/L) 50–80 140–300 150–180 250–730
Specific power (W/kg) 75–300 250–1000 150–200 250–340
Power density (W/L) 10–400 80–300 220–300 100–210
Cycle lifetime (cycles) 100–1000 500–1200 2500+ 400–1200
Energy capacity cost ($/kWh) 150–400 150–1500 100–200 500–2500
Power capacity cost ($/kW) 175–600 150–1500 150–300 175–4000
148 5 Electric and Hybrid Electric Powertrains

Table 5.3 Top HEV models in 2020 [178]

Minimum (p.u.) Fuel economy in (mpg) (city/ Wheel-drive
Vehicle model cost highway) architecture
Honda Insight 0.22 55/49 Front
Toyota Avalon 0.36 43/44 All
Hybrid
Hyundai Ioniq 0.22 55/58 Front
Hybrid
Toyota Prius Prime 0.27 55/53 Front or all
Hybrid
Toyota Prius V 0.23 54/50 Front or all
Honda Accord 0.23 48/47 Front
Hybrid
Kia Optima Hybrid 0.28 21/30 Front
Toyota Camry 0.26 44/47 Rear
Hybrid
Lincoln MKZ Hybrid 0.35 20/31 Front or all
Toyota Prius C 0.22 48/44 Front
Hyundai Sonata 0.27 50/54 Front
Hybrid
Ford Fusion Energi 0.34 109/97 Front
Chevrolet Volt 0.39 43/42 Front
Toyota Prius 0.24 54/50 Rear or all
Ford Fusion Hybrid 0.27 43/41 Front
Note: 1 p.u. cost equals 103,500 US dollars

onboard ICE power rating, where the HEV has a smaller and less power rating
battery pack compared with the ICE power rating.
Fuel consumption or fuel economy plays a key factor in future road vehicle
production since now many countries are using gasoline vehicle fuel economy new
standards to direct their automotive industry to meet certain miles per gallon (mpg)
numbers. Table 5.3 lists various commercial HEVs, their fuel economy, per-unit
price, and wheel-drive architecture, while Fig. 5.4 shows pictures of four hybrid EV
models. In addition to the improvement in the driving range, which can be governed
by fuel tank size, HEVs and plug-in HEVs showed a clear cost advantage over EVs
while generating less emission for similar ICE-powered vehicles’ performance.

5.3.1 HEV Powertrain Configurations

The surveyed hybrid EVs, as in Table 5.3, might have powertrain components
similar to what is shown in Fig. 5.5. The common simplified hybrid EV powertrain
configurations that consist of two energy sources, such as battery and ICE, are
illustrated in Fig. 5.6. The parallel hybrid electric vehicle format, as in Fig. 5.6a, is
5.3 Overview of HEVs 149

Fig. 5.4 Several HEV recent models [178] (a) Honda Insight (By Kevauto—Own work, CC
BY-SA 4.0) (b) Toyota Prius V (By Vauxford—Own work, CC BY-SA 4.0) (c) Lincoln MKZ
Hybrid (By Kevauto—Own work, CC BY-SA 4.0) (d) Hyundai Sonata Hybrid (By Chu—Own
work, CC BY 4.0)
These vehicle photos are not sponsored, endorsed, or promoted by American Honda Motors Co.,
Inc., Toyota Motors North America, Inc., Ford Motor Co., Hyundai Motor America, or their
affiliates

Fig. 5.5 Detailed plug-in HEV powertrain components [179]

generally the favored scheme for most vehicle suppliers to date since it allows them
to continue to use their existing powertrain components while accommodating an
additional power input, thus requiring lower investment and minimizing system risk.
However, the series hybrid electric vehicle, as illustrated in Fig. 5.6b, can offer more
flexibility where the ICE energy source is replaced by alternatively fueled combus-
tion engines (using petroleum products, methanol, hydrogen) or fuel cell systems
[180]. Here, the ICE is mechanically disconnected from the powertrain and is hence
independent of vehicle road wheel speed. Energy conversion is then from the
150 5 Electric and Hybrid Electric Powertrains

Fig. 5.6 Different simplified hybrid EV powertrain configurations (a) Parallel hybrid (b) Series
hybrid (c) Series-parallel hybrid (power-split hybrid)

onboard fuels’ chemical energy to kinetic energy via the ICE, and then from kinetic
to electrical energy via an engine-mounted or -coupled generator, Fig. 5.6b [181]. In
Fig. 5.6c, another possible powertrain implementation is shown, where three drive
modes, such as all-electric, hybrid, and engine drive, can be realized. There is,
therefore, an interest in onboard auxiliary power units that would serve as an energy
input to the vehicle powertrain for suburban or highway driving.
5.5 Series Hybrid Electric Vehicle (SHEV) 151

5.4 Vehicle Driving Cycles

The fuel economy and emission certification tests can be applied onboard or
off-board standard light-duty and heavy-duty vehicles. These tests are derived
from the total fuel consumption of fuel over approved driving cycle using a dyna-
mometer and gaseous particular matter portable emission measurement system
known as AVL PEMS. In some labs, such as for the off-board economy and
emission test, the AVL PEMS can remain on a cart and placed where the sample
line can reach the vehicle exhaust pipe for laboratory testing, as shown in Fig. 5.7.
For example, in the United States, a highway driving cycle (HWFET) is used for the
driving test, as shown in Fig. 5.8a. In Europe, a new European driving cycle (NEDC)
is used. The NEDC consists of four urban (ECE15) cycles and one suburban cycle,
as shown in Fig. 5.8b. In Japan, two standard drive cycles, known as 10–15 mode
and JC08 cycles, are used for driving fuel economy and emission measurement
tests [183].

5.5 Series Hybrid Electric Vehicle (SHEV)

The previously discussed hybrid EV powertrain configurations, illustrated in

Fig. 5.6, tend to combine the engine start and generator functionalities into one
machine, a philosophy of reduced component count. However, the start and gener-
ation operational requirements are quite different. Here, the machine has to provide
high peak power transients during starting and engine braking. More importantly,
the starter/alternator requires a voltage source inverter for active power conversion
between starter/alternator and the ICE; hence, the power converter needs to have
active switching and high silicon VA rating. However, most of the machine opera-
tion is in the steady power generation mode, which is usually over a limited speed

Fig. 5.7 Actual laboratory

AVL PEMS [182]
152 5 Electric and Hybrid Electric Powertrains

Fig. 5.8 US and European standard driving cycles [95] (a) US highway driving cycle (HWFET)
(b) NEDC driving cycle

range and at lower power levels than for the system starting transient. Thus, if the
starting and steady power generation are realized by two machines, one a traditional
starter and the second a lower power generator operating over a constrained speed
range, the installed machines could be simpler in form and the power electronic
conversion requirements reduced.
Compared with series HEVs, the parallel HEV format is generally the favored
electrified scheme to date for most vehicle suppliers since it allows the original
equipment manufacturers (OEM) to utilize their existing products and gradually
increase their investments over time. However, with the series HEV topology, the
combustion engine is disconnected from the system, as shown in Fig. 5.6b, its size is
reduced, and its rotational speed can be made dependent on or independent of the
5.5 Series Hybrid Electric Vehicle (SHEV) 153

Fig. 5.9 Different ICE/generator machine and power conversion system implementations for
auxiliary power unit in SHEVs. (a) ICE/PM machine and passive power converter (b) ICE/PM
machine and active power converter (c) ICE/HPM machine and passive power converter

vehicle road speed, essentially to provide a fixed or variable speed to the generator.
Therefore, in the series hybrid electric vehicle (SHEV) format, the battery pack can
provide the peak power for the vehicle DC-link while the combustion engine,
powered from fuel tank, provides the vehicle average power using an onboard
generator. Here, there are a number of ways in which the output voltage of the PM
(or other machine technology) generator may be facilitated, as illustrated schemat-
ically in Fig. 5.9. The PM generator output (three-phase in this case) can be passively
rectified, and the output voltage magnitude can be regulated by speed control of the
ICE prime mover, Fig. 5.9a.
However, voltage transients are usually much faster than the control dynamics of
ICEs, and hence an active rectification scheme is usually required and the ICE speed
effectively fixed, as shown in Fig. 5.9b. This adds power electronic complexity but
gives additional control functionality. Alternatively, the ICE can be used as the
prime mover to the HPM machine, discussed in Chaps. 3 and 4, the output of which
is connected to a simple passive rectification stage, as illustrated in Fig. 5.9c,
showing the HPM machine (again three-phase) in the powertrain system schematic.
As discussed in previous chapters in this book, the hybrid PM (HPM) machine is
predominantly a permanent magnetic machine, but with some facility to control its
output open-circuit voltage via a secondary field excitation source. Hence, the hybrid
terminology refers to the hybridization of permanent magnet and wound field
excitation in one machine package.
The function of HPM machines in SHEV powertrains is illustrated schematically
in Fig. 5.10. Here, the ICE acts as the prime mover to an electrical three-phase HPM
generator that is specified to match the vehicle terminal voltage transients, as shown
in Fig. 5.10, while maintaining constant HPM power output to the vehicle DC-link.
154 5 Electric and Hybrid Electric Powertrains

Fig. 5.10 Series HEV powertrain schematic with a three- or nine-phase HPM generator supplying
power to the vehicle DC-link

Whether the generator is a HPM topology or not, the quality of the generated DC
voltage and power will impact other system components in the vehicle powertrain, in
particular, the requirement for additional DC-link capacitance at the input to the
traction machine (TM) power electronic converter (traction drive), which is predom-
inantly a voltage source inverter. The three-phase HPM machine terminal AC
voltage is converted to DC via a passive rectification stage, shown as AC/DC in
Fig. 5.10. The HPM wound field excitation modifies the machine output voltage due
to the permanent magnets, again as illustrated in Fig. 5.10, via a simple excitation
control loop, as will be discussed in more detail in Chap. 6.

5.5.1 ZEBRA Battery

Battery-specific energy (Wh/kg) depends on the battery technology, where the

discharge and charge efficiency does not only depend on battery technology but
5.5 Series Hybrid Electric Vehicle (SHEV) 155

Table 5.4 ZEBRA battery parameters [95]

Capacity (Ah) 32
Rated energy (kWh) 17.6
Max. open circuit voltage (V) 575
Max. Regen. voltage (V) 670
Max. discharge current (A) 112
Cell type and ML3C
No. of cells 216
No. of strings 1
Specific power (W/kg) 178.5
Peak power (kW) at 2/3rd open-circuit voltage; for 30 s or until max temperature rated 32
335  C
Cooling Forced
air

also on the battery state-of-charge (SOC). Note, SOC is a key quantity to measure
the amount of electric energy stored in or released from the battery during charging
and discharging, respectively, and it is calculated based on (5.1).

QA
SOC ¼ ð5:1Þ
QT

where QA is the actual battery charge and QT is the total battery charge.
Usually, the battery discharge efficiency is at its highest value at high SOC
values, and it decreases with the SOC. The ZEBRA battery considered here is
capable of tolerating 5%–10% cell failure and discharging to full capacity without
degradation [184]. The ZEBRA battery has passed various tests, including crash test
at 50 km/h, short circuit, vibration, underwater, and external fire tests [184–
187]. Compared to Lithium-ion (Li-ion), the ZEBRA battery has a comparable
specific energy; however, its specific power is lower [184]. The SOC for the
ZEBRA battery considered here varies between 0.1 and 1, while the battery cell
open-circuit voltage varies between 2.32 V and 2.66 V. Table 5.4 lists specifications
of a 216-cell ZEBRA battery while Fig. 5.11 shows the battery open-circuit voltage
curve over a driving cycle.
There are three regions of battery operation, which are specified as Region 1: the
battery system has the highest charge (575 V), Region 2: the battery voltage is at its
nominal value (555 V), and Region 3: the battery voltage decreases to minimum
operating voltage from its nominal value. Furthermore, Fig. 5.12 shows the battery
terminal voltage transient and SOC of Region 2 in Fig. 5.11. Note, during operation,
a battery management interface (BMI) supervises the battery status, that is, voltage,
current, and SOC, and ensures that all the battery parameters are within the normal
operating states.
156 5 Electric and Hybrid Electric Powertrains

Fig. 5.11 Characteristics of ZEBRA battery [95]

Fig. 5.12 Region 2 of Fig. 5.11 terminal voltage and SOC for ZEBRA battery during NEDC [95]

5.5.2 Internal Combustion Engine (ICE)

Generally, ICEs demonstrate high efficiency and provide a low specific fuel con-
sumption over a relatively small region of their engine power-speed characteristics,
as shown Fig. 5.13. They also demonstrate particularly high fuel consumption and
emissions during transient engine operation. Hence, ICE operating point can be
adjusted by controlling its output power at a predefined fixed speed. The selection of
an adequate operating point assures an efficient and optimized ICE operation in
SHEVs. Here, a benchmark powertrain is used to model vehicle physical parameters,
required traction torque, and the energy demand, while a Toyota Prius 43 kW engine
actual test data [188] is used to model the ICE. Note that although the Prius peak
demand may reach 43 kW, the average power for a Toyota Prius over New European
Driving Cycle (NEDC) is around 3.2 kW [188]. The Toyota Prius considered in this
section is a 1.5-ton vehicle. Figure 5.14 shows that the scaled-down engine used
carbon dioxide (CO2) emission data of the Toyota Prius 43 kW engine based on both
5.5 Series Hybrid Electric Vehicle (SHEV) 157

Fig. 5.13 Efficiency map of a typical ICE [189]

Fig. 5.14 Toyota Prius engine emission for different power and speed [95]

an actual taxi engine and the calculated peak power requirements calculated for the
proposed driving cycles’ emissions based on CO2 for an average power varying from
0.5 kW to 3.2 kW and for a speed range of 477 RPM (50 rad/s) to 4770 RPM
(500 rad/s). As seen, the emissions are lowest when the ICE operates at 3000 RPM
(314 rad/s) while delivering 3–3.2 kW power. Therefore, in the analyzed SHEV
powertrain, the ICE is to be operated at its optimum speed of 3000 RPM to deliver a
fixed power to the HPM generator system.
Furthermore, the SHEV fuel consumption (fuel economy) based on ICE output
power (PICE) and efficiency (ηICE) can be calculated as
158 5 Electric and Hybrid Electric Powertrains

FCR  PICE  t
VFC ¼ ð5:2Þ
EC  ηICE

where FCR represents the fuel consumption rate, t is the instantaneous driving time,
and EC is the energy content of gasoline in consumed fuel rate. The VFC is used to
calculate the mass of the SHEV fuel tank (in kg). Note that the mass is based on the
maximum capacity of the vehicle fuel tank.

5.5.3 Engine-Mounted Multiphase HPM Generator

In the SHEV considered here, the vehicle is powered by two sources of energy, a
battery (ZEBRA) system as a main power source, and an ICE driving a HPM
generator as an auxiliary power source. In the scheme shown in Fig. 5.15, the ICE
is mechanically decoupled from the main SHEV powertrain using a multiphase
HPM generator. Therefore, mechanically decoupling the ICE from the vehicle
powertrain (and hence the road speed) and operating the engine at a fixed speed
and output power offers the potential for reduced engine size, fuel consumption, and
emissions. As discussed in the previous chapters, the HPM generator has two rotor
elements: a PM rotor with fixed excitation and a WF rotor with adjustable excitation.
At a given speed, the PM rotor induces a fixed stator voltage while the WF rotor
induces a variable but controlled stator voltage. Therefore, the HPM generator’s total
output voltage, which is the sum of induced PM and WF voltages, is controlled over
a prescribed range. The SHEV DC-link voltage may vary from its nominal value as
the vehicle accelerates/decelerates. The DC-link voltage profile shown in Fig. 5.15 is
measured when the vehicle operates on the NEDC driving cycle. As observed, the
voltage transiently exceeds its nominal value (555 V) during vehicle deceleration.
However, when the vehicle accelerates, the voltage significantly drops (to 350 V)
due to a sudden increase in the current required for the traction motor to drive the
vehicle to the acceleration. The higher current flow results in a higher equivalent
impedance voltage drop, hence transiently reducing the DC-link voltage. The stron-
ger the acceleration/deceleration, the higher the voltage drop/increase.
The HPM generator is interfaced to this variable voltage DC-link; therefore, its
rectified output voltage is required to follow the variations. Traditionally, the ICE
provides a variable speed to the shaft, hence facilitating the HPM generator output
voltage control. However, as previously discussed, the ICE speed is maintained fixed
to maintain an optimum efficiency. Therefore, as the electrical system dynamics are
much faster than those of the ICE, the HPM generator output voltage is controlled by
the WF rotor excitation to match the HPM generator rectified voltage to the DC-link
voltage.
5.6 Electric Vehicle Range Extender 159

Fig. 5.15 The SHEV powertrain components and DC-link voltage variation

5.6 Electric Vehicle Range Extender

5.6.1 Introduction

To extend the range of all-electric vehicles, a larger battery may be used. However,
alternatively, this can be achieved by using auxiliary power sources such as ICE or
fuel cell together with the main battery, as shown in Fig. 5.16. Various energy
management and control strategies for onboard and off-board range extender func-
tions for SHEV powertrain topology have been studied. In the following section, a
literature review of EV range extender studies will be discussed along with a detailed
evaluation of an ICE/HPM generator that is used as an onboard range extender in the
SHEV powertrain.
160 5 Electric and Hybrid Electric Powertrains

Fig. 5.16 Gasoline-powered ICE and fuel cell technologies for EV range extension (a) Lotus range
extender engine-generator system [190] (b) ALP-5 fuel cell range extender [191]

5.6.2 Literature Review of EV Range Extender Studies

Different control strategies for HEVs to optimize the operation of SHEVs for fuel
economy and emissions are investigated in [192–204]. The approach presented in
[192] optimizes the SHEV fuel consumption in the presence of three energy sources:
an ICE-generator set, a battery, and ultra-capacitors. The authors in [192] offer a
torque distribution strategy between the traction motors in the SHEV that minimizes
energy usage. The study in [193] conducted on a sports class SHEV utilizes the
battery system to reduce the power demand instead of the battery acting as a primary
energy source. A control strategy proposed in [193] minimizes the size of the battery
system while considering the driving range. In [194], the results show that by a
proper selection of battery size and operating the SHEV on combination of ICE and
on-board energy storge the fuel consumption is minimized. A study of cost map for a
bus utilizing SHEV powertrain is presented in [195], where the proposed control
strategy optimizes an ICE-generator set while targeting the desired emission. The
approach in [195] is based on a two-step algorithm for size reduction and demand
determination. A power distribution control for the SHEV energy sources, that is, a
hybrid thermostat strategy, is proposed in [196], where a thermostat strategy is
combined with traditional power flow to improve efficiency and achieve target
performance. Optimization of a small ICE in terms of fuel consumption and emis-
sion by maintaining a fixed ICE speed and a variable output torque is discussed in
[197]. However, the downsized ICE in [197] does not operate at its maximum
efficiency. A two-step strategy to maximize the fuel economy of a SHEV powertrain
configuration is proposed in [198], where a genetic algorithm is used to determine
component sizing and a dynamic programming is employed for supervisor control
optimization. Analysis of a 10-kW ICE-PM generator system supplying a constant
DC-link power for a small HEV is presented in [199, 200], while [201] proposes an
5.6 Electric Vehicle Range Extender 161

Fig. 5.17 Electric vehicle with an off-board range extender on a trailer (courtesy of EP Tender)
prototype [203]

Fig. 5.18 Plug-in Nissan e-NV200 van with a fuel cell range extender [205]
This vehicle photo is not sponsored, endorsed, or promoted by Nissan North America, Inc., or their
affiliates

optimized control for a range extension in SHEVs based on power demand predic-
tion. A simulation powertrain platform, as in [202], combines two energy sources: an
energy-dense ZEBRA and a power-dense supercapacitor. The ZEBRA battery is
specified to fulfill the vehicle range requirements, while the supercapacitor provides
the peak power for acceleration and regenerative braking. A control approach is
proposed in [203], where it manages the energy between a battery and a rental range
extender to minimize the end-use incurred costs. The range extender comprises
mainly an ICE coupled with an electric generator.
Furthermore, [203] investigates different operational scenarios and presents ben-
efits of convex optimization compared with dynamic programming. The authors in
[204] propose a hydrogen-generator system that can use a mix of gasoline and
hydrogen as a range extender for EV powertrain. An example of an actual
off-board and onboard EV range extender functions is shown in Figs. 5.17 and
5.18. Figure 5.17 shows an actual off-board ICE-generator system used as a range
extender for EV, while Fig. 5.18 shows a plug-in HEV that has an onboard fuel cell
energy source.
162 5 Electric and Hybrid Electric Powertrains

5.7 ICE/HPM Generator Range Extender in SHEVs

The structure of a SHEV powertrain with an ICE/HPM generator system is shown in

Fig. 5.19b, while Fig. 5.19a presents the nine-phase HPM generator prototype that
has previously been developed in Chap. 4.
A 2.5-ton vehicle [185] is chosen as a benchmark system based on which two
powertrain configurations are studied: (i) an all-electric powertrain utilizing a high
energy density battery system acting as a standalone energy source, here a commer-
cially available ZEBRA battery technology, and (ii) a SHEV powertrain featuring a
primary source, that is, a ZEBRA battery and an ICE/HPM generator system acting
as an axillary energy source. Figure 5.19 shows a schematic view of the SHEV
powertrain under study. The ICE considered here is based on available data from the

Fig. 5.19 Series hybrid electric vehicle powertrain (a) Nine-phase HPM generator prototype (b)
System configuration
5.7 ICE/HPM Generator Range Extender in SHEVs 163

Toyota Prius engine [188]. In the SHEV, Fig. 5.19, the ICE is used as a constant
energy source and, together with the HPM generator, acts as a range extender to the
SHEV powertrain. Therefore, the ICE operates at a fixed speed and spins a HPM
generator, the output of which is rectified and connected to the vehicle DC-link.

5.7.1 Vehicle Traction Machine Torque

The future of electrified drivetrain lies on the new landscape that considers compat-
ibility, reliability, efficiency, costs, and high torque density, a major requirement of
old and new types of electric machine topologies along with their drive schemes.
Brushless DC machines with nonsinusoidal (more-trapezoidal) back-EMF wave-
form are considered one of the established candidates for the growing field of
traction motor drive system [206–208]. In general, motor control is achieved via
either sensorless [209, 210] or sensored [208, 211, 212] control algorithm. Sensored
types incorporate the use of encoders, resolvers, or Hall effect sensors to always
detect rotor position with respect to the stator to achieve a proper voltage source
converter (VSC) commutation, while the sensorless control removes any sensored
components from the design. Note that it is hard to have a sensorless position
control, where torque control is difficult to be applied sensorlessly. Therefore, a
simplified brushless DC machine with an innovative motor drive system is of interest
to many researchers.
Here, for the upcoming analysis, the prime mover for the vehicle is a three-phase
brushless DC motor acting as a traction machine, with an integrated gear reduction
and differential drive to the vehicle back-axle. The efficiency of the gear system is
assumed to be constant, while the brushless DC traction machine efficiency is
calculated from the machine efficiency map. The brushless DC motor is controlled
via a three-phase VSC, the DC supply to which is provided by the onboard battery
pack, an ICE/HPM generator combination, using conventional vector control.
Moreover, the traction machine torque equation can be extracted from the road
vehicle kinematics, which is used to estimate the dynamic tractive requirement of the
vehicle drivetrain. According to the basic principles of physics, any type of vehicle
motion can be determined by analyzing the forces acting on it in the direction of
motion, as illustrated in Fig. 5.20. Hence, the vehicle resultant force governs the
motion based on Newton’s second law. Here, when the vehicle moves, it encounters
different resistive forces that try to retard its motion, such as tire to road power loss or
rolling resistance (Fr), aerodynamic resistance or aerodynamic drag (Fd), resistive
force related to road gradient or uphill resistance (Fg), and transient force required to
accelerate or retard the vehicle (Fa).

F r ¼ K r mg cos θ ð5:3Þ
164 5 Electric and Hybrid Electric Powertrains

Fig. 5.20 Forces acting on

a vehicle while in motion

1
F d ¼ ρC d A f v2 ð5:4Þ
2

F g ¼ mg sin θ ð5:5Þ

dv
Fa ¼ m ð5:6Þ
dt

where Kr is the rolling resistance coefficient, m is the mass of the vehicle, θ is the
road gradient, g is the gravitational constant, ρ is the air density, Cd is the drag force
coefficient, Af is the vehicle frontal area, and v is the vehicle linear velocity.
Therefore, the net traction force is defined in Eq. (5.7), and based on the equation
of motion, the road wheel torque equation can be calculated as in (5.8), where Jw,
ωw, and rw are the wheel inertia, angular velocity, and mean radius, respectively, and
df is a distribution factor proportion to torque distribution on the vehicle rear axle.

F net ¼ F r þ F d þ F g þ F a ð5:7Þ

dωw
T w ¼ Jw  þ d f  r w  F net ð5:8Þ
dt

The output torque of the traction machine can be modified if a gear stage is
included in the drivetrain system, which is related to the road wheel torque by the
5.7 ICE/HPM Generator Range Extender in SHEVs 165

total transmission gear ratio (ntg), transmission efficiency (ηt), and the machine rotor
inertia (Jm). By combining these components into (5.8) yields a general expression
for traction machine torque:

dωm 1
Tm ¼ Jm  þ T ð5:9Þ
dt ntg ηt w

Furthermore, the wheel and traction machine angular velocity can be expressed in
terms of vehicle linear velocity as

v
ωw ¼ ð5:10Þ
rw

v
ωm ¼ ntg ð5:11Þ
rw

Hence, the traction machine torque equation can be expressed in terms of vehicle
linear velocity by substituting (5.7), (5.8), (5.10), and (5.11) into (5.9) [213] as
shown in (5.12). Note that the parameters included in Eq. (5.12) depends on the
vehicle type, and the required mechanical power (Pm) to satisfy the prescribed
driving cycle is given in (5.13).
     
ntg ηt Jw d f rw m dv d f r w
Tm ¼ þ þ þ ðF þ F g þ F d Þ ð5:12Þ
rw ntg ηt r w ntg ηt dt ntg ηt r

Pm ¼ T m ωm ð5:13Þ

5.7.2 Hybridization Ratio

In the SHEV powertrain, the relative size of the battery and the ICE is an important
design criterion as it defines how strongly the energy sources are combined. There-
fore, for the scheme presented in Fig. 5.18, a hybridization ratio is characterized to
measure the power share between the battery and the ICE/HPM generator. This
power share depends on the driving cycle peak and average power requirements.
One of the design scenarios in the SHEV is the prevision of a steady power by
ICE/HPM generator system while the dynamic power is provided by the battery. The
ICE in this case is smaller, produces less emission, and is more efficient as it is
designed to provide a constant power at its optimum operating region. Therefore, in
the SHEV scheme presented in Fig. 5.18, the ICE is set to supply a constant vehicle
average power, spinning the HPM generator at a constant speed, while the ZEBRA
166 5 Electric and Hybrid Electric Powertrains

battery supplies the peak power as the vehicle accelerates. In the event of vehicle
idling or deceleration, the ICE/HPM generator system is used to charge the battery.
A constant power source that provides the vehicle energy, E (Wh), during a
driving cycle with T (hrs) duration is calculated as

E
Pconst ¼ ð5:14Þ
T

For the SHEV configuration, the constant power source is the maximum allow-
able ICE-HPM generator output power, hence,

Pconst ¼ PICE=HPM ð max Þ ð5:15Þ

Therefore, the minimum battery power required to provide vehicle total demand
is calculated as

Pbatð min Þ ¼ Ppeak  PICE=HPM ð max Þ ð5:16Þ

where Ppeak is the peak power demand. Thus, the hybridization ratio for the SHEV is
calculated as

Ppeak  PICE=HPM ð max Þ

HR ¼  100 ð5:17Þ
Pbat

Using equation (5.17), a percentage power share between the vehicle battery and
the ICE/HPM generator is defined.

5.7.3 Range Extender Sizing in SHEV Powertrain

Searching for the adequate ICE/HPM generator size depends on the electric vehicle’s
desired driving range. A sequential algorithm can be implemented in a simulation
platform to size the primary power source, here the ZEBRA battery, and axillary
power unit, here the ICE/HPM generator, in the SHEV powertrain. The algorithm
consists of two sections: (i) an all-electric and (ii) a hybrid model that searches for an
adequate onboard energy source and a hybridization ratio subject to the energy
source’s output power (kW), mass (kg), and a desired extended range (km), as
illustrated in Fig. 5.21. The input to the platform includes driving cycles, vehicle
physical parameters, and profile of electrified powertrain components for any type of
road vehicle. This repetitive process is intended for the hybrid mode only, which is
utilized to obtain an adequate average power percentage share (X) that lies between
(0 and 100%) of the maximum allowable ICE/HPM generator fixed power output, as
illustrated inside the red dotted line in Fig. 5.21. Note that the average vehicle power
demand over a specified driving cycle is set as a maximum allowable power to be
5.7 ICE/HPM Generator Range Extender in SHEVs 167

Fig. 5.21 Energy management algorithm for EV driving range extension

Table 5.5 Example of power requirements for a 2.5-ton vehicle over two driving profiles
Driving cycle profile Average power (kW) Peak power (kW) Cycle times (s)
NEDC 18 63.2 400
HWFET 20 62.0 765

provided by the axillary onboard power unit. The fuel consumption and emissions
are calculated based on power demand over the selected driving range and the
calculated hybridization ratio. Note that the ICE/HPM generator constant power is
maintained via a HPM generator excitation field control function, which will be
discussed in detail in Chap. 6.
Here, for two driving cycle profiles, such as NEDC and HWFET, there is a wide
disparity in the peak-to-average power requirements for a 2.5-ton vehicle, as sum-
marized in Table 5.5. For the NEDC suburban driving cycle, as shown in Table 5.5,
the vehicle peak power demand is 63.2 kW. Using the developed simulation
platform, the energy consumed during the driving cycle and time is calculated and,
based on Eq. (5.14), an 18-kW constant power source is set to produce all that
energy. Hence, the minimum battery power calculated based on Eq. (5.16) is
168 5 Electric and Hybrid Electric Powertrains

Table 5.6 Two different Parameters 3 kW 14.5 kW

HPM generator ratings
PM flux per single turn (mWb) 0.25 1.22
Axial length (mm) 40 176
Mass (kg) 10.7 52.1
Volume (mm3) 1.5 6.9
Phase number of turns 54 11
Parallel conductor per turn 1 8
Phase synchronous inductance (mH) 2.3 0.43
Stator resistance (Ω) 2.5 0.17
Rotor resistance (Ω) 16.3 34.4
Note: Rotor turns ¼ 80 turns, number of poles ¼ 32, speed ¼ 3000
RPM, peak EMF voltage ¼ 269 V, specific power ¼ 280 W/kg,
and power density ¼ 2.1 MW/m3 (for lamination volume exclud-
ing casing)

Fig. 5.22 Hybridization

ratio between pure
all-electric and pure gasoline
powertrain

45.2 kW. Therefore, to switch from an all-electric to a SHEV powertrain, the

maximum battery power required is 63.2 kW (i.e., a total of two 216-cell ZEBRA
batteries listed in Table 5.4), while the minimum battery power is 45.2 kW. There-
fore, the ICE/HPM generator is set as the constant power source whose size varies
from zero in an all-electric to 18 kW for the SHEV powertrain. Similarly, for a
62-kW vehicle peak power in HWFET driving cycle (Table 5.5), a 20-kW constant
power source, ICE/HPM generator, is calculated to be sufficient to produce all the
required energy while the minimum battery power is calculated as 42 kW.
Therefore, to turn the benchmark 2.5-ton pure EV into a SHEV, an axillary
ICE/HPM generator unit is set to provide an average fixed power while spinning
at a fixed speed of 3000 RPM. As discussed, this will result in ICE minimum fuel
consumption and maximum efficiency. Table 5.6 lists the parameters for 3-kW and
14.5-kW HPM generator ratings, which are calculated for 95% and 77% powertrain
hybridization rations, respectively, as illustrated in Fig. 5.22.
5.7 ICE/HPM Generator Range Extender in SHEVs 169

5.7.4 Study Cases

To examine the ICE/HPM generator system, two case studies are considered here.
Case 1 is composed of two scenarios that compare a pure EV and a SHEV with a
hybridization ratio (HR) of 95% based on the 3-kW ICE/HPM generator system over
both NEDC and HWFET driving cycles. Case 2 is composed of two scenarios that
compare a pure EV and a SHEV with a HR of 77% based on a 14.5-kW ICE/HPM
generator system that is scaled from the 3-kW prototype unit. Case 2 attains
the desired minimum range of driving between some destinations without passing
the allowable maximum ICE/HPM generator system output power. Note that the
ZEBRA battery size in the pure EV in Case 1 is calculated based on a mass
equivalent to the total weight of 3-kW ICE/HPM generator and full fuel tank. For
Case 2, the same approach is taken, albeit for a 14.5-kW ICE/HPM generator.
The analysis results of the suburban driving cycle case studies are shown in
Table 5.7. The results illustrate that by utilizing the proposed SHEV configuration
compared with all-electric powertrain, the driving range can be extended by 18% and
23% for HWFET and NEDC cycles, respectively, in Case 1, and the driving range
can be extended by 34% and 52% for HWEFT and NEDC, respectively, in Case
2. However, using the combination of Case 1, the vehicle could not reach the
proposed destination, but the energy used was better than the pure battery scenario.
In Case 2, where equal energy storage mass for both scenarios was taken into
account, the new rating of the ICE/HPM generator system in the vehicle powertrain
reached the specified destination and passed it by 102 km with less CO2 emission,
compared with the European Commission proposal 2015–2021, which is

Table 5.7 Comparison between all-electric (pure EV) and SHEV over two driving cycles
Cases Case 1 Case 2
SHEV (ZEBRA SHEV (ZEBRA
Scenarios Pure EV (ZEBRA battery and 3.0- Pure EV (ZEBRA battery and 14.5-
(1–2–3-4) battery) kW ICE/HPM) battery) kW ICE/HPM)
Driving HWFET NEDC HWFET NEDC HWFET NEDC HWFET NEDC
cycle
Range (km) 103 89.3 121.8 110.3 154.3 124.2 361.9 437.5
Fuel use (l) 0 0 1.57 1.72 0 0 21.7 30.9
Fuel use 0 0 0.013 0.016 0 0 0.06 0.071
(l/km)
Fuel use 0 0 220.7 181.1 0 0 47.1 40.0
(m/g)
SOC 0.1 0.18 0.1 0.149 0.1 0.24 0.1 0.1
CO (g/km) 0 0 0.56 0.65 0 0 1.87 2.20
CO2 (g/km) 0 0 25.3 30.14 0 0 86.5 102
HC (g/km) 0 0 0.136 0.159 0 0 0.456 0.537
HC (g/km) 0 0 0.265 0.31 0 0 0.89 1.05
Note: The energy content of gasoline in 1 liter ¼ 32.2 MJ and the gram to liter conversion
factor ¼ 737.22
170 5 Electric and Hybrid Electric Powertrains

Fig. 5.23 Battery terminal voltage for all-electric and hybrid mode for a 14.5-kW ICE/HPM
generator over NEDC case (a) All-electric mode (b) Hybrid mode

120–130 g/km [214]. The impact on the driving range due to the proposed ICE/HPM
generator size in SHEVs is shown in Fig. 5.23. Such that, Fig. 5.23a presents the
battery terminal voltage for all-electric and Fig. 5.23b presents the hybrid mode of
Case 2 for the NEDC suburban driving scenario. (Fig. 5.23).

5.8 Conclusion

In this chapter, an overview of different EVs and HEVs based on powertrain

configuration, per-unit cost, fuel economy, driving range, and wheel architecture is
presented. Various battery technologies used in electric and hybrid electric vehicle
powertrains are discussed. Electric vehicle range extension, fuel economy, and
emissions are evaluated over HWFET and NEDC driving cycles at different vehicle
powertrain hybridization ratios, which are characterized by the ICE/HPM generator
and ZEBRA battery system in SHEV configuration.
It is shown that by employing a 14.5 kW ICE/HPM generator system, as
discussed in Sect. 5.7.4, the requirements of emissions lower than the 2015–2021
European Commission proposal are achieved. The battery energy can be used more
efficiently, and an improved battery life is suggested by virtue of the reduced
transient battery loadings.
Chapter 6
Operation and Characterization
of Multiphase HPM Generator in SHEV
Powertrain

6.1 Introduction

Generally, ICEs demonstrate a low specific fuel consumption over a relatively small
region of their engine power-speed characteristics. They also demonstrate particu-
larly high fuel consumption and emissions during transient engine operation. There-
fore, mechanically decoupling the ICE from the vehicle powertrain and operating the
engine at a fixed speed and output power offers the potential for reduced engine size,
fuel, and emissions. For hybrid EV topologies, a significant saving in energy can be
realized by leveling of the vehicle energy demand via the inclusion of a peak power
buffer in the vehicle powertrain, downsizing the ICE power capability and operating
the ICE over the most optimum region of the engine power-speed characteristics.
Therefore, ICE acts as the prime mover to an electrical generator specified to satisfy
the vehicle average power demands, while a battery system supplies the main peak
power demands.
Hence, one of the control objectives is to keep the ICE/HPM generator output
power constant while minimizing speed variation via regulation of the HPM wound
field excitation. Recent studies have shown that brushless permanent magnet gener-
ators with a secondary source of excitation are an interesting topology for imple-
mentation in hybrid EV powertrain [94–96, 106–108, 215]. For HEVs, the use of
dual power and energy sources requires a detailed analysis of the drivetrain in order
to optimize component specifications and energy management strategies [216]. In
particular, the vehicle powertrain DC-link voltage may typically have a 30% voltage
variation during acceleration and deceleration due to vehicle power transients, an
example of which is illustrated in Fig. 6.1 [217, 218]. Hence, a form of voltage
management is required for the ICE/electrical generator system, thus the interest in
HPM generators that can be designed to match the 30% voltage variation.
It is widely recognized in the literature that, in terms of higher efficiency and
power density, PM machines with an active rectification stage represent a better
candidate for use in SHEVs than classical wound rotor machines with passive

© Springer Nature Switzerland AG 2022 171

A. S. Al-Adsani, O. Beik, Multiphase Hybrid Electric Machines,
https://doi.org/10.1007/978-3-030-80435-0_6
172 6 Operation and Characterization of Multiphase HPM Generator in SHEV Powertrain

Fig. 6.1 Schematic of a series hybrid EV powertrain and typical DC-link voltage variation

rectification [36, 37]. Afjei et al. present a new configuration of a switch reluctance
motor with field assistance that represents a hybrid generator topology [36, 37]. A
major drawback of the active rectification scheme is the cost of power electronics,
typically accounting for two-thirds of the total generator system cost.
Yoo et al. discuss a power flow management scheme for a series hybrid electric
vehicle, proposed mainly for military applications [219]. The powertrain discussed
by Yoo et al. comprises a diesel engine linked to an interior PM (IPM) generator and
passive rectification stage, lead-acid battery, and supercapacitor bank. The primary
voltage source is the engine-IPM generator, designed with a terminal voltage equal
to the rated DC-link voltage. The output of the generator is controlled by varying the
engine speed via the engine throttle. The lead-acid battery provides additional
energy storage, while the supercapacitor is used to supply short-term power
fluctuations [219].
The improvement of DC-link voltage variation by replacing part of the PM
machine excitation with different wound field configurations has attracted some
research efforts in recent years. For example, Bin He et al. discuss the average
6.2 HPM Machine Back-EMF Control Strategy 173

modeling and control of a diesel-based auxiliary power unit (APU) that consists of a
diesel engine, wound field synchronous generator, and three-phase passive rectifier
stage [220, 221]. The diesel APU provides the primary energy source for the vehicle
while the battery acts as a transient power buffer.
In the Bin He papers [220, 221], the vehicle DC-link voltage is supplied and
regulated during power transients by the diesel engine fuel injection and synchro-
nous generator field excitation. However, fuel consumption and pollutant emissions
will not be minimized for this system due to the varied engine operating speed. A
different operating philosophy is considered in this chapter. Here, it is assumed that
the ICE is operated within a very tight speed envelope delivering a constant power of
3 kW to the vehicle powertrain. An average power demand of 3 kW is typical of
small, 1.0–1.2 ton, urban vehicles and normally supplements the battery energy or
charges the battery when the vehicle is at rest. Here, the battery technology consid-
ered is ZEBRA [222, 223], and the vehicle is assumed to operate over repetitive
NEDC and ECE-15 driving cycles, these being automotive industry standard cycles
for vehicle powertrain performance evaluation [224]. Note that this choice of battery
technology or vehicle powertrain is not claimed to be the most optimum or appro-
priate solution but is considered purely to investigate the HPM generator excitation
requirements and operating philosophy. The choice of 3 kW arises from the average
power required to drive a typical family 1.5-ton vehicle over the ECE-15 driving
cycle [225].
The HPM machine wound field is modified to facilitate the control of the total
generated back-EMF. This chapter proposes a guideline for the method and imple-
mentation of a suitable control strategy and investigates the HPM machine stator coil
number of turns, where design decisions posed by higher system DC-link voltages
will be discussed. Given the available data for higher voltage ZEBRA battery-
powered vehicles (i.e., 587 VDC), the initial vehicle system study is based on such
a high voltage vehicle powertrain. Furthermore, characterization and operational
envelope of the multiphase ICE/HPM generator system at low DC-link voltage
winding (18-turn) will be discussed in detail over NEDC driving cycle.

6.2 HPM Machine Back-EMF Control Strategy

Dynamic models that can be used to analyze the system behavior of an

ICE/generator system, power electronic converter, and controller scheme in HEV
applications have been widely reported in the literature [226, 227]. This chapter
utilizes a mixture of machine data obtained by FEA, vehicle DC-link voltage
variation during driving cycle operation, and Matlab/Simulink vehicle simulation
data [202] to link the generator dynamic model with real data collected by a vehicle
data acquisition system. Part of the control of this dynamic system is subjected to
input and output constraints, such as the ICE/HPM generator, which is controlled via
an adaptive means of field excitation adjustment to present the overall system
dynamic behavior in offline simulation.
174 6 Operation and Characterization of Multiphase HPM Generator in SHEV Powertrain

6.2.1 Control Strategy Analysis

To deduce the HPM machine total back-EMF control strategy, the Matlab/Simulink
model presented in Chap. 4 and illustrated in Fig. 4.15 is expanded to include the
vehicle DC-link voltage variation during dynamic driving and the WF back-EMF
regulation function via a flux-linkage feedback loop. An overview of the expanded
model is illustrated in Fig. 6.2, showing the HPM generator, rectifier stage, and
variable DC-link voltage potentials, for the case of VRDC > VDC-link, which supports
the derivation of the system power flow equation. When delivering a constant power
into the vehicle DC-link, the rectified DC voltage (VRDC) due to the back-EMF of the
HPM generator has to be slightly greater than the DC-link voltage to overcome the
DC-link resistance (RDC-link) and other voltage drops while facilitating the desired
DC power flows.
The HPM machine back-EMF control strategy calculations are divided into three
stages. The first stage calculates the rectified output voltage for the required power
output to the vehicle DC-link. The initial calculation steps are

V RDC ¼ I DC RDClink þ V DClink ð6:1Þ

V RDC  V DClink
I DC ¼ ð6:2Þ
RDClink

PDCD ¼ I DC V DClink ð6:3Þ

Fig. 6.2 Simplified schematic diagram of HPM generator system

6.2 HPM Machine Back-EMF Control Strategy 175

RDClink PDCD
V RDC ¼ V DClink þ ð6:4Þ
V DClink

where IDC is the DC current and PDCD is the desired DC output power. To account
for any offset between the desired and demand DC-link output power, an error signal
is added to (6.4):
 
R P
V RDC ¼ V DClink þ DClink DCD þε
V DClink

or
 
PDCD RDClink þ V 2DClink
V RDC ¼ þε ð6:5Þ
V DClink

where the error signal is given by

ε ¼ ðPDCD  PDCM Þk ð6:6Þ

and PDCM is the actual DC output power and k is the error tuning gain.
Furthermore, the second stage should provide the required excitation current
value based on the previously obtained VRDC. Finally, the third stage utilizes the
required excitation current, as specified in (6.7), to calculate the total HPM machine
flux-linkage per phase, and this in turn is substituted into the phase back-EMF
equation (4.39) in Chap. 4.

If var
λHPM ¼ λPM  λWF var ¼ λPM  λWF rated ð6:7Þ
If rated

where λPM and λWF _ var are fixed PM and variable WF flux-linkages, respectively,
λWF _ rated is the rated WF flux-linkage, and If _ rated and If _ var are rated and variable
WF currents, respectively.
The model of Fig. 6.2 is now represented by the open-loop single line diagram
illustrated in Fig. 6.3, as applied to the 18-turn, 9-phase HPM machine that was
described in Chap. 4 and the model analyzed over incremental time periods
representing a few electrical periods of the HPM machine operation, where the
DC-link voltage is assumed constant. This analysis facilitates the development of
the HPM machine design and control philosophy while providing benchmark data
and waveforms.
Note that in order to investigate the full system response characteristic, the
DC-link voltage was varied over a much wider range than the 30% variation
expected during vehicle operation, as illustrated in Fig 6.4, where the system
DC-link voltage is varied from 40 to 250 V. The utility of this approach will be
demonstrated later in this chapter where different voltage ranges, winding turns, and
excitation schemes are considered against the options for field excitation control.
176 6 Operation and Characterization of Multiphase HPM Generator in SHEV Powertrain

Fig. 6.3 HPM generator open-loop control system block diagram

Fig. 6.4 DC-link voltage and peak phase back-EMF with respect to the WF excitation current at
constant DC-link output power (3 kW)

It was shown in Chap. 4 that the maximum DC-link power, which can be
delivered by the HPM machine system at a fixed speed and hence fixed back-EMF
with zero wound field excitation and at the thermal limit, was 3.15 kW. Control is
applied to the system model to ensure that the HPM maintains a constant DC-link
output power of 3 kW with minimal DC-link voltage ripple. Figure 6.4 illustrates the
total (PM and WF) peak phase back-EMF and the DC-link voltage variation at this
constant power of 3 kW for the 18-turn, 9-phase HPM machine with respect to
wound field excitation current. Because of the fixed power output requirement, the
field excitation current options are constrained, as illustrated by Fig. 6.4, where, for
this specific machine, the HPM machine peak phase back-EMF can only be regu-
lated over a field current range from 0.4 to 4.2 A. This result is a function of the
6.2 HPM Machine Back-EMF Control Strategy 177

Table 6.1 Operational points data of Fig. 6.4

Operational VDC-link Vterm (peak) Back-EMF (peak) Is (peak)
points If (A) (V) (V) (V) (A)
(a) 6.83 40 23.9 138.5 28.5
(b) 2.36 55 32.1 106.7 20.7
(c) 0 76 44.1 89.9 14.9
(d) 0.4 95 54.6 87.0 11.9
(e) 0 114 64.9 89.9 10.2
(f) 2.46 160 90.5 107.4 8.2
(g) 5.04 200 112.7 125.8 7.2
(h) 6.6 225 125.7 136.9 6.7

machine stator turns and DC-link power flow demand, as will be demonstrated later.
As can be noted from Fig. 6.4, the back-EMF boost mode represents the dominant
operating mode for this choice of HPM machine winding power demand and
operating speed. Data for the points illustrated in Fig 6.4 are given in Table 6.1.
When the DC-link voltage is low, the DC-link current must be high to maintain
the same output power. Therefore, the field excitation current and hence total
machine back-EMF are high since the total machine back-EMF has to compensate
for the machine reactive voltage drop and the voltage drops in the rectification
scheme. Essentially, the HPM machine is operating at what equates to a poor lagging
power factor, as illustrated by Fig. 6.5a showing machine terminal voltage, total
back-EMF, and phase current when generating into a DC-link voltage of 40 V. This
generating condition is also identified on Fig. 6.4 as Point (a). As the DC-link
voltage is varied from 40 to 225 V, that is, Points (a) to (h) on Fig. 6.4, the HPM
phase current reduces and hence the machine and rectification voltage drops reduce.
Referring to Figs. 6.4, 6.5, and 6.6 illustrate phase terminal voltage back-EMFs and
current as the DC-link voltage varies from 40 VDC, Fig. 6.5a, to 225 VDC, Fig. 6.6h,
and the HPM machine field excitation is varied from a maximum positive current, to
a maximum negative current of 0.4 A and again to a maximum positive current, to
maintain the fixed generated power demand of 3 kW. Figure 6.7 illustrates the same
waveforms for points (a), (d), and (h), but with magnified axes to show the phase
current detail and respective phase relationships more clearly. As can be seen, the
boost mode at higher DC-link voltages results in the phase current becoming
discontinuous, Fig. 6.7h.
Note, the simulation results of Figs. 6.5 and 6.6 are not in steady state until 
0.4 ms, hence the asymmetry of the waveforms during this time period. The
simulation settles at around 0.4 ms, as inferred from the resulting waveforms.
178 6 Operation and Characterization of Multiphase HPM Generator in SHEV Powertrain

Fig. 6.5 Phase voltage and current waveforms for four DC-link voltage levels for the 18-turn,
9-phase HPM machine at constant DC-link output power (3 kW). (a) VDC-link ¼ 40 V, If ¼ 6.83 A
(b) VDC-link ¼ 55 V, If ¼ 2.36 A. (c) VDC-link ¼ 76 V, If ¼ 0 A (d) VDC-link ¼ 95 V, If ¼ 0.4 A

6.2.2 DC-link Design Options

The result of Fig. 6.4 presents a number of control options if only a 30% variation in
the vehicle system DC-link voltage is required. The questions to be resolved are
• Should the PM component of the HPM output voltage be reduced by the WF
back-EMF component (i.e., bucked)?
• Should it be increased by the WF back-EMF component (i.e., boosted)?
• Should some element of buck-boost be considered for design?
Therefore, having considered the analysis of the 18-turn machine over a much
wider range of DC-link voltages, machines having the same lamination and PM
design but with increased turns per phase were analyzed over wide voltage ranges,
but still delivering the required 3 kW at 3 kRPM. Table 6.2 gives the machine
winding electrical parameters for winding designs from 18 to 67 turns. Figure 6.8
illustrates the trend in DC-link voltage and peak phase back-EMF characteristics
with respect to wound field excitation current for HPM machines having turns per
phase ranging from 18 to 67. The results for the 18-turn machine are as per Fig. 6.4,
while the results for turns from 62 to 67 come from winding design optimization for
a vehicle system having a nominal battery open-circuit DC-link voltage of
6.2 HPM Machine Back-EMF Control Strategy 179

Fig. 6.6 Phase voltage and current waveforms for four DC-link voltage levels for the 18-turn,
9-phase HPM machine at constant DC-link output power (3 kW). (e) VDC-link ¼ 113.5 V, If ¼ 0 A (f)
VDC-link ¼ 160 V, If ¼ 2.46 A. (g) VDC-link ¼ 200 V, If ¼ 5.04 A (h) VDC-link ¼ 225 V, If ¼ 6.6 A

600 V. The results for the 53-turn design are included as a check to verify the trend in
the characteristic curves from 18- to 67-turn. As with the 18-turn machine design, the
options to buck, boost, or buck-boost are determined by matching the HPM machine
output to the required specified DC-link variation by choice of an appropriate
number of turns per phase.
Having simulated open-loop characteristics for the HPM as a function of winding
turns (Fig. 6.8), a curve fitting function based on a sixth-order polynomial regression
method is applied to the characteristics to calculate the desired nonlinear high order
equations that represent the required wound field excitation current for a specified
DC-link power demand, HPM machine parameters, and DC-link voltage variation.
The open-loop model of Fig. 6.3 is now closed and includes a block with a defined
curve fit function and a limiter block that sets the upper and lower limits to the
DC-link voltage excursions. The excitation current calculation then inputs to the
flux-linkage calculation block, as illustrated in Fig. 6.9, showing the complete HPM
generator closed-loop control system block diagram. The second stage of the HPM
system analysis procedure now calculates the desired HPM-rectified DC output
voltage, VRDC, from the difference in actual and demand power, and system
DC-link voltage, VDC-link, as illustrated in Fig. 6.9.
180 6 Operation and Characterization of Multiphase HPM Generator in SHEV Powertrain

Fig. 6.7 Magnified views

of Figs. 6.5a and d and
Fig. 6.6h illustrating
waveform phase
relationships. (a) VDC-
link ¼ 40 V, If ¼ 6.83 A (d)
VDC-link ¼ 95 V, If ¼ 0.4 A
(h) VDC-link ¼ 225 V,
If ¼ 6.6 A
6.3 HPM Machine Output Power Control 181

Table 6.2 HPM machine winding electrical parameters as a function of stator turns
nstc (turns) 18 53 62 63 65 67
Rs (Ω) 0.245 1.466 2.006 2.039 2.167 2.373
LS (mH) 1.16 9.67 13.23 13.73 14.64 15.49
λPM (mWb-turn) 4.468 13.155 15.389 15.637 16.134 16.630

Fig. 6.8 DC-link voltage and peak phase back-EMF with respect to the WF excitation current for
varying turns per phase and constant DC-link output power (3 kW) (a) DC-link voltage versus WF
excitation current (b) Back-EMF versus WF excitation current

6.3 HPM Machine Output Power Control

6.3.1 Introduction

It is well known that the dynamic behavior of vehicle power requirements presents
some control issues to the operational philosophy of hybrid EVs. Therefore, analysis
of the vehicle powertrain is required in order to optimize component specifications
and develop and understand the impact of varied energy management strategies. To
182 6 Operation and Characterization of Multiphase HPM Generator in SHEV Powertrain

Fig. 6.9 HPM generator closed-loop control system block diagram

investigate the function of a HPM generator operating in a hybrid EV powertrain and

thus be in a position to optimize or improve the machine electromagnetic design,
case study designs must be undertaken. Here, the primary energy source of the
hybrid powertrain is a ZEBRA battery having maximum and minimum terminal
voltage levels of 585 and 405 V, respectively. The ICE-driven HPM generator
system is used as a constant energy input to the vehicle powertrain DC-link,
essentially acting in a range extending function.
The electrical system dynamics are much faster than those of the ICE. Thus, the
function of the HPM generator excitation is to match the generator output voltage to
the varying DC-link voltage of the vehicle system while delivering a fixed power
input to the DC-link. Constant ICE/HPM generator output power therefore requires
control of the HPM generator field excitation current, which is calculated by
feedback signals that monitor the DC-link voltage and current variations, HPM
generator output power, and the desired output power, as previously discussed.
Thus, the feedback signals provide continuous control over the field current that
forces the generator terminal voltage to be larger than the DC-link voltage by a
certain small value to maintain constant generator output power. As a result of this
auxiliary hybrid power source control, an extended vehicle range will be achieved
while maintaining maximum ICE efficiency. In the case studied, vehicle range is
extended by more than 50%, for both repetitive NEDC and ECE-15 driving cycles,
by the addition of a 3-kW range extender to the vehicle powertrain [164, 213]. The
vehicle is a 2.4-ton electric vehicle operating as a public taxi in urban city areas. The
range extender is a downsized ICE driving the HPM. The additional range could be
realized by a larger ICE/HPM generator combination, as discussed previously in
Chap. 5.
For the ZEBRA battery technology, the battery open-circuit voltage, as illustrated
in Fig. 6.10, shows full SOC (1.0 p.u.) (a), a gradually reducing region from around
0.95 to 0.2 SOC (b), a reduction in open-circuit voltage at 0.2 SOC, which is only
6.3 HPM Machine Output Power Control 183

Fig. 6.10 ZEBRA battery open-circuit voltage versus SOC characteristics [40]

Fig. 6.11 Battery terminal voltage during repetitive NEDC driving cycles (a) All-electric case (b)
Hybrid case

characteristic of this battery technology (c), and the reducing voltage to fully
discharged (0.2 to 0.0 SOC) (d) [40].
Figure 6.11 illustrates the vehicle system DC-link voltage variation during
repetitive NEDC cycling showing the case for all-electric operation (a) and for series
hybrid operation (b), or range extender mode where the range extender adds an
additional constant 3 kW to supplement the battery energy. Similarly, Fig. 6.12
illustrates the vehicle system DC-link voltage variation during repetitive ECE-15
cycling showing the case for all-electric operation (a) and for series hybrid operation
(b), or range extender mode as before. In both figures, the battery terminal voltage
reductions are caused by acceleration current demands and the consequential voltage
drop due to the battery internal resistance. Likewise, the peaks are due to regener-
ative current braking. Note that the mean voltage envelope is indicative of the
ZEBRA battery open-circuit voltage characteristic, as shown in Fig. 6.10 [213].
184 6 Operation and Characterization of Multiphase HPM Generator in SHEV Powertrain

Fig. 6.12 Battery terminal voltage during repetitive ECE-15 driving cycles (a) All-electric case (b)
Hybrid case

Figure 6.13 illustrates examples of the 3 kW HPM generator during part of the
duty cycles of Fig. 6.11, showing the rectified DC-link output voltage (a), the
generator DC-link current output (b), and the rotor excitation current (c).

6.3.2 HPM Generator Operating Scenarios

A number of operating scenarios can be considered for the ICE/HPM generator

system control. To investigate and understand the functionality of the ICE/HPM
generator in a series hybrid EV powertrain, different operating scenarios relating to
the battery terminal voltage levels are considered. Five case studies are made based
on predefined HPM machine open-circuit back-EMF voltages with zero wound field
excitation. With regard to the battery terminal voltage of the vehicle system, as
illustrated in Fig. 6.11b, five operating scenarios are considered, as illustrated in
Fig. 6.14. For each of the five scenarios, the vehicle system voltage varies from
420 VDC [40], requiring a predominantly boosting function from the WF back-EMF
component, to 588 VDC [40], where the WF back-EMF function is essentially
bucking that of the PM back-EMF.
To determine the intermediate voltage levels of interest, and hence the HPM
system operational philosophy, a further three regions (that are comparable to the
battery open-circuit voltage characteristic previously illustrated in Fig. 6.10) were
chosen for study, as illustrated in Fig. 6.14, showing the HPM PM back-EMF
voltage levels for the five operating scenarios considered. There is therefore a
decision to be made as to the operating voltage of the HPM generator with zero
WF excitation, that is, whether the field current excitation increases or decreases the
net terminal voltage of the generator to match the DC-link voltage variation at
constant output power, as illustrated schematically in Fig. 6.14 by five-scenario
concepts.
6.3 HPM Machine Output Power Control 185

Fig. 6.13 HPM generator

system variables variation
for 3 kW at 3000 RPM (a)
DC-link voltage (b)
Generator DC-link current
(c) Rotor excitation current
186 6 Operation and Characterization of Multiphase HPM Generator in SHEV Powertrain

Fig. 6.14 Five operating scenarios of the HPM generator with respect to the DC-link voltage levels
in volts
6.3 HPM Machine Output Power Control 187

6.3.3 Energy Loss Prediction for Two Driving Cycles

The HPM generator system model is solved using ZEBRA battery terminal voltage
variation from repetitive NEDC and ECE-15 driving cycles provided from
[213]. Key simulation data for the NEDC and ECE-15 driving cycles are presented
in Tables 6.3 and 6.4, respectively. These tables summarize the HPM machine
energy loss over the full battery discharge regime from 1.0 to 0.0 SOC. It is assumed
that the machine iron, winding, and frictional losses are constant for the scenarios
considered and that the machine copper losses form the dominant energy loss
mechanism. In terms of HPM generator system energy losses, the results presented
in Tables 6.3 and 6.4 show that scenarios 3 and 4 produce the lowest total HPM
machine energy losses over the full battery discharge regime. This is further illus-
trated in Fig. 6.15, showing the minimum energy loss solution for both the NEDC
and ECE-15 driving cycles.
Referring to Tables 6.3 and 6.4, scenario 1 shows the highest average rotor copper
and total energy losses; however, the stator copper loss is less than that for the other
scenarios due to the stator winding turns. For scenarios 2, 3, 4, and 5, loss of the
HPM wound field excitation by any means during vehicle operation would result in
(high) fault current being generated into the vehicle system DC-link. Therefore, the
HPM machine and rectifier stage should be fuse protected to prevent such a system
fault. Alternatively, although scenario 1 produces the highest energy loss in the HPM
generator system, it represents the most robust in terms of WF failure, a decision
ultimately for the vehicle system designer. Note that, as discussed, the data in
Tables 6.3 and 6.4 are obtained from the simplified simulation model. However,
data for the excitation current (ΔIf) obtained from the detailed simulation model are

Table 6.3 HPM machine rotor excitation options for repetitive NEDC driving cycles
Scenario Scenario
Parameters Scenario 1 2 Scenario 3 4 Scenario 5
Voltage due to PM (V) 420 515 535 555 575
DC voltage range due to WF 167 72 52 32 +12
(V) (ΔVDC ¼ 167V) 0 95 115 135 155
Excitation voltage range ΔVf +46 +19.2 +16 +8.6 +3.5
(V) 0 26.6 29.5 37.5 42.5
Excitation current range ΔIf 2.624 0.9 0.7574 0.375 0.1257
(A) (2.939) (0.9207) (0.75268) (0.42828) (0.117)
0 1.25 1.353 1.63 1.795
(0.01396) (0.98) (1.0324) (1.138) (1.2487)
RMS excitation current (A) 1.565 0.424 0.411 0.523 0.654
RMS phase current (A) 1.956 1.852 1.844 1.825 1.796
Average stator copper loss (W) 50.49 61.89 62.41 64.94 68.85
Average rotor copper loss (W) 39.97 2.93 2.75 4.47 6.99
Stator energy loss (kJ) 571.5 687.0 692.8 720.8 764.2
Rotor energy loss (kJ) 443.7 32.5 30.5 49.6 77.6
Total energy loss (kJ) 1015.2 719.5 723.2 770.4 841.9
188 6 Operation and Characterization of Multiphase HPM Generator in SHEV Powertrain

Table 6.4 HPM machine rotor excitation options for ECE-15 driving cycles
Scenario Scenario
Parameters 1 2 Scenario 3 Scenario 4 Scenario 5
Voltage due to PM (V) 420 515 535 555 575
DC voltage range due to WF +168 73 53 33 +12
(V) (ΔVDC ¼ 199V) 31 126 146 166 187
Excitation voltage range ΔVf (V) 45 18.5 15 7.89 2.5
9 37 40.5 47 51.5
Excitation current range ΔIf (A) 2.56 0.86 0.7 0.35 +0.15
(2.939) (0.9208) (0.7527) (0.42828) (0.1169)
0.52 1.75 1.9 2.03 2.18
(0.706) (1.247) (1.2535) (1.2762) (1.3067)
RMS excitation current (A) 1.682 0.369 0.319 0.368 0.503
RMS phase current (A) 1.925 1.821 1.816 1.797 1.766
Average stator copper loss (W) 48.9 59.86 60.54 62.95 66.63
Average rotor copper loss (W) 46.2 2.226 1.659 2.204 4.127
Stator energy loss (kJ) 988.3 1209.8 1223.5 1272.3 1346.6
Rotor energy loss (kJ) 933.7 45.0 33.5 44.6 83.4
Total energy loss (kJ) 1922.0 1254.8 1257.1 1316.8 1430.0

Fig. 6.15 Total energy

losses for the NEDC and
ECE-15 driving cycles

also presented, for comparison, as shown in parentheses. Although there is some

inconsistency in the excitation current range between the actual and simplified HPM
generator model, this inconsistency can be considered a worst-case calculation;
hence, the obtained results are considered acceptable regarding the choice of best
operating scenario. Thus, for the lowest energy losses, scenarios 2 and 3 present the
preferable operating scenarios. However, scenario 3 was selected due to the lower
rotor energy loss compared with scenario 2.
Figure 6.16 illustrates the advantage of scenario 3 over the other scenarios in
terms of average wound rotor excitation power loss, while Fig. 6.17 illustrates the
HPM generator wound field excitation current variation during repetitive NEDC
driving cycles for the five operating scenarios.
6.3 HPM Machine Output Power Control 189

Fig. 6.16 HPM machine

rotor power loss for the five
operating scenarios and two
driving cycles (a) Repetitive
NEDC driving cycles (b)
Repetitive ECE-15 driving
cycles

6.3.4 Solving Final Choice with Full Simulation Model

To finalize the HPM machine design study, the 63-turn HPM design is analyzed at
fixed DC-link voltage levels using the full dynamic model as a cross-check of the
simplified model and to illustrate actual phase voltage and current waveforms.
Figure 6.18 illustrates the total (PM and WF) peak phase back-EMF for the
9-phase, 63-turn HPM machine and the DC-link voltage variation while delivering
a constant power of 3 kW with respect to wound field excitation current. As can be
noted from Fig. 6.18, the back-EMF bucking mode represents the dominant
190 6 Operation and Characterization of Multiphase HPM Generator in SHEV Powertrain

Fig. 6.17 HPM generator WF excitation current variation for repetitive NEDC driving cycles

operating mode for this choice of HPM machine winding, power demand, and
operating speed. Data for the points identified in Fig. 6.18 are given in Table 6.5
for completeness. Again, because of the fixed power output requirement, the field
excitation current options are constrained. For this specific machine and the
6.3 HPM Machine Output Power Control 191

Fig. 6.18 DC-link voltage and peak phase back-EMF with respect to the WF excitation current for
63-turn case and a constant DC-link output power (3 kW)

Table 6.5 Operational point data of the 63-turn HPM machine

Operational VDC-link Vterm (peak) Back-EMF (peak) Is (peak)
points If (A) (V) (V) (V) (A)
(a) 0.753 589 309.5 333.3 2.14
(b) 0 530 282.9 314.5 2.33
(c) 1.27 360 202.9 282.8 3.34

considered driving cycles, the HPM machine peak phase back-EMF can only be
regulated over a field current range from 1.27 to 0.753 A.
Referring to Figs. 6.18 and 6.19 illustrates phase terminal voltage, back-EMFs,
and current as the DC-link voltage varies from 589 VDC, Fig. 6.19a, to 360 VDC,
Fig. 6.19c, and the HPM machine field excitation is varied from a maximum positive
current, to a maximum negative current of 1.27 A, and then again to a maximum
positive current of 0.753 A, to maintain the fixed generated power demand of 3 kW.
Figure 6.20 illustrates the same waveforms for points (a), (b), and (c), but with
magnified axes to show the phase current detail and respective phase relationships
more clearly. Note that, as before, the simulation results of Fig. 6.19 are not in steady
state until  0.4 ms, hence the asymmetry of the waveforms during this time period.
The simulation settles at around 0.4 ms, as inferred from the resulting waveforms.

6.3.5 Thermal Analysis Results of the Investigated HPM

Machine

As a final check on the machine design solution, the general thermal analysis of
Chap. 4 is now resolved for the stator and rotor losses specified to the vehicle
application considered and the 9-phase, 63-turn HPM machine. The method utilized
in Chap. 4 to predict dynamic and steady-state temperature distributions is used
192 6 Operation and Characterization of Multiphase HPM Generator in SHEV Powertrain

Fig. 6.19 Phase voltage

and current waveforms for
eight DC-link voltage levels
for the 63-turn, 9-phase
HPM machine at constant
DC-link output power
(3 kW) (a) VDC-link ¼ 589 V,
If ¼ 0.753 A (b) VDC-
link ¼ 530 V, If ¼ 0 A (c)
VDC-link ¼ 360 V,
If ¼ 1.27 A

again to predict the machine steady-state temperature based on the NEDC driving
cycle loads. The steady-state thermal analysis results are detailed in Table 6.6 and
illustrated in Figs. 6.21 and 6.22, showing much lower temperature distributions.
Hence, the thermal analysis results are considered to be satisfactory and acceptable.
6.3 HPM Machine Output Power Control 193

Fig. 6.20 Magnified views

of Fig. 6.19 that illustrates
waveform phase
relationships (a) VDC-
link ¼ 589 V, If ¼ 0.753 A
(b) VDC-link ¼ 530 V, If ¼ 0
A (c) VDC-link ¼ 360 V,
If ¼ 1.27 A
194 6 Operation and Characterization of Multiphase HPM Generator in SHEV Powertrain

Table 6.6 HPM machine thermal modal results

PM section WF section
Node name Temperatures ( C) Temperature ( C)
Ambient 25.0 25.0
Housing 30.1 28.4
Rotor lamination pole – 81.3
Rotor magnet 41.4 –
Rotor lamination yoke 40.7 76.8
Rotor surface 41.7 80.1
Rotor copper – 79.5
Stator surface 55.0 52.8
Stator yoke 49.3 42.1
Stator tooth 45.3 51.2
Stator tooth tip 55.1 52.6
Stator winding average 55.3 55.7

6.4 HPM Machine Characterization Using Brushless

Excitor

Excitation systems, which are generally referred to as schemes that provide DC

voltage and current to the wound field of electric generators, are either rotating or
static [228, 229], where the rotating schemes can be brushed or brushless [230]. Due
to issues of brush maintenance, commutators, and slip rings, the brushless exciters
are more attractive and offer improved reliability and better performance
[231, 232]. Brushless excitation schemes can be implemented using the concept of
contactless energy transfer as rotating transformers [233–235]. In these schemes,
there usually exist two sets of windings, stationary and rotating, and the power from
the stationary winding is induced to the rotating section. Both single- and three-
phase rotary transformers have been proposed [235–237]. Here, the brushless DC
excitation schemes are used to provide the field current and voltage to the wound
field of hybrid excitation machines where applicable [33, 238, 239]. In [239],
two-stage brushless excitation scheme for a hybrid excitation synchronous generator
is proposed where the input power of the exciter is supplied from the output of the
main generator. Dynamics and control of the exciter contribute to the system
stability and operation and, as such, there have been studies focused on this area
[240, 241]. This section discusses the characterization of the HPM generator in
SHEVs for low DC-link voltage (85–125 V) using a brushless exciter scheme. This
is achieved by (i) integration of HPM generator and brushless exciter in one machine
housing as shown in Fig. 6.23, (ii) proposing a control strategy for the integrated
HPM generator–exciter system in a SHEV, (iii) materializing the conceptual study in
[241] by proposing a practical design for brushless exciter suitable for integrated
system, (iv) characterizing the system using the previously discussed control strat-
egy while the SHEV operates under NEDC, and (v) providing an analysis of the
6.4 HPM Machine Characterization Using Brushless Excitor 195

Fig. 6.21 Steady-state

temperature distribution for
PM section of 63-turn HPM
generator (a) Radial cross
section (b) Axial cross
section
196 6 Operation and Characterization of Multiphase HPM Generator in SHEV Powertrain

Fig. 6.22 Steady-state

temperature distribution for
WF section of 63-turn HPM
generator (a) Radial cross
section (b) Axial cross
section
6.4 HPM Machine Characterization Using Brushless Excitor 197

Fig. 6.23 Sketch of the HPM generator with the brushless excitation scheme (a) Exciter and HPM
generator in one machine housing (b) Cross-sectional view of the HPM generator and brushless
exciter rotor/stator laminations

integrated system by presenting operating performance curves, operational enve-

lope, and efficiency mapping under prescribed operating conditions.

6.4.1 32-Phase Brushless Excitation Scheme

The brushless exciter rotor has 32 poles where the stator is constrained to lamina-
tions with 36 slots, as shown in Fig 6.23, which have to accommodate a DC winding
supplied from the rectified output of the HPM generator via a buck-boost converter.
The saturation of steel core constraints the air-gap flux density. Therefore, when
considering the magnetic design of a brushless exciter with a given lamination, the
198 6 Operation and Characterization of Multiphase HPM Generator in SHEV Powertrain

saturation of steel core constraints the air-gap flux density. As the number of
magnetic poles reduces, the flux density in the back-iron increases bringing the
operating point closer to saturation. In addition, for the brushless exciter, the number
of poles is ultimately limited by the number of stator teeth since the wound field rotor
acts as secondary winding. Therefore, to study a different number of poles (mini-
mum 2 and maximum 36) and in order to avoid steel saturation, two different pole
combinations that is, 18- and 36-poles for the brushless exciter are chosen. To form
18-pole configurations, two different winding arrangements for the stator lamina-
tions can be considered: (i) Design 1: 18 concentric coils wound around every other
stator tooth (single layer) and (ii) Design 2: distributed winding with every other
tooth having two coil sides (double layer). However, for the 36-pole configurations
(Design 3), there are 36 coils wound around each stator tooth, forming a double layer
concentrated winding.
The exciter air-gap flux density and stator slot current density with respect to
stator excitation current are presented in Fig. 6.24a. Therefore, the excitation current
is chosen such that the air-gap flux density approaches saturation, equating to an
excitation current of 300 Ampere-turns (A-t) and corresponding to a peak air-gap
flux density of 0.89 T, 0.74 T, and 0.94 T and a stator slot current density of 2.49
(A/mm2), 4.97 (A/mm2), and 4.97 (A/mm2) for Design 1, Design 2, and Design
3, respectively. Here, Design 3, that is, 36-pole with 36 concentrated stator coils
connected in series is selected as the brushless exciter stator configuration. Note that
the HPM generator is a 32-pole machine, and this is not to be confused with the
36-pole brushless exciter. Figure 6.24b illustrates the rotor coil flux-linkage and
induced back-EMF per turn per unit axial length versus electrical angle of rotation
for a stator excitation current of 300 A-t and a rated speed of 3000 RPM. Note that
the more trapezoidal voltage waveform in Fig. 6.24b is beneficial as it results in
lower rectified voltage ripple at the output of exciter rotating rectifier, hence reduced
filtering requirements.
The brushless exciter rotor is constrained to laminations with 32 slots, the same as
HPM WF, into which an AC winding has to be accommodated. A multiphase design
approach for the brushless exciter rotor winding is considered for high power density
and a high-quality DC at the output of rotating rectifier and to minimize/eliminate
passive capacitances and improved reliability. Assuming there are m number of rotor
phases, the number of slots per pole per phase, that is, phase belt, to accommodate
36 poles would be 32/(36 m). No matter how many phases are chosen, the phase belt
ratio is always less than unity, resulting in 32 concentrated coils wound on each rotor
tooth. However, due to the symmetry of the magnetic design, teeth that are
180 degrees mechanically displaced from each other experience the same flux
density in the same direction. Given the same number of coil turns, back-EMFs of
diametrically opposite rotor coils have the same magnitude and phase shift. There-
fore, for 32 rotor coils, a maximum of 16 phases can be achieved.
To form 16 phases, a series or parallel connection of rotor coils can be realized.
For the series connection, as illustrated in Fig. 6.25a, two coils are series wound with
one connection taken to the main star point and the other to a 16-leg rotating rectifier.
In the parallel connection shown in Fig. 6.25b, one end of each coil is connected to
6.4 HPM Machine Characterization Using Brushless Excitor 199

Fig. 6.24 Brushless exciter characteristics (a) Air-gap flux density and stator slot current density
study (b) Per-unit rotor flux linkage and EMF at 3000 RPM

the main star point and the other ends are taken to a 32-leg rotating rectifier. Note that
although there are 32 coils and a 32-leg rotating rectifier, the winding is still a
16-phase winding arrangement. The benefit of the 16-phase parallel connection with
a 32-leg rotating rectifier is minimized interconnections, which improves the
manufacturing process, an important feature for mass production. Thus, in the
200 6 Operation and Characterization of Multiphase HPM Generator in SHEV Powertrain

Fig. 6.25 Two brushless excitor rotor winding designs (a) 16-phase series configuration (b)
16-phase parallel configuration

Table 6.7 HPM generator 32-phase brushless exciter

brushless exciter
Stator outer diameter (mm) 230.4
specifications
Rotor outer diameter (mm) 122
Air-gap thickness (mm) 0.4
Axial length (mm) 4
No. of stator slots 36
No. of rotor slots 32
No. of poles 36
No. of turns per rotor coil (p.u.) 73
No. of turns per stator coil (p.u.) 50
No. of exciter rotor phases 16
Conductor diameter (mm) 0.56
Conductor area (mm2) 0.25
Copper area (mm2) 35.62

upcoming analysis, the 16-phase parallel option is the adopted brushless exciter rotor
winding design, and this specification is illustrated in Table 6.7.

6.4.2 Performance Curves

The multiphase HPM generator system is designed to deliver a rated power of 3 kW

at 3000 RPM, a detailed design of which is given in Chap. 4. The nominal voltage of
the vehicle battery system interfaced to the DC-link, Fig. 5.19, in this analysis
becomes 120 V at the open circuit but varies around 30% from its rated voltage,
that is, from 85 to 125 V during load transients throughout the driving cycle as
shown in Fig. 6.26a. The HPM generator is designed to interface this varying
DC-link voltage by modulation of the machine WF element. As discussed, the
6.4 HPM Machine Characterization Using Brushless Excitor 201

Fig. 6.26 Control strategy of the multiphase HPM generator via brushless exciter (a) Vehicle
DC-link voltage variation (b) Control block diagram schematic

HPM generator WF element facilitates the control of the total generated back-EMF
and hence machine rectified output DC voltage. The HPM generator is required to
maintain a constant power input to the vehicle DC-link during the entire driving
cycle. Indeed, the control objective is to maintain the ICE/HPM generator system
output power constant at a given value while minimizing speed variation via
regulation of the HPM wound field excitation. Thus, a DC-link voltage increase
beyond the nominal value requires a WF current increase at a given speed, a scenario
of boosting capability. Similarly, in a buck functionality, the WF current is decreased
to reduce the machine output voltage. A general schematic diagram of the HPM
generator system along with its feedback control is illustrated in Fig. 6.26b. The
brushless exciter stator is fed from the output of the HPM generator via a DC/DC
converter. Here, a WF excitation current linear function that uses interpolation-
exportation is utilized as a lookup method to produce a duty cycle (D) for the
brushless exciter DC/DC converter. By controlling the duty cycle, the rectified
202 6 Operation and Characterization of Multiphase HPM Generator in SHEV Powertrain

Fig. 6.27 Brushless HPM generator performance curves at 3000 RPM for different output power

voltage at the output of the exciter rotating rectifier is adjusted to provide the desired
WF current.
However, in order to characterize the system under a prescribed control strategy,
the HPM machine performance curves are deduced for a rated power range of
0.5–4.2 kW with an extended DC-link voltage variation, that is, beyond the actual
voltage of 85–125 V. The HPM generator is capable of delivering 3.2 kW contin-
uously; however, the 4.2 kW rated power is chosen as the upper limit due to the
limitation on machine current density and thermal performance under a short period
of time. Figure 6.27 presents the brushless HPM generator performance curves for
different output power, DC-link voltage, and WF excitation current. Utilizing the
previously described control strategy while maintaining a maximum stator current
density (S.C.D.), a minimum DC-link voltage is derived for each power curve at a
fixed speed of 3000 RPM. The dotted line in Fig. 6.27 denotes the performance
curve limit due to the maximum stator current density. For an excitation WF current
limit of 4 A, chosen to limit the teeth and back-iron saturation and the WF rotor slot
current density, the 0.5-kW and 1-kW performance curves present a linear behavior
as the DC-link voltage varies. However, for powers greater than 1 kW, the curves
follow a nonlinear cubic function behavior. This is because at some WF excitation
currents, there exist two DC-link voltages that result in the same output power. For
instance, for an excitation current of 1 A, both 52 V and 104 V DC-link voltages
result in the same rated power of 2.5 kW. As observed from Fig. 6.27, for rated
output powers greater than 2.5 kW, the WF excitation current is always positive,
hence simplifying requirements for the DC/DC power electronics converter,
Fig. 6.25b, that otherwise would be necessary to reverse the WF excitation current.
Furthermore, as the power increases, the minimum DC-link voltage limit is
increased while the maximum DC-link voltage limit is reduced.
For instance, the minimum DC-link voltage for a rated power of 2.5 kW is 52 V
while this value for the 3 kW is 60 V. This is due to the effect of the machine
synchronous inductance and hence power factor effect that is worsened as the power
6.4 HPM Machine Characterization Using Brushless Excitor 203

Table 6.8 Comparison of HPM machine system characteristics at 3000 RPM

Item DC-link voltage HPM WF back-EMF
Percentage (%) Percentage (%)
Region P (kW) If (A) Base (V) Buck () Boost (+) Buck () Boost (+)
Underrated 0.5 4 124 37 38 33.3 33.3
1 4 114 44 46 33.3 33.3
1.5 2.7 106 71 49 22 33.3
2 1.3 97.2 58 52 11 33.3
NCR 2.5 0 72.3 29 98 0 33.3
3 1.1 79.4 23 73 0 33.3
3.2 1.6 82 19 63 0 33.3
PCR 3.5 2.15 87 19 51 0 33.3
4 3.15 91 9 34 0 33.3
4.2 3.5 95 11 24 0 33.3
Note: Maximum excitation field current is 4 A.

Fig. 6.28 HPM WF excitation current and DC-link voltage at 3000 RPM for a range of HPM
generator output power

is increased. Note that these curves are not to be confused with the V-curves of the
conventional wound field synchronous machines. Table 6.8 lists the brushless HPM
generator buck and boost capability at different WF current and machine output
power. For each power level, a base DC-link voltage is considered where the WF
current is zero. However, if the performance curve lies only in the first quadrant, the
base DC-link voltage is defined at minimum WF current, that is, the cube base of the
curve. It is seen from Table 6.8 that in all cases except for 4.2 kW, the range of
DC-link voltage variations, that is, buck-boost capability, is larger than the back-
EMF variations. Therefore, for a rated power of 3 kW, the HPM generator’s
maximum nine-phase back-EMF boost capability is 33.3% while the DC-link
voltage can be 23% and 73% bucked and boosted, respectively.
204 6 Operation and Characterization of Multiphase HPM Generator in SHEV Powertrain

Figure 6.28 shows base, minimum, and maximum DC-link voltage for a range of
power, that is, 0.5–4.2 kW at 3000 RPM along with the required WF excitation
current for each power. Three regions are recognized: (i) nominal continuous power
rating (NCR) 2.5–3.2 kW, (ii) peak continuous power rating (PCR) 3.2–4.2 kW, and
(iii) underrated power region 0.5–2.5 kW. As it is observed from Fig. 6.28, the
highest DC-link boost capability of the HPM generator occurs in the NCR region.
Indeed, for the 2.5 kW, 3 kW, and 3.2 kW, the DC-link boost is 98%, 73%, and 63%,
respectively, as seen from Table 6.8. Therefore, if the HPM generator output power
is maintained in the NCR region, the electric vehicle DC-link voltage is allowed to
sustain the highest variation. In the NCR region, the field current is always positive,
hence simplifying the requirements for the brushless exciter DC/DC converter,
Fig. 6.26b. Hence, the 3-kW demand average power lies in the NCR region, a
preferred control region due to the highest boost capability for the HPM machine.
Therefore, the brushless HPM generator system is controlled to deliver 3 kW at a
rated speed of 3000 RPM.

6.4.3 Efficiency Mapping

In this section, efficiency maps of the HPM generator system is investigated for a
1.0-ton SHEV with an average power demand of 3 kW under NEDC driving cycle
and for a 85–25 V DC-link voltage variation. The operational envelopes are
constrained by the WF excitation current, prime mover speed, and average
DC-link power. Here, the HPM generator iron loss is predicted using the curve
fitting method, as detailed in Chap. 3 (Sect. 3.5.3), which is expanded to include
higher speed losses. The manufacture’s iron loss curves combine hysteresis, eddy
current, and excess losses and are given in watts per kilogram, that is, loss density for
different fundamental flux densities. Figure 6.29 compares the predicted and

Fig. 6.29 Predicted and measured iron loss versus flux density at different speeds
6.4 HPM Machine Characterization Using Brushless Excitor 205

Fig. 6.30 HPM generator torque-speed and efficiency mapping. Note that the efficiencies include
the passive rectification stage losses (a) Efficiency map for different HPM generator speed at 3 kW
(b) Torque-speed characteristics

measured iron loss density for a range of speeds (0.5–7.0 kRPM). The comparison
shows a good agreement, hence the accuracy of the loss analysis.
Figure 6.30a illustrates HPM generator efficiency (including the passive rectifi-
cation stage) for different delivered power. As seen, higher positive WF currents
result in improved system efficiency making the boosting functionality of the HPM
generator system a favorite control strategy. Figure 6.30b presents the HPM gener-
ator torque-speed characteristics along with efficiencies at different machine speeds.
As seen from Fig. 6.30b, for the speeds between 2000 and 3500 RPM, the HPM
generator and rectification system results in the highest efficiency with an output
power of 3 kW while satisfying WF current constraints. As previously discussed, the
206 6 Operation and Characterization of Multiphase HPM Generator in SHEV Powertrain

ICE speed is maintained within a tight envelope that results in optimum engine
efficiency. The HPM generator system operating speed region is therefore within the
ICE speed envelope.

6.5 Conclusion

This chapter discusses the operation of the HPM machine in a series hybrid EV
powertrain, where a dynamic has been developed to include the DC-link voltage
variation during dynamic driving and the WF back-EMF regulation function via a
flux-linkage feedback loop. The HPM machine back-EMF control strategy to supply
a constant DC-link power demand has been discussed, along with a justification of
the desired wound field excitation current regulation function for different machine
stator winding turns. The analysis has highlighted the importance of choosing the
most suitable number of stator winding turns for a given variation of the vehicle
system terminal voltage. Moreover, five operating scenarios for the HPM machine in
a series hybrid EV powertrain have been investigated. The five analyzed operating
scenarios gave a decision regarding the most appropriate operating scenario that can
be adopted for minimal HPM machine with regard to total energy and rotor copper
loss (scenario 3) or system failure modes (scenario 1).
For low DC-link voltage (80–125 V), characterization of the multiphase HPM
generator system is also addressed utilizing a 32-phase brushless excitation scheme.
Different cases, such as normal, boost, and buck functionality of HPM machine
operation, are analyzed, and a choice of the most appropriate operation mode has
been selected to regulate the total back-EMF via a WF excitation current control.
Several system power performance curves have been generated, and the most
suitable continuous machine power rating range is concluded. The selected NCR
power region supports the proposed machine safe continuous operation, which is
70% of the PCR power region with 74% efficiency. The dynamic model simulation
study illustrates the buck/boost percentage regulation capability of the HPM
machine back-EMF and vehicle DC-link based on the DC excitation field voltage
and current limits.
References

1. S.J. Chapman, Electric Machinery Fundamentals, McGraw-Hill Series in Electrical Engineer-

ing Power and Energy, 2nd edn. (1991)
2. T. Gönen, Electrical Machines (Power International Press, 1998)
3. [Online], Magnets Assemblies, Available: https://www.goudsmitmagnets.com
4. A.E. Fitzgerald, C. Kingsley Jr., S.D. Umans, Electric Machinary, McGraw-Hill Series in
Electrical Engineering Power and Energy, 5th edn. (1990)
5. P.C. Sen, Principles of Electric Machines and Power Electronics, 2nd edn. (Wiley, 1996)
6. M.T. York, High Power Permanent Magnet Hybrid Alternator Rotor (U.S. Patent 6744165,
June 1/2004)
7. A.D. Akemakou, Hybrid Alternator with Axial End Retainers for Permanent Magnets
(U.S. Patent 6784586, Aug. 31/2004)
8. C.D. Syverson, Hybrid Alternator (U.S. Patent 5397975, Mar. 14/1995)
9. U. Schüller, J. Brandes, Hybrid-Energized Synchronous Electric Machines (U.S. Patent
5731647, Mar 24/1998)
10. M.S. Shah, G.B. Kliman, M.V.K. Chari, Hybrid Salient Pole Machine (U.S. Patent 6784585,
Aug. 31/2004)
11. M.R. Shah, G.B. Kliman, Hybrid Synchronous Machines Comprising Permanent Magnates
and Excitation Winding in Cylindrical Elements Slots (U.S. Patent 6509664, Jan. 21/2003)
12. A. Detela, Hybrid Synchronous Machine with Transverse Magnetic Flux (U.S. Patent
5712521, Jan. 27/1998)
13. H. Arita, M. Inoue, Y. Asoa, T. Yamamoto, Hybrid-Excitation Rotating Machine, and Vehicle
with the Hybrid-Excitation Rotating Machine (U.S. Patent 2007/0090713, Apr. 26/2007)
14. K. Nagayama, T. Mizuno, T. Ashikaga, Method and Apparatus for Controlling Hybrid
Excitation Type Permanent Magnet Synchronous Motor (U.S. Patent 5689166, Nov. 18/1997)
15. J. Lucidarme, M. Gabsi, M. Lecrivain, H. Ben Ahmed, Rotating Electric Machine with
Improved Double Excitation (U.S. Patent 6462449, Oct. 8/2002)
16. Z. Chao-hui, Y. Yang-guang, A review of development of hybrid excitation synchronous
machine. Proceeding of the IEEE International Symposium on Industrial Electronics (ISIE
2005) 2, 857–862 (2005)
17. Y. Dou, H. Chen, A design research for hybrid excitation rare earth permanent magnet
synchronous generator. Proceedings of the Fifth International Conference on Electrical
Machines and Systems (ICEMS 2001) 2, 856–859 (2001)
18. Z. Chao-hui, G. Huanqiu, W. Xin-wei, The study of structure and characteristic of a new type
HECPSG. IEEE International Symposium on Industrial Electronics (ISIE 2007), 1097–1100
(2007)

© Springer Nature Switzerland AG 2022 207

A. S. Al-Adsani, O. Beik, Multiphase Hybrid Electric Machines,
https://doi.org/10.1007/978-3-030-80435-0
208 References

19. E. Spooner, S.A.W. Khatab, N.G. Nicolaou, Hybrid Excitation of AC and DC Machines
(Fourth International on Electrical Machines and Drives Conference (EMDC’89), 1989),
pp. 48–52
20. A. Novinschi, N.L. Brown, E. Spooner, A. Mebarki, L. Haydock, Finite Element Analysis of a
Novel Axial Flux PM Synchronous Machine with Excitation Control, vol 1 (Second Interna-
tional Conference on Power Electronics, Machines and Drives, (PEMD), 2004), pp. 420–425.,
31 March–2 April 2004
21. J.F. Eastham, P.D. Evans, P.C. Coles, M. Ibrahim, Double disc alternators with hybrid
excitation. IEEE Trans. Magn. 28, 3039–3041 (1992)
22. N. Naoe, T. Fukami, Trial production of a hybrid excitation type synchronous machine. IEEE
International on Electric Machines and Drives Conference (IEMDC), 545–547 (2001)
23. L. Xiaogang, T.A. Lipo, A synchronous/permanent magnet hybrid AC machine. IEEE Trans-
action on Energy Conversion 15, 203–210 (2000)
24. J.A. Tapia, F. Leonardi, T.A. Lipo, Consequent Pole Permanent Magnet Machine with Field
Weakening Capability (IEEE International Conference on Electric Machines and Drives
(IEMDC), 2001), pp. 126–131
25. J.A. Tapia, F. Leonardi, T.A. Lipo, Consequent-pole permanent-magnet machine with
extended field-weakening capability. IEEE Trans. Ind. Appl. 39, 1704–1709 (2003)
26. M. Aydin, H. Surong, T.A. Lipo, A new axial flux surface mounted permanent magnet
machine capable of field control. The 37th IAS Annual Conference on Industry Applications
2, 1250–1257 (2002)
27. T.A. Lipo, M. Aydin, Filed Weakening of Permanent Magnet Machines: Design Approaches
(EPE Power Electronics and Motion Control (EPE-PEMC’04)0, Riga, Latvia, Sept. 2004)
28. M. Aydin, J. Yao, E. Kayikci, S. Huang, T.A. Lipo, Design Considerations and Experimental
Results of an Axial Flux PM Motor with Field Control (16th International Conference on
Electrical Machines (ICEM’04), Poland, Sept. 2004, 2004)
29. M. Aydin, S. Huang, T.A. Lipo, Performance Evaluation of an Axial Flux Consequent Pole
PM Motor Using Finite Element Analysis, vol 3 (IEEE International Conference on Electric
Machines and Drives (IEMDC’03), Madison, WI, USA, 2003), pp. 1682–1687., 1–4 June
2003
30. Y. Amara, J. Lucidarme, M. Gabsi, M. Lecrivain, A.H.B. Almed, A.D. Akemakou, A new
topology of hybrid synchronous machine. IEEE Trans. Ind. Appl. 1, 1273–1281 (2001)
31. L. Vido, Y. Amara, M. Gabsi, M. Lecrivain, F. Chabot, Compared performances of homopolar
and bipolar hybrid excitation synchronous machines. Fourtieth IAS Annual Meeting and
Industry Applications Conference 3, 1555–1560 (2005)
32. L. Vido, M. Gabsi, M. Lecrivain, Y.A.A.Y. Amara, F.A.C.F. Chabot, Homopolar and bipolar
hybrid excitation synchronous machines Homopolar and bipolar hybrid excitation synchro-
nous machines. IEEE International Conference on Electric Machines and Drives (IEMD’05),
1212–1218 (2005)
33. D. Fodorean, A. Djerdir, I.A. Viorel, A. Miraoui, A double excited synchronous machine for
direct drive application – Design and prototype tests. IEEE Transaction on Energy Conversion
22, 656–665 (2007)
34. D. Fodorean, I.A. Viorel, A. Djerdir, A. Miraoui, Double-excited synchronous motor with
wide speed range: Numerical and experimental results. Iranian Journal of Electric and Com-
puter Engineering 5 (2006, Winter-Spring)
35. I.A. Viorel, R. Munteanu, D. Fodorean, L. Szabo, On the possibility to use a hybrid synchro-
nous machine as an integrated starter-generator. IEEE International Conference on Industrial
Technology (ICIT), 1195–1200 (2006)
36. E. Afjei, H. Toliyat, H. Moradi, A Novel Hybrid Brushless Dc Motor/Generator for Hybrid
Vehicles Applications (International Conference on Power Electronics, Drives and Energy
Systems (PEDES’06), 2006, 12–15 Dec), pp. 1–6
37. E. Afjei, O. Hashemipour, H. Toliyat, A new hybrid reluctance motor/field-assisted generator.
IEEE Electric Machines and Drives Conference (IEMDC’07) 1, 543–547 (2007, 3–5 May)
References 209

38. K. Chau, Y. Li, J. Jiang, S. Niu, Design and control of a pm brushless hybrid generator for
wind power application. IEEE International Magnetics Conference (INTERMAG), 318–318
(2006, 8–12 May)
39. X. Liu, H. Lin, Z.Q. Zhu, C. Yang, S. Fang, J. Guo, A novel dual-stator hybrid excited
synchronous wind generator. IEEE Trans. Ind. Appl. 45(3), 947–953 (2009, May–June)
40. A.S. Al-Adsani, Hybrid Permanent Magnet Machines for Electric Vehicles (Ph.D. thesis,
University of Manchester, 2011, September)
41. N. Schofield, S.A. Long, D. Howe, M. McClelland, Design of a Switched Reluctance Machine
for extended speed operation. IEEE Trans. Ind. Appl. 45(1), 116–122 (2009). https://doi.org/
10.1109/TIA.2008.2009506
42. N. Schofield, S.A. Long, Generator operation of a switched reluctance starter/generator at
extended speeds. IEEE Trans. Veh. Technol. 58(1), 48–56 (2009). https://doi.org/10.1109/
TVT.2008.924981
43. EU Framework V RTD Project, New High Efficiency Electric Machines Solutions For Mild
Hybrid Applications (ELMAS) (Contract No: ENK6-CT1999-00017, 2001-04)
44. Website: https://www.wardsauto.com/news-analysis/rotary-engine-finally-could-find-home-
future-hybrids, as visited on April 2020
45. A.S. Al-Adsani, N. Schofield, Hybrid permanent magnet generators for electric vehicle
applications. IEEE International Electric Machines and Drives Conference, 1754–1761. ,
Miami, FL (2009)
46. A.S. Aladsani, O. Beik, Design of a Multiphase Hybrid Permanent Magnet Generator for
series hybrid EV. IEEE Transactions on Energy Conversion 33(3), 1499–1507 (Sept. 2018)
47. Website: http://femm.foster-miller.net/wiki/HomePage, as visited on January 2007
48. D. Meeker, Finite Element Method Magnetics (Version 4.0, User’s Manual, 2006, January)
49. Z. Kolondzovski, Evaluation of Methods for Calculation Synchronous Generator Reactances
(International PhD-seminar on Computation of Electromagnetic Fields, Budva, Serbia and
Montenegro, 2004)
50. G.R. Slemon, Electric Machines and Drives (Book, 1992)
51. J.J. Cathey, Electric Machines: Analysis and Design Applying Matlab(R) (Book, 2001)
52. F. Meier, J. Soulard, PMSM with Non-Overlapping Concentrated Windings: Design Guide-
lines and Model References (Proceeding (EVER’09), Monaco, 2009, March)
53. Z.Q. Zhu, D. Howe, Influence of design parameters on cogging torque in permanent magnet
machines. IEEE Transactions on Energy Conversion 15(4), 407–412 (2000)
54. C.-M. Ong, Dynamic Simulation of Electric Machinery (New Jersey, US, 1998). ISBN
0-13-723785-5
55. M.-F. Hsieh, Y.-C. Hsu, D.G. Dorrell, K.-H. Hu, Investigation on end winding inductance in
motor Stator windings. IEEE Trans. Magn. 43(6), 2513–2515 (2007)
56. Website: https://www.albmagnets.com, as visited on August 2020
57. D. Ishak, Z.Q. Zhu, D. Howe, Eddy-current loss in the rotor magnets of permanent-magnet
brushless machines having a fractional number of slots per pole. IEEE Trans. Magn. 41(9),
2462–2469 (2005)
58. A. Boglietti, A. Cavagnino, M. Lazzari, M. Pastorelli, Two simplified methods for the iron
losses prediction in soft magnetic materials supplied by PWM inverter. IEEE International on
Electric Machines and Drives Conference (IEMDC), 391–395 (2001)
59. L. Li, X. Huang, B. Kao, B. Yan, Research of core loss of permanent magnet synchronous
motor (PMSM) in AC servo system. International Conference on Electrical Machines and
Systems (ICEMS), 602–607 (2008)
60. G. Bertotti, A. Boglietti, M. Chiampi, D. Chiarabaglio, F. Fiorillo, M. Lazzari, An improved
estimation of iron losses in rotating electrical machines. IEEE Trans. Magn. 27(6), 5007–5009
(1991)
61. G.R. Slemon, X. Liu, Core losses in permanent magnet motors. IEEE Trans. Magn. 26(5),
1653–1655 (1990)
210 References

62. Lamination Steel Third Edition CD-ROM, Electric Motor Education and Research Founda-
tion, ISBN 0–9714391–3-3, Second Printing 2009
63. P.H. Mellor, D. Roberts, D.R. Turner, Lumped parameter thermal model for electrical
machines of TEFC design. IEE Proceedings on Electric Power Applications 138, 205–218
(1991)
64. Z.J. Liu, D. Howe, P.H. Mellor, M.K. Jenkins, Thermal analysis of permanent magnet
machines. Sixth International Conference on Electrical Machines and Drives (IEMDC) (1993)
65. R. Ibtiouen, S. Mezani, O. Touhami, N. Nouali, M. Benhaddadi, Application of lumped
parameters and finite element methods to the thermal modeling of an induction motor. IEEE
International on Electric Machines and Drives (IEMDC) (2001)
66. D. Staton, A. Boglietti, A. Cavagnino, Solving the more difficult aspects of electric motor
thermal analysis in small and medium size industrial induction motors. IEEE Transaction on
Energy Conversion 20, 620–628 (2005)
67. A. Boglietti, A. Cavagnino, D.A. Staton, TEFC induction motors thermal models: A parameter
sensitivity analysis. IEEE Trans. Ind. Appl. 41, 756–763 (2005)
68. E. Belicova, V. Hrabovcova, Analysis of an axial flux permanent magnet machine (AFPM)
based on coupling of tow separated simulation models (electrical and thermal ones). J. Electr.
Eng. 58(1) (2007)
69. M.M. Baggu, H.L. Hess, K. Rink, Thermal modelling of direct lamination cooling (DLC)
induction motor for hybrid electric vehicle applications. IEEE Conference on Vehicle Power
and Propulsion (VPPC’05) (2005)
70. A.M. El-Refaie, N.C. Harris, T.M. Jahns, K.M. Rahman, Thermal analysis of multibarrier
interior PM synchronous machine using lumped parameter model. IEEE Transaction on
Energy Conversion 19, 303–309 (2004)
71. F. Sahin, Design and Development of a High-Speed Axail-Flux Permanent-Magnet Machine
(PhD Thesis, Eindhoven Technische Universiteit Eindhoven, ISBN 90-386-1380-1, 2001)
72. D.A. Staton, A. Cavagnino, Convection Heat Transfer and Flow Calculations Suitable for
Analytical Modelling of Electric Machines (The 32nd IEEE Annual Conference on Industrial
Electronics (IECON 2006), 2006)
73. J.P. Holman, Heat Transfer, 7th edn. (1990)
74. D. A. Staton, Thermal Analysis of Electrical Motors and Generators. www.motor-design.com,
2008, March
75. D.C. White, H.H. Woodson, Electromechanical Energy Conversion, 1st edn. (Wiley,
New York, 1959)
76. E.A. Klingshirn, High phase order induction motors – Part II-experimental results. IEEE
Transactions on Power Apparatus and Systems PAS-102(1), 54–59 (1983)
77. G.K. Singh, K. Nam, S.K. Lim, A simple indirect field-oriented control scheme for multiphase
induction machine. IEEE Trans. Ind. Electron. 52(4), 1177–1184 (2005)
78. E. Levi, Multiphase Electric Machines for Variable-Speed Applications. IEEE Trans. Ind.
Electron. 55(5), 1893–1909 (2008)
79. R. Bojoi, A. Cavagnino, A. Tenconi, A. Tessarolo, S. Vaschetto, Multiphase electrical
machines and drives in the transportation electrification, in 2015 IEEE 1st International
Forum on Research and Technologies for Society and Industry Leveraging a Better Tomorrow
(RTSI), Turin, (2015), pp. 205–212
80. R. Bojoi, S. Rubino, A. Tenconi, S. Vaschetto, Multiphase electrical machines and drives: A
viable solution for energy generation and transportation electrification, in 2016 International
Conference and Exposition on Electrical and Power Engineering (EPE), Iasi, (2016),
pp. 632–639
81. O. Lopez, J. Alvarez, J. Doval-Gandoy, F.D. Freijedo, Multilevel multiphase space vector
PWM algorithm. IEEE Trans. Ind. Electron. 55(5), 1933–1942 (2008)
82. S. Williamson, S. Smith, Pulsating torque and losses in multiphase induction machines. IEEE
Trans. Ind. Appl. 39, 986–993 (2003)
References 211

83. S. Brisset, D. Vizireanu, P. Brochet, Design and optimization of a nine-phase axial-flux PM

synchronous generator with concentrated winding for direct-drive wind turbine. IEEE Trans.
Ind. Appl. 44(3), 707–715 (2008)
84. E. Levi, M. Jones, S.N. Vukosavic, H.A. Toliyat, Operating principles of a novel multiphase
multimotor vector-controlled drive. IEEE Transactions on Energy Conversion 19(3), 508–517
(2004)
85. A.S. Thomas, Z.Q. Zhu, R.L. Owen, G.W. Jewell, D. Howe, Multiphase flux-switching
permanent-magnet brushless machine for aerospace application. IEEE Transaction on Industry
Applications 45(6) (2009)
86. T. Mazzucca, M. Torre, The FREMM architecture: A first step towards innovation. IEEE
SPEEDAM, 574–579 (2008)
87. F. Scuiller, J. Charpentier, E. Semail, Multi-star multi-phase winding for a high power naval
propulsion machine with low ripple torques and high fault tolerant ability. IEEE-VPPC, 1–5
(2010)
88. G. Sulligoi, A. Tessarolo, V. Benucci, A.M. Trapani, M. Baret, F. Luise, Shipboard power
generation: Design and development of a medium-voltage dc generation system. IEEE Ind.
Appl. Mag. 19(4), 47–55 (2013)
89. A.S. Nanoty, A.R. Chudasama, Design of multiphase induction motor for electric ship
propulsion. ESTS, 283–287 (2011)
90. B. Gamble, G. Snitchler, T. MacDonald, Full power test of a 36.5 MW HTS propulsion motor.
IEEE Trans. Appl. Supercond. 21(3), 1083–1088 (2011)
91. J. Dai, S.W. Nam, M. Pande, G. Esmaeili, Medium-voltage current-source converter drives for
marine propulsion system using a dual-winding synchronous machine. IEEE Trans. Ind. Appl.
50(6), 3971–3976 (2014)
92. G. Sulligoi, A. Tessarolo, V. Benucci, M. Baret, A. Rebora, A. Taffone, Modeling, simulation,
and experimental validation of a generator system for medium-voltage DC integrated power
systems. IEEE Trans. Ind. Appl. 46(4), 1304–1310 (2010)
93. O. Beik, N. Schofield, A. Al-Adsani, Variable Speed Brushless Hybrid Permanent Magnet
Generator for Hybrid Electric Vehicles (IEEE Transportation Electrification Conference and
Expo. (ITEC 2014), Dearborn, MI, USA, 2014), pp. 1–6. , 15–18 June 2014
94. A.S. Al-Adsani, O. Beik, Characterization of a hybrid PM generator using a 32-phase
brushless excitation scheme. IEEE Transactions on Energy Conversion 34(3), 1391–1400
(Sept. 2019)
95. A.S. Al-Adsani, A. Jarushi, O. Beik, ICE/HPM generator range extender for a series hybrid
EV powertrain. IET Electrical Systems in Transportation 10(1), 96–104 (2020)
96. A.S. Al-Adsani, O. Beik, Efficiency and Operational Envelop of a Hybrid PM Machine System
for HEVs (10th IET International Conference on Power Electronics, Machines and Drives
(PEMD 2020), Online Conference, UK, 2020), pp. 1–6. , 15–17 Dec. 2020
97. R. Bojoi, A. Cavagnino, M. Cossale, A. Tenconi, S. Vaschetto, Design trade-off and exper-
imental validation of multiphase starter generators for 48 V mini-hybrid powertrain. IEEE-
IEVC, 1–7 (2014)
98. C.C. Chan, J.Z. Jiang, G.H. Chen, X.Y. Wang, K.T. Chau, A novel polyphase multipole
square-wave permanent magnet motor drive for electric vehicles. IEEE Trans. Ind. Appl. 30
(5), 1258–1266 (1994)
99. M.G. Simoes, P. Vieira, A high-torque low-speed multiphase brushless machine - a perspec-
tive application for electric vehicles. IEEE Trans. Ind. Electron. 49(5), 1154–1164 (2002)
100. S. Waikar, T. Gopalaranthnam, H.A. Toliyat, J.C. Moreira, Evaluation of multiphase brushless
permanent magnet (BPM) motors using finite element method (FEM) and experiments. Pro-
ceedings IEEE APEC, Dallas, TX, 396–402 (1999)
101. M.G. Simoes, N.N. Franceschetti, P. Vieira Jr., Design and evaluation of a polyphase
brushless DC-machine direct drive systems, in Conf. Rec. IEEE IAS Annu. Meeting, Chicago,
IL, (2001), pp. 835–824
212 References

102. T. Gopalarathnam, S. Waikar, H.A. Toliyat, M.S. Arefeen, J.C. Moreira, Development of low
cost multi-phase brushless DC (BLDC) motors with unipolar current excitations, in Conf. Rec.
IEEE IAS Annu. Meeting, Phoenix, AZ, (1999), pp. 173–179
103. O. Beik, N. Schofield, Hybrid Generator for Wind Generation Systems (IEEE Energy Con-
version Congress and Exposition (ECCE), Pittsburgh, PA, 2014), pp. 3886–3893
104. O. Beik, N. Schofield, High voltage generator for wind turbines. 8th IET International
Conference on Power Electronics, Machines and Drives (PEMD), Glasgow, 1–6 (2016)
105. O. Beik, N. Schofield, An offshore wind generation scheme with a high-voltage hybrid
generator, HVDC interconnections, and transmission. IEEE Transactions on Power Delivery
31(2), 867–877 (2016)
106. O. Beik, N. Schofield, High-voltage hybrid generator and conversion system for wind turbine
applications. IEEE Trans. Ind. Electron. 65(4), 3220–3229 (2018)
107. O. Beik, A.S. Al-Adsani, Parallel nine-phase generator control in a medium-voltage DC wind
system. IEEE Trans. Ind. Electron. 67(10), 8112–8122 (2020)
108. O. Beik, A.S. Al-Adsani, Wind turbine multiphase operational trajectory in an all-DC wind
generation system. IET Renewable Power Generation 14(15), 2916–2923 (2020)
109. O. Beik, A.S. Al-Adsani, DC Wind Generation Systems: Design, Analysis, and Multiphase
Turbine Technology (Springer, 2020)
110. S. Kato, Y. Inui, M. Michihira, A Comparative Study of Steady-State Characteristics of
Permanent Magnet Synchronous Generator Systems (European Conference on Power Elec-
tronics and Applications, Aalborg, 2007), pp. 1–10
111. M.G. Simoes, P. Vieira, A high-torque low-speed multiphase brushless machine-a perspective
application for electric vehicles. IEEE Trans. Ind. Electron. 49(5), 1154–1164 (2002)
112. D. Vizireanu, S. Brisset, P. Brochet, Design and optimization of a 9-phase axial-flux
PMsynchronous generator with concentrated winding for direct-drive wind turbine, in Con-
ference Record IEEE IAS Annu. Meeting, Tampa, FL, (2006), pp. 1912–1918
113. D. Vizireanu, S. Brisset, X. Kestelyn, P. Borchet, Y. Milet, D. Laloy, Investigation on multi-
star structures for large power direct-drive windgenerator. Electric Power Components and
Systems 35(2), 135–152 (2007)
114. Z. Xiang-Jun, Y. Yongbing, Z. Hongtao, L. Ying, F. Luguang, Y. Xu, Modelling and control
of a multi-phase permanent magnet synchronous generator and efficient hybrid 3L-converters
for large direct-drive wind turbines. IET Electr. Power Appl. 6(6), 322–332 (2012)
115. O. Beik, A.S. Al-Adsani, Proposed Wind Turbine Limited- and High-Speed Operation (IEEE
Electric Power and Energy Conference (EPEC 2020), Online Conference, Edmonton, AB,
Canada, 2020), pp. 1–6, 9–10 Nov. 2020
116. I. Gonzalez, M.J. Duran, H.S. Che, E. Levi, J. Aguado, Fault-Tolerant Efficient Control of
Six-Phase Induction Generators in Wind Energy Conversion Systems with Series-Parallel
Machine-Side Converters (7th IET International Conference on Power Electronics, Machines
and Drives (PEMD), Manchester, 2014), pp. 1–6
117. B.C. Mecrow, A.G. Jack, D.J. Atkinson, S.R. Green, G.J. Atkinson, A. King, B. Green, Design
and testing of a four-phase fault-tolerant permanent magnet machine for an engine fuel pump.
IEEE Trans. On Energy Conversion 19 (2004)
118. G.J. Atkinson, B.C. Mecrow, A.G. Jack, D.J. Atkinson, P. Sangha, M. Benarous, Comparison
of Two Different Fault-Tolerant Switched Reluctance Machines for Fuel Pump Drive in
Aircraft (IEMDC, 2005)
119. C. Xiaoyuan, D. Zhiquan, P. Jingjing, L. Xiangsheng, Comparison of two different fault-
tolerant switched reluctance Machines for Fuel Pump Drive in aircraft. IEEE-IPEMC,
2086–2090 (2009)
120. C. Gerada, K. Bradley, X. Huang, A. Goodman, C. Whitley, G. Towers, A 5-phase fault-
tolerant brushless permanent magnet motor drive for an aircraft thin wing surface actuator.
IEMDC, 1643–1648 (2007)
References 213

121. X. Huang, A. Goodman, C. Gerada, Y. Fang, Q. Lu, Design of a Five-Phase Brushless DC

Motor for a Safety Critical Aerospace Application. IEEE Trans. Industrial Electronics 59(9),
3532–3541 (2012)
122. R.H. Nelson, P.C. Krause, Induction machine analysis for arbitrary displacement between
multiple winding sets. IEEE Transactions on Power Apparatus and Systems PAS-93(3),
841–848 (1974)
123. T.A. Lipo, A d-q model for six phase induction machines, in Proc. Int. Conf. Electrical
Machines (ICEM), Athens, Greece, (1980), pp. 860–867
124. L.A. Pereira, C.C. Scharlau, L.F.A. Pereira, J.F. Haffner, Model of a five-phase induction
machine allowing for harmonics in the air-gap field, part I: Parameter determination and
general equations, in Proc. IEEE IECON, Busan, Korea, (2004), pp. 98–103
125. L.A. Pereira, C.C. Scharlau, L.F.A. Pereira, J.F. Haffner, Model of a five-phase induction
machine allowing for harmonics in the air-gap field, part II: Transformation of co-ordinates
and d-q models, in Proc. IEEE IECON, Busan, Korea, (2004), pp. 1682–1687
126. L.A. Pereira, C.C. Scharlau, L.F.A. Pereira, J.F. Haffner, General model of a five-phase
induction machine allowing for harmonics in the air gap field. IEEE Trans. Energy Conver-
sion. 21(4), 891–899 (2006)
127. J. Huang, M. Kang, J. Yang, H. Jiang, D. Liu, Multiphase machine theory and its applications.
International Conference on Electrical Machines and Systems, Wuhan, 1–7 (2008)
128. J.M. Apsley, S. Williamson, A.C. Smith, M. Barnes, Induction motor performance as a
function of phase number. IEEE Proceedings - Electric Power Applications 153(6) (2006)
129. H.A. Toliyat, L. Xu, T.A. Lipo, A five-phase reluctance motor with high specific torque. IEEE
Trans. Ind. Appl. 28(3), 659–667 (1992)
130. H.A. Toliyat, S.P. Waikar, T.A. Lipo, Analysis and simulation of five-phase synchronous
reluctance machines including third harmonic of airgap MMF. IEEE Trans. Ind. Appl. 34(2),
332–339 (1998)
131. F. Scuiller, J. Charpentier, E. Semail, S. Clenet, Comparison of two 5-phase permanent magnet
machine winding configurations. Application on naval propulsion specifications. IEEE Inter-
national Electric Machines & Drives Conference, Antalya, 34–39 (2007)
132. L. Parsa, H.A. Toliyat, A. Goodarzi, Five-phase interior permanent-magnet motors with low
torque pulsation. IEEE Trans. Ind. Appl. 43(1), 40–46 (2007)
133. L. Parsa, T. Kim, Reducing Torque Pulsation of Multi-Phase Interior Permanent Magnet
Machines (Conference Record of the 2006 IEEE Industry Applications Conference Forty-First
IAS Annual Meeting, Tampa, FL, 2006), pp. 1978–1983
134. J. Figueroa, J. Cros, P. Viarouge, Generalized transformations for polyphase phase-
modulation motors. IEEE Trans. Energy Conversion 21(2), 332–341 (2006)
135. S. Waikar, T. Gopalarathnam, H.A. Toliyat, J.C. Moreira, Evaluation of Multiphase Brushless
Permanent Magnet (BPM) Motors Using Finite Element Method (FEM) and Experiments, vol
1 (APEC’99. Fourteenth Annual Applied Power Electronics Conference and Exposition,
Conference Proceedings Dallas, TX, USA, 1999), pp. 396–402
136. L. Parsa, T. Kim, Reducing torque pulsation of multi-phase interior permanent magnet
machines, in Conference Record IEEE IAS Annual Meeting, Tampa, FL, (2006),
pp. 1978–1983
137. F. Locment, E. Semail, F. Piriou, Design and study of a multiphase axial-flux machine. IEEE
Trans. Magnetics. 42(4), 1427–1430 (2006)
138. M.H. Rashid, Power Electronics Circuits, Devices, and Applications, 2nd edn., U.S.A., ISBN
0-13-678996-X (1993)
139. N. Mohan, M.T. Undeland, P. Robbins, Power Electronics Converters, Applications, and
Design, 2nd edn., ISBN 0-471-58408-8 (1995)
140. Website: http://www.Cree.com, as visited on 10-11-2010
141. Y. Burkhardt, A. Spagnolo, P. Lucas, M. Zavesky, P. Brockerhoff, Design and Analysis of a
Highly Integrated 9-Phase Drivetrain for EV Applications (2014 International Conference on
Electrical Machines (ICEM), Berlin, 2014), pp. 450–456
214 References

142. D.W. Novotny, T.A. Lipo, Vector Control and Dynamics of AC Drives (Clarendon, Oxford,
1998)
143. N. Bernard, A.H. Ben, B. Multon, Semi-Analytical Inductance Calculation on an Axial-Field
Synchronous Machine for a Flywheel Storage System Using Surfacic Permeances (IEEE
International Conference on Electric Machines and Drives (IEMDC 2001), 2001),
pp. 382–390
144. A.M. EL-Refaie, T.M. Jahns, Optimal flux weakening in surface PM machines using fractional
slot-slot concentrated windings. IEEE Trans. Ind. Appl. 41(3) (2005)
145. A.M. El-Refaie, T.M. Jahns, D.W. Novotny, Analysis of surface permanent magnet machines
with fractional-slot concentrated windings. IEEE Transactions on Energy Conversion 21(1),
34–43 (2006)
146. A.M. EL-Refaie, T.M. Jahns, Impact of winding layer number and magnet type on synchro-
nous surface PM machines designed for wide constant-power speed range operation. IEEE
Transactions on Energy Conversion 23(1) (2008)
147. A.M. EL-Refaie, Z.Q. Zhu, T.M. Jahns, D. Howe, Winding inductances of fractional slot
surface-mounted permanent magnet brushless machines. IEEE IAS Annual Meeting, Edmon-
ton, AB, Canada, 1–8. , 2–5 Oct. 2008 (2008)
148. Chee-Mun Ong, “Dynamic Simulation of Electric Machinery.”, ISBN 0-13-723785-5, New
Jersey, US, 1998
149. N. Schofield, Brushless Permanent Magnet Traction Machines with Extended Speed Capa-
bility (Ph.D. thesis, 1997)
150. S. Funda, Design and Development of a High-Speed Axial-Flux Permanent Magnet Machine.
ISBN 90-386-1380-1, 2001
151. J.L. Guardado, K.J. Cornick, Calculation of machine winding electrical parameters at high
frequencies for switching transient studies. IEEE Transactions on Energy Conversion 11(1),
33–40 (1996)
152. M.-F. Hsieh, Y.-C. Hsu, D.G. Dorrell, K.-H. Hu, Investigation on end winding inductance in
motor Stator windings. IEEE Trans. Magn. 43(6), 2513–2515 (2007)
153. O. Chiver, L. Petrean, L. Neamt, Z. Erdei, Study Regarding End Winding Inductance of Three
Phase a.C. Windings in a Single Layer (International conference on Energy and Environment
Technologies and Equipment, ISBN 978-960-474-181-6, 2010)
154. Y. Amara, K. Oujehani, E. Hoang, M. Gabsi, M. Lecrivain, A.H. Ben Ahmed, S. Derou, Flux
Weakening of Hybrid Synchronous Machines (IEEE International on Electric Machines and
Drives Conference (IEMDC 2001), 2001), pp. 367–373
155. J.W. Kelly, E.G. Strangas, J.M. Miller, Multiple space vector pulse width modulation. IEEE
Transaction on Energy Conversion 18(2) (2003)
156. R. Krishnan, PM Synchronous and Brushless DC Motor Drives (Taylor and Francis Group.,
ISBN: 978082475384, 2010)
157. Website: http://www.electronic-tutorials.ws/diode, as visited on 10-11-2010
158. Website: http://www.Evoxrifa.com/europe/index.html, as visited on 20-4-2009
159. G. Dusan, P. Marco, IGBT power losses calculation using the data-sheet parameters. Appli-
cation Note on Automotive Power V1(1) (2009)
160. F. Soltani, N. Debbache, Integration of converter losses in the modelling of hybrid
photovoltaic-wind generating system. European Journal of Scientific Research ISSN 1450-
216X 21(4), 707–718 (2008)
161. S. Bernet, R. Teichmann, A. Zuckerberger, P. Steimer, Comparison of High Power IGBTs and
Hard Driven GTOs for High Power Inverters, vol 2 (Thirteenth Annual Conference Pro-
ceedings on Applied Power Electronics and Exposition (APEC’98), 1998), pp. 711–718. ,
15–19 Feb. 1998
162. F. Casanellas, Losses in PWM inverters using IGBTs. IEEE Proceeding on Electronic Power
Applications 141(5) (1994)
163. A.J. Brown, An Assessment of the Efficiencies of Soft and Hard-Switched Inverters for
Application in Power Electric Vehicles, 2000
References 215

164. N. Schofield, H.T. Yap, C.M. Bingham, A H2 PEM Fuel Cell and High Energy Dense Battery
Hybrid Energy Source for an Urban Electric Vehicle (International Electric Machines and
Drives Conference (IEMDC 2005), San Antonio, Texas, USA, 2005), pp. 1–8, IEEE Cat.
05EX1023C, ISBN: 0-7803-8988-3, 15-18, May 2005
165. Website: https://cars.usnews.com/cars-trucks/best-electric-cars, as visited on December 2020
166. Website: http://rusautonews.com/2020/08/05/russian-electric-car-market-decreased-by-18-
within-the-first-half-of-2020, as visited on December 2020
167. Website: https://leandesign.com/tearing-down-tesla-segment-2-tesla-model-3-vs-bmw-i3-
with-rex-system, as visited on December 2020
168. Website: https://pamplinmedia.com/wh/481309–388157-2020-chevy-bolt-ev-premier-still-
breaking-new-ground-pwoff, as visited on December 2020
169. Website:https://upload.wikimedia.org/wikipedia/commons/1/14/2018_Tesla_Model_S_75D.
jpg, as visited on December 2020
170. Congressional Research Service, Informing the legislative debate since 1914, “Electrification
May Disrupt the Automotive Supply Chain”, 2019
171. Website: https://afdc.energy.gov, as visited on December 2020
172. A. Emadi, Advanced Electric Vehicles (Taylor and Francis Group, 2015)
173. Website: https://commons.wikimedia.org/w/index.php?curid¼401224, as visited on
December 2020
174. Website: https://commons.wikimedia.org/w/index.php?curid¼3391964, as visited on Jan.
2021
175. Website: https://commons.wikimedia.org/w/index.php?curid¼3666593, as visited on Jan
2021
176. Website: https://commons.wikimedia.org/w/index.php?curid¼3392127, as visited on
December 2020
177. R. Yatheesha, A. Anarghya, B. Ranjith, et al., Extended range electric vehicle (EREV).
International Journal of Engineering and Advance Technology (IJEAT), ISSN 2249–8958 4
(1), 68–73 (2014)
178. Website: https://commons.wikimedia.org/w/index.php, as visited on December 2020
179. Website: https://afdc.energy.gov/vehicles/how-do-plug-in-hybrid-electric-cars-work, as vis-
ited on December 2020
180. H.T. Yap, N. Schofield, Test Characterisation of a H2 PEM Fuel Cell (IEEE Conference on
Vehicle Power and Propulsion (VPPC’07), 2007), pp. 551–558. , 9-12 Sept 2007
181. C.W. Ayers, J.S. Hsu, L.D. Marlino, C.W. Miller, G.W. Ott, C.B. Oland, Evaluation of 2004
Toyota Prius Hybrid Electric Drive System Interim Report (Oak Ridge National Laboratory)
ORNL,TM-2004/247
182. Website: https://ww3.arb.ca.gov/msprog/onroadhd/avlpemssop.pdf, as visited on December
2020
183. Website: https://www.car-engineer.com/the-different-driving-cycles, as visited on December
2020
184. J. Dixon, I. Nakashima, E.F. Arcos, M. Ortuzar, Electric vehicle using a combination of
Ultracapacitors and ZEBRA battery. IEEE Trans. Ind. Electron. 57(3), 943–949 (2010)
185. N. Schofield, H.T. Yap, C.M. Bingham, A H/sub 2/ PEM fuel cell and high energy dense
battery hybrid energy source for an urban electric vehicle. IEEE International Conference on
Electric Machines and Drives, San Antonio, TX, 1793–1800 (2005)
186. O. Veneria, C. Capassoa, S. Patalanob, Experimental study on the performance of a ZEBRA
battery based propulsion system for urban commercial vehicles. Appl. Energy 185, 2005–2018
(2017)
187. D.K. Bhatt, M. El Darieby, An Assessment of Batteries Form Battery Electric Vehicle
Perspectives (IEEE International Conference on Smart Energy Grid Engineering, Oshawa,
Canada, 2018), pp. 255–259
216 References

188. ADVISOR data file for fuel consumption and emissions, generated by computer models
developed by Feng An and calibrated by using real test data of 1.5L (43 kW) Toyota Prius
engine, 1999
189. M. Ehsani, G. Yimin, E.G. Sebastien, E. Ali, Modern Electric, Hybrid Electric, and Fuel Cell
Vehicles: Fundamentals, Theory, and Design, 1st edn. (Boca Raton, CRC Press. ,
395 p. ISBN: 0-8493-3154-4, 2005)
190. Website: https://www.greencarcongress.com/2009/09/lotus-range-extender.html, as visited on
Jan. 2021
191. Website: https://hydrogentoday.info/news/5329, as visited on Jan. 2021
192. Y. Ko, J. Lee, H. Lee, A supervisory control algorithm for a series hybrid vehicle with multiple
energy sources. IEEE Trans. Veh. Technol. 64(11), 4942–4953 (2015)
193. M. Shahverdi, M.S. Mazzola, Q. Grice, M. Doude, Bandwidth-based control strategy for a
series HEV with light energy storage system. IEEE Trans. Veh. Technol. 66(2), 1040–1052
(2017)
194. S. Neglur, M. Ferdowsi, Effect of Battery Capacity on the Performance of Plug-in Hybrid
Electric Vehicles (IEEE vehicle Power and Propulsion Conference, Dearborn, MI, 2009),
pp. 649–654
195. V. Sezer, M. Gokasan, S. Bogosyan, A novel ECMS and combined cost map approach for
high-efficiency series hybrid electric vehicles. IEEE Trans. Veh. Technol. 60(8), 3557–3570
(2011)
196. M. Kim, D. Jung, K. Min, Hybrid thermostat strategy for enhancing fuel economy of series
hybrid Intracity bus. IEEE Trans. Veh. Technol. 63(8), 3569–3579 (2014)
197. C. Polini, F. Crescimbini, A. Di Napoli, A. Lidozzi, L. Solero, Power Flow Analysis for Plug
in Hybrid Electric Vehicles (International Conference on Electrical Machines, Rome, 2010),
pp. 1–6
198. A. Ivanco, Z. Filipi, Framework for Powerpack Optimization in the SHEV: Two Stage
Optimization for Best Efficiency in the Hybrid Electric Vehicle Powertrain Design (IEEE
Vehicle Power and Propulsion Conference (VPPC), Montreal, QC, 2015), pp. 1–4
199. P. Capaldi, A. Dannier, I. Spina, A simple on-board electric generator for road electric vehicles
based on single-cylinder engine and PM-brushless generator. International Symposium on
Power Electronics Power Electronics, Electrical Drives, Automation and Motion, Sorrento,
374–378 (2012)
200. C. Wang, M. Jin, J. Shen, C. Yuan, A permanent magnet integrated starter generator for
electric vehicle onboard range extender application. IEEE Trans. Magn. 48(4), 1625–1628
(2012)
201. A. Styler, A. Sauer, I. Nourbakhsh, H. Rottengruber, Learned Optimal Control of a Range
Extender in a Series Hybrid Vehicle (IEEE International Conference on Intelligent Transpor-
tation Systems, Las Palmas, 2015), pp. 2612–2618
202. A. Jarushi, N. Schofield, Modelling and analysis of energy source combinations for electric
vehicles. World Electric Vehicle Journal 3 (2009)
203. J. Guanetti, S. Formentin, S.M. Savaresi, Energy management system for an electric vehicle
with a rental range extender: A least costly approach. IEEE Trans. Intell. Transp. Syst. 17(11),
3022–3034 (2016)
204. L.J. Kere, S. Kelouwani, K. Agbossou, Y. Dubé, Improving efficiency through adaptive
internal model control of hydrogen-based Genset used as a range extender for electric vehicles.
IEEE Trans. Veh. Technol. 66(6), 4716–4726 (2017)
205. Website: https://insideevs.com/news/330318/nissan-unveils-worlds-first-solid-oxide-fuel-
cell-e-nv200-video, as visited on Jan. 2021
206. Z.Q. Zhu, W.Q. Chu, Y. Guan, Quantitative comparison of electromagnetic performance of
electrical machines for HEVs/EVs. CES Transactions on Electrical Machines and Systems 1
(1) (2017)
References 217

207. H. Seol, J. Lim, D. Kang, J.S. Park, J. Lee, Optimal design strategy for improved operation of
IPM BLDC motors with low-resolution hall sensors. IEEE Trans. Ind. Electron. 64(12),
9758–9766 (2017)
208. A.S. Al-Adsani, M.E. AlSharidah, O. Beik, BLDC Motor Drives: A Single Hall Sensor
Method and a 160o Commutation Strategy (IEEE Transactions on Energy Conversion, early
access, 2020)
209. G. Liu, C. Cui, K. Wang, B. Han, S. Zheng, Sensorless control for high-speed brushless DC
motor based on the line-to-line Back EMF. IEEE Trans. Power Electron. 31(7), 4669–4683
(2016)
210. X. Zhou, X. Chen, M. Lu, F. Zeng, Rapid self-compensation method of commutation phase
error for low- inductance BLDC motor. IEEE Transactions on Industrial Informatics 13(4),
1833–1842 (2017)
211. H. Wu, M. Wen, C. Wong, Speed Control of BLDC Motors Using Hall Effect Sensors Based
on DSP (2016 International Conference on System Science and Engineering (ICSSE), Puli,
2016), pp. 1–4
212. B.P. Reddy, A. Murali, SoC FPGA-Based Field Oriented Control of BLDC Motor Using Low
Resolution Hall Sensor (IECON 2016 – 42nd Annual Conference of the IEEE Industrial
Electronics Society, Florence, 2016), pp. 2941–2941
213. A.M. Jarushi, Analysis and Modelling of Energy Source Combinations for Electric Vehicles
(Ph.D. thesis, 2010)
214. Website: https://ec.europa.eu/clima/policies/transport/vehicles/cars_en, as visited on
December 2020
215. O. Beik, A.S. Al-Adsani, Active and passive control of a dual rotor wind turbine generator for
DC grids. IEEE Access 9, 1987–1995 (2021)
216. N. Schofield, H.T. Yap, C.M. Bingham, Hybrid Energy Sources for Electric and Fuel Cell
Vehicle Propulsion (IEEE Conference on Vehicular Power and Propulsion (VPPC’05), Chi-
cago, USA, CD, ISBN: 0-7803-9281-7, 7-9 Sept. 2005, 2005)
217. N. Schofield et al., Develop High Energy Battery and High Power Supercaps for all Electric
Range Van Evaluation (DESERVE) (Technology Strategy Board, Low Carbon Vehicles
Innovation Platform, Project No. TP/9/LCV/6/I/S0037C, Aug. 2008–Feb. 2011, 2011)
218. H.T. Yap, N. Schofield, C.M. Bingham, Hybrid energy/power sources for electric vehicle
traction systems. IEE PEMD Conference, 1–6 (2004)
219. H. Yoo, S.-K. Sul, Y. Park, J. Jeong, System integration and power-flow Management for a
Series Hybrid Electric Vehicle Using Supercapacitors and Batteries. IEEE Trans. Ind. Appl. 44
(1), 108–114 (2008)
220. B. He, M. Ouyang, Average Modelling of Diesel Auxiliary Power Unit for Series Hybrid
Electric Vehicle, vol 4 (The 62nd IEEE Vehicular Technology Conference, (VTC-2005-Fall.
2005), 2005), pp. 2372–2376. , 25–28 Sept. 2005
221. B. He, M. Yang, Robust LPV control of diesel auxiliary power unit for series hybrid electric
vehicles. IEEE Trans. Power Electron. 21(3), 791–798 (2006)
222. J.L. Sudworth, Zebra batteries. J. Power Sources 51, 105–114 (1994)
223. R.C. Galloway, S. Haslam, The ZEBRA electric vehicle battery: Power and energy
improvements. J. Power Sources 80, 164–170 (1999)
224. H.T. Yap, N. Schofield, C.M. Bingham, Hybrid energy/power sources for electric vehicle
traction systems. IEE PEMD Conference, 1–6 (2004)
225. N. Schofield, H.T. Yap, C.M. Bingham, Hybrid Energy Sources for Electric and Fuel Cell
Vehicle Propulsion (IEEE Vehicular Power and Propulsion Conference (VPPC’05), Chicago,
USA, CD, ISBN: 0-7803-9281-7, 2005), pp. 1–8. , 7-9 Sept. 2005
226. S. Gui-Jia, L. Tang, X. Huang, Control of Two Permanent Magnet Machines Using a Five-Leg
Inverter for Automotive Applications, vol 3 (Conference Record of the 41st IAS Annual
Meeting, 2006), pp. 1606–1612. , 8–12 Oct. 2006
227. Z. Zhou, W. Knapp, J. MacEnri, H.C. Sorensen, M.E. Friis, I. Masters, P. Igic, Permanent
Magnet Generator Control and Electrical System Configuration for Wave Dragon MW Wave
218 References

Energy Take-off System (IEEE International Symposium on Industrial Electronics (ISIE 2008),
2008), pp. 1580–1585
228. N. Jiao, W. Liu, T. Meng, J. Peng, S. Mao, Design and control of a two-phase brushless exciter
for aircraft wound-rotor synchronous starter/generator in the starting mode. IEEE Trans.
Power Electron. 31(6), 4452–4461 (2016)
229. C.J. Wu, Y.S. Chuang, C.P. Huang, Y.P. Chang, Tuning of static excitation systems to
improve dynamic performance for generators serving electric arc furnaces loads. IEEE Trans-
actions on Energy Conversion 22(2), 350–357 (2007)
230. I. Boldea, Synchronous Generators (CRC Press/Taylor and Francis, 2006)
231. R.C. Schaefer, Applying static excitation systems. IEEE Ind. Appl. Mag. 4(6), 41–49 (1998)
232. J.K. Nøland, F. Evestedt, J.J. Pérez-Loya, J. Abrahamsson, U. Lundin, Design and character-
ization of a rotating brushless outer pole PM exciter for a synchronous generator. IEEE Trans.
Ind. Appl. 53(3), 2016–2027 (2017)
233. K.D. Papastergiou, D.E. Macpherson, An airborne radar power supply with contactless
transfer of energy—Part I: Rotating transformer. IEEE Trans. Ind. Electron. 54(5),
2874–2884 (2007)
234. R. Trevisan, A. Costanzo, A 1-kW contactless energy transfer system based on a rotary
transformer for sealing rollers. IEEE Trans. Ind. Electron. 61(11), 6337–6345 (2014)
235. A. Abdolkhani, A.P. Hu, N.K.C. Nair, A double Stator through-hole type contactless Slipring
for rotary wireless power transfer applications. IEEE Transactions on Energy Conversion 29
(2), 426–434 (2014)
236. H. Zhong, L. Zhao, X. Li, Design and analysis of a three-phase rotary transformer for doubly
fed induction generators. IEEE Trans. Ind. Appl. 51(4), 2791–2796 (2015)
237. H.R. Mohabati, J.S. Moghani, S.T. Boroujeni, Complex vector modelling and sequence
analysis of the integrated three-phase rotating transformer for design of a symmetrical struc-
ture. IET Electr. Power Appl. 10(7), 649–657 (2016)
238. S. Zhu, C. Liu, K. Wang, X. Yuan, Y. Hu, Y. Ning, Magnetic field distribution and operating
characteristics of a hybrid excitation generator based on integrated brushless excitation. IEEE
Trans. Magn. 52(9), 1–12 (2016)
239. Z. Shushu, L. Chuang, N. Yinhang, T. Jie, A two-stage brushless excitation method for hybrid
excitation synchronous generator. IEEE Trans. Magn. 51(6), 1–11 (2015)
240. A. Darabi, C. Tindall, Brushless exciter modeling for small salient pole alternators using finite
elements. IEEE Transactions on Energy Conversion 17(3), 306–312 (2002)
241. D.C. Aliprantis, S.D. Sudhoff, B.T. Kuhn, A brushless exciter model incorporating multiple
rectifier modes and Preisach’s hysteresis theory. IEEE Transactions on Energy Conversion 21
(1), 136–147 (2006)

Further Reading

242. O. Beik, N. Schofield, A Brushless Exciter Design for a Hybrid Permanent Magnet Generator
Applied to Series Hybrid Electric Vehicles (7th IET International Conference on Power
Electronics, Machines and Drives (PEMD 2014), Manchester, 2014)
243. O. Beik, A. Dekka, M. Narimani, A New Modular Neutral Point Clamped Converter with
Space Vector Modulation Control (2018 IEEE International Conference on Industrial Tech-
nology (ICIT), Lyon, 2018), pp. 591–595
244. O. Beik, A. Al-Adsani, Wind Turbine Productivity and Wind Energy Assessment: An Ontario
Case Study (2020 IEEE Electric Power and Energy Conference (EPEC), Edmonton, AB,
Canada, 2020), pp. 1–5
References 219

245. O. Beik, R. Yang, A. Emadi, Design of a 5-Phase IPM Machine for Electric Vehicles (2018
IEEE Transportation Electrification Conference and Expo (ITEC), Long Beach, CA, 2018),
pp. 168–172
246. O. Beik, Design optimization of a PM motor: A practical approach for mass production. IEEE
Transactions on Energy Conversion 35(4), 1849–1858 (2020)
247. D. Schumacher, O. Beik, A. Emadi, Standalone integrated power electronics system: Appli-
cations for off-grid rural locations. IEEE Electrification Magazine 6(4), 73–82 (2018)
248. O. Beik, A. Emadi, Toward integrated digital aircraft control systems: For electrified-system
optimized size, cost, and efficiency. IEEE Electrification Magazine 5(4), 46–52 (2017)
249. N. Schofield, X. Niu, O. Beik, Multiphase Machines for Electric Vehicle Traction (2014 IEEE
Transportation Electrification Conference and Expo (ITEC), Dearborn, MI, 2014), pp. 1–6
250. O. Beik, Operational envelope of an electric transmission auxiliary system. in IEEE Trans-
actions on Energy Conversion 35(3), 1513–1521 (2020)
251. A. Dekka, O. Beik, M. Narimani, Modulation and voltage balancing of a five-level series-
connected multilevel inverter with reduced isolated direct current sources. IEEE Trans. Ind.
Electron. 67(10), 8219–8230 (2020)
252. M.R. Aghamohammadi, A. Beik-Khormizi, Small Signal Stability Constrained Rescheduling
Using Sensitivities Analysis by Neural Network as a Preventive Tool (IEEE PES T&D 2010,
New Orleans, LA, 2010), pp. 1–5
253. A.B. Khormizi, A.S. Nia, Damping of Power System Oscillations in Multi-Machine Power
Systems Using Coordinate Design of PSS and TCSC (2011 10th International Conference on
Environment and Electrical Engineering, Rome, 2011), pp. 1–4
254. A. Al-Adsani, N. Schofield, Comparison of Three and Nine-Phase Hybrid Permanent Magnet
Generator Designs (The 35th Annual Conference of the IEEE Industrial Electronic Society
(IECON 2009), Porto, Portugal. , 3–5 Nov. 2009, 2009)
255. A. Al-Adsani, A. Jarushi, N. Schofield, Operation of a Hybrid PM Generator in a Series
Hybrid Electric Vehicle (The 35th Annual Conference of the IEEE Industrial Electronic
Society (IECON 2009), Porto, Portugal, 2009), pp. 3898–3903. , 3–5 Nov. 2009
256. A. Al-Adsani, A.M. Jarushi, N. Schofield, An ICE/HPM Generator Range Extender in a Series
Hybrid Electric Vehicle (5th IET International Conference on Power Electronics, Machines
and Drives (PEMD 2010), Brighten, UK, 2010), pp. 1–6. , 19–21 April 2010
257. N. Schofield, A. Al-Adsani, Operation of a Hybrid PM Generator in a Series Hybrid EV
Power-Train (The 7th IEEE Vehicular Power and Propulsion Conference (VPPC’11), Chi-
cago, USA, 2011), pp. 1–6. , 6–9 Sept. 2011
258. N. Xin, N. Schofield, A. Al-Adsani, Comparison of 3- and 9-Phase Brushless PM Drive
Systems for Aerospace Actuation (6th IET International Conference on Power Electronics,
Machines and Drives (PEMD 2012), Bristol, UK, 2012), p. 13., , 27–29 March 2012
259. O. Beik, An HVDC off-Shore Wind Generation Scheme with High Voltage Hybrid Generator
(Ph.D. dissertation, McMaster University, 2016)
260. N. Schofield, O. Beik, Wind Turbine High Voltage Generator and System (The 15th Interna-
tional Workshop on Large-Scale Integration of Wind Power into Power Systems, Vienna,
Austria, 2016)
261. O. Beik, Electrified Landing Gear Adaptor (US Patent US 2020/0377062 A1, Dec 03, 2020,
2020)
262. M.R. Aghamohammadi, A. Beik Khormizi, M. Rezaee, Effect of Generator Parameters
Inaccuracy on Transient Stability Performance (2010 Asia-Pacific Power and Energy Engi-
neering Conference, Chengdu, 2010), pp. 1–5
263. A.B. Khormizi, M. Hajati, M.R. Aghamohammadi, An analytical method to parameter
estimation by dynamic performance evaluation. International Journal of Energy and Power
(IJEP) 2(1), 1–7 (2013)
Index

A Consequent pole permanent magnet hybrid

AC L-L voltage, 100 excitation machine
Afjei design, 40 (CPPM), 27, 29, 33
Alternating current (AC), 6 Convection heat transfer, 84–86
Ampere’s law, 13 Conventional gasoline/diesel-fueled
AVL PEMS, 151 vehicles, 143
Axillary ICE/HPM generator unit, 168

D
B DC excitation, 194, 206
Back-EMF, 101–106 DC-link voltage, 130, 132–134, 158, 159, 172
Battery management interface (BMI), 155 Diode conduction losses, 138
Battery technology, 146, 147, 170 Direct current (DC) machines, 7–10
Bipolar hybrid excitation synchronous dq mathematical model, 115–118
machines, 35 Driving cycle profiles, 167
Brushed DC and AC machines, 7–10 Dual-rotor machine, 30, 33
Brushless AC machines, 10, 11, 163 Dual-stator hybrid excited synchronous wind
Brushless DC traction machine efficiency, 163 generator (DSHESG), 40, 43
Brushless excitor, characterization
using, 194, 197
efficiency mapping, 204–206 E
performance curves, 200–204 Efficiency mapping, 204–206
32-phase brushless excitation scheme, Electric machine fundamentals, 5, 6
197–200 Electric machine theory
classical electric machine topologies, 6
brushed DC and AC machines, 7–10
C brushless AC machines, 10, 11
Carbon dioxide (CO2) emission data, 156 switch reluctance machines, 11, 12
Classical electric machine topologies, 6 electric machine fundamentals, 5, 6
brushed DC and AC machines, 7–10 ferromagnetic materials and magnetization
brushless AC machines, 10, 11 curve, 3–5
switch reluctance machines, 11, 12 magnetic circuit principles, 1
Commercial HEVs, 148 magnetic field distribution and flux
Conduction heat transfer, 82–84 density, 1–3
Conduction losses, 136

© Springer Nature Switzerland AG 2022 221

A. S. Al-Adsani, O. Beik, Multiphase Hybrid Electric Machines,
https://doi.org/10.1007/978-3-030-80435-0
222 Index

Electric machine theory (cont.) H

WF and PM synchronous machine Highway driving cycle (HWFET), 151, 152
excitation fields, 13 Homopolar hybrid excitation synchronous
magnetic flux path representation, 13–15 machines, 35
Electric vehicles (EVs), 172, 204 HPM generator, 158
automotive market, 143 HPM generator excitation field control
battery technology, 146, 147 function, 167
companies, 143, 144 HPM generator open-loop control system block
driving range, 143 diagram, 176
high-voltage battery system, 143 HPM generator ratings, 168
powertrain configuration, 145 HPM machine, 153
vehicle gross weight, 143 HPM machine back-EMF control strategy, 173
Electrified powertrain components, 166 control strategy analysis, 174–178
Energy conversion, 149 DC-link design options, 178, 179, 181
Energy management algorithm, 167 HPM machine final design model analysis, 75,
Energy management and control 76
strategies, 159 core loss prediction, 77–81
Engine-mounted multiphase HPM rotor PM demagnetization, 75–78
generator, 158 HPM machine output power control, 181–185
Equivalent magnetic circuit (EMC) model operating scenarios, 184, 186
PM machine analysis via, 56–61 solving final choice with full simulation
European Commission proposal, 170 model, 189–191, 193
EV models, 144 thermal analysis, 191, 194–196
EVs powertrain configuration, 145 two driving cycles, energy loss prediction
EVs range extender for, 187–189
auxiliary power sources, 159 HPM machine thermal model, 86, 88, 89
battery system, 160 HPM machine wound field, 173
control strategies, 160 HPM machines, 18
energy sources, 160 HPM machines prototype
hybrid thermostat strategy, 160 construction of, 111–113
ICE coupled electric generator, 161 HPM topology, 154
ICE/HPM generator, 159 HWFET driving cycle, 168
off-board and onboard functions, 161 Hybrid electric machine concept
operational scenarios, 161 classification, 17
requirements, 161 history, 17
simulation powertrain platform, 161 surveyed literature on, 40, 43, 44
supervisor control optimization, 160 topologies, 18
EVs vs. SHEV, 169 consequent pole permanent magnet
Excitation systems, 194 hybrid excitation machine, 27, 29
dual-rotor machine, 30, 33
dual-stator hybrid excited synchronous
F wind generator, 40, 43
Fast Fourier transform (FFT), 103 field controlled torus-NS machine, 30,
Ferromagnetic materials 33
(FERMMs), 3–5 hybrid excitation synchronous machine,
Field controlled torus-NS (FCT-NS) 23, 25
machine, 30, 33 imbricated hybrid excitation machine,
Finite element method 32, 34, 36, 37
program, 52–55 PM synchronous machine with claw
Flux density, 1–3 pole field excitation, 18, 19, 21
Fuel cell range extender, 161 series double excited synchronous
machine, 36, 38, 39
Index 223

switch reluctance machine with stator WF rotor designs, comparative analysis

field assistance, 39, 40, 42 of, 65–70
synchronous permanent magnet hybrid WF to PM split ratio, 64, 65
AC machine, 25, 26, 28, 29 Hybrid PM (HPM), 153
toroidal-stator transverse-flux Hybrid terminology, 153
machine, 21–25 Hybrid thermostat strategy, 160
Hybrid electric vehicles (HEVs) Hybridization ratio (HR)
battery charging time, 146 calculation, 166
commercial, 148 constant power source, 166
fuel consumption/fuel economy, 148 dynamic power, 165
models, 148 ICE/HPM generator, 166
plug-in HEVs, 146 minimum battery power, 166
powertrain components, 148, 150 power share measurement, 165
topology, 152 ZEBRA battery, 165
vehicle range issue, 143, 146
Hybrid EV powertrain configurations, 151
Hybrid excitation field regulation I
topologies, 17 ICE energy source, 149
Hybrid excitation synchronous generator ICE/HPM generator fixed power
(HESG), 19 output, 166
Hybrid excitation synchronous machine ICE/HPM generator range extender, SHEVs
(HESM), 23, 25 algorithm, 166
Hybrid permanent magnet machine design, 45 benchmark system, 162
HPM machine final design model analysis, case studies, 169
75, 76 HWFET and NEDC cycles, 169
core loss prediction, 77–81 hybridization ratio, 165
rotor PM demagnetization, 75–78 ICE, 162
HPM machine parameters sizing, 166
synchronous inductance and winding structure, 162
resistance, 73–75 vehicle traction machine
torque prediction torque, 163
and saturation, 69, 71, 72 ICE/HPM generator system output
machine thermal model power, 169
conduction heat transfer, 82–84 ICE/PM machine, 153
convection heat transfer, 84–86 Imbricated hybrid excitation machine (IHEM),
HPM machine thermal model, 86, 88, 90 32, 34, 36, 37
lumped parameter method, principle of, Induction machine (IMs), 10
81, 82 Interior PM (IPM) generator, 172
radiation heat transfer, 86 Internal combustion engine (ICE), 143
machine volume envelope consideration, 46 actual test data, 156
finite element method program, 52–55 driving cycles emissions, 157
machine back-EMF prediction, 55–57 efficiency, 156, 157
PM machine analysis via EMC model, FCR, 158
56–61 operating point, 156
PM machine dimensions, 47, 48 SHEV fuel consumption, 157
PM machine stator winding layout,
47–50
stator winding fill factor and resistance, L
50, 51 Lead-acid battery technology, 146
PM and four HPM machine topologies, Lithium-ion (Li-ion), 146
90–93 Loss audit
WF machine, 61 multiphase HPM generator
rotor design, 62–64 systems, 134, 135
224 Index

Loss audit (cont.) HPM machines prototype, construction

core loss prediction, 136, 137 of, 111–113
passive and active converter loss, 136, nine-phase winding layout and back-
138–141 EMF, 101–103
Lumped parameter method, 59, 65, 81, 82 resistance and inductance
measurements, 114, 115
synchronous inductance prediction,
M 107–111
Machine back-EMF prediction, 55–57 rectified voltage, 96, 98–100
Magnetic circuit principles, 1 three- and nine-phase HPM generator
magnetic field distribution and flux system studies, 119–121
density, 1–3 DC-link voltage quality, 130, 133, 134
Magnetic field distribution, 1–3 synchronous inductance and rectifier,
Magnetization curves, 3–5 impact on, 121–126
Matlab/Simulink SimPower model, 103 system sensitivity, 126–128, 130, 131
Motor-CAD, 85, 86 Multiphase windings principles, 96, 97
Motor control, 163
Multiphase HPM generator, 171
brushless excitor, characterization using, N
194, 197 NEDC driving cycle, 158
32-phase brushless excitation scheme, NEDC suburban driving cycle, 167, 170
197–200 Neodymium–iron–boron (NdFeB), 75
efficiency mapping, 204, 205 New European driving cycle (NEDC), 151, 156
performance curves, 200–204 Nickel–metal hydride (Ni-MH), 146
HPM machine back-EMF control Nine-phase HPM generator parameters, 101
strategy, 173 back-EMF and torque waveform harmonics
control strategy analysis, 174–178 prediction, 103–106
DC-link design options, 178, 179, 181 HPM machines prototype, construction
HPM machine output power control, of, 111–113
181–185 nine-phase winding layout and back-EMF,
operating scenarios, 184, 186 101–103
solving final choice with full simulation resistance and inductance measurements,
model, 189–191, 193 114, 115
thermal analysis, 191, 194–196 synchronous inductance prediction,
two driving cycles, energy loss 107–111
prediction for, 187–189 Nine-phase HPM generator prototype, 162
HPM machine wound field, 173 Nine-phase HPM generator system studies,
hybrid EV powertrain, 172 119–121
power flow management scheme, 172 DC-link voltage quality, 130, 132–134
vehicle DC-link voltage, 173 synchronous inductance and rectifier,
Multiphase HPM generator systems, 95, 96 impact on, 121–126
analysis models system sensitivity, 126–131
general dq mathematical model, Nine-phase HPM generator systems, 96,
115–118 98–100
simulation model, 118, 119 Nine-phase stator winding, 118
loss audit, 134, 135 Nine-phase winding layout, 101–103
core loss prediction, 136, 137
passive and active converter
loss, 136, 138–141 O
multiphase windings principles, 96, 97 Off-board ICE-generator system, 161
nine-phase HPM generator parameters, 101 Off-the-shelf machines, 95
back-EMF and torque waveform Original equipment manufacturers (OEM), 152
harmonics prediction, 103–106 Output power ratio curves, 127
Index 225

P Simulation model
Parallel HEV format, 152 multiphase HPM generator systems, 118,
Parallel hybrid electric vehicle format, 148 119
Passive and active rectification Squirrel-cage machines, 11
stage losses, 141 Standard drive cycles, 151
Performance curves, 200–204 Starter/alternator, 151
Permanent magnet machine dimensions, 47, 48 State-of-charge (SOC), 155
Plug-in HEV powertrain components, 146, 149 Stator winding fill factor and resistance, 50, 51
PM and four HPM machine topologies, 90–93 Steady power generation mode, 151
PM machine analysis via EMC model, 56–61 Suburban driving cycle, 169
PM machine stator winding layout, 47–50 Switch reluctance machine with stator field
PM synchronous machine with claw pole field assistance, 39, 40, 42
excitation (PSCPF), 18, 19, 21 Switch reluctance machines (SRMs), 11, 12
PM synchronous machines Synchronous inductance and winding
magnetic flux path representation of, 14, 15 resistance, 73–75
Power flow management scheme, 172 Synchronous inductance prediction, 107–111
Power losses, 136 Synchronous permanent magnet hybrid AC
Powertrain hybridization rations, 168 machine (SynPM), 25, 26, 28, 29

R T
Radiation heat transfer, 86 32-phase brushless excitation scheme, 197,
Rectified voltage, 96, 98–100 198, 200
Reverse recovery loss, 138 Thermal model, HPM
Revolutions per minute (RPM), 9 conduction heat transfer, 82–84
Rotor PM demagnetization, 75–78 convection heat transfer, 84–86
HPM machine thermal model, 86, 88, 90
lumped parameter method, principle
S of, 81, 82
Sensorless control, 163 radiation heat transfer, 86
Series double excited synchronous machine Three-phase HPM generator system studies,
(SDESM), 36, 38, 39 119–121
Series HEV powertrain schematics, 154 DC-link voltage quality, 130, 133, 134
Series hybrid electric vehicle (SHEV) synchronous inductance and rectifier,
battery pack, 153 impact on, 121–126
excitation control loop, 154 system sensitivity, 126–128, 130, 131
HPM generator, 158 Three-phase HPM generator system studies
HPM machine, 153 synchronous inductance and
ICEs, 153, 156 rectifier, impact on, 124
PM output voltage, 153 Three-phase HPM generator systems, 96,
ZEBRA battery (see ZEBRA battery 98–100
technology) Three-phase HPM machine, 48, 73, 75
Series hybrid electric vehicle Three-phase HPM machine terminal AC
powertrain, 162 voltage, 154
SHEV DC-link voltage, 158 Three-phase stator winding, 118
SHEV fuel consumption, 160 TM power electronic converter, 154
SHEV powertrain configuration, 160 Toroidal-stator transverse-flux machine
SHEV powertrain topology, 159 (TSTFM), 21–25
Simplified hybrid EV powertrain Torque prediction and saturation, 69, 71, 72
configurations, 148, 150 Torque waveform harmonics prediction,
Simplified powertrain configuration, 145 103–106
SimPower, 141 Total transmission gear ratio, 164
SimPower model, 130, 134 Traction machine (TM), 154
226 Index

Turn-off losses, 136, 138 W

Two driving cycles WF machine, 61, 62, 69, 72, 82
energy loss prediction for, 187, 188, 190 rotor design, 62–64
WF rotor designs, comparative analysis
of, 65–70
U WF to PM split ratio, 64, 65
US and European standard driving cycles, 152 WF rotor designs
comparative analysis of, 65–69
WF synchronous machine excitation
V fields, 13
Vacuum pressure impregnation (VPI), 113 magnetic flux path representation, 13, 14
Vehicle DC-link voltage, 173 WF synchronous machines
Vehicle driving cycle magnetic flux path representation
AVL PEMS, 151 of, 13, 14
fuel consumption, 151
NEDC, 151
Vehicle linear velocity, 165 Z
Vehicle traction machine torque ZEBRA battery technology, 182
angular velocity, 165 BMI, 155
driving cycle, 165 characteristics, 156
electric machine topologies, 163 commercial, 162
equation, 163, 164 comparable specific energy, 155
forces, 163, 164 Na-Ni-Cl battery technology, 146
motor control, 163 operation regions, 155
Newton’s second law, 163 parameters, 155
output torque, 164 SOC, 155, 156
three-phase brushless DC motor, 163 specific energy, 146
vehicle linear velocity, 165 specifications, 155
Voltage source converter (VSC), 163 tests, 155