Gallardo Sánchez - César Tesis

University of Concepción
Department of Electrical Engineering

Doctorate in Engineering Sciences with a major in Electrical Engineering
César Gallardo Sánchez
DEVELOPMENT OF DESIGN TECHNIQUES TO

OPTIMIZE THE PERFORMANCE OF SYNCHRONOUS
RELUCTANCE MACHINES WITH ANISOTROPIC
ROTOR STRUCTURE
Thesis for the degree of Doctor of Science at the

University of Concepcion, Concepción, Chile
cgallardos@udec.cl
i
Supervisor: Professor Juan A. Tapia Ladino

Department of Electrical Engineering
University of Concepción, Concepción
Chile
Professor Michele Degano

Department of Electrical and Electronic Engineering
University of Nottingham, Nottingham
United Kingdom
January, 2024
© 2023 by César Gallardo Sánchez is licensed under CC BY 4.0
ii
Abstract
César Gallardo Sánchez

Title: “Development of Design Techniques to optimize the performance of Synchronous Reluctance
Machines with Anisotropic Rotor Structure”
Concepción 2024
82 pages
This thesis provides a comprehensive understanding of Synchronous Reluctance Machines
(SynRM), addressing both the fundamental principles of operation and the development of various
design techniques. To expedite the sizing stage, a precise analytical model was developed which
combines two methods: to calculate the air-gap flux density and average torque, the magnetic
potential of the rotor and stator were used, and the torque ripple was calculated using the energy
stored in the air-gap. This model is extended to machines with multiple flux barriers. Comparisons
with the Finite Element Analysis (FEA) yield promising results, both in terms of air-gap flux density
and electromagnetic torque. However, a harmonic analysis reveals that the analytical model tends
to overestimate the air-gap flux density and torque due to underlying assumptions made during its
development. Despite this, the model offers valuable capabilities, including the ability to extract
machine parameters in the d-q reference frame, facilitating a preliminary control strategy analysis.
In pursuit of further enhancing SynRM performance, an asymmetric rotor topology was
introduced. This feature achieves a significant reduction in torque ripple and an increase in
maximum internal power factor. While both designs exhibit similar efficiency, the asymmetric design
excels by offering a wider constant power speed range (CPSR). Simultaneously, a comprehensive
study on discrete skew methodology was conducted. The proposed method provides deeper
understandings into the impact of skew angle on torque ripple. This method introduces an indicator
to assess the potential reduction achievable by selecting various skew steps and angles. Validation
through FEA in both two and three dimensions reinforces its applicability.
These techniques were used to design two SynRMs using the same stator with two different
rotor topologies, one symmetrical and the other asymmetrical. The design was carried out by
optimization using a multi-objective genetic algorithm (MOGA), coupled with the techniques
developed throughout the thesis. The preliminary results highlight the superiority of the asymmetric
design in terms of performance indices. However, it is worth noting a significant reduction in these
indices when skew is applied, underscoring the importance of design choices. These two rotor
topologies were manufactured to validate the improvement through experimental measurements.
Keywords: analytical model, asymmetric rotor, design guidelines, electromagnetic torque,

finite element analysis, step skew, saliency ratio, synchronous reluctance machine, tolerances.
iii
Acknowledgements
I would like to express my gratitude to my supervisors, Professor Juan A. Tapia, and Professor
Michele Degano, for giving me the opportunity to work on this project. I have been collaborating
with them since 2019, and I thank them for their support and guidance on some difficult tasks over
these years.
I also want to thank the LEME staff for their valuable support, including Carlos Madariaga
for his results discussions and invaluable help with reading and correcting my grammar in several
papers.
I would like to offer my acknowledgements to the PEMC group at The University of
Nottingham for their warm welcome at the time I stay there during my internship.
I am grateful to my parents and family for their love and sacrifice to support my education. I
also appreciate the support of our friends here and at home.
I would like to express my love and gratitude to my wife, Yenisley, for her infinite support and
patience during these challenging years. It was difficult being apart, but her courage and
responsibility taught me how to be a better professional. I love you.
Lately, I would like to express my thanks to the Agencia Nacional de Investigación y
Desarrollo (ANID) and the University of Concepcion for provided financial support for my studies
in part by grant ANID-PFCHA/Doctorado Nacional/2020-21200527, project FONDECYT REGULAR
1230670, project FONDEF ID21I10099 and the Fondo de Internacionalización UCO 1866.
CONTENTS
ABSTRACT ......................................................................................................................................................................II
ACKNOWLEDGEMENTS .......................................................................................................................................... III
LIST OF FIGURES ........................................................................................................................................................ VI
LIST OF TABLES............................................................................................................................................................. X
NOMENCLATURE AND ACRONYMS ................................................................................................................... XI
INTRODUCTION ............................................................................................................................................................ 1
I. HYPOTHESIS............................................................................................................................................................ 1
II. OBJECTIVES ............................................................................................................................................................. 2
III. OUTLINE OF THE DOCTORAL DISSERTATION .......................................................................................................... 2
IV. CONTRIBUTIONS OF THE DOCTORAL DISSERTATION .............................................................................................. 3
V. LIST OF PUBLICATIONS AND FUNDINGS.................................................................................................................. 3
CHAPTER I: SYNCHRONOUS RELUCTANCE MACHINE: AN OVERVIEW. .................................................. 5
1.1. OPERATING PRINCIPLE ....................................................................................................................................... 5
1.2. ELECTROMAGNETIC TORQUE .......................................................................................................................... 6
1.3. INTERNAL POWER FACTOR AND POWER FACTOR ................................................................................... 8
1.4. CONTROL STRATEGY ........................................................................................................................................... 9
1.5. INFLUENCE OF SATURATION AND CROSS-COUPLING EFFECT ......................................................... 11
1.6. EFFICIENCY ............................................................................................................................................................ 13
1.7. DESIGN TECHNIQUES. MACHINE PERFORMANCE IMPROVEMENT ................................................ 15
1.7.1. Analytical models vs. Finite elements methods ............................................................................................... 15

1.7.2. Rotor structure optimization employing asymmetries .................................................................................... 16
1.7.3. Rotor skewing .................................................................................................................................................. 17
1.8. SUMMARY .............................................................................................................................................................. 19
CHAPTER II: ACCURATE ANALYTICAL MODEL FOR SYNCHRONOUS RELUCTANCE MACHINE .. 20
2.1. MAIN ASSUMPTIONS ......................................................................................................................................... 20
2.2. AIR-GAP FLUX DENSITY CALCULATION ..................................................................................................... 21
2.3. TORQUE CALCULATION .................................................................................................................................... 24
2.4. EXTENSION OF THE METHOD TO MACHINES WITH A LARGER NUMBER OF FLUX BARRIERS
PER POLE ........................................................................................................................................................................ 28
2.5. PARAMETERS ON THE D-Q REFERENCE FRAME....................................................................................... 32
2.6. SUMMARY .............................................................................................................................................................. 34
CHAPTER III: ROTOR ASYMMETRIC IMPACT ON SYNCHRONOUS RELUCTANCE MACHINE

PERFORMANCE ............................................................................................................................................................ 35
3.1. MODELLING OF THE SYNRMS UNDER ANALYSIS ................................................................................... 35
3.1.1. Sizing method .................................................................................................................................................. 35

3.1.2. Optimization process ....................................................................................................................................... 36
3.1.3. Optimum Designs ........................................................................................................................................... 38
3.2. RESULTS ANALYSIS ............................................................................................................................................. 39
3.2.1. Saliency ratio ................................................................................................................................................... 39

3.2.2. Maximum internal power factor ..................................................................................................................... 40
3.2.3. Electromagnetic torque .................................................................................................................................... 40
3.2.4. Torque ripple ................................................................................................................................................... 41
3.2.5. Additional discussion on efficiency and CPSR ............................................................................................... 42
3.2.6. Sensitive analysis ............................................................................................................................................ 44
3.3. SUMMARY .............................................................................................................................................................. 45
CHAPTER IV: A METHOD TO DETERMINE THE TORQUE RIPPLE HARMONIC REDUCTION IN

SKEWED SYNCHRONOUS RELUCTANCE MACHINES ................................................................................... 46
4.1. SELECTED MACHINES ........................................................................................................................................ 46
4.2. ANALYTICAL METHOD DERIVATION FOR DISCRETE SKEWING ....................................................... 47
4.3. FINITE ELEMENT VALIDATION: RESULTS AND DISCUSSION ............................................................. 49
4.3.1. FEA evaluation original designs (skewless machines) .................................................................................... 50

4.3.2. Evaluation of torque ripple reduction by means of two-step discrete skew ..................................................... 51
4.3.3. Evaluation of torque ripple reduction by means of three-step discrete skew ................................................... 52
4.4.4. Evaluation of torque ripple reduction by means of four-step discrete skew..................................................... 54
4.4. COMPARISON, ANALYSIS, AND RECOMMENDATIONS ........................................................................ 56
4.5. SUMMARY .............................................................................................................................................................. 58
CHAPTER V: PROPOSED DESIGN OF A SYNCHRONOUS RELUCTANCE MACHINE ............................ 59
5.1. SET-UP OF THE OPTIMIZATION PROCESS .................................................................................................. 59
5.2. RESULTS FROM THE OPTIMIZATION PROCESS ....................................................................................... 61
5.3. STRUCTURAL ANALYSIS ................................................................................................................................... 63
5.4. OPTIMAL DESIGN ADJUSTMENT................................................................................................................... 64
5.5. IMPACT OF ROTOR SKEWING ON THE PERFORMANCE OF THE MACHINE .................................. 75
5.6. MANUFACTURED PROTOTYPE ....................................................................................................................... 77
5.7. SUMMARY .............................................................................................................................................................. 78
6. CONCLUSIONS AND FUTURE WORK ............................................................................................................... 80
6.1. CONCLUSION ........................................................................................................................................................ 80
6.2. FUTURE WORK ...................................................................................................................................................... 81
REFERENCES ................................................................................................................................................................. 82
vi
List of Figures
Figure 1.1. Sketch representation of the reluctance principle. (a) isotropic object; (b) anisotropic object.
Figure 1.2. Sketch of a recent geometry for a 4-pole SynRM.
Figure 1.3. Phasor diagram for a SynRM at steady-state considering losses.
Figure 1.4. Influence of the current and load angle on the electromagnetic torque for a SynRM. (a) current
angle vs. electromagnetic torque; (b) load angle vs. electromagnetic torque.
Figure 1.5. Influence of the current and load angle on the IPF for a SynRM. (a) current angle vs. power
factor; (b) load angle vs. power factor.
Figure 1.6. Maximum power factor for a SynRM for different values of saliency ratio.
Figure 1.7. Behavior of the power factors in a six pole-three flux barriers SynRM analyzed through FEA.
Figure 1.8. Machine operating points under vector control techniques.
Figure 1.9. Influence of the current and the current angle on the machine inductances in a SynRM. (a)
material with infinite permeability; (b) behavior of the material according to B-H curve; (c)
M350-50A B-H curve; (d) mean flux density in the machine, six points were considered:
barrier carrier, radial bridge, two points in the stator yoke and two in the teeth. The analysis
was carried out in FEA.
Figure 1.10. Core loss distribution in a 6-pole SynRM for low speed and 5 A/mm2.
Figure 1.11. Influence of the current and load angles on the efficiency in a SynRM. (a) current angle vs.
efficiency; (b) load angle vs. efficiency.
Figure 1.12. 3D sketch of different types of skew. (a) reference skewless rotor; (b) continuous skew; (c)
discrete skew, also mentioned as step skew.
Figure 2.1. Sketch of a general rotor geometry for a SynRM.
Figure 2.2. Stator and rotor magnetic potential in the air-gap.
Figure 2.3. Flux barrier geometry and rotor magnetic potential in a rotor pole.
Figure 2.4. Magnetic potential and air-gap flux density without slotting effect for 𝜃𝑚 = 0. (a) stator and
rotor magnetic potential; (b) air-gap flux density.
Figure 2.5. Air-gap flux density with slotting effect for 𝜃𝑚 = 0. (a) air-gap flux density waveform; (b)
harmonic distribution of the air-gap flux density.
Figure 2.6. Electromagnetic torque waveform for two torque ripple periods with slotting effect. (a) torque
waveform; (b) torque harmonic components.
Figure 2.7. Harmonic torque analysis when considering slotting effect for two torque ripple periods.
Figure 2.8. Electromagnetic torque waveform for two torque ripple periods with slotting effect. (a) torque
waveform; (b) torque harmonic components.
Figure 2.9. Analytical method workflow.
Figure 2.10. Air-gap flux density waveform for 𝜃𝑚 = 0 for different rotor configuration and the harmonic
components. (a, f) two barriers per pole; (b, g) three barriers per pole; (c, h) four barriers per
pole; (d, i) five barriers per pole; (e, j) six barriers per pole torque ripple periods.
Figure 2.11. Electromagnetic torque waveform for different rotor configuration and the harmonic
components. (a, f) two barriers per pole; (b, g) three barriers per pole; (c, h) four barriers per
pole; (d, i) five barriers per pole; (e, j) six barriers per pole torque ripple periods.
Figure 2.12. Winding function for a four pole SynRM.
Figure 2.13. Flux linkage for a four pole SynRM. (a) flux linkage for the abc reference frame; (b) flux linkage
for the d-q reference frame.
Figure 3.1. Machine topology. a) stator and rotor; b) rotor parametrization.
Figure 3.2. Optimization process workflow.
Figure 3.3. Designs obtained from the optimization process for a 4-pole machine with two flux barriers
per pole. The current was fixed to ~10 A/mm2 and the current angle was defined in the
vii
optimization algorithm for MTPA. (a) symmetric design; (b) asymmetric design. The selected
designs are highlighted on the figure.
Figure 3.4. Saliency ratio as a function of the current density and the current angle. (a) symmetric design
counter-clockwise; (b) asymmetric design counter-clockwise; (c) symmetric design clockwise;
(b) asymmetric design clockwise.
Figure 3.5. Internal power factor as a function of the current density and the current angle. (a) symmetric
design counter-clockwise; (b) asymmetric design counter-clockwise; (c) symmetric design
clockwise; (b) asymmetric design clockwise.
Figure 3.6. Electromagnetic torque as a function of the current density and the current angle. MTPA
trajectory is highlighted in red. (a) symmetric design counter-clockwise; (b) asymmetric
design counter-clockwise; (c) symmetric design clockwise; (b) asymmetric design clockwise.
Figure 3.7. Torque ripple as a function of the current density and the current angle. MTPA trajectory is
highlighted in red. (a) symmetric design counter-clockwise; (b) asymmetric design counter-
clockwise; (c) symmetric design clockwise; (b) asymmetric design clockwise.
Figure 3.8. Electromagnetic torque and power factor as a function of the current density and the current
angle for counter-clockwise rotation. (a) asymmetric design; (b) symmetric design.
Figure 3.9. Efficiency map when the machine is operated a MTPA for a maximum current density of 5
A/mm2 for counter-clockwise rotation. (a) symmetric design; (b) asymmetric design.
Figure 3.10. Percentage of contribution of each term to the global variance. Influence of barriers position
and opening on torque ripple for both designs.
Figure 4.1. 3D sketches of (a) four-pole synchronous reluctance motor with two barriers per pole; (b) six-
pole synchronous reluctance motor with two barriers per pole. Three-phase stator windings
are highlighted in red, green, and blue, corresponding to each phase. Rotor in both machines
is shown as an exploded view.
Figure 4.2. Schematics of the proposed methodology to calculate the skew angle 𝜃skew to mitigate a
selected harmonic of order 𝑤. 2 step and 3-step discrete skew are presented, although the
method can be used for any number of slides.
Figure 4.3. Electromagnetic torque waveform and spectrum of skewless reference machine; (a) torque
waveform of 4-pole machine with two barriers per pole; (b) torque spatial harmonic content
of 4-pole machine with two barriers per pole; (c) torque waveform of 6-pole machine with two
barriers per pole; (d) torque spatial harmonic content of 6-pole machine with two barriers per
pole. Additionally, 2D and 3D simulation results are compared.
Figure 4.4. Comparison of electromagnetic torque waveform and harmonic content of skewless machine
vs 2-step skewed machine; (a) torque waveform of 4-pole machine with two barriers per pole;
(b) torque spatial harmonic content of 4-pole machine with two barriers per pole; (c) torque
waveform of 6-pole machine with two barriers per pole; (d) torque spatial harmonic content
of 6-pole machine with two barriers per pole. 2D and 3D simulation results are compared.
viii
Figure 4.7. Flux density distribution in a 6-pole machine with two barriers per pole for 𝐽=10 A/ mm2 and
𝜃𝑟,𝑒=0. (a) skewless machine; (b) 2-step skew machine; (c) 3-step skew machine; (d) 4-step
skew machine.
Figure 5.1. Sketch of the rotor of a SynRM with three flux barriers per pole. The parameters subject to
optimization are properly defined on both sides of the q-axis of the rotor.
Figure 5.2. Designs obtained from the optimization process for a 4-pole machine with three flux barriers
Figure 5.4. Behavior of the average torque and torque ripple in the symmetrical design for different radii
of the duct located in the first rotor island when 𝛼𝑖𝑒 = 60° elect. degree. (a) average torque;
(b) torque ripple.
of the duct located in the first rotor island when 𝛼𝑖𝑒 = 60° elect. degree. (a) average torque;
(b) torque ripple.
Figure 5.6. Von Mises stress distribution for the optimal designs at 10,000 rpm. (a) symmetric design; (b)
asymmetric design. Graphical results are magnified x10 for easy viewing.
Figure 5.7. Total deformation distribution for the optimal designs at 10,000 rpm. (a) symmetric design;
(b) asymmetric design. Graphical results are magnified x10 for easy viewing.
Figure 5.8. Saliency ratio for the symmetric and asymmetric design. (a) symmetric design counter-
clockwise rotation; (b) asymmetric design counter-clockwise rotation; (c) symmetric design
clockwise rotation; (b) asymmetric design clockwise rotation.
Figure 5.9. Internal power factor for the symmetric and asymmetric design. (a) symmetric design
counter-clockwise rotation; (b) asymmetric design counter-clockwise rotation; (c) symmetric
design clockwise rotation; (b) asymmetric design clockwise rotation.
Figure 5.10. Torque ripple for the symmetric and asymmetric design. (a) symmetric design counter-
Figure 5.11. Mean torque for the symmetric and asymmetric design. (a) symmetric design counter-
Figure 5.12. Harmonic components of the electromagnetic torque for the optimum symmetric and
asymmetric design counter-clockwise rotation; (a) current density 5 A/mm2; (b) current
density 7.5 A/ mm2; (c) current density 10 A/ mm2; (d) current density 12.5 A/ mm2; (e) current
density 15 A/ mm2.
asymmetric design clockwise rotation; (a) current density 5 A/mm2; (b) current density 7.5 A/
mm2; (c) current density 10 A/ mm2; (d) current density 12.5 A/ mm2; (e) current density 15 A/
mm2.
asymmetric design applying four step skew and counter-clockwise rotation; (a) current
density 5 A/mm2; (b) current density 7.5 A/ mm2; (c) current density 10 A/ mm2; (d) current
density 12.5 A/ mm2; (e) current density 15 A/ mm2.
asymmetric design applying four step skew and clockwise rotation; (a) current density 5
ix
A/mm2; (b) current density 7.5 A/ mm2; (c) current density 10 A/ mm2; (d) current density 12.5
A/ mm2; (e) current density 15 A/ mm2.
Figure 5.16. Saliency ratio for the symmetric and asymmetric design with four step skew is applied. (a)
symmetric design counter-clockwise rotation; (b) asymmetric design counter-clockwise
rotation; (c) symmetric design clockwise rotation; (b) asymmetric design clockwise rotation.
Figure 5.17. Internal power factor for the symmetric and asymmetric design with four step skew is
applied. (a) symmetric design counter-clockwise rotation; (b) asymmetric design counter-
clockwise rotation; (c) symmetric design clockwise rotation; (b) asymmetric design clockwise
rotation.
Figure 5.18. Torque ripple for the symmetric and asymmetric design with four step skew is applied. (a)
Figure 5.19. Mean torque for the symmetric and asymmetric design with four step skew is applied. (a)
Figure 5.20. Superposition of the contour curves of the main performance indices over a certain range of
current density and current angle with counter-clockwise rotation. Two operating points were
defined for MTPA for 7.5 A/mm2 and 10 A/mm2. (a) symmetric design; (b) asymmetric design.
current density and current angle with clockwise rotation. Two operating points were defined
for MTPA for 7.5 A/mm2 and 10 A/mm2. (a) symmetric design; (b) asymmetric design.
Figure 5.22. Symmetric and asymmetric design electromagnetic torque waveform for MTPA when the
current density is 7.5 A/mm2 and 10 A/mm2 and the machine rotated counter-clockwise. (a)
skewless designs; (b) 5-step skew designs.
Figure 5.23. Magnetic flux density distribution for the optimal designs for MTPA at 7.5 A/mm2. (a)
rotation.
Figure 5.24. Magnetic flux density distribution for the optimal design for MTPA at 10 A/mm 2. (a)
rotation.
Figure 5.25. Performance indices for the symmetric design for MTPA for different current density levels.
(a) current angle; (b) saliency ratio; (c) maximum internal power factor; (d) average torque;
(e) torque ripple; (f) efficiency.
Figure 5.26. Performance indices for the asymmetric design for MTPA for different current density levels.
(a) current angle; (b) saliency ratio; (c) maximum internal power factor; (d) average torque;
(e) torque ripple; (f) efficiency.
Figure 5.27. CPSR performance for the symmetric and asymmetric design for maximum current density
of 10A/mm2. (a) torque vs. speed for the symmetric design; (b) power vs. speed for the
symmetric design; (c) torque vs. speed for the asymmetric design; (d) power vs. speed for the
asymmetric design.
Figure 5.28. Efficiency map when the machine is operated a MTPA for a maximum current density of
10A/mm2. (a) skewless symmetric design; (b) skewless asymmetric design; (c) 5-step skew
symmetric design; (d) 5-step skew asymmetric design.
Figure 5.29. Exploded view of the motor assembly. (a) asymmetric rotor; (b) asymmetric rotor; (c) stator
lamination and winding assembled in the housing (jacket water); (d) shaft.
x
List of tables
Table 1.1. Output parameters for a case study 6-pole SynRM through FEA.
Table 2.1. Machine main parameters.
Table 2.2. Average torque and torque ripple values for harmonic analysis.
Table 2.3. Average torque and torque ripple for different rotor configurations.
Table 2.4. Mean torque comparison between FEM and AM.
Table 3.1. Design parameters
Table 3.2. Input variables range.
Table 3.3. Geometric values for optimal symmetrical and asymmetrical designs.
Table 3.4. Performance indices for the operating points defined in Figure 3.8 and Figure 3.9 for the
symmetrical and asymmetrical designs when the machine rotated counter-clockwise.
Table 4.1. Main data of the selected machines.
Table 4.2. Skew angle to reduce a specific electromagnetic torque harmonic order by discrete skew.
Table 4.3. Torque ripple harmonic component reduction as a result of 2-step skewing. 3D results are
considered, and relevant harmonic components are analyzed.
Table 4.4. Average torque and torque ripple comparison when applying 2-step skewing, by means of
2D and 3D FEA simulations.
considered and relevant harmonic components are analyzed.
Table 5.1. Main data of the machine subject to optimization.
Table 5.2. Objective function for the optimization process.
Table 5.3. Input variables range for the optimization process for 4-pole and 6-pole SynRM.
Table 5.4. Skew angle to reduce a specific harmonic component of the electromagnetic torque.
Table 5.5. Value of the mitigation factor when the first harmonic component is selected to reduce (𝑤 =
6).
Table 5.6. Value of the mitigation factor when the second harmonic component is selected to reduce (𝑤
= 12).
Table 5.7. Value of the mitigation factor when the third harmonic component is selected to reduce (𝑤 =
18).
Table 5.8. Performance indices for the operating points defined in Figure 6.20 for the symmetrical and
asymmetrical designs when the machine rotated counter-clockwise.
asymmetrical designs when the machine rotated clockwise.
xi
Nomenclature and acronyms
Latin alphabet
𝐵g Flux density in the air-gap
𝑏𝑠𝑠 Slot opening
𝑏t Tooth width
𝐷ri Rotor inner diameter
𝐷ro Rotor outer diameter
𝐷si Stator inner diameter
𝐷so Stator outer diameter
𝑔 Air-gap length
𝑔r Air-gap function
ℎt Tooth height
𝑖d d-axis current
𝑖q q-axis current
𝐼rms Supply current
𝑖s Stator current
𝐽 Current Density
𝑘air Insulation ratio
𝐾p Ratio between the barrier angles
𝑘rs,𝑣 Skew mitigation factor for harmonic v
𝐾s Electric loading
𝑘w Winding factor
𝐾w Ratio between the barrier opening angles
𝑙axial Axial length
𝐿aa a-phase self-inductance
𝐿ac a-phase and c-phase mutual-inductance
𝐿ab a-phase and b-phase mutual-inductance
𝐿bb b-phase self-inductance
𝐿ba b-phase and a-phase mutual-inductance
𝐿bc b-phase and c-phase mutual-inductance
𝐿cc c-phase self-inductance
𝐿ca c-phase and a-phase mutual-inductance
𝐿cb c-phase and b-phase mutual-inductance
𝐿d d-axis self-inductance
𝐿q q-axis self-inductance
𝐿dq d-axis and q-axis mutual-inductance
𝐿qd q-axis and d-axis mutual-inductance
𝑙st Stack length
𝑚 Number of phases
𝑁i Winding Function for “i" phase
𝑁s Turns per slot
𝑁t Turns per pole per phase
𝑃 Air-gap permeance
𝑝 Pole pair
Pcu Joule losses
Pedd+hys+add Core losses (eddy-current, hysteresis and additional)
𝑞 Number of slots
xii
𝑅 Air-gap radius
ℛb Reluctance
Rs Stator resistance
𝑇avg Mean torque
𝑇em Electromagnetic torque
𝑇rp Torque ripple
𝑈s Stator magnetic potential
𝑈r Rotor magnetic potential
𝑣d d-axis voltage
𝑣q q-axis voltage
𝑣s Stator voltage
𝑊b Energy stored in the flux barrier
𝑊g Energy stored in the air-gap
𝑊m Magnetic energy
Greek alphabet
𝛼ie Current angle
δ Load angle
𝜃s Position along the stator periphery
θskew Skew angle
𝜃m Rotor periphery
𝜃r Rotor position
𝜃s Stator periphery
𝜆d d-axis flux linkage
𝜆q q-axis flux linkage
𝜆i Flux linkage for “i" phase
𝜇0 Air permeability
𝜉 Saliency ratio
𝜏p Polar pitch
𝜏s Slot pitch
∅b Magnetic flux
𝜔 Electrical speed
Abbreviations
2D Two-dimensional
3D Three-dimensional
AM Analytical Model
CF Constant Flux
CPSR Constant Power Speed Region
DC Direct Current Machine
FEA Finite Element Analysis
FEM Finite Element Method
FW Flux Weakening
IPF Internal Power Factor
IPFmax Maximum Internal Power Factor
IM Inductor Machine
IPM Interior Permanent Magnet Machine
MMF Magnetic Motive Force
MOGA Multi-Objective Genetic Algorithm
MTPA Maximum Torque per Ampere
MTPV Maximum Torque per Volt
xiii
PM Permanent Magnet
PMa-SynRM Permanent Magnet assisted Synchronous Reluctance Machine
PMSM Permanent Magnet Synchronous Machine
SRM Swishes Reluctance Machine
SynRM Synchronous Reluctance Machine
1
Introduction
Interest in Synchronous Reluctance Machines (SynRMs) is growing due to their many

advantages over other machine types. SynRMs are single-excited machines that offer acceptable
torque density [1], [2], high efficiency [3], [4], high‐speed suitability [5], [6], fault tolerance capability
[7], [8], and low cost [9], [10]. This makes them a promising alternative to induction machines (IM)
[10] and permanent magnet synchronous machines (PMSM) [9], [10]. The absence of rotor windings
or rotor cage in SynRMs eliminates rotor ohmic losses, leading to lower temperature and higher
efficiency compared to IMs [11]. In addition, the absence of permanent magnets (PMs) simplifies the
manufacturing process and reduces the cost compared to PMSMs.
SynRMs are used in various applications and are emerging as a substitute for induction motors
in industry. However, their use is also beginning to be studied in the field of vehicle electrification.
Nevertheless, these comparatives advantages come at the cost of two critical drawbacks: high torque
ripple [12], [13] and low power factor [9], [13]. One of the main contributors for the large torque
ripple in SynRM is the interaction of spatial harmonics of the magnetic motive force (MMF)
generated by stator currents and rotor geometry [14]. High torque ripple can lead to undesired
mechanical vibrations and potential acoustic noise, as well as impact the current harmonics.
Moreover, in high-performance applications, low torque ripple is strictly required [15] and,
consequently, different techniques have been developed to reduce its magnitude as much as possible
such as: (i) modify and optimize the winding configuration [15], [16], and (ii) optimize the rotor
structure, in terms of geometry and dimensions [17], [18]; i.e.: using skewed rotor, adopting two
different flux-barrier geometries in the same lamination and asymmetries rotor structures.
Despite these two drawbacks, SynRMs have continued to attract interest for two main reasons:
(i) the increasing cost of rare earth permanent magnets and (ii) the increasing demand for high-
efficiency motors. However, considering the main requirements of different applications, these
machines topology are under intense research and development work at the moment. This work
aims to further the research and development of SynRM by developing new design techniques. The
work will focus on modifying the rotor structure by incorporating asymmetrical flux barriers and
see how these modify the performance of the machine. Other modifications to the rotor were also
evaluated to reduce the torque ripple, such as a detailed skew technique. An accurate analytical
model will be developed for a fast-sizing stage which incorporate the slotting effect in the torque
waveform. The studies carried out will serve as design guidelines for SynRM, reducing design times
and improving machine performance.
I. Hypothesis
The hypothesis that was tested during this doctoral thesis is:
An asymmetric rotor structure, which features a different position of the flux barriers relative to the
rotor's q-axis, in a Synchronous Reluctance Machine can increase magnetic anisotropy in the rotor structure.
This improvement enables an increase in saliency ratio, power factor, efficiency, and average torque, while also
reducing torque ripple in the machine operation.
2
II. Objectives
General objective:
❖ Enhance the performance of Synchronous Reluctance Machines through the
development of design techniques and guidelines focused on increasing rotor
anisotropy.
Specific objectives:
1. Develop and validate an accurate analytical model for Synchronous Reluctance
Machines.
2. Investigate the impact of rotor asymmetries on the performance of Synchronous
Reluctance Machines
3. Analyze and compare the performance of Synchronous Reluctance Machines with
enhanced rotor anisotropy by including asymmetries against traditional designs under
various operating conditions.
4. Validate the practical feasibility of the optimized Synchronous Reluctance Machines
designs with rotor asymmetries through Finite Element Analysis and experimental
testing and validation.
III. Outline of the doctoral dissertation
This doctoral dissertation focuses on the different design techniques for SynRM we put some
efforts in apply some asymmetric features on the rotor structure.
• Chapter I provides a brief overview of the basic theory, development history, model, vector
diagram, and main characteristics of the SynRM. The non-linear behavior is also discussed. The
chapter concludes by discussing the different design technique to improve the performance of
SynRM.
• Chapter II presents the linear analytical model of SynRM with multiples flux barriers per
rotor pole and the slotting effect is take into account. This model is based on equivalent lumped-
parameter magnetic network and aims to compute stator and rotor magnetic potential, air-gap flux
density and electromagnetic torque.
• Chapter III explores the benefits of asymmetrical rotor geometry. It analyzes various
performance indices across a range of operating points. The behavior of the machine is compared
between a symmetrical model and an asymmetrical model, both optimized through Finite Element
Method (FEM).
• Chapter IV introduces a generalized analytical expression for multi-step discrete skewing in
SynRM, examining its impact on torque harmonic components. It introduces a method to visualize
how to reduce specific harmonic content. Two SynRM designs are analyzed through finite element
analysis to compare the effectiveness of each approach and validate the optimal step skewing
predicted by the analytical formulation.
• Chapter V employs the methods discussed throughout the manuscript to design a SynRM
prototype through optimization using genetic algorithms and FEM simulations. The optimization
results in two designs, a symmetrical one and an asymmetrical one, which are compared in terms of
average torque, torque ripple, saliency ratio, and power factor. The optimal designs go through
further evaluation from a mechanical standpoint to ensure the rotor integrity.
3
IV. Contributions of the doctoral dissertation
The specific contributions of this Thesis are summarized as follows:

The Main Contribution of Chapter II
Development a fast and accurate analytical model to determine the air-gap flux density
distribution and the torque waveform for different combination of pole/barrier and take into account
the slotting effect.
Provide the equations to obtain from the air-gap flux density the flux linkages in the q-d
reference frame.
The Main Contribution of Chapter III
A comprehensive study on the influence of the benefits and disadvantages of applying
asymmetries in the rotor structure of a SynRM for different operating points.
The Main Contribution of Chapter IV
Present a comprehensive analytical expression for multi-step discrete skewing in SynRM.
The method calculates a specific skew angle to eliminate a specific harmonic component and a
mitigation factor to showcase the impact of skewing on each harmonic component.
The Main Contribution of Chapter V
Design and manufacturing of 24kW prototype with two different rotors (symmetric and
asymmetric) for laboratory level examinations in order to validate the design methods at University
of Nottingham.
V. List of publications and fundings
This doctoral dissertation contains material from the following papers.

Journal publications
❖ C. Gallardo, J. A. Tapia, M. Degano and H. Mahmoud, "Accurate Analytical Model for
Synchronous Reluctance Machine With Multiple Flux Barriers Considering the Slotting
Effect," in IEEE Transactions on Magnetics, vol. 58, no. 9, pp. 1-9, Sept. 2022, Art no.
8107709.
❖ C. Gallardo, C. Madariaga, J. A. Tapia, and M. Degano, “A Method to Determine the
Torque Ripple Harmonic Reduction in Skewed Synchronous Reluctance Machines,” Applied
Sciences 2023, vol. 13, no. 5, pp. 2949, Feb. 2023.
❖ C. Madariaga, C. Gallardo, J. A. Tapia, W. Jara, A. Escobar, and M. Degano, "Fast
Assessment of Rotor Barrier Dimensional Allowances in Synchronous Reluctance Machines,"
in IEEE Access, vol. 11, pp. 58349-58358, 2023.
❖ C. Madariaga, C. Gallardo, N. Reyes, J. A. Tapia, W. Jara, and M. Degano, " Design and
Evaluation of Matrix Rotor Induction Motor for High-Torque Low-Speed Applications," in
IEEE Transactions on Energy Conversion, under review.
Conference publications
4
❖ C. Gallardo, J. A. Tapia, M. Degano, H. Mahmoud and A. E. Hoffer, "Rotor Asymmetry

Impact on Synchronous Reluctance Machines Performance," 2022 International Conference
on Electrical Machines (ICEM), pp. 848-854, 2022.
❖ C. Gallardo, C. Madariaga, V. Rodriguez-Merchan and J. A. Tapia, "Comparative
Analysis of a Synchro nous Reluctance and a Solid-Rotor Induction Machine for High-Speed
Applications," 2022 IEEE ANDESCON, 2022.
❖ C. Gallardo, C. Madariaga, J. A. Tapia and M. Degano “Effects of Rotor Asymmetries on
d-q Axis Parameters in Synchronous Reluctance Machines,” 2023 IEEE International Electric
Machines & Drives Conference (IEMDC), San Francisco, CA, USA, 2023.
❖ C. Madariaga, C. Gallardo, J. A. Tapia, W. Jara and D. Riquelme, "Quick Comparison of
the Cogging Torque Severity in Permanent Magnet Synchronous Machines with Segmented
Stator Core," 2022 IEEE International Conference on Automation/XXV Congress of the
Chilean Association of Automatic Control (ICA-ACCA), 2022.
❖ F. Ortiz, C. Gallardo, C. Madariaga and J. A. Tapia “Design and Evaluation of Axial-Flux
Permanent Magnet Machine with Enhanced Saliency,” 2023 IEEE International Electric
Machines & Drives Conference (IEMDC), San Francisco, CA, USA, 2023.
❖ N. Reyes, M. Jiménez, A. Hoffer, C. Madariaga, C. Gallardo, J. A. Tapia, W. Jara, A.
Escobar, and M. Degano, " Structural Feasibility and Electromagnetic Analysis of an Ironless
Rotor Axial Flux Permanent Magnet Synchronous Machine," 2023 IEEE Energy Conversion
Congress and Exposition (ECCE), Nashville, TN, USA, 2023.
❖ C. Gallardo, C. Madariaga, J. A. Tapia and M. Degano “Impact of Rotor Step Skew on the
Performance of Synchronous Reluctance Machines,” 2023 IEEE CHILEAN Conference on
Electrical, Electronics Engineering, Information and Communication Technologies,
Valdivia, Chile, 2023.
❖ C. Madariaga, C. Gallardo, J. A. Tapia, W, Jara, N. Reyes and F. Santacruz “Comparison
of Matrix-Rotor Induction Motor and Permanent Magnet Machine for Low-Speed High-Torque
Applications,” 2023 IEEE CHILEAN Conference on Electrical, Electronics Engineering,
Information and Communication Technologies, Valdivia, Chile, 2023.
❖ F. Santacruz, C. Madariaga, C. Gallardo and J. A. Tapia “Electromagnetic Sizing
Validation of Double Cage Induction Motor for Electric Vehicles Using Finite Element
Simulation,” 2023 IEEE CHILEAN Conference on Electrical, Electronics Engineering,
Information and Communication Technologies, Valdivia, Chile, 2023.
Fundings
❖ Fondo de Internacionalización UCO 1866, University of Concepcion.

❖ ANID-PFCHA/Doctorado Nacional/2020-21200527.
❖ VIU23P0053, "Diseño de máquinas eléctricas para aplicaciones industriales: software de
análisis y simulación para la formación de competencias profesionales"
5
Chapter I: Synchronous Reluctance Machine: an overview.
This chapter introduces an overview of the basic principles of operation for SynRMs. It covers
the mathematical model and includes the saturation and cross-coupling effect between the d-q axes.
Additionally, several design techniques that play a major role in improving performance during the
SynRM design phase are highlighted.
1.1. Operating principle
The first theoretical and technical introduction of a motor with reluctance torque production
and sinusoidal MMF using a conventional IM stator was made by Kostko in 1923 [19]. This led over
the years to the concept of a SynRM evolving from a simple rotor structure to more complex design
structures. However, it was not until the 1990s that it was demonstrated that the SynRM can be
controlled using vector control techniques [20]–[23] and thus improve its performance, making this
machine topology more attractive.
SynRMs develop electromagnetic torque on the shaft based on the variation of the reluctance.
In Figure 1.1 the object (a) has the same geometrical dimensions in all axes, so it is isotropic, when
exposed to a magnetic field it has the same reluctance in all axes. Then, the object (b) in Figure 1.1
has different dimensions in the axes (anisotropic), so when exposed to the same magnetic field a
reluctance difference appears its axes. As a result, an electromagnetic torque develops in the object
(b) which tends to align the d-axis to the position of less reluctance; this torque will be present as long
as there is an angular difference between the d-axis and the magnetic field to which the object is
exposed. The angle between the d-axis and the magnetic field produced by the stator is known as the
load angle (δ). The stator current is responsible for both magnetization (main field) and torque
production [21].
q d q d
λ
δ
Reluctance
torque
(a) (b)
Figure 1.1. Sketch representation of the reluctance principle. (a) isotropic object; (b) anisotropic object.
In general, the SynRM is similar to the conventional salient-pole synchronous motor, but it
does not have an excitation winding in the rotor structure. The SynRM rotor is composed of flux
carriers (electrical steel) and flux barriers (insulating material, usually air), as shown in Figure 1.2.
The main advantage of the SynRM lies in the absence of copper losses in the rotor, which allows a
higher continuous torque than other IMs of the same size [24], [25]. To meet these requirements, the
rotor geometry of a SynRM must be designed so that the inductance on the d-axis is maximum and
minimum on the q-axis. This ratio between the d-axis and q-axis inductances is known as the saliency
ratio (𝜉) and is calculated as
6
𝐿d
𝜉= . (1.1)
𝐿q
Stator
Reluctance windings
rotor
Flux barrier
Stator
Flux carrier
Figure 1.2. Sketch of a recent geometry for a 4-pole SynRM.
1.2. Electromagnetic torque

The design of an electrical machine must be such that it produces torque efficiently. In the
SynRM, the electromagnetic torque is produced by the interaction between the flux in the air gap
and their respective magnetizing current. Different expressions for the analytical calculation of the
electromagnetic torque in a SynRM were delivered when the saturation and cross-coupling effect is
neglected.
3
𝑇em = 𝑝(𝜆d 𝑖q − 𝜆q 𝑖d ) (1.2)
2
where 𝑖d , 𝑖q are the current in the d-axis and q-axis, respectively; and 𝜆d and 𝜆q are the flux linkage
in the d-axis and q-axis, respectively.
From the torque equations, it is clear that for a specified operating current and speed the
electromagnetic torque developed on the shaft depends on the saliency ratio (ξ); this reaching its
maximum value for a given load angle (𝛿) or current angle (𝛼𝑖𝑒 ). The vector diagram in the d-q
reference frame is shown in Figure 1.3 at steady-state, including the total iron losses, from which the
equations relating load angle and current angle can be derived.
𝜆q
δ = tan−1 ( ) (1.3)
𝜆d
1 𝑖q
δ = tan−1 ( ) (1.4)
𝜉 𝑖d
𝑖q
𝛼ie = tan−1 ( ) (1.5)
𝑖d
1 1
δ = tan−1 [ tan(𝛼ie )] 𝑜𝑟 tan(δ) = tan(𝛼ie ) (1.6)
𝜉 𝜉
7
q
V Vq
isRs
emf δ
is
ic
iq
im
φ
φi
λq λ
αie
δ
d
Vd id λd
Figure 1.3. Phasor diagram for a SynRM at steady-state considering losses.
In the torque equations the machine inductances and flux linkages are constant since saturation
is not considered, while the actual parameters depend on the current and the current angle. In fact,
the electromagnetic torque shown in (1.2) does not consider the variation of the co-energy with the
variation of the rotor position. Therefore, we would only obtain the average torque, it is more
accurate to obtain the torque waveform using FEM simulations [26].
Combining the relationship between the current angle and the load angle it is possible to
express the torque equation as a function of the load angle as
3𝑝
𝑇𝑒𝑚 = − 𝐿 (𝜉 − 1)𝑖s2 sin{2 tan−1 [𝜉 tan(δ)]}. (1.7)
22 q
Figure 1.4 shows the influence of current angle and load angle on the electromagnetic torque.
It can be observed that as the saliency ratio increases, the machine develops a higher torque value,
which is directly dependent on the current and load angles.
(a) (b)
Figure 1.4. Influence of the current and load angle on the electromagnetic torque for a SynRM. (a)
current angle vs. electromagnetic torque; (b) load angle vs. electromagnetic torque.
8
1.3. Internal Power Factor and Power Factor

The vector diagram in Figure 1.3 allows to identify the internal power factor (IPF) and the
terminal power factor (PF) of the machine. This characteristic can be expressed in different forms
[27]–[29].
𝜆𝑑 𝑖𝑞
+
𝜋 𝜆𝑞 𝑖𝑑
IPF = cos 𝜑𝑖 = cos ( + 𝛿 − 𝛼ie ) = cos tan−1 (1.8)
2 𝜉−1
[ ( )]
𝑃𝑖𝑛
PF = cos 𝜑 = (1.9)
𝑆𝑖𝑛
where 𝑃𝑖𝑛 is the real power and 𝑆𝑖𝑛 is the apparent power.
Figure 1.5 shows the influence of current angle and load angle on the IPF for an ideal case,
when the saturation and losses are neglected. It can be observed that as the saliency ratio increases,
the machine develops a higher IPF, which is directly dependent on the current and load angles. The
IPF is higher for higher current angles, but the same is not true for the load angle, where a higher IPF
is obtained for lower load angles.
(a) (b)
Figure 1.5. Influence of the current and load angle on the IPF for a SynRM. (a) current angle vs. power
factor; (b) load angle vs. power factor.
Equation (1.8) shows that IPF depends on the current angle and the magnetic saliency of the
machine. The maximum internal power factor (IPFmax ) can be expressed by (1.10) [28], confirming
that a higher saliency ratio ensures a higher IPF. It observed in the Figure 1.6 that to obtain an
acceptable power factor the saliency ratio must be greater than 10 [30]. In practice, it is almost
impossible to obtain such high saliency values, usually around ~5. Assisted permanent magnet
synchronous reluctance machines (PM-SynRMs) with ferrite magnets are emerging as an alternative
to achieve a suitable power factor.
𝜉−1
IPFmax = (1.10)
𝜉+1
9
Figure 1.6. Maximum power factor for a SynRM for different values of saliency ratio.
When considering all physical phenomena in the electromagnetic analysis, the IPFmax is not
constant, as depicted in Figure 1.7. The saliency changes due to saturation and cross-coupling effects,
leading to higher values for the IPFmax at larger current angles. IPFmax represents the limit for the PF
of the machine. In any operating point, the machine will exhibit a lower PF than the IPFmax .
Regarding to the IPF and PF, for any operation point the PF will be a slightly lower than the IPF due
the saturation and losses. In an ideal scenario, the PF would be equal to the IPF, but practical
considerations introduce some deviations.
Figure 1.7. Behavior of the power factors in a six pole-three flux barriers SynRM analyzed through
FEA.
1.4. Control strategy

SynRMs can be controlled by vector control techniques, for which it is necessary to know the
voltage and current equations in the d-q plane. The equations below are obtained by using the vector
diagram in Figure 1.3 as a reference and neglected losses.
𝑣d = −𝜔𝐿q 𝑖q = −𝜔𝜆q (1.11)
𝑣q = 𝜔𝐿d 𝑖d = 𝜔𝜆d (1.12)
2
𝑣s = 𝜔√(𝐿q 𝑖q ) + (𝐿d 𝑖d )2 (1.13)
𝑖d = 𝑖s cos 𝛼ie (1.14)
𝑖q = 𝑖s sin 𝛼ie (1.15)

10
𝑖d2 + 𝑖q2 = 𝑖s2 (1.16)
where 𝜔 is the electrical angular speed.

In addition, it is necessary to define the voltage and current limits, which are normally defined
by the converter or the insulation characteristics of the machine. The current limit is defined by (1.17)
as a circle with center at the origin, whose radius is defined by the current limit. On the other hand,
the voltage limit is defined as an ellipse centered at the origin, (1.18), where the length of its axes
decreases with increasing machine operating speed.
𝑖d2 + 𝑖q2 = 𝑖lim
2
(1.17)
2 𝑣lim 2
(𝐿q 𝑖q ) + (𝐿d 𝑖d )2 = ( ) (1.18)
𝜔
q
Speed MTPV
increase 1 p.u
Torque
increase
1 p.u
B MTPA
A
Constant torque
O d
Current
limited
Voltage
limited
Figure 1.8. Machine operating points under vector control techniques.
The literature shows that a SynRM can operate efficiently in different operating modes
according to speed and load [31]–[36], such as: maximum torque per ampere (MTPA) from zero
speed until the machine reaches rated speed, flux weakening (FW) at medium and high speed, and
maximum torque per volt (MTPV) at high speed. The operating points for any SynRM are shown in
Figure 1.8; where, with increasing speed the operating points move along the OABO line. At OA the
machine operates at MTPA, at AB it operates at FW and at BO it operates at MTPV.
11
1.5. Influence of saturation and cross-coupling effect

Magnetic saturation is normally not considered in the model when it is analyzed by the
equations in the d-q reference frame. However, it is necessary to consider this effect when it is desired
to implement a high-performance control strategy or maximizing the machine capabilities.
Saturation, resulting from the nonlinear characteristic of the ferromagnetic material, leads to the fact
that the inductances and flux linkages in the d-q axes are not constant, but rather a function of the
currents [33], [37]–[40] and can therefore be expressed as
𝜆d = 𝜆d (𝑖d , 𝑖q ) (1.19)
𝜆q = 𝜆q (𝑖d , 𝑖q ). (1.20)
There are two ways to calculate the flux linkage equations considering the influence of
saturation and cross-coupling effects. The first one and the more accurate is by means of FEA
simulations in order to obtain the flux linkage per phase and, subsequently, apply the Clarke-Park
transformation as follows.
2 2𝜋 4𝜋
𝜆d = [𝜆a cos(𝜃m ) + 𝜆b cos (𝜃m − ) + 𝜆c cos (𝜃m − )] (1.21)
3 3 3
2 2𝜋 4𝜋
𝜆q = [𝜆a sin(𝜃m )+𝜆b sin (𝜃m − ) + 𝜆c sin (𝜃m − )] (1.22)
3 3 3
where 𝜆a , 𝜆b and 𝜆c are the flux linkage of each phase, p is the pole pairs and 𝜃m is the angular
coordinate referring to the stator reference frame and is calculated as follow
𝜃m = 𝑝(𝜃r + 𝜃0 ) (1.23)
where 𝜃r is the rotor position and 𝜃0 is the initial rotor position.

Another way to obtain the flux linkage expressions for (1.19) and (1.20) if the inductance per
phase is known, is to consider the following transformation
𝐿dq0 = 𝑃(𝜃m )[𝐿abc (𝜃m )]𝑃(𝜃m )−1 , (1.24)
where
𝐿d 𝐿dq 𝐿d0
𝐿dq0 = [𝐿qd 𝐿q 𝐿q0 ], (1.25)
𝐿0d 𝐿0q 𝐿00
𝐿aa (𝜃m ) 𝐿ab (𝜃m ) 𝐿ac (𝜃m )

𝐿abc (𝜃m ) = [𝐿ba (𝜃m ) 𝐿bb (𝜃m ) 𝐿bc (𝜃m )], (1.26)
𝐿ca (𝜃m ) 𝐿cb (𝜃m ) 𝐿cc (𝜃m )
and
12
2𝜋 4𝜋
cos(𝜃m ) cos (𝜃m − ) cos (𝜃m − )
3 3
2𝜋 4𝜋
𝑃(𝜃m ) = sin(𝜃m ) sin (𝜃m − ) sin (𝜃m − ) . (1.27)
3 3
1 1 1
[ √2 √2 √2 ]
Considering the stationary reference frame, flux linkage expressions are

𝜆d 𝐿d 𝐿dq 𝐿d0 𝑖d
[𝜆dq0 ] = [𝜆q ] = [𝐿qd 𝐿q 𝐿q0 ] [𝑖q ]. (1.28)
𝜆0 𝐿0d 𝐿0q 𝐿00 𝑖0
It is assumed that the machine is being supplied by balanced three phase currents, reason why
the zero-sequence component of the stator currents is zero. Thus, the d-q axes flux linkage
expressions of (1.19) and (1.20) can be calculated as
𝜆d = 𝐿d 𝑖d + 𝐿dq 𝑖q , (1.29)
𝜆q = 𝐿qd 𝑖d + 𝐿q 𝑖q . (1.30)
From this perspective, and once the effects of saturation and cross-coupling are included in the
analysis, the torque equation (1.2) becomes
3
𝑇em = 𝑝[𝑳𝐝𝐪 (𝒊𝟐𝐪 − 𝒊𝟐𝐝 ) − (𝑳𝐪 − 𝑳𝐪 )𝒊𝐝 𝒊𝐪 ]. (1.31)
2
The first term is the torque produced because of the magnetic coupling between d-q axes
circuits and the second term is the reluctance torque. At low current the first term could be neglected
since the magnetic coupling between axis has a low influence on the machine.
Figure 1.9 shows the influence of the current and the current angle on the machine inductances.
It can be observed that the inductances when the permeability is infinite, the saturation is neglected,
remain constant. In this case, the saliency of the machine is greater than eight and remains constant
for every analyzed operating point. On the other hand, when saturation and the cross-coupling effect
are considered, the inductances tend to decrease when the phase current increase.
The saturation of the ferromagnetic material causes a change in the saliency ratio of the
machine. This is because the reluctance in the d-axis increases, resulting in a decrease in the d-axis
inductance until it approaches the value of the q-axis inductance. For a fixed current, the q-axis
inductance experiences a minor variation with the change in current angle. When the machine
operates past the knee of the B-H curve, the q-axis inductance is considered constant, while the d-
axis inductance decreases. However, the d-axis inductance experiences greater variations as the
current angle changes; at a higher current angle, the d-axis inductance increases. Thus, as the current
angle increases, the q-axis inductance remains constant, the d-axis inductance increases, and the
machine's saliency ratio increases. Operating the machine with a large current angle can improve its
performance. However, this approach moves the machine away from the MTPA trajectory and could
lead to increased torque ripple.
13
(a) (b)
(c) (d)
Figure 1.9. Influence of the current and the current angle on the machine inductances in a SynRM. (a)
material with infinite permeability; (b) behavior of the material according to B-H curve; (c) M350-50A B-H
curve; (d) mean flux density in the machine, six points were considered: barrier carrier, radial bridge, two
points in the stator yoke and two in the teeth. The analysis was carried out in FEA.
1.6. Efficiency
Under the assumption that joule losses in the stator windings are dominant among other losses
in the machine at low speeds, the efficiency can be calculated by [41], [42]
𝑃cu −1
𝜂 ≈ [1 + ] . (1.32)
𝜔𝑇em
This equation can be expanded with some modification as

−1
1
𝜂 ≈ [1 + ] (1.33)
𝜔 𝑇em
( 2 )
3𝑅s 𝑖s
𝑇em
where R s is stator winding resistance. It can be deduced from (1.33) that increasing the ratio of ( )
𝑖s2
tends to increase the machine efficiency. However, this is only valid when core losses and additional
losses are negligible compared to the stator resistive losses.
The torque generated by a SynRM is not constant due to space harmonics in the air gap, leading
to torque ripple and contributing to core losses. Maximum machine efficiency cannot be achieved
through optimization of rotor structure alone, as core losses occur both in the stator and rotor. The
distribution of core losses in a standard SynRM is shown in Figure 1.10, with maximum values near
the air-gap and minimal losses in the inside part of the rotor due to minimal changes in magnetic
flux density over time.
14
Figure 1.10. Core loss distribution in a 6-pole SynRM for low speed and 5 A/mm2.
The output parameters of a 6-pole SynRM, supplied at 50Hz with a current density of 5 A/mm2,
are shown in Table 1.1. The efficiency, obtained from FEA, is 91.4% while the efficiency calculated
using equation (1.33) is 92.2%. There is a small error in the calculation due the core losses was
neglected, but this data can be considered a first approximation of efficiency in the early design stage.
Table 1.1. Output parameters for a case study 6-pole SynRM through FEM.
Parameter Symbol Quantity Unit

Torque Tem 49.5 Nm
Stator current Is 20 A
Rotor angular speed ω 104.7 rad/s
Winding losses Pcu 432.7 W
Core losses Pedd+hys+add 58.3 W
Figure 1.11 shows the impact of current angle and load angle on efficiency. As the saliency ratio
increases, the efficiency of the machine increases. The efficiency is maximum when the machine
develops the maximum torque and this depend on the current angle and load angle.
(a) (b)
Figure 1.11. Influence of the current and load angles on the efficiency in a SynRM. (a) current angle vs.
efficiency; (b) load angle vs. efficiency.
15
1.7. Design techniques. Machine performance improvement

1.7.1. Analytical models vs. Finite elements methods
Finite Element Analysis (FEA) is one the most popular tool to design and optimize SynRM [9]
thanks to the ability to incorporate all the electromagnetic phenomena present in the machines with
high accuracy. Although FEM generally produces accurate results, it is subject to a high
computational cost, especially during an optimization process where numerous iterations are
required to obtain a good design. Consequently, a great deal of research has been conducted to build
analytical models that are capable to preserve good accuracy, compared to FEM and achieve results
significantly faster [43], [44].
The work done in [43] presents an analytical model based on Maxwell's equations to calculate
the flux density in the air-gap. The effects of rotor and stator slots are considered by conformal
mapping; however, this does not include the derivation of torque and torque ripple. In [44], an
analytical model where the reluctance of the flux barriers is calculated from the conformal mapping
is implemented, but the effect of stator slotting is neglected. A simplified method is proposed in [45]
from developing a magnetic circuit for a SynRM, the impact of stator slots is neglected. On the other
hand, extensive work has been done evaluating the magnetic flux density in the air-gap and the
torque as a function of the rotor magnetic potential and the stator electric loading [17], [46]–[50]. The
air-gap magnetic flux density is defined as a function of the rotor magnetic potential produced by
the stator electric loading, and the torque as an integrated function of the stator electric loading and
the rotor magnetic potential. Similarly, the effect of stator slotting is neglected, with the consequent
absence of the slot harmonics that are a fundamental component to be considered to capture the
torque fluctuations.
A further analytical model for SynRM and its validation through FEM is presented in [51]. This
work calculates the air-gap flux density as a function of stator and rotor magnetic potential and the
torque by integrating the Lorentz’s force density along the air gap surface; but the study neglecting
the slotting effect. Using the same principle as in [51], a SynRM with asymmetric flux barrier is
designed in [52] to minimize the torque ripple. To obtain higher fidelity, this analysis considers the
slotting effect with the conformal mapping technique. This technique requires a hard algebraical
work which becomes critical as the number of flux barrier increases.
An accurate Analytical Model (AM) for SynRMs can help the reduction of time spent to derive
the preliminary design phase. These models, besides being fast, must deliver sufficiently accurate
results; but most importantly, they need to enable the correlation between different parameters
defining machine performance, such as electrical, magnetic, and geometrical. The analytical
equations allow to study the behavior of certain performance indexes of the machine including the
variation of the design parameters and their relationships.
Amongst the features that an AM for SynRM should include, the interaction among the flux
barriers and the stator teeth is the most important; the relative position between them causing the
appearance of spatial harmonics in the air-gap flux density. These harmonics are one of the primary
sources of torque ripple in this type of machine. This issue is solved in[53] where an accurate AM for
SynRMs is presented. This work succeeds in predicting the air gap flux density and electromagnetic
torque considering the slotting effect.
16
1.7.2. Rotor structure optimization employing asymmetries

Lately, studies have been conducted to investigate the impact of asymmetric structures on the
performance of electrical machines [54]–[57], particularly Interior Permanent Magnet Synchronous
Machines (IPM). This research has revealed an improvement in average torque and reduction in
cogging torque when the rotor structure is designed to be asymmetric. The application of asymmetric
designs is believed to lead to an increase in average torque, a decrease in cogging torque and torque
ripple. Historically, electrical machines have been designed with symmetrical structures and
windings for each pair of poles, with the same electrical and magnetic characteristics being repeated
in both the rotor and stator structures.
The stator of an electric machine comprises of the teeth and windings as its main active
elements, which are typically designed to be identical and symmetrical. Recent studies have focused
on the asymmetrical design of the teeth and slots in the stator structure, with the goal of reducing
cogging torque. One example of this is found in [58] where it was shown that asymmetrical design
of the stator teeth can lead to a reduction in cogging torque. The impact of unequal tooth width in
the stator structure has also been studied in [59], which found that adjusting the width can result in
similar torque values while reducing stator current values, leading to a reduction in copper losses
and an increase in machine efficiency.
The opening of the slots in a stator may vary and may not be precisely centered in each slot.
Additionally, variations in the dimensions of the teeth can be used to achieve optimal
electromagnetic design [60]. However, it is typical for all poles in the rotor to be designed
symmetrically from the centerline of each pole or identically from pole to pole. Asymmetric design
elements can be applied to the rotor, such as the positioning of magnets in PMSMs or the design of
flux barriers in IPMSs and SynRMs. These changes can improve the performance of the machine, as
evidenced by the numerous patents [61]–[64] and articles [17], [54]–[58], [65]–[75] that have been
published on the subject in recent years.
There is a significant gap in the number of studies addressing asymmetries in the stator and
rotor structures. Research in [56], [65], [66] have shown that in IPMs with an asymmetric rotor can
lead to the two torque components, the excitation and reluctance torque, to achieving their maximum
values for the same stator current angle. On the other hand, in SynRMs, the rotor structure plays a
crucial role in their performance. The optimal design of flux barriers can increase the saliency ratio,
improve average torque, power factor, and efficiency, while also reducing torque ripple as reported
in [17], [75].
In [56], an analysis was conducted on IPM with radial magnets (spoke type). Another study,
referenced in [70], examined IPMS with V-type rotors. Previous research has shown that applying
asymmetries to the rotor structure can improve the average torque. Also, in these IPMs, the two
torque components reach their maximum value at the same current angle. A new asymmetric rotor
design, with additional flux barriers on the interpolar axis, was compared with a V-type IPM in [57].
This design resulted in an increase in average torque, decrease in torque ripple, and improved
constant power characteristic.
The design of PMa-SynRM is presented in reference [71]. The rotor design's asymmetry is
achieved through a combination of magnet positioning and flux barrier design. In [65], a rotor
structure was developed combining the advantages of both a SynRM and a PMSM. The rotor
assembly is designed to achieve the maximum of both torque components at the same current angle
and significantly reduce torque ripple. Reference [75] presented a rotor with two flux barriers, with
17
the position of the barriers relative to the stator teeth being different at each rotor pole, creating an
asymmetrical design. This design reduces torque ripple and increases average torque.
A comprehensive study on the impact of asymmetries in the rotor structure of a SynRM was
conducted in reference [55], evaluating various rotor topologies without radial ribs to enhance
average torque and decrease torque ripple. The study compares a symmetric structure with three
asymmetric designs. All three asymmetric designs showed an improvement in average torque,
however, only two achieved a reduction in torque ripple, highlighting those asymmetric structures
are a promising approach in the design of electrical machines.
It is apparent that most research on asymmetric structures has been focused on IPMs, with
relatively few studies on the benefits of asymmetries in SynRMs. The most extensive analysis on this
subject can be found in [18], where the impact of asymmetric positioning of flux barriers relative to
the q-axis of the rotor on machine performance is examined for a specific operating point. However,
it is important to note that an electrical machine does not function continuously at the same operating
point, as variations in the load will cause the operating point to change. Thus, it is crucial to
understand how the machine responds to these changes.
1.7.3. Rotor skewing
Skewing techniques can be classified into two main categories: continuous skewing and
discrete skewing (also called step skewing). The first one involves rotating each lamination of the
stator or rotor core in regular angular distribution, between the first and the last slice equal to the
skew angle [76], as depicted in Figure 1.12 (b). This can drastically reduce torque ripple but
complicates and makes the manufacturing process more expensive since each lamination of the rotor
has a different position with respect to a symmetry axis, thus requiring specific tooling. In turn, the
second category, considers the division of the rotor stack into a few discrete segments, as shown in
Figure 1.12 (c).
(a) (b) (c)

Figure 1.12. 3D sketch of different types of skew. (a) reference skewless rotor; (b) continuous skew; (c)
discrete skew, also mentioned as step skew.
A considerable number of works have addressed the rotor skewing technique using the
conventional one-slot-pitch skew angle. These works can be found in references [77]–[84]. In [77], the
performance of a SynRM is compared when it is operated without skewing to when the rotor is
skewed by one stator tooth pitch. Reference [78] investigates the effect of rotor skewing on reducing
slot harmonic torques, using a conventional skew angle of one stator slot pitch, but this reduces the
average torque as well. This was confirmed in [79] and [80]. In [81], the equation to calculate the skew
angle to suppress the stator slot harmonic component (one slot pitch) was presented, considering the
18
number of slices in the calculation. The novel forced feasibility concept was introduced in [82] to
improve optimization convergence and reduce overall optimization time in a SynRM design. A rotor
skew was chosen as the best suited torque-ripple mitigation option by skewing the rotor at an angle
of one stator slot. Recently in [83], a SynRM with salient pole rotor was continuously skewed by one
stator slot pitch to improve energy conversion and reduce torque ripple. The impact of rotor skewing
on torque ripple in SynRMs was analyzed in recent research [84], comparing both continuous and
segmented rotor skewing. Post-optimization simulations were performed for both methods and
yielded similar results, with a slight advantage to segmented rotor skewing due to the increased cost
of continuous skewing. In both cases, the total skew angle was equal to one stator slot. As it may be
noted from the dates of these works, the one-slot-pith skewing trend has been dominant as a post
optimization process, applicable to several machine topologies up to date.
In turn, several papers have discussed the skewing technique for reducing torque ripple [85]–
[91], but they lack information on the selection of the skew angle or the skewing parameters used. In
[85], a rotor design with an asymmetric flux barrier was created to reduce torque ripple by splitting
the rotor into two step-skewed parts. However, the paper does not mention the method used to
determine the skew angle, which is assumed to be equal to the slot pitch. Reference [88] evaluates
the suitability of SynRMs for electric traction applications. Skew is applied to reduce torque ripple
by using a 2.5° mechanical skew angle between three stacks, resulting in an angle close to the slot
pitch. In [89], the torque ripple was reduced by dividing the machine into three layers using step
skew, but the skewing angle is not specified. A SynRM was optimized using topology optimization
in [90], which increased the torque compared to a model optimized with parameters, but also
increased the torque ripple. The skewing technique was used to reduce the torque ripple by dividing
the rotor into two slices, but the specific skew angle and its determination method are not reported.
In [91], the goal was to reduce torque ripple through rotor skewing while maintaining a power factor
through optimization. The study found that the optimum mechanical skew angle across all machines
was 2.5 mechanical degrees, very close to the slot pitch. All these works seem to match the one-slot-
pith skewing trend.
Despite the prevalence of the slot pitch angle as the optimal skew angle in literature, references
[73], [92] question its effectiveness in minimizing torque ripple. Reference [73] demonstrates that a
torque ripple of less than 3.0% can be achieved by applying rotor skew. The optimum rotor skew
angles, ranging from 60-70% of a slot pitch angle for the 24-slot machine and 30-80% for the 36-slot
machine, have as an outcome this achievement. The analysis highlights that the ideal rotor skew
angle heavily depends on both the stator configuration and the rotor topology. In [92], a comparison
between continuous skewing discrete skewing was performed over a SynRM. The results indicated
that the torque ripple was significantly reduced even with two stacks, while only slightly decreasing
in the average torque when high order torque harmonics were produced. Some results showed that
the optimal skew angle differed from the traditional one-slot pitch, being either higher or lower,
depending on the type of skew technique applied (continuous or discrete).
As a result, recent trends have been investigated to improve the effectiveness of skewing in
different topologies [93]–[95]. A new unconventional magnet step-skew method for PMSM is
introduced in [93]. It involves varying both the length of the magnet and the skew angle between
magnet segments, in contrast to the constant stack length and step-skew angle in conventional PM
motors. A semi-FEA algorithm is developed showing improved performance compared to
conventional step-skew. However, it comes with increased magnet manufacturing cost. In [94] a new
19
method for parameterizing the flux barrier profiles of SynRMs and PMa-SynRMs was introduced.
To reduce torque ripple the skew angle is obtained through a parametric FEA analysis. A discrete
rotor skewing of 4° in 3-step pieces is applied which deviates from the conventional 10° angle for a
36-slot machine based on one slot pitch. In [95], the impact of step skewing on the output torque and
motor inductance in a 30-slot/4-pole SynRM configuration was examined. A comparison of step
skewing was made theoretically, and the study considered the effect of two harmonic orders but did
not account for the number of steps in the skewing angle calculation. Moreover, it was discovered
that the average torque reduction resulting from skewing is dependent on the machine operating
point.
This topic has brought increased attention in the literature. A better understanding of how the
optimal skewing can be achieved at the design stage is an important factor in the sizing of a SynRM.
It is necessary a generalized multi-step skewing approach for SynRM which considered a torque
ripple harmonic content and not the traditional one-slot-pitch skewing technique. This aims to serve
as an input of post-optimization processes in the design stage of SynRMs.
Reducing the difficulties and cost of the manufacturing process and providing the possibility
of mitigating relevant harmonic content in the electromagnetic torque may be achieved when
applying discrete multi-step skewing in a SynRM. Of course, the manufacturing complexity and
performance benefits may be adversely affected as the number of steps increases, but this is yet to be
disclosed. As a result, in [96] is propose some generalized analytical expressions for multi-step
discrete skewing on SynRM in order to mitigate undesired harmonic content of the electromagnetic
torque and its influence on other harmonic components.
1.8. Summary
This chapter provides an overview of the basic principles of operation for SynRMs. It covers
the mathematical model, vector diagram, and including the saturation and cross-coupling effect
between d-q axes. Additionally, a first approximation of the efficiency calculation is developed. Key
design features of the machine are described and how their affect the machine’s performances. As
well, the electrical, mechanical, magnetic, and geometrical parameters and their impact on the first
stage of the machine’s design were discussed. Additionally, a few designs technique that play a
major role in the performance improvement in the design step of SynRM were highlighted, such as
the use of asymmetrical barriers on the rotor and the use of the skew technique to reduce torque
ripple.
20
Chapter II: Accurate Analytical Model for Synchronous Reluctance Machine
This chapter introduces an Analytical Model (AM) for predicting air-gap flux density and
electromagnetic torque in SynRMs with good accuracy. The impact of stator slotting and flux barriers
on the torque and magnetic flux density in the air-gap is analyzed. Two methods for calculating
torque are studied, and their limitations in accounting for the effect of slotting are emphasized. The
method will first be developed for a rotor with one flux barrier and then expanded to multiple
barriers. The AM's results are compared with FEM results to validate the model.
2.1. Main assumptions
Figure 2.1 shows a general rotor geometry for a SynRM. The following assumptions are made
for the implementation of the analytical method:
• the tangential iron bridges of the rotor barriers are highly saturated under normal working
conditions, the relative permeability in these parts is negligible. Therefore, the regions of the rotor
iron bridges are considered as air [43], [44], [50], [51], [97],
• the rotor structure has no radial iron bridges,
• the permeability of the ferromagnetic material tends to infinity,
• and the tangential flux density is not considered for the torque calculation.
Tangential iron bridges
Rotor islands
Radial iron bridges
Flux barriers
Figure 2.1. Sketch of a general rotor geometry for a SynRM.
The above assumptions have been considered in most of the analytical models described so
far, e.g., [43], [44], [50], [51], [98]. These assumptions also limit the proposed method to low-speed
applications. In high-speed applications the radial and tangential bridges must be thick enough to
provide the rotor with the required structural integrity. This causes the bridges to not fully saturate
and provide a low reluctance path for the q-axis flux, penalizing the saliency ratio of the machine.
The fact that the bridges are wide causes the cross-coupling phenomenon between the d-axis and q-
axis to be present to a greater extent in high-speed SynRM. The above consideration means that in
high-speed SynRM, bridges must be taken into account in the electromagnetic analysis.
21
2.2. Air-gap flux density calculation

The approach for calculating the magnetic flux density consists mainly of two steps: first, the
magnetic potential across the flux barriers is determined. Secondly, the magnetic potential is
combined with Ampere's Law to calculate the flux density distribution of the air-gap.
It is necessary to start by calculating the MMF, which can be derived from Ampere's Law;
according to [99], it can be defined by (2.1) for a three-phase winding.
∞
3 𝑘w,6𝑛+1
𝑀𝑀𝐹g (𝜃𝑠 , 𝜃m ) = 𝑀𝑀𝐹f ∑ 𝑠𝑖𝑛[(6𝑛 + 1)𝜃s 𝑝 − 𝑝𝜃m − 𝛼ie ] (2.1)
2 6𝑛 + 1
𝑛=−∞
2
𝑀𝑀𝐹f = 𝑁 √2𝐼rms (2.2)
𝜋 t
where 𝑘w,6𝑛+1 is the winding factor, 𝑝 is the pole pair of the machine, 𝜃s is the position along the
stator periphery, 𝜃m is the rotor position, 𝛼ie is the current angle, 𝑁t is the turns per pole per phase,
𝐼rms is the supply current and 𝑛 = 0, ±1, ±2 … ± ∞.
AM’s for the calculation of the flux density is presented by several authors in [48]–[50], [85],
[100], where the flux density in the air-gap is calculated from the magnetic potential of the stator and
rotor in the air-gap like see in Figure 2.2, as in
𝜇0 [−𝑈s (𝜃s , 𝜃m ) + 𝑈r (𝜃m )]
𝐵g (𝜃s , 𝜃m ) = . (2.3)
𝑔
where 𝑈s (𝜃s , 𝜃m ) is the stator magnetic potential, which is equivalent to the drop of the 𝑀𝑀𝐹(𝜃s , 𝜃m ),
𝑈r (𝜃m ) is the rotor magnetic potential, 𝑔 is the air-gap length and 𝜇0 the air permeability.
The electric loading produces a magnetic potential distributed along the inner circumference
of the stator. The magnetic flux (green lines in Figure 2.2) crossing the flux barrier has a magnetic
potential in the rotor, calculated by (2.4). The rotor potential will be considered constant in each
magnetic "island" of the rotor, bordered by the air-gap and the flux barrier, and zero in the remainder,
this can be seen in Figure 2.2.
q
Us
θr d
θs
Ur
θm
Air-gap
lb
tb ϕb phase a-axis
Rb
Rotor island
Figure 2.2. Stator and rotor magnetic potential in the air-gap.
22
In this research, the behavior of the flux barrier opening is the same as the behavior of the slot.
Therefore, just as the slot opening is considered in the stator magnetic potential from the winding
factor, it will also be considered in the rotor magnetic potential. This behavior is considered from a
wave between the points (𝛼r , 𝐷ro /2) and (𝑙br , 𝑈r ) in Figure 2.3.
d q
Ur
αr lbr lbl αl
Dro/2
0 π/2p π/p θr
Figure 2.3. Flux barrier geometry and rotor magnetic potential in a rotor pole.
For the slot opening on the right side the curve will have positive slope and for the left side
negative slope.
𝑈r (𝜃m ) = ∅b (𝜃m )ℛb (2.4)
𝑡b
ℛb = (1.5)
𝜇0 𝑙b 𝑙axial
where ∅b (𝜃m ) is the magnetic flux crossing the flux barrier, ℛb is the reluctance of the flux barrier,
𝑡b and 𝑙b are the width and length of the flux barrier, respectively.
The flux crossing the flux barrier is calculated by (2.6), integrating the magnetic flux density in
the air-gap from the limits defined by the position of the flux barriers. A single flux barrier per pole
structure will be analyzed to explain the procedure correctly.
𝜋
−𝛼
𝑝 l −𝐵g (𝜃s , 𝜃m )𝑙axial 𝐷ro (2.6)
∅b (𝜃m ) = ∫ 𝑑𝜃s
𝛼r 2
where 𝐷ro is the rotor outer diameter.

Substituting (2.3), (2.4), and (2.5) into (2.6) and performing algebraic work, it is possible to
derive the magnetic potential of the rotor as
𝜋/𝑝−𝛼l
𝐷ro 𝑡b 𝜋 (2.7)
𝑈r (𝜃m ) = [ ] ( )∫ 𝑈s (𝜃s , 𝜃m ) 𝑑𝜃s − [ − (𝛼r + 𝛼l )] 𝑈r (𝜃m )
2𝑔 𝑙b 𝛼r 𝑝
Simplifying (2.7)
𝜋/𝑝−𝛼l
𝑈r (𝜃m ) = 𝑎 ∫ 𝑈s (𝜃s , 𝜃m ) 𝑑𝜃s (2.8)
𝛼r
where
23
𝐷 𝑡
{[ ro ] ( b )}
2𝑔 𝑙b
𝑎= (2.9)
𝐷 𝑡 𝜋
{1 + + [[ ro ]] ( b ) [ − (𝛼r + 𝛼l )]}
2𝑔 𝑙b 𝑝
Figure 2.4 shows the results of applying the procedure described above for a SynRM with two
poles pairs. The stator has 48 slots and distributed three-phase windings; the other design data are
summarized in Table 2.1.
Table 2.1. Machine main parameters.

Stator outer diameter 𝐷so 346.7 mm
Stator inner diameter 𝐷si 245.2 mm
Rotor outer diameter 𝐷ro 244 mm
Rotor inner diameter 𝐷ri 70 mm
Air-gap length 𝑔 0.6 mm
Radio barrier 𝑅b 45 mm
Axial length 𝑙axial 172 mm
Slot pitch 𝜏s 7.5 mechanical degrees
Slot opening 𝑏ss 1.4 mechanical degrees
Flux barrier width 𝑡b 15 mm
Flux barrier length 𝑙b 194.5 mm
Flux barrier opening 𝑙b(r=l) 7.08 mechanical degrees
Flux barrier angle 𝛼(r=l) 10.5 mechanical degrees
Turns per slot 𝑁s 10 turns
Current Angle αei 45 electrical degrees
(a) (b)
Figure 2.4. Magnetic potential and air-gap flux density without slotting effect for 𝜃m = 0. (a) stator and
rotor magnetic potential; (b) air-gap flux density.
̂s ) fundamental value is the 8.4 A/mm. Figure 2.4

For this analysis, the of the electric loading (𝐾
(b) compares the magnetic flux density in the air-gap calculated by analytical model, which is
represented by the acronym AM on the curve, and obtained in FEM. It is important to note that only
the effect of the rotor flux barriers is considered; the influence of the stator slots is not considered.
Equation (2.3) is modified to consider the effect of the flux barriers on the air-gap flux density, where
24
𝑃(𝜃s , 𝜃m ) is the air-gap permeance (2.11) and 𝑔r (𝜃s , 𝜃m ) is the air-gap function, it is calculated from
the infinite slot assumption [101].
𝐵g (𝜃s , 𝜃m ) = 𝑃(𝜃s , 𝜃m )[−𝑈s (𝜃s , 𝜃m ) + 𝑈r (𝜃m )] (2.10)
𝜇0
𝑃(𝜃s , 𝜃m ) = (2.11)
(𝑔 + 𝑔r (𝜃s , 𝜃m ))
Similarly, the effect of stator slotting can be included in the result of magnetic flux density in
the air-gap under the same assumption of the infinite slot, as clear from the results shown in Figure
2.5. Both effects, slotting and barriers, are correctly processed by the proposed AM. In Figure 2.5 (a)
the flux density waveform has the same behavioral trend for both analyses, FEM and AM. Figure 2.5
(b) shows the harmonic analysis for the air-gap flux density. As expected, the AM overestimates the
flux density. This can be seen in Figure 2.5 (b), where the fundamental component is slightly larger
in the comparison with the results obtained by FEM. This difference is mainly due to the assumptions
that were adopted during the development of the analytical model, i.e., by not considering the
nonlinearities in the machine. The remaining harmonic components in the AM show the same
behavior as those obtained from FEM. This flux density distribution in the air-gap is used to calculate
the electromagnetic torque waveform in the following section.
(a) (b)
Figure 2.5. Air-gap flux density with slotting effect for 𝜃𝑚 = 0. (a) air-gap flux density waveform; (b)
harmonic distribution of the air-gap flux density.
2.3. Torque calculation

The electromagnetic torque waveform is essential to evaluate torque ripple; in basic terms,
torque ripple can be defined mathematically, as shown in (2.12), it expressed in Nm (peak to peak).
𝑇rp = max(𝑇em ) − min (𝑇em ) (2.12)
Historically, the torque of a SynRM was derived from two methods. Several researchers mainly
focus on evaluating the torque as a function of stator electric loading and rotor magnetic potential
[48]–[50], [85], [100]. The torque can be obtained by integrating Lorentz's force (𝐵g (𝜃s )𝐾s (𝜃s )) along
the air-gap surface, as in
2𝜋
𝑇em (𝜃m ) = 𝑅2 𝑙axial ∫ 𝐵g (𝜃s , 𝜃m )𝐾s (𝜃s , 𝜃m )𝑑𝜃s . (2.13)
0
25
where 𝑙axial is the axial length of the machine, 𝑅 is the air-gap radius and 𝐾s (𝜃s , 𝜃m ) is the electric
loading.
In [99], the effect of the 𝑀𝑀𝐹 in the air-gap field is represented with sufficient accuracy by a
current charge distributed over the area of the slot opening, in other words, by the electric loading.
Therefore, the 𝑀𝑀𝐹 can also be calculated from (2.14), from where the electric loading can be
obtained.
𝐾s (𝜃s , 𝜃m )𝐷si
𝑀𝑀𝐹g (𝜃s , 𝜃m ) = ∫ 𝑑𝜃s (2.14)
2
The electromagnetic torque for both flux densities calculated in the previous section is shown
in Figure 2.6. These curves were performance by moving the rotor 60 mechanical degrees, which
corresponds to two torque ripple periods for three–phase machine [102]. When the effect of stator
slotting is ignored, Figure 2.6. (a), the comparison with FEM is satisfactory; while when the influence
of slotting is evaluated, the results show considerable differences, Figure 2.6. (b). Therefore, the above
method effectively determines the electromagnetic torque waveform when the stator slotting effect
is neglected.
(a) (b)
Figure 2.6. Electromagnetic torque waveform for two torque ripple periods. (a) without slotting effect;
(b) with slotting effect.
Harmonic analysis was performed to analyze this fact. The magnetic flux density is correct; at
any instant of time studied, the comparison with FEM is accurate. Additionally, the electric loading
is relatively easy to obtain and is the same for both cases investigated, including the slotting effect or
not. Figure 2.7 shows the torque when considering specific harmonics of the magnetic flux density.
It is observed that considering only the first harmonic, the average torque of the analytical model
and FEM are matching well; the exact values are shown in Table 2.2. However, as the number of
harmonics of the flux density increases, torque ripple appears, and the average torque decreases.
This effect becomes more predominant when harmonics above the 50th are included (green curve in
Figure 2.7). It is important to note that the flux density's harmonics significantly influence the final
torque result.
26
Figure 2.7. Harmonic torque analysis when considering slotting effect for two torque ripple periods.
It is well known that the stator slotting introduces the higher order harmonics in the magnetic
flux density. Therefore, by not considering the higher order harmonics, it again ignores the slotting
effect, which is not the objective of the research. Thus, the calculation method for electromagnetic
torque from Lorentz's force is effective only for finding the average torque when the stator slotting
effect is considered.
Based on the above, the magnetic flux density in the air-gap can be decomposed as a sum of
harmonics as
𝐵g (𝜃s ) = ∑|𝐵g (𝜈)| cos(𝜈𝜔0 𝜃s + 𝜑(𝜈)). (2.15)

𝜈=1
Table 2.2. Average torque and torque ripple values for harmonic analysis.
Analysis method & harmonic order 𝐓𝐚𝐯𝐠 (𝐍𝐦) 𝐓𝐫𝐩 (𝐍𝐦)

FEM 21.03 15.3
AM-1st harmonic 21.33 1.14
AM-30th harmonic 21.29 5.76
AM-50th harmonic 18.49 7.48
AM-full order 11.07 4.22
In the same way, the electric loading can be decomposed as follows,

∞
𝐾s (𝜃s ) = ∑|𝐾s (𝜂)| cos(𝜂𝜔0 𝜃s + 𝜑(𝜂)). (2.16)

𝜂=1
Therefore, the average torque is calculated by (2.17) when 𝜈 = 𝜂 = 1;

2𝜋
𝑇avg (𝜃m ) = −𝑅2 𝑙axial ∫ |𝐵g (𝜈)||𝐾s (𝜂)| cos(𝜈𝜔0 𝜃s + 𝜑(𝜈)) cos(𝜂𝜔0 𝜃s + 𝜑(𝜂)) 𝑑𝜃s (2.17)
0
Furthermore, the principal of the torque pulsation relative to slot harmonics lays on the fact
that, in SynRM the rotor is forced to be aligned with the minimum reluctance. Any change on this
27
alignment causes variations of stored energy at the slots opening area in the rotor and stator, results
in changes in their equivalent co-energy. These variations which depend on the rotor angle and them
are the potential sources for torque ripple [103], [104]. Discarding the magnetic saturation, the torque
ripple is calculated from the total energy stored in the machine [105]. Under the hypothesis of infinite
permeability, no energy is stored in the ferromagnetic core. Therefore, the magnetic energy (𝑊m ) is
𝑊m (𝜃m ) = 𝑊g (𝜃m ) + 𝑊b (𝜃m ). (2.18)
where 𝑊g (𝜃m ) is the energy stored in the air-gap and 𝑊b (𝜃m ) is the energy stored in the flux barrier,
(2.19) and (2.20), respectively.
𝑙axial 𝑅𝑔 2𝜋 2
𝑊g (𝜃m ) = ∫ [𝐵g (𝜃s , 𝜃m )] 𝑑𝜃s (2.19)
2𝜇0 0
2
𝑊b (𝜃m ) = 𝑝 [𝑅b (∅b (𝜃m )) ] (2.20)
In this case, the torque is calculated by (2.21). The first term is the average torque, calculated
by (2.17). The second term represents the torque ripple, the pulsating component of the torque,
obtained from the magnetic energy stored in the machine, (2.22).
𝑇em (𝜃m ) = 𝑇avg (𝜃m ) + 𝑇rp (𝜃m ) (2.21)
𝜕𝑊m
𝑇rp (𝜃m ) = − (2.22)
𝜕𝜃m
As can be seen, the pulsating component depends on the magnetic flux density distribution in
the air-gap, which is a function of the position of the rotor flux barriers and the stator slotting.
Therefore, it is possible to analyze the effect of both parameters on the torque waveform from the
stored energy method, which is compared to the one resulting from FEM simulations as shown in
Figure 2.8. It can be noted from Figure 2.8 (a) that the oscillation trends obtained in AM is consistent
with FEM. In the same way that the flux density in the air-gap was analyzed, the harmonic
distribution for the torque was obtained. From Figure 2.8 (b), the AM overestimates the fundamental
harmonic component of the electromagnetic torque, which was to be expected since the torque is
obtained from two methods that depend on the air-gap flux density. The behavior of the rest of the
harmonic components in AM is the same as that obtained from FEM.
Figure 1.8 shows that the model can predict the ripple torque with good accuracy. Table III
shows that the absolute error for the average torque and the ripple torque is 0.303 N·m and 0.903
N·m, respectively. These errors can be considered acceptable. The main cause of this difference is the
assumptions made throughout the development of the AM and the fact that the effect of tangential
flux density is neglected in the torque calculation.
28
(a) (b)
Figure 2.8. Electromagnetic torque waveform for two torque ripple periods with slotting effect. (a)
torque waveform; (b) torque harmonic components.
2.4. Extension of the method to machines with a larger number of flux barriers per pole
An increased number of flux barriers per pole leads to greater rotor anisotropy, which results
in a higher magnetic saliency, the primary metric for evaluating the performance of a SynRM. The
saliency refers to the ratio between the inductance of the direct and quadrature axes. By increasing
the number of barriers, the reluctance of the quadrature axis is altered, which modifies the saliency
and enhances machine performance.
• Geometric parameters
Start
• Excitation
Magnetic Relative
Potential Permeance
(Us Ur) (P)
Equation 2.10
Airgap
Flux
Density
(Bg)
Average Torque Ripple

Torque (Magnetic
(Lorentz's Energy Equation 2.21
Force) Stored)
(Tavg) (Trp)
Figure 2.9. Analytical method workflow.
The extension of the method is necessary to evaluate machines with better performance
characteristics and to generalize the approach to any number of flux barriers. It is presented
following the procedure described above for one flux barrier per pole, considering that the magnetic
potential of each barrier is the result of the flux crossing the barrier, and this is only present in the
barrier island. For the sake of clarity, a more intuitive representation of the method adopted, the
29
workflow adopted is shown in Figure 2.9. Following this procedure, the air-gap magnetic flux
density and torque were calculated when the rotor structure has two, three, four, five, and six flux
barriers per pole. It is chosen to compute up to six flux barriers per pole because this is the optimum
number of flux barriers per pole for a stator configuration with 48 slots [3], [106]. The comparison
with FEM is shown in Figure 2.10 for flux density and torque. It is important to note that the higher
the number of flux barriers the more parameters must be analyzed, and therefore the model's
accuracy may decrease.
The flux density shows a good agreement with FEM for any of the cases analyzed. Figure 2.10
shows the comparison for 𝜃𝑚 = 0, but for any rotor position analyzed, the results are equally
satisfactory. The flux density curves clearly show the effect of flux barriers and stator slots.
On the other hand, Figure 2.11 shows that the torque behavior is agreement to the one obtained
in FEM; the average torque values, torque ripple, and absolute error are shown in Table 2.3. This
metric compares the absolute difference between the FEA values and the values predicted by the
analytical model. The torque ripple results are less accurate than the average torque, as indicated by
the mean absolute error. However, the analytical model achieves an error of less than one with
respect to the FEA results, which indicates a good prediction of the phenomenon.
Table 2.3. Average torque and torque ripple for different rotor configurations.
Barrier number Analysis method & error 𝐓𝐚𝐯𝐠 (𝐍𝐦) 𝐓𝐫𝐩 (𝐍𝐦)
FEM 21.03 15.37
1 AM 21.33 14.50
Absolute error 0.30 0.90
FEM 21.49 14.03
2 AM 22.82 13.61
FEM 21.49 24.66
3 AM 22.84 20.90
FEM 21.44 13.14
4 AM 22.18 13.02
FEM 20.70 13.03
5 AM 21.61 12.86
FEM 20.72 12.45
6 AM 21.21 11.92
Mean Absolute Error 0.85 0.98
30
(a) (f)
(b) (g)
(c) (h)
(d) (i)
(e) (j)
Figure 2.10. Air-gap flux density waveform for θm = 0 for different rotor configuration and the harmonic
components. (a, f) two barriers per pole; (b, g) three barriers per pole; (c, h) four barriers per pole; (d, i) five
barriers per pole; (e, j) six barriers per pole torque ripple periods.
31
(a) (f)
(b) (g)
(c) (h)
(d) (i)
(e) (j)
Figure 2.11. Electromagnetic torque waveform for different rotor configuration and the harmonic
components. (a, f) two barriers per pole; (b, g) three barriers per pole; (c, h) four barriers per pole; (d, i) five
barriers per pole; (e, j) six barriers per pole torque ripple periods.
32
Despite this, the AM proposed shows an acceptable behavior; of the six rotor structures
analyzed, the rotor with three flux barriers per pole shows the highest error in the torque calculation.
Moreover, the variation trend of the torque waveform obtained by both methods is consistent. It is
also observed from Table 2.3 that in all cases the average torque value is higher in the analytical
model. The assumption that the permeability of the ferromagnetic material is infinite leads to ignore
the magnetic saturation of the cores iron and hence the average torque will be overestimated. This
occurs because in the FEM model the permeance has a very large value to simulate the infinite
permeance assumed in the analytical model, but despite this, it is a finite value and small drops of
potential are affecting the results.
2.5. Parameters on the d-q reference frame
According to [107] the Winding Function (WF) for an arbitrary phase "i" and its corresponding
Fourier Series can be calculated from equation (2.25); for this analysis the windings are assumed to
be stationary, only depending on 𝜃s . Figure 2.12 shows the winding function for each phase of the
machine under analysis.
∞
4𝑁t 𝑘w𝑛 2𝜋
𝑁𝑖 (𝜃s ) = ∑ cos [𝑝𝑛 (𝜃s − 𝑘 )] (2.25)
𝑝𝑛𝜋 𝑝𝑚
𝑛=1,3,5…
where k=0, 1, 2…. for phases A, B, C…, p represents the pole pairs, m the number of phases and n the
order of the harmonic, 𝑁𝑡 the number of turns per phase per pole, 𝑘w𝑛 is the winding factor for
harmonic n and it is calculated as follows,
𝑘w𝑛 = 𝑘d𝑛 𝑘p𝑛 𝑘x𝑛 (2.26)
𝑛𝜋
sin ( )
𝑘d𝑛 = 2𝑚
𝑛𝜋 (2.27)
𝑞 sin ( )
2𝑞𝑚
𝜋𝑊
𝑘p𝑛 = sin (𝑛 ) (2.28)
2 𝜏p
𝑛𝑏ss
( )
𝑘x𝑛 = sin 2 (2.29)
𝑛𝑏ss
2
where q is the number of slots per phase per pole, W is the width of the winding, 𝜏p is the polar pitch
y 𝑏ss is the slot opening.
33
Figure 2.12. Winding function for a four pole SynRM.
The flux linkages can be calculated as the multiplication of the flux density by the cross-
sectional area, generalizing this to a three-phase machine the flux linkages for an arbitrary phase “i"
can be calculated as
2𝜋
𝜆𝑖 (𝜃m ) = ∫ 𝐵g (𝜃m , 𝜃s )𝑁𝑖 (𝜃s )𝑑𝐴 (2.23)
0
𝑑𝐴 = 𝑅𝑙axial 𝑑𝜃s (2.24)
When the Clarke-Park’s transformation is applied to these flux linkages like in the Chapter I,
the flux linkage on the d-q axes is obtained as follows
2 2𝜋 4𝜋
𝜆d = [𝜆a cos(𝜃m ) + 𝜆b cos (𝜃m − ) + 𝜆c cos (𝜃m − )] (2.30)
3 3 3
2 2𝜋 4𝜋
𝜆q = [𝜆a sin(𝜃m )+𝜆b sin (𝜃m − ) + 𝜆c sin (𝜃m − )]. (2.31)
3 3 3
Once the flux linkages are obtained in d-q reference frame the inductances and the average
torque developed by the machine can calculate by (1.4). The comparison with FEM is shown in Figure
2.13 showing a correct prediction of the AM for the flux linkage in both reference frame. On the other
hand, the mean torque values are shown in Table 2.4. The main differences in the results are due to
the assumptions that were considered during the development of the AM and the stator leakage
inductance are neglected.
(a) (b)
Figure 2.13. Flux linkage for a four pole SynRM. (a) flux linkage for the abc reference frame; (b) flux
linkage for the d-q reference frame.
34
Table 2.4. Mean torque comparison between FEM and AM.
Analysis method 𝐓𝐞𝐦 (𝐍𝐦)

FEM 20.5
AM 21.5
2.6. Summary
This chapter presents the development of an AM to evaluate the performance of a SynRM that
combines two methods proposed by the research community as possible ways to be adopted for the
analysis of SynRMs. The method including the slotting effect in a relatively simple way in the torque
wave and with equations that allow to analyze how each parameter affects the result. In addition, it
provides the possibility that the parameters can be freely and easily modified for further
optimization. The analysis process is explained in detail for a machine configuration with one flux
barrier per pole and subsequently extended for machines with a larger number of barriers.
Even though the assumptions were made throughout the development of the method, the
comparison with FEM shows satisfactory results, both for air-gap flux density and torque. To
validate the method, several FEM simulations were performed for different machine configurations
where the dimensions and the electrical steel used were changed. In all analyses the AM shows a
correct prediction of the machine behavior until the machine starts to operate above the saturation
knee of the B-H curve. The analyses carried out allow us to affirm that the analytical model gives
sufficiently accurate results when the machine is operated in the linear zone of the B-H curve.
The harmonic analysis showed that the AM overestimates the air-gap flux density and the
electromagnetic torque. This difference is mainly due to the assumptions that were adopted during
the development of AM, i.e., by not considering the nonlinearities in the machine. The remaining
harmonics components in the AM shows the same behavior as that obtained from the FEM. There is
another factor that also affects the torque results and that is that the tangential flux density is
neglected, being the model with the highest deficiency the one with three flux barriers per pole.
It is also important to mention that for the application of the AM it was necessary to assume
certain conditions that are mentioned at the beginning of the chapter, which limits its application to
low-speed purposes. In practice, radial and tangential barrier bridges are necessary for mechanical
reasons. Considering these as air is a good assumption, but ignoring these bridges also leads to a
small overestimation of the average torque.
35
Chapter III: Rotor asymmetric impact on Synchronous Reluctance Machine Performance
This chapter examines the impact of an asymmetric rotor design on the performance of a
SynRM using FEM. The analyzed rotor has a simple structure, featuring two flux barriers per pole,
optimized through a multi-objective optimization algorithm. The rotor design exhibits asymmetry
along the q-axis, but each machine pole is identical, resulting in circumferential symmetry.
3.1. Modelling of the SynRMs under analysis
3.1.1. Sizing method
In this study, a 12kW@1500rpm SynRM with a two-pole-pair was designed to investigate the
potential benefits of an asymmetrical rotor design. The machine was sized using a specific
methodology [108], and features a stator with 48 slots and three-phase distributed windings (Figure
3.1), as well as a rotor with two flux barriers per pole. The ferromagnetic material used in the stator
and rotor cores is M19-29G. The main dimensions of the machine are outlined in Table 3.1. It is well
known that torque ripple in SynRMs is influenced by the interaction between the spatial harmonics
of the magnetomotive force caused by stator currents and the rotor geometry. Research [109] and
[102] have shown that the position of the flux barriers is a key factor in reducing torque ripple in
SynRMs, and thus is the primary variable to be analyzed in this study. The topology of the machine
under study is shown in Figure 3.1 together with the rotor parameterization.
Stator
q
Flux barrier bridge
Kp > 1
Kp < 1
Rso Flux barrier
αl1 αr1
d d
αl2 tb1dr αr2
Rsi
Air-gap tb1dl
tb1q
tb2dl wc tb2dr
tb2q
Rotor
Rb
Rri Rro
(a) (b)
Figure 3.1. Machine topology. a) stator and rotor; b) rotor parametrization.
36
Table 3.1. Design parameters

Stator outer diameter 𝐷so 346.7 mm
Rotor outer diameter 𝐷ro 244 mm
Axial length 𝑙axial 172 mm
Slot pitch 𝜏s 7.5 mechanical degree
Slot open 𝑏ss 1.4 mechanical degree
Turns per slot 𝑁s 10 -
Slots numbers 𝑄s 48 -
3.1.2. Optimization process

From the preliminary sizing described above, a design refinement stage is carried out to
optimize two objective functions: maximize torque production and minimize torque ripple. The air-
gap length, the inner and outer diameters of the stator and rotor are geometrical constraints.
The objective variables are set only on the rotor geometry, such as the position of flux barriers
concerning the d-axis (𝛼r𝑛 , 𝛼l𝑛 ), flux barrier opening (𝑙br𝑛 , 𝑙bl𝑛 ), flux barrier width (𝑡b𝑛 ), the
separation between the barriers (𝑤c ) and the radius to the center the most inner barrier (𝑅b ).
The optimization was implemented in ANSYS commercial software package using a Multi-
Objective Genetic Algorithm (MOGA). Rotor geometry was parameterized in a CAD software
included in the ANSYS package and FEM analysis was performed in ANSYS Electronics Desktop.
MOGA is defined with an initial number of 150 samples with a maximum of 35 iterations; 100 designs
were analyzed per iteration. The symmetric design converged after 2087 evaluations, and the
asymmetric after 2623. This analysis allows to obtain the anisotropic structure of the rotor that fulfills
the objective functions.
Figure 3.2 shows the optimization process workflow, and the range of input variables is shown
in Table 3.2. The model was supplying with 10 A/mm2 during the optimization; this current value
corresponds to approximately 2 times the rated current. Choosing a current value between 2 and 3
times the rated current ensures that the torque ripple is less sensitive to load variations [110], [111].
On the other hand, the current angle is adjusted according to the objective functions for MTPA.
The insulation ratio, which is defined as the ratio between thickness of total insulation (𝑡b1 +
𝑡b2 ) over total length ((𝐷ro − 𝐷ri )/2 ) inside the rotor, by (3.1). This parameter attempts to represent
the feature of the machine anisotropic structure quality. The insulation ratio has a great impact on
the torque production; therefore, in order to determine the performance characteristics of the
machine are only affected by the possible asymmetry in the rotor structure, this parameter is the
same for both designs in the optimization process.
𝑡b1 + 𝑡b2
𝑘air = (3.1)
(𝐷ro − 𝐷ri )/2
37
Start
Set of
New variables Geometry
parameters and
assignment constructions
variables
FE Analysis
Multi
NO Objective
Genetic
Algorithm
Are the objective Optimum

YES
functions satisfied? design
End
Figure 3.2. Optimization process workflow.
Table 3.2. Input variables range.
Boundaries
Parameter Symbol Unit
Lower Upper
Flux barrier angle 1 (right) 𝛼r1 1 30 mechanical degree
Flux barrier angle 2 (right) 𝛼r2 1 30 mechanical degree
Flux barrier angle 1 (left) 𝛼l1 1 30 mechanical degree
Flux barrier angle 2 (left) 𝛼l2 1 30 mechanical degree
Flux barrier opening 1(right) 𝑙br1 2 10 mechanical degree
Flux barrier opening 2(right) 𝑙br2 2 10 mechanical degree
Flux barrier opening 1(left) 𝑙bl1 2 10 mechanical degree
Flux barrier opening 2(left) 𝑙bl2 2 10 mechanical degree
Flux barrier width 1 𝑡𝑏1 10 22 mm
Flux barrier width 2 𝑡b2 10 22 mm
Width between barriers 1-2 𝑤c 2 10 mm
Current Angle 𝛼ie 45 75 electrical degree
Inner barrier radius 𝑅b 40 50 mm
38
3.1.3. Optimum Designs

For the symmetric model, the position and opening of the flux barriers on the left side remain
the same as those on the right side, while in the asymmetric model, they can take any value within
the set limits. The optimization results are presented in Figure 3.3 for the symmetric (a) and
asymmetric (b) models. The rotor structure data and the current angle for the optimal designs are
shown in Table 3.3.
(a) (b)
Figure 3.3. Designs obtained from the optimization process for a 4-pole machine with two flux barriers
per pole. The current was fixed to ~10 A/mm2 and the current angle was defined in the optimization algorithm
for MTPA. (a) symmetric design; (b) asymmetric design. The selected designs are highlighted on the figure.
Table 3.3. Geometric values for optimal symmetrical and asymmetrical designs.
Parameter Symbol Symmetric design Asymmetric design Unit

Flux barrier position index 𝐾p 1 1.42 -
Flux barrier opening index 𝐾w 1 1.65 -
Flux barrier width 1 𝑡b1 21.16 17.90 mm
Flux barrier width 2 𝑡b2 11.81 14.95 mm
Width between barriers 1-2 𝑤c 8.39 9.25 mm
Deeper radio barrier 𝑅b 48.77 47.70 mm
Insulation ratio 𝑘air 0.38 0.38 -
Current angle 𝛼ei 60.56 61.95 °
To characterize the level of asymmetry of the designs, the following coefficients, 𝐾𝑝 and 𝐾𝑤 ,
are introduced. The first index (𝐾𝑝 ) is the ratio between the barrier angles and the second (𝐾𝑤 ) is the
ratio between the barrier opening angles. Equations (3.2) and (3.3) show how these indices are
calculated for a i-th barrier.
𝛼r𝑖
𝐾p𝑖 = (3.2)
𝛼𝑙𝑖
𝑙r𝑖
𝐾w𝑖 = (3.3)
𝑙l𝑖
When the machine is symmetrical these indices are equal to one, both sides of the magnetic
pole have the same dimension. If the design is asymmetrical, both 𝐾p𝑖 and 𝐾w𝑖 can be greater or less
than one. To characterize the possible displacement of the machine's q-axis because of asymmetry, it
will be classified as positive or negative according to the value of 𝐾p𝑖 . If 𝐾p𝑖 > 1 the asymmetry will
39
be positive, and the q-axis will be displaced in a leftward direction and if 𝐾p𝑖 < 1 the symmetry will
be negative, and the q-axis will be displaced to the right as shown in Figure 3.1 (b).
To increase the anisotropy in SynRMs, it is common that the rotor has more than one flux
barrier per pole, in this case the indexes are calculated as follows, where 𝑛 is the number of flux
barriers per pole of the rotor.
∑𝑛𝑖=1 𝐾p
𝑖 (3.4)
𝐾p =
𝑛
∑𝑛𝑖=1 𝐾w 𝑖
𝐾w = (3.5)
𝑛
3.2. Results analysis

Both designs were analyzed using ANSYS Electronics Desktop, a commercial software
package. The analysis involved variations in rotation direction (clockwise and counter-clockwise)
and operating conditions (current angle ranging from 25-85 degrees and current density from 0-10
A/mm2) in order to evaluate the behavior of the designs under different scenarios.
3.2.1. Saliency ratio
The variation of the saliency ratio for symmetrical and asymmetrical designs as a function of
current angle and current density is shown in Figure 3.4. It can be observed that the symmetric design
shows higher saliency values than the asymmetric design for a value of the current angle less than
60 electrical degrees, when exceeding this value, the asymmetric design shows better saliency. In
general, the highest saliency values are achieved for both designs when the machine begins to be
saturated and for high current angle values.
(a) (b)
(c) (d)
Figure 3.4. Saliency ratio as a function of the current density and the current angle. (a) symmetric design
counter-clockwise; (b) asymmetric design counter-clockwise; (c) symmetric design clockwise; (b) asymmetric
design clockwise.
40
The saliency ratio (ξ) is used to evaluate the SynRM performance. As mentioned in Chapter I,
for a fixed value of current the q-axis inductance presents a small variation associated with the change
in the current angle. When the machine is operated beyond the knee of the B-H curve the q-axis
inductance can be considered constant, while the d-axis inductance continues to reduce. On the other
hand, the d-axis inductance presents greater variations as the current angle changes, for a higher
current angle the value of the d-axis inductance will be higher. Therefore, if the current angle
increases the q-axis inductance remains constant and the d-axis inductance increases, and the saliency
ratio of the machine increases as the current angle increases. This behavior in the inductance is caused
by the saturation and cross-coupling between the d-axis and q-axis.
3.2.2. Maximum internal power factor
Figure 3.5 shows the IPF behavior in both designs for different values of current and current
angle. Given the dependence between IPF and saliency ratio (equation 1.8), the behavior of both
indices is similar, the symmetric design shows higher power factor than the asymmetric design for a
value of the current angle less than 60 electrical degrees, when exceeding this value, the asymmetric
design shows better power factor reaching values above 0.8.
(a) (b)
(c) (d)
Figure 3.5. Internal power factor as a function of the current density and the current angle. (a) symmetric
design counter-clockwise; (b) asymmetric design counter-clockwise; (c) symmetric design clockwise; (b)
asymmetric design clockwise.
3.2.3. Electromagnetic torque

Figure 3.6 illustrates the behavior of electromagnetic torque for both symmetrical and
asymmetrical designs as a function of current density and current angle. Neglecting saturation, a
SynRM achieves its maximum torque value at a current angle of 45 electrical degrees. However, as
saturation occurs, the current angle for maximum torque increases above 45 electrical degrees. The
impact of saturation can be easily observed in the torque contour curves (Figure 3.6) for both designs.
As the current density increases, the current angle for maximum torque also increases.
41
(a) (b)
(c) (d)
Figure 3.6. Electromagnetic torque as a function of the current density and the current angle. MTPA
trajectory is highlighted in red. (a) symmetric design counter-clockwise; (b) asymmetric design counter-
clockwise; (c) symmetric design clockwise; (b) asymmetric design clockwise.
In the asymmetric design this effect is a is slightly more evident because the asymmetry
generates a shift in the q-axis of the machine. In a symmetric design, d-axis inductance is obtained
when aligning the a-phase symmetry axis with the rotor d-axis and supplying the a-phase with
current that has a current angle of 45 electrical degrees (αei = 45°) and the maximum torque is
obtained. In the case of an asymmetric rotor, repeating this procedure does not provide the maximum
torque since the q-axis and d-axis are shifted as a result of the asymmetry. Accordingly, for the
proposed asymmetrical design, the maximum torque is obtained for a larger current angle than in
the symmetrical design. This increase in the current angle to obtain the maximum torque makes it
possible that when the machine operates at MTPA the salience presented by the machine is higher,
despite this advantage, the asymmetric model is not able to develop a higher torque than the
symmetric design since the 𝑠𝑖𝑛2𝛼𝑖𝑒 begins to decrease when the current angle is greater than 45
electrical degrees.
3.2.4. Torque ripple
The primary advantage of the asymmetric design over the symmetric design is the significant
reduction in torque ripple. Figure 3.7 illustrates the torque ripple behavior for both designs at various
current and current angle values. When following an MTPA trajectory, the asymmetric design
consistently exhibits the lowest torque ripple value, as shown in Figure 3.7. Adhering to this control
strategy, the asymmetric design consistently demonstrates a torque ripple lower than 20%, while the
symmetric design exhibits a torque ripple higher than 20%, reaching as high as 30% at specific
operating points. These torque ripple values are further discussed in the following section.
One of the main contributors for the large torque ripple in SynRM is the interaction of spatial
harmonics of the MMF generated by stator currents and rotor geometry [14]. High torque ripple
generates vibrations that can cause higher acoustic noise and impact the current harmonics.
Moreover, in high-performance applications, low torque ripple is strictly required [15]. In this sense,
42
the reduction in torque ripple is mostly due to the asymmetric position of the rotor flux barriers,
which allow some degree of mitigation of the spatial harmonics appearing in the air-gap.
(a) (b)
(c) (d)
Figure 3.7. Torque ripple as a function of the current density and the current angle. MTPA trajectory is
highlighted in red. (a) symmetric design counter-clockwise; (b) asymmetric design counter-clockwise; (c)
symmetric design clockwise; (b) asymmetric design clockwise.
3.2.5. Additional discussion on efficiency and CPSR

In this section the designs will only compare when they rotate counter-clockwise, since, as it
was observed in the previous sections, there are no considerable differences in the performance
indexes according to the change of the direction of rotation. Figure 3.8 show the intersection of the
contour curves of the internal power factor, torque ripple and the mean torque over a certain range
of current density and current angle for the symmetric and asymmetric design, respectively. Two
operating points in the MTPA trajectory corresponding to 5 A/mm2 for the inverted blue triangle and
10 A/mm2 for the red triangle have been highlighted in the figure. Table 3.4 provides the values of
performance indexes for the operating points defined in Figure 3.8. The mean torque values are
relatively similar between the symmetric and asymmetric designs. The asymmetric design
demonstrates a 10% increase in power factor compared to the symmetric design. The most notable
distinction between the symmetric and asymmetric designs can be found in torque ripple, where the
asymmetric design offers a reduction of approximately 35% with respect to the symmetric one.
Symmetric Asymmetric
Operating point IPF Tavg [Nm] Trp [%] P [kW] IPF Tavg [Nm] Trp [%] 𝑃 [kW]
0.7 80.4 24.2 12.6 0.8 78.2 16.2 12.3
0.6 175.1 29.9 27.5 0.7 176.2 15.7 27.6
43
(a) (b)
Figure 3.8. Electromagnetic torque and power factor as a function of the current density and the current
angle for counter-clockwise rotation. (a) asymmetric design; (b) symmetric design.
To evaluate the performance of the designs under examination, an efficiency map was created
for each design using the ANSYS Electronics Desktop's Machine-Toolkit. The analysis was
performed using a line-to-line voltage of 380 V with a star connection and operated at MTPA with a
maximum current density of 5 A/mm2. Figure 3.9 shows the result for the symmetrical and
asymmetrical design for counter-clockwise rotation.
The most widely accepted approaches acknowledge that the constant power speed range
(CPSR) largely depends on the saliency ratio [112]. Therefore, designing SynRM with enhanced
saliency ratio can expand the application of this motor topology in various fields, such as electric
vehicles. Figure 3.9 shows that the asymmetric design maintains nominal torque up to the nominal
speed, whereas in the symmetric design, the torque begins to decrease at 1200 rpm, resulting in an
8% reduction of torque at the nominal speed. This is primarily due to the increased saliency ratio in
the asymmetric design. In terms of efficiency, both designs exhibit a similar behavior, reaching 90%
efficiency at the rated speed.
(a) (b)
Figure 3.9. Efficiency map when the machine is operated a MTPA for a maximum current density of 5
A/mm2 for counter-clockwise rotation. (a) symmetric design; (b) asymmetric design.
44
3.2.6. Sensitive analysis

After optimizing a design, it is important to conduct a sensitivity analysis to determine which
variables have the greatest impact on the target functions. This analysis provides insight about
potential errors that may occur during the manufacturing of a prototype. The ModeFRONTIER
optimization program can be used to conduct the sensitivity analysis, which illustrates the relative
significance of each component of the model based on its percentage contribution to overall variance.
The sensitivity analysis identifies the most crucial input variables by evaluating the main and
interaction effects of factors on responses. Factors are inputs that affect changes in responses, while
responses are outputs that are often dependent variables. Based on the sensitivity analysis results,
certain variables may be eliminated in future optimizations to reduce computational effort and gain
a deeper model understanding.
The key factors of the rotor structure that affect the average torque and torque ripple are the
length, width, and location of the flux barriers [113]. The insulation ratio and current angle play a
significant role in the average torque developed by the machine. Since the insulation ratio is kept
constant between the two designs and the current angle is adjusted for maximum torque, the average
torque is the same for both designs and no sensitivity analysis was conducted for this performance
metric.
Figure 3.10 shows the relative significance of different terms on the torque ripple, specifically,
the percentage of contribution of each term to the overall variance. The analysis focuses solely on the
torque ripple and how the position and opening of the flux barriers affect it. As shown in this figure,
the position of the flux barriers has the most significant influence on torque ripple in both designs.
The increased degree of freedom in the asymmetric model allows for a smaller contribution of the
position of each barrier to the overall variance than in the symmetric model. This results in a
significant reduction of torque ripple in an asymmetric design, as well as an improvement in other
performance indices such as average torque, power factor and CPSR.
(lbl2)
(lbl1)
(lbr2)
(lbr1)
(lb2) (αl2)
(lb1) (αl1)
(α2) (αr2)
(α1) (αr1)
50 40 30 20 10 0 10 20 30 40 50
Contribution (%)
Symmetric design Asymmetric design
Figure 3.10. Percentage of contribution of each term to the global variance. Influence of barriers position
(𝛼𝑟𝑛 and 𝛼𝑙𝑛 ) and opening (𝑙𝑟𝑛 and 𝑙𝑙𝑛 ) on torque ripple for both designs.
45
3.3. Summary
The proposed asymmetrical rotor topology for a SynRM demonstrates improved performance,
specifically a 35% reduction in torque ripple at a specific operating point without a decrease in
average torque compared to the symmetrical design. The asymmetric design also presents a lower
torque ripple throughout the entire MTPA trajectory. FEM analysis confirms that the improvement
is given by the asymmetrical positioning of the rotor flux barriers, which enhances the saliency ratio
of the machine and reduces space harmonics in the air-gap. Additionally, the IPF increases by 10%
in the asymmetric design in comparison to the symmetric design. The efficiency values for both
designs are similar, but the asymmetric design can achieve the rated torque up to the rated speed,
resulting in a higher CPSR than the symmetric model. These performance improvements indicate
that incorporating asymmetry in the rotor's q-axis is a promising technique for reducing torque
ripple and increasing power factor, which are significant drawbacks in this electric machine
configuration.
46
Chapter IV: A method to determine the torque ripple harmonic reduction in skewed
Synchronous Reluctance Machines
This section presents a general analytical expression for multi-step discrete skewing in SynRM,
which considers the impact on all torque harmonic components and offers a way to visualize the
reduction of specific undesired harmonic content. To evaluate the expression's accuracy, two SynRM
designs are analyzed through both two-dimensional (2D) and three-dimensional (3D) finite element
analysis (FEA). The results show that 2D FEA is effective in predicting the optimal step skewing and
the proposed expression is validated by the comparison between the 2D and 3D analysis. The
proposed equations are applied to 2-step, 3-step, and 4-step skewing and yield promising results,
demonstrating the ability to mitigate selected undesired harmonics and reduce other harmonic
content as a side effect.
4.1. Selected machines
With the aim of giving insight of the procedures required to use the proposed N-step skewing
analytical expressions, two SynRM machines are considered as a case study and assessed in this
work, as presented in Figure 4.1(a) and Figure 4.1(b), respectively. Single-layer distributed windings
are considered for both machines.
(a) (b)
Figure 4.1. 3D sketches of (a) four-pole synchronous reluctance motor with two barriers per pole; (b) six-
pole synchronous reluctance motor with two barriers per pole. Three-phase stator windings are highlighted in
red, green, and blue, corresponding to each phase. Rotor in both machines is shown as an exploded view.
Several geometrical parameters of the rotor structure of SynRM can affect to different levels
the performance of a SynRM. There are a number of design guidelines established in the literature
to choose the number of flux barriers and poles. The number of parameters increase exponentially
as the number of flux barriers per pole and pole pairs increase. SynRM is designed to maximize d-
axis inductance and minimize q-axis inductance as this ensures that the machine's saliency ratio is
large enough for the machines to achieve the required torque performance. On the one hand, to
obtain a good saliency ratio, a small number of pole pairs is preferred, and the literature recommends
to adopting two or three pole pairs [45]. On the other hand, the optimum number of flux barriers is
defined according to the number of stator slots. For the case of a 36-slot machine, some authors do
47
not encourage to adopt more than three flux barriers. A greater number of barriers could jeopardize
the mechanical integrity of the rotor or make the design process more complex [3], [106].
Therefore, in this paper two machines with two and three pole pairs are considered, and each
pole features two barriers. The common data for all machines are presented in Table 4.1.
Table 4.1. Main data of the selected machines.
Parameter Symbol Value Unit

Stator outer diameter 𝐷so 245 mm
Rotor outer diameter 𝐷ro 160.4 mm
Tooth height ℎt 22.8 mm
Tooth width 𝑏t 9 mm
Stack length 𝑙st 120 mm
Turns per slot 𝑁s 20 -
Number of slots 𝑄s 36 -
Speed 𝑛 3000 rpm
Current Density 𝐽 10 A/mm2
Current angle 𝛼ie 60 °
4.2. Analytical method derivation for discrete skewing

The main period of torque ripple of three-phase winding machines is 60 electrical degrees [102],
which therefore dictates the period of the harmonic of the torque ripple of order v, given in electrical
degrees by:
360°
T𝑣,elec = . (4.1)
𝑣
The aim of the discrete skewing to consider N machine slices, rotated with respect to each other
by a specific angle, so that each slice contributes to different torque harmonics, that will ultimately
modulate the torque waveform. This superposition has to be adjusted with the aim of mitigating
undesired harmonic components of the resulting torque waveform. The 𝑣-th harmonic of the torque
ripple can be expressed as a term of the Fourier series expansion of the electromagnetic torque as:
𝜋
Τripple,𝑣 (𝜃𝑟,𝑒 ) = 𝐴𝑣 ∙ cos ( 𝜃 𝑣 + 𝜙𝑣 ), (4.2)
180 𝑟,𝑒
where 𝐴𝑣 is the amplitude of the 𝑣-th harmonic of the torque ripple, 𝜃𝑟,𝑒 is the rotor position in
electrical degrees, 𝜙𝑣 is the phase shift of the torque waveform, and 𝑣 = 1,2,3, …
In this sense, and following the concept of balanced multiphase systems, if the 𝑤-th harmonic
of the torque ripple wants to be mitigated, then the electrical angle between each one of the rotor
sectors (N) of the machine is proposed as:
360°
θ𝑤,elec = . (4.3)
𝑁𝑤
48
For example, if a two-step skew is adopted, then the machine rotor will be comprised of two
sectors (two halves), each one contributing with torque waveforms that are shifted one with respect
to the other. If the electrical shift is θ𝑤,elec, then the positive semi-cycles of the 𝑤-th harmonic of the
torque ripple of one half of the rotor will compensate the negative semi-cycles of the other half, hence
mitigating the undesired component.
The electrical angle between the N slices of the machine is translated into the skew angle
(𝜃skew ), which corresponds to the mechanical angle in which two consecutive rotor slices are rotated
one with respect to the other so as to mitigate the 𝑤-th harmonic of the torque ripple.
360°
θskew = (4.4)
𝑝𝑁𝑤
In Figure 4.2, the proposed methodology is schematized, comprising the decomposition of the
electromagnetic torque waveform in harmonic components, the selection of a high-magnitude
undesired component, and the calculation of θs to mitigate that undesired harmonic depending on
the adopted number of slides N.
FFT
Electromagnetic torque
Skew angle must be calculated depending on the step Harmonic component of interest is selected.
count.
𝜃skew
𝜃skew
2‐step skew
𝜃skew 𝜃skew 𝜃skew
𝜃skew
3‐step skew
Figure 4.2. Schematics of the proposed methodology to calculate the skew angle 𝜃skew to mitigate a
selected harmonic of order 𝑤. 2 step and 3-step discrete skew are presented, although the method can be used
for any number of slides.
49
Considering the proposed skew angle, (4.2) can be expressed in terms of a sum of the
contributions of the N slices of the machine as
N
A𝑣 𝜋 2𝜋𝑣
Τrs,𝑣 (𝜃𝑟,𝑒 ) = ∑ cos ( 𝜃𝑟,𝑒 𝑣 − (𝑖 − 1)). (4.5)
𝑁 180 𝑁𝑤
𝑖=1
The resulting reduction on each component of the torque ripple, considering idealized
conditions and electromagnetic independency between adjacent slices can be obtained by means of
the phasor representation and consequent analysis of (4.5). Then:
𝑁
A𝑣 2𝜋𝑣
𝔑𝔢{Τrs,𝑣 (𝜃𝑟,𝑒 )} = ∑ cos (− (𝑖 − 1)) (4.6)
𝑁 𝑁𝑤
𝑖=1
𝑁
A𝑣 2𝜋𝑣
ℑ𝔪{Τrs,𝑣 (𝜃𝑟,𝑒 )} = ∑ sin (− (𝑖 − 1)) (4.7)
𝑁 𝑁𝑤
𝑖=1
In consequence, the magnitude of the 𝑣-th harmonic of the torque ripple after applying discrete
skewing following the design guidelines of (4.3) and (4.4) can be expressed as
2 2
A𝑣 2𝜋𝑣 2𝜋𝑣
|Τrs,𝑣 (𝜃𝑟,𝑒 )| = √(cos (− (𝑖 − 1))) + (sin (− (𝑖 − 1))) . (4.8)
𝑁 𝑁𝑤 𝑁𝑤
Finally, a skew mitigation factor (𝑘rs,𝑣 ) can be devised by obtaining the ratio between (4.2) and
(4.8), given by:
2 2
1 2𝜋𝑣 2𝜋𝑣
𝑘rs,𝑣 = √(cos (− (𝑖 − 1))) + (sin (− (𝑖 − 1))) . (4.9)
𝑁 𝑁𝑤 𝑁𝑤
This factor is meant to be used after the calculation of the skew angle as per (4.4), and it allows
to estimate the resulting amplitude of each torque ripple component (of order v) after applying the
skew. Simplified equations for 𝑁 = 2, 3 and 4 are provided in the following section, which are further
verified and analyzed by means of 2D and 3D finite element analysis.
4.3. Finite Element validation: results and discussion
This section shows the results obtained by applying the analytical method described in section
4.2 for both 4-pole and 6-pole SynRMs. In order to provide the results in a clear fashion, this section
is divided into four subsections. The first presents the evaluations of the machines without
considering any type of skew, hereby called skewless machines; and the following three sections
address machines with different slide number (𝑁 = 2, 𝑁 = 3 and 𝑁 = 4). All the results are obtained
by means of 2D and 3D FEA simulations carried out in the commercial package ANSYS Electronic
Desktop. The simulation time was chosen to evaluate a whole period of the machine's torque ripple.
50
4.3.1. FEA evaluation original designs (skewless machines)

Figure 4.3 presents the electromagnetic torque waveform and spatial harmonic spectrum of
both the 4-pole and the 6-pole SynRMs. Both 2D and 3D results are shown for comparison. From the
results, it can be seen that the evaluation of the 2D and 3D models provide similar values, showing
expected small differences, in accordance with the findings of [114]. The harmonic components of
the electromagnetic torque for each machine are detailed in Figure 4.3(b) and Figure 4.3(d) for the 4-
pole and 6-pole machine respectively. As expected, the largest harmonic component in both designs
corresponds to the one generated by the stator slotting effect. Specifically, the highest-magnitude
harmonic of the 4-pole machine corresponds to 𝑣 = 3, and that of the 6-pole machine corresponds to
𝑣 = 2. Therefore, and according to the methodology proposed in Section 3, 𝑤 = 3 for the 4-pole
SynRM and 𝑤 = 2 for the 6-pole machine. Table 4.2 shows the skew angle that should be considered
to discrete skew the machine according to (4.4), depending on the desired step number and aiming
to mitigate the highest-magnitude harmonic component. The following sections evaluate the impact
of the proposed skew methodology on mitigating torque ripple for 𝑁 = 2, 𝑁 = 3 and 𝑁 = 4.
(a) (c)
52.2 56.1
51.5 55.1
(b) (d)
2D skewless 3D skewless
Figure 4.3. Electromagnetic torque waveform and spectrum of skewless reference machine; (a) torque
waveform of 4-pole machine with two barriers per pole; (b) torque spatial harmonic content of 4-pole machine
with two barriers per pole; (c) torque waveform of 6-pole machine with two barriers per pole; (d) torque spatial
harmonic content of 6-pole machine with two barriers per pole. Additionally, 2D and 3D simulation results are
compared.
Table 4.2. Skew angle to reduce a specific electromagnetic torque harmonic order by discrete skew.
Harmonic Mechanical angle Mechanical angle Mechanical angle

order for 2-step skew for 3-step skew for 4-step skew
2p2b 18th 5° 3.33° 2.5°
3p2b 12nd 5° 3.33° 2.5°
51
4.3.2. Evaluation of torque ripple reduction by means of two-step discrete skew

The comparison of the electromagnetic torque for the 4-pole and 6-pole machine is shown in
Figure 4.4 when two-step skewing is applied. It can be observed that there is a significant reduction
in the harmonic torque component to be mitigated, to around 10% of its original value.
(a) (c)
52.2 51.2 56.1 53.2

51.5 49.9 55.1 51.8
(b) (d)
2D skewless 3D skewless 2D 2-step skew 3D 2-step skew

vs 2-step skewed machine; (a) torque waveform of 4-pole machine with two barriers per pole; (b) torque spatial
harmonic content of 4-pole machine with two barriers per pole; (c) torque waveform of 6-pole machine with
two barriers per pole; (d) torque spatial harmonic content of 6-pole machine with two barriers per pole. 2D and
3D simulation results are compared.
This agrees with the estimations obtained from the reduction factor proposed in (4.9), which
can be simplified when evaluating 𝑁 = 2 to:
√2 𝑣
𝑘rs,𝑣 = √1 + cos (𝜋 ). (4.10)
2 𝑤
According to (4.10), the reduction of the main component of the electromagnetic torque (which
has order 𝑤 = 18) should be maximum (𝑘rs,𝑤 = 0). The difference lies in the fact that (9) considers
magnetically independent rotor slices, neglecting their interaction. Regardless of this, a significant
reduction was observed in other harmonic components. In addition, for the 4-pole machine it was
found the torque ripple harmonic component with 𝑣 = 12 was also reduced to ~45% of its original
value, whilst other relevant harmonics of order 𝑣 = 6 and 𝑣 = 36 were not significantly affected. This
agrees with the estimated reduction factor proposed in (4.10), since 𝑘rs,6 = 0.87, 𝑘rs,12 = 0.5 and
𝑘rs,36 = 1. In turn, for the 6-pole machine it was observed the relevant torque ripple harmonic
component with 𝑣 = 24 is slightly reduced (to 87% of the original value), and that other relevant
harmonics of order 𝑣 = 6 and 𝑣 = 36 are reduced to 77% and 14% of their original magnitude,
respectively. This agrees with the estimated reduction factor proposed in (4.10), since 𝑘rs,6 = 0.7,
52
𝑘rs,24 = 1 and 𝑘rs,36 = 0. In addition, the trend of other less-relevant harmonic components of both
machines matches closely with the estimations of (4.10). These findings are summarized in Table 4.3.
considered, and relevant harmonic components are analyzed.
Harmonic order 4-pole SynRM 6-pole SynRM

𝑘rs,𝑣 (analytical) 𝑘rs,𝑣 (FEA) 𝑘rs,𝑣 (analytical) 𝑘rs,𝑣 (FEA)
𝑣=6 0.87 0.95 0.71 0.77
𝑣 = 12 0.50 0.45 0.00 0.06
𝑣 = 18 0.00 0.10 - -
𝑣 = 24 - - 1.00 0.87
𝑣 = 30 - - - -
𝑣 = 36 1.00 1.02 0.00 0.14
As a consequence of the discrete two-step skewing, the peak-to-peak value of the torque ripple
was reduced. Table 4.4 summarizes the torque ripple as a percentage of the mean torque for the 2D
and 3D simulations results, respectively. The torque ripple reduction is greater than 70% for the 4-
pole machine and up to 55% for the 6-pole design, and an expected slight reduction of the average
torque was also obtained.
Skewless 2D Skewless 3D 2-step skew 2D 2-step skew 3D

𝑇avg 51.0 Nm 50.2 Nm 50.0 Nm 48.7 Nm
4-pole SynRM
𝑇rp 73.9 % 72.1 % 20.2 % 22.6 %
𝑇avg 54.7 Nm 53.8 Nm 51.9 Nm 50.5 Nm
6-pole SynRM
𝑇rp 49.2 % 45.9 % 21.4 % 19.2 %
In this specific case, it can be observed that the 3D evaluation shows worse results than the 2D
assessment, since in the 3D the ripple reduction is lower, and the mean torque reduction is increased,
with respect to the 2D simulations. This can be ascribed to the fact that 2D simulations on ANSYS
consider each slice of the machine as independent machines, when discrete skewing is applied, and
the interface between slices is not taken into account, resulting in an idealization of the problem. On
the other hand, the 3D simulation evaluates the rotor as a whole unit comprised of physically rotated
slices, hence considering effects within the adjacent slices interface. This is further addressed in
section 4.5.
4.3.3. Evaluation of torque ripple reduction by means of three-step discrete skew
Figure 4.5 when adopting a three-step discrete skew strategy.
53
(a) (c)
52.2 51.1 56.1 52.8

51.5 49.9 55.1 52.3
(b) (d)

As in the case of the two-step skewing, there is a significant reduction of the highest-magnitude
harmonic component of the torque ripple, of up to 93%. This is concordance with the reduction factor
derived in (4.9), expression that can be further simplified for cases with 𝑁 = 3 as per:
1 2𝜋𝑣
𝑘rs,𝑣 = [1 + 2 cos ( )] (4.11)
3 3𝑤
According to (4.11), the reduction of the highest-magnitude component of the electromagnetic

torque, when 𝑣 = 𝑤, is maximum (𝑘rs,𝑤 = 0). Additionally, for the 4-pole machine it was found the
torque ripple harmonic components with 𝑣 = 12 and 𝑣 = 36 were reduced to ~45% and ~6% of their
original value, respectively, whilst the other relevant harmonic of order 𝑣 = 6 is not significantly
affected. This agrees with the reduction factor proposed in (11) for 𝑁 = 3, since 𝑘rs,6 = 0.84, 𝑘rs,12 =
0.45 and 𝑘rs,36 = 0. On the other hand, for the 6-pole machine it was observed the harmonic
components with 𝑣 = 6 and 𝑣 = 24 were reduced to ~70% and ~12% of their original value,
respectively, whilst the harmonic of order 𝑣 = 36 is not visibly reduced. This agrees with the values
provided by (4.11), since 𝑘rs,6 = 0.67, 𝑘rs,24 = 0 and 𝑘rs,36 = 1. Although they are not relevant
contributors, the trend of other harmonic components of both machines are in good agreement with
the expression presented in (4.11). These findings are summarized in Table 4.5.
54

𝑣=6 0.84 0.95 0.67 0.70
𝑣 = 12 0.45 0.45 0.00 0.06
𝑣 = 18 0.00 0.09 - -
𝑣 = 24 - - 0.00 0.12
𝑣 = 30 - - - -
𝑣 = 36 0.00 0.06 1.00 0.90
As a result of the discrete three-step skewing, the peak-to-peak value of the torque ripple was
considerably reduced. Table 4.6 presents the obtained mean torque and torque ripple (as a
percentage of the mean torque) for the 2D and 3D simulations respectively. It may be noted that the
torque ripple reduction is greater than 75% in both designs analyzed and there is an expected slight
reduction of the average torque.
Similar to the case of two-step skewing, it can be appreciated that the 3D evaluation shows a
worse outcome than the 2D assessment, since a lower ripple reduction and a higher mean torque
reduction are achieved.

𝑇avg 51.0 Nm 50.2 Nm 49.9 Nm 48.7 Nm
4-pole SynRM
𝑇rp 73.9 % 72.1 % 15.6 % 18.3 %
𝑇avg 54.7 Nm 53.8 Nm 51.5 Nm 51.0 Nm
6-pole SynRM
𝑇rp 49.2 % 45.9 % 11.8 % 12.7 %
4.4.4. Evaluation of torque ripple reduction by means of four-step discrete skew

Figure 4.6 when four-step skewing is applied. It can be observed that there is a significant reduction
in the harmonic torque component to be mitigated, to around 10% of its original magnitude, in a
similar extent to the 2-step and 3-step skewing.
The information in Figure 4.8 agrees with the estimations obtained from the reduction factor
proposed in (4.9), which, for the specific case of 𝑁 = 4 can be simplified to:
√2 𝜋𝑣 𝜋𝑣
𝑘rs,𝑣 = cos ( ) √1 + cos ( ) (4.12)
2 2𝑤 2𝑤
55
(a) (c)
52.2 51.0 56.1 52.6

51.5 50.1 55.1 52.7
(b) (d)

Based on (4.12), the reduction of the highest-magnitude component of the electromagnetic

torque, when 𝑣 = 𝑤, is maximum (𝑘rs,𝑤 = 0). Additionally, for the 4-pole machine it was found the
torque ripple harmonic components with 𝑣 = 12 and 𝑣 = 36 were reduced to ~35% and ~5% of their
original value, respectively, whilst the other relevant harmonic of order 𝑣 = 6 is not affected. This
agrees with the reduction factor proposed in (4.12) for 𝑁 = 3, since 𝑘rs,6 = 0.84, 𝑘rs,12 = 0.43 and
𝑘rs,36 = 0. On the other hand, for the 6-pole machine it was observed the harmonic components with
𝑣 = 6, 𝑣 = 24 and 𝑣 = 36 were reduced to ~74%, ~12% and ~6% of their original value, respectively.
This agrees with the values provided by (4.12), since 𝑘rs,6 = 0.65, 𝑘rs,24 = 0 and 𝑘rs,36 = 0. Although
they are not relevant contributors, the trend of other harmonic components of both machines are in
good agreement with the expression presented in (4.12). The summary of these results is presented
in Table 4.7.

𝑣=6 0.84 1.01 0.65 0.74
𝑣 = 12 0.43 0.35 0.00 0.06
𝑣 = 18 0.00 0.08 - -
𝑣 = 24 - - 0.00 0.12
𝑣 = 30 - - - -
𝑣 = 36 0.00 0.05 0.00 0.06
56
Four-step skewing resulted into a severe torque ripple reduction, as summarized in Table 4.8,
which presents the results obtained by means of the 2D and 3D evaluations, respectively. The torque
ripple reduction is greater than 75% in both designs analyzed and there is an expected slight
reduction of the average torque.

𝑇avg 51.0 Nm 50.2 Nm 49.8 Nm 48.7 Nm
4-pole SynRM
𝑇rp 73.9 % 72.1 % 15.0 % 18.3 %
𝑇avg 54.7 Nm 53.8 Nm 51.4 Nm 51.0 Nm
6-pole SynRM
𝑇rp 49.2 % 45.9 % 11.2 % 12.7 %
Close to the case of two-step and three-step skewing, it may be appreciated that the 3D
evaluation shows a worse outcome than the 2D assessment, since a lower ripple reduction and a
higher mean torque reduction are achieved.
4.4. Comparison, analysis, and recommendations

The flux density distribution for the 6-pole SynRM is show in Figure 4.7 for the skewless rotor,
and for the 2, 3 and 4-steps skew models. It can be observed that the tangential bridges of the barriers
are well saturated. From the 3D plots of the flux density distribution the number of steps used for
the skewing is clearly highlighted. The step used for the skew of each case becomes smaller as the
number of steps increases. SynRMs are affected by the leakage flux and the cross-coupling effect
between the dq-axes. The cross-coupling between the different modules causes a leakage between
each slice, which is only taken into account by 3D simulations. All the above-mentioned phenomena
are affected by the skew angle: a smaller skew angle causes have less influence on this effect, and
thus on the performance of the machine.
When comparing the results of the 2D with those of 3D FEA simulations, a slight difference
can be observed. First, the average torque reduction due to skew in 3D simulation is greater than the
2D simulation, which occurs because the step-step leakage contributes to the reduction of the average
torque. Nevertheless, the torque ripple reduction in both cases is quite similar. As a result, it can be
said that during the preliminary design stage 2D simulations can be used to save time and
understand the effects of discrete skew method on torque performance.
57
(a) (b)
(c) (d)
Figure 4.7. Flux density distribution in a 6-pole machine with two barriers per pole for 𝐽 = 10 A/mm2
and 𝜃𝑟,𝑒 = 0. (a) skewless machine; (b) 2-step skew machine; (c) 3-step skew machine; (d) 4-step skew machine.
It is possible to observe that, when N=2 and N=3 is compared, there is a considerable
improvement in the reduction of the torque ripple, and the behavior is different for N=3 and N=4,
where the reduction is the same. For all cases, N=2, N=3 and N=4, the average torque remains
relatively constant. Therefore, when assessing close-to-purely-sinusoidal electromagnetic torque
waveforms in a SynRM, then 2-step skew may be sufficient as a single ripple component must be
mitigated. Conversely, if there are several preponderant harmonic components, then it is worth
taking a multi-step skewing approach, to mitigate multiple harmonics at once. In consequence, it is
necessary to correctly analyze the harmonic distribution of the electromagnetic torque to properly
choose N, as increasing the number of steps does not always guarantee a significant reduction in
torque ripple and could lead to other different manufacturing costs.
58
4.5. Summary
A study on the discrete-skew methodology was conducted and a method was proposed to
better understand the impact of skewing angle and its determination during the design phase of
SynRM. After reviewing relevant literature, a reduction factor for each harmonic component was
introduced and derived in general form to estimate the torque ripple component amplitudes as a
function of skew. This allowed for the calculation of the overall torque ripple waveform. The
proposed method was validated by evaluating two SynRMs and achieving a torque ripple reduction
of up to 70%. Results from the analysis and FEA evaluation showed good agreement. The results
suggest that 2D FEA is preferred over computationally intensive 3D simulations to assess the
performance of discrete skew. The harmonic distribution of torque ripple can be used to select the
best skewing strategy: a 2-step skewing is recommended for mostly-purely-sinusoidal waveforms,
while a multi-step skewing is advised for machines with multiple high-magnitude harmonic
components.
59
Chapter V: Proposed design of a Synchronous Reluctance Machine
This chapter utilizes the techniques outlined in previous chapters to design a SynRM prototype
through an optimization process based on genetic algorithms and FEM simulations. The
optimization yields two designs, one symmetrical and one asymmetrical, which are evaluated for
average torque, torque ripple, saliency ratio, and power factor. The optimal designs are further
scrutinized mechanically to ensure they meet established standards for rotor integrity.
5.1. Set-up of the optimization process
Figure 5.1 presents the rotor parameterization for one pole of a SynRM with three flux barriers
per pole. The optimization process was carried out using ANSYS commercial software package
(Electronic Desktop for electromagnetic analysis, DesignModeler for geometry parameterization,
and Workbench for optimization), combining MOGA and FEM simulations. The optimization
considered two machine configurations: a 4-pole and a 6-pole. Only rotor geometry was optimized
in both cases because torque ripple is more sensitive to rotor structure variations [115] as showed in
sensitivity analysis on Chapter 3. The number of poles was chosen because average torque is
inversely proportional to the number of poles [45], [116], making machines with a low number of
poles preferable.
wbel12 lbl1 lbr1 wber12

d wbel23 lbl2 lbr2 wber23 d
lbl3
lbr3
αbl wrb1 αbr
tb1
wbc12
wrb2 tb2
Rro wbc23 wtb
wrb3 tb3
rb
βq-axis
Rri
Figure 5.1. Sketch of the rotor of a SynRM with three flux barriers per pole. The parameters subject to
optimization are properly defined on both sides of the q-axis of the rotor.
The optimization process involves two steps. First, a symmetrical design is found where the
rotor parameters on both sides of the q-axis are identical. The second step involves optimization for
an asymmetrical design with complete freedom for the rotor parameters on either side of the q-axis.
60
Radial bridges are excluded from the electromagnetic optimization because they would significantly
prolong the optimization phase. Each step of the optimization follows the workflow outlined in
Figure 3.2 of Chapter III. The optimization starts with defining the geometry, then it is analyzed
through FEM simulation to obtain the desired performance indices.
The main data of the machine is presented in table 5.1. During optimization, the machine was
supplied with 20 A/mm2, which corresponds to approximately 3 times the rated current listed in
Table 5.1. Selecting a current between 2 and 3 times the rated current improves the torque ripple's
insensitivity to load variations [110], [111].
Table 5.1. Main data of the machine subject to optimization.
Parameter Symbol Value

Stator outer diameter 𝐷so 246 mm
Rotor outer diameter 𝐷ro 160.4 mm
Tooth height ℎt 22.8 mm
Toot width 𝑏t 9 mm
Stack length 𝑙st 120 mm
Number of slots 𝑄s 36
Number of turns 𝑁s 6
Number of pole pairs 𝑝 2-3
Synchronous speed 𝑛 5000 rpm
Rated current density 𝐽n 7.5 A⁄mm2
Stacking factor 𝑘s 0.95
Lamination thickness 𝑒 0.35 mm
The objective functions are set to minimize the torque ripple, maximize the power factor, and
maintain the average torque above a specific value (100 Nm), as shown in Table 5.2. The primary
constraint during optimization was to maintain the machine's insulation ratio within the limits
defined in literature [117], between 0.35 and 0.45. Geometric constraints were established to ensure
that feasible rotor geometries are obtained and to prevent errors or unrealistic solutions during the
optimization process. The upper and lower limits for the rotor's geometrical parameters (objective
variables) are properly defined and displayed in Table 5.3.
Table 5.2. Objective function for the optimization process.
Parameter Symbol Operator Value Objective Unit

Torque Ripple Trp - - min -
Average Torque Tavg ≥ 100 Nm
Maximum Internal Power Factor IPFmax - - max -
The optimization process was adjusted by defining a minimum quality mesh that would yield
proper torque waveform values, enabling effective extraction of average torque and torque ripple.
Iron losses were not analyzed in this stage for faster optimization. Only one torque ripple period was
61
simulated, equivalent to 60 electrical degrees in three-phase machines [102], and the sample number
was selected based on the Nyquist criteria for accurate FFT calculation.
5.2. Results from the optimization process
The optimization results are presented in Figures 5.2 and 5.3 for the 4-pole and 6-pole
machines, respectively. It can be concluded from the results that the 6-pole machine delivers a higher
average torque and a lower torque ripple compared to the 4-pole machine. The optimal design is
selected based on a trade-off between average torque and torque ripple as it is not possible to achieve
the lowest torque ripple and highest average torque simultaneously.
Table 5.3. Input variables range for the optimization process for 4-pole and 6-pole SynRM.
4-pole 6-pole
Symbol Lower Upper Lower Upper Unit
𝛼br 1 35 1 25 mech. degree
𝛼bl 1 35 1 25 mech. degree
𝑙br1 3 10 2.5 10 mech. degree
𝑙bl1 3 10 2.5 10 mech. degree
𝑤ber12 2 10 2 10 mech. degree
𝑤bel12 2 10 2 10 mech. degree
𝑤ber23 2 10 2 10 mech. degree
𝑤bel23 2 10 2 10 mech. degree
𝑡b1 5 15 5 15 mm
𝑡b2 5 15 5 15 mm
𝑡b3 5 15 5 15 mm
𝑤bc12 2 10 2 10 mm
𝑤bc23 2 10 2 10 mm
𝑤tb 0.5 2 0.5 2 mm
𝑟b 40 50 40 50 mm
𝛼ie 50 70 50 70 elect. degree
(a) (b)
62
(a) (b)
The optimized design incorporates ducts in the first rotor island to decrease the island mass
and centrifugal force. Figures 5.4 and 5.5 depict the impact of duct radius and current density on
average torque and torque ripple with a fixed current angle of 60 electrical degree. The ducts have
minimum effect on average torque, but a slight effect on torque ripple. Thus, ducts were included in
the final design to reduce island mass and the width of radial bridges, leading to quicker saturation
and fewer flux lines crossing the flux barriers. The mechanical analysis of these optimal designs will
be conducted in the next subchapter.
(a) (b)
of the duct located in the first rotor island when 𝛼𝑖𝑒 = 60° elect. degree. (a) average torque; (b) torque ripple.
(a) (b)
of the duct located in the first rotor island when 𝛼𝑖𝑒 = 60° elect. degree. (a) average torque; (b) torque ripple.
63
5.3. Structural analysis

To validate the rotor's mechanical robustness, 2D structural analysis was conducted using
ANSYS Mechanical at a maximum speed of 10,000 rpm. Radial iron bridges were added to the rotor
structure and their thicknesses were determined based on the rotor island mass, as described in [118].
Figure 6.6 shows the considered rib distribution, with innermost, middle, and outermost barriers
having thicknesses of 1.8mm, 1.5mm, and 1.4mm, respectively, for both symmetrical and
asymmetrical designs. Both designs have a tangential rib thickness of ~1.3 mm. The effects of
temperature were neglected in this analysis. The symmetrical design has a safety factor of 1.78, and
the asymmetrical design has a safety factor of 1.64, based on the 465 MPa yield stress of M250-35A
laminated steel. Figure 5.6 shows some areas of the rotor structure that are subject to high von-Mises
stress in both designs.
Figure 5.7 displays the ANSYS Mechanical results for 10,000 rpm M250-35A steel rotor show a
deformation of ~0.013 mm (~3% of the air-gap length) in the air-gap zona. This deformation falls
within the acceptable range for many applications, indicating that the rotor is likely to perform
reliably and efficiently under these operating conditions. However, it's important to keep in mind
that this deformation could be an indicator of excessive loading or wear on the rotor and require
further evaluation to ensure its continuity and safety in operation.
(a) (b)
Figure 5.6. Von Mises stress distribution for the optimal designs at 10,000 rpm. (a) symmetric design; (b)
asymmetric design. Graphical results are magnified x10 for easy viewing.
64
(a) (b)
Figure 5.7. Total deformation distribution for the optimal designs at 10,000 rpm. (a) symmetric design;
(b) asymmetric design. Graphical results are magnified x10 for easy viewing.
5.4. Optimal design adjustment

The optimal designs were analyzed for their saliency ratio, power factor, average torque, and
torque ripple across various operating points. These analyses were conducted for both clockwise and
counter-clockwise rotations. Figures 5.8, 5.9, 5.10 and 5.11 display the contour maps for the saliency
ratio, internal power factor, ripple torque, and average torque, respectively.
(a) (b)
(c) (d)
Figure 5.8. Saliency ratio for the symmetric and asymmetric design. (a) symmetric design counter-
clockwise rotation; (b) asymmetric design counter-clockwise rotation; (c) symmetric design clockwise rotation;
(b) asymmetric design clockwise rotation.
65
(a) (b)
(c) (d)
Figure 5.9. Internal power factor for the symmetric and asymmetric design. (a) symmetric design
counter-clockwise rotation; (b) asymmetric design counter-clockwise rotation; (c) symmetric design clockwise
rotation; (b) asymmetric design clockwise rotation.
(a) (b)
(c) (d)
Figure 5.10. Torque ripple for the symmetric and asymmetric design. (a) symmetric design counter-
66
(a) (b)
(c) (d)
Figure 5.11. Mean torque for the symmetric and asymmetric design. (a) symmetric design counter-
The greatest differences are observed in the average torque and torque ripple. The
asymmetrical design presents a slightly larger zone of maximum power factor and saliency ratio,
compared to the symmetrical design. However, the symmetrical design has a lower torque ripple at
low current density levels. Despite both designs reaching a torque ripple of 20% in their respective
areas of lowest torque ripple, for electromobility applications a ripple of less than 5% is desired [15].
To choose a suitable skew angle for reducing the torque ripple to acceptable values, a FFT was
performed on the electromagnetic torque. The harmonic components of both designs under different
load conditions are shown in Figure 5.12 and Figure 5.13 for counter-clockwise and clockwise
rotation, respectively.
(a) (b) (c)
(d) (e)
asymmetric design counter-clockwise rotation; (a) current density 5 A/mm2 ; (b) current density 7.5 A/mm2 ; (c)
current density 10 A/mm2 ; (d) current density 12.5 A/mm2 ; (e) current density 15 A/mm2 .
67
(a) (b) (c)
(d) (e)
asymmetric design clockwise rotation; (a) current density 5 A/mm2 ; (b) current density 7.5 A/mm2 ; (c) current
density 10 A/mm2 ; (d) current density 12.5 A/mm2 ; (e) current density 15 A/mm2 .
Both rotational directions show similar magnitudes of harmonic components. When the
current density is below 10 A/mm2, the 12-nd harmonic is the dominant one, which is caused by the
stator slotting. When the current density is above 10 A/mm2, the 18-th harmonic is dominant in the
symmetrical design and the 24-th harmonic in the asymmetrical design. Table 5.4 shows the skew
angle that can be used to mitigate specific harmonics component, which was calculated using
equation 4.4. This analysis considered the 6-th, 12-nd, and 18-th harmonics for mitigation using 2-
step, 3-step, 4-step, and 5-step skewing.
Table 5.4. Skew angle to reduce a specific harmonic component of the electromagnetic torque.
Mechanical angle Mechanical angle Mechanical angle Mechanical angle

Harmonic order
for 2-step skew for 3-step skew for 4-step skew for 5-step skew
6th 10° 6.66° 5° 4°
12nd 5° 3.33° 2.5° 2°
18th 3.33° 2.22° 1.66° 1.33°
The harmonic distribution of the electromagnetic torque increases with increasing current
density, so it is desirable to apply skew to reduce these harmonics. The mitigation factor (defined in
Chapter IV by equation 4.9) can be used to determine the reduction of each harmonic component.
Tables 5.5, 5.6, and 5.7 report the mitigation factor when the skew is applied to the 6-th, 12-nd, and
18-th harmonics, respectively.
68
Table 5.5. Value of the mitigation factor when the first harmonic component is selected to reduce
(𝑤 = 6).
𝒌𝐫𝐬,𝒗
Harmonic order
6th 0 0 0 0
12nd 1 0 0 0
18th 0 1 0 0
24th 1 0 1 0
Table 6.6. Value of the mitigation factor when the second harmonic component is selected to reduce
(𝑤 = 12).
𝒌𝐫𝐬,𝒗
Harmonic order
6th 0.7 0.6 0.6 0.6
12nd 0 0 0 0
18th 0.7 0.3 0.3 0.2
24th 1 0 0 0
Table 6.7. Value of the mitigation factor when the third harmonic component is selected to reduce
(𝑤 = 18).
𝒌𝐫𝐬,𝒗
Harmonic order
6th 0.8 0.8 0.8 0.8
12nd 0.5 0.4 0.4 0.4
18th 0 0 0 0
24th 0.5 0.3 0.2 0.2
The analysis shows that the best option for applying skew is to mitigate the 6-th harmonic
using a 5-step skew. This eliminates the harmonic components theoretically, but in practice, the
reduction is not complete. Figures 5.14 and 5.15 show the harmonic components of the optimal
symmetrical and asymmetrical designs for counter-clockwise and clockwise rotation, respectively,
with a 5-step skew applied.
69
(a) (b) (c)
(d) (e)
asymmetric design applying four step skew and counter-clockwise rotation; (a) current density 5 A/mm2 ; (b)
current density 7.5 A/mm2 ; (c) current density 10 A/mm2 ; (d) current density 12.5 A/mm2 ; (e) current density
15 A/mm2 .
(a) (b) (c)
(d) (e)
asymmetric design applying four step skew and clockwise rotation; (a) current density 5 A/mm2 ; (b) current
density 7.5 A/mm2 ; (c) current density 10 A/mm2 ; (d) current density 12.5 A/mm2 ; (e) current density 15 A/mm2 .
In both rotational directions, the magnitude of the harmonic components is similar when a 5-
step skew is applied. When the current density is below 10 A/mm2, the predominant harmonic
component is the 12-nd, corresponding to the stator slotting. However, when the current density is
greater than 10 A/mm2, the predominant harmonic component remains the 12-nd, differing from the
behavior of the skewless model.
Although the mitigation factor suggests that the first to fourth harmonic components should
be reduced to zero, in practice this is not achieved. The reduction of the harmonic components is
approximately 50%, and the reduction is greater in the asymmetric design. The greater reduction in
70
the harmonic component in the asymmetric design is due to the asymmetrical positioning of the flux
barriers, which provides additional contribution in mitigating the spatial harmonic components in
the electromagnetic torque.
Figures 5.16, 5.17, 5.18 and 5.19 show the contour maps for the optimal symmetric and
asymmetric designs, respectively, for counter-clockwise and clockwise rotation, with a 5-step skew
applied, for the main performance indexes.
(a) (b)
(c) (d)
Figure 5.16. Saliency ratio for the symmetric and asymmetric design with four step skew is applied. (a)
symmetric design counter-clockwise rotation; (b) asymmetric design counter-clockwise rotation; (c) symmetric
(a) (b)
(c) (d)
Figure 5.17. Internal power factor for the symmetric and asymmetric design with four step skew is
applied. (a) symmetric design counter-clockwise rotation; (b) asymmetric design counter-clockwise rotation;
(c) symmetric design clockwise rotation; (b) asymmetric design clockwise rotation.
71
(a) (b)
(c) (d)
Figure 5.18. Torque ripple for the symmetric and asymmetric design with four step skew is applied. (a)
(a) (b)
(c) (d)
Figure 5.19. Mean torque for the symmetric and asymmetric design with four step skew is applied. (a)
The greatest differences are observed in the average torque and torque ripple. The asymmetric
design has a larger zone of maximum internal power factor and saliency ratio and can develop higher
mean torque than the symmetrical design with lower torque ripple. In both rotational directions, the
region with torque ripple less than 10% is larger in the asymmetric design, reaching less than 5%
when rotated counter-clockwise.
72
Figure 5.20 depict the overlay of the contour curves of internal power factor, torque ripple, and
mean torque over a specific range of current density and current angle for counter-clockwise rotation
in symmetric and asymmetric designs, respectively. Two operating points in the MTPA trajectory
corresponding to 5 A/mm2 for the inverted blue triangle and 10 A/mm2 for the red triangle have been
highlighted in the figure. The performance indices for the defined operation points are listed in table
5.8. The internal power factor and efficiency values are comparable between the symmetric and
asymmetric designs. The main difference between the symmetric and asymmetric designs is found
in the average torque and torque ripple, where the asymmetric design yields a 25% reduction in
torque ripple and a 5% increase in average torque compared to the symmetric design.
(a) (b)
current density and current angle with counter-clockwise rotation. Two operating points were defined for
MTPA for 7.5 A/mm2 and 10 A/mm2. (a) symmetric design; (b) asymmetric design.
Operation Tavg Trp P (Pcu + Pfe ) η Tavg Trp 𝑃 (Pcu + Pfe ) η
IPF IPF
point [Nm] [%] [kW] [kW] [%] [Nm] [%] [kW] [kW] [%]
0.49 28.7 6.7 15.0 1.7 88.3 0.51 30.2 4.9 15.8 1.8 88.4
0.51 43.6 7.5 22.8 2.3 89.4 0.51 45.9 5.4 24.0 2.4 89.6
Figure 5.21 show the overlap of contour curves for internal power factor, torque ripple, and
mean torque within a specified range of current density and current angle for the symmetric and
asymmetric design for clockwise rotation, respectively. Two operating points in the MTPA trajectory
corresponding to 5 A/mm2 for the inverted blue triangle and 10 A/mm2 for the red triangle have been
highlighted in the figure. The performance indices for the defined operation points are listed in table
5.9. Similar to counter-clockwise rotation, the power factor and efficiency values are nearly equal
between symmetrical and asymmetrical designs. The major distinction lies in the average torque and
73
torque ripple, where the asymmetrical design offers a 20% reduction in torque ripple and a 5%
increase in average torque compared to the symmetrical design.
(a) (b)
current density and current angle with clockwise rotation. Two operating points were defined for MTPA for
7.5 A/mm2 and 10 A/mm2. (a) symmetric design; (b) asymmetric design.
asymmetrical designs when the machine rotated clockwise.
Operation Tavg Trp P (Pcu + Pfe ) η Tavg Trp 𝑃 (Pcu + Pfe ) η
IPF IPF
point [Nm] [%] [kW] [kW] [%] [Nm] [%] [kW] [kW] [%]
0.5 28.7 6.5 15.0 1.7 88.3 0.51 30.3 5.5 15.8 1.8 88.3
0.51 43.6 7.5 22.8 2.3 89.4 0.51 45.9 5.8 24 2.4 89.6
The electromagnetic torque waveform as a function of rotor position is shown in Figure 5.22
for both symmetrical and asymmetrical designs in counter-clockwise rotation. It is evident from the
figure that the asymmetric design produces a higher average torque than the symmetrical one. When
skew is not applied, the torque ripple in both designs is comparable. However, with the application
of skew, the asymmetric design developed lower torque ripple than the symmetrical one, but the
average torque is compromised in both designs.
74
(a) (b)
Symmetric J=7.5A/mm2 Symmetric J=10A/mm2 Asymmetric J=7.5A/mm2 Asymmetric J=10A/mm2
Figure 5.22. Symmetric and asymmetric design electromagnetic torque waveform for MTPA when the
current density is 7.5 A/mm2 and 10 A/mm2 and the machine rotated counter-clockwise. (a) skewless designs;
(b) 5-step skew designs.
The magnetic field distribution of the designs is shown in Figures 5.23 and 5.24 for 7.5 A/mm2
and 10 A/mm2, respectively. It is seen that the highest levels of saturation are found in the end-point
parts of the barriers and ducts what is expected in this topology of machines. In general, both
machines operate near the saturation knee of the B-H curve, taking advantage of the magnetic
properties of the material.
(a) (b)
Figure 5.23. Magnetic flux density distribution for the optimal designs for MTPA at 7.5 A/mm 2. (a)
symmetric design counter-clockwise rotation; (b) asymmetric design counter-clockwise rotation.
75
(a) (b)
Figure 5.24. Magnetic flux density distribution for the optimal design for MTPA at 10 A/mm 2. (a)
symmetric design counter-clockwise rotation; (b) asymmetric design counter-clockwise rotation.
5.5. Impact of rotor skewing on the performance of the machine

Figures 5.25 and 5.26 illustrate the impact of applying skew on different performance indices
of a SynRM, for both symmetric and asymmetric designs, as a function of the supply current density.
The results reveal that applying skew significantly reduces all machine indices in both scenarios. In
a SynRM, saliency ratio is a critical parameter that directly affects electromagnetic torque, power
factor, and efficiency. However, the saliency ratio decreases by more than 20% for all current values
when skew is applied to the rotor, leading to a decrease in IPF and mean torque of the machine, as
shown in Figures 5.25 and 5.26 (c) and (d), respectively. Furthermore, torque ripple is significantly
reduced for both designs, achieving a reduction of around 60% and 70% for the symmetric and
asymmetric designs, respectively, at 10 A/mm2. The asymmetric design offers a lower torque ripple
for all current density levels when skew is applied. The efficiency is also slightly affected by the
application of skew in both designs.
(a) (b) (c)
(d) (e) (f)

Skewless 5-step skew
Figure 5.25. Performance indices for the symmetric design for MTPA for different current density levels.
(a) current angle; (b) saliency ratio; (c) maximum internal power factor; (d) average torque; (e) torque ripple;
(f) efficiency.
76
(a) (b) (c)
(d) (e) (f)

Figure 5.26. Performance indices for the asymmetric design for MTPA for different current density
levels. (a) current angle; (b) saliency ratio; (c) maximum internal power factor; (d) average torque; (e) torque
ripple; (f) efficiency.
The development of SynRM with a higher saliency ratio is a critical factor for expanding the
motor's application in various fields, including electric vehicles, as it enables the development of a
larger CPSR. Typically, the CPSR is dependent on the saliency ratio, which underscores the
importance of designing a motor with an improved saliency ratio to broaden the application of
SynRM. Figure 5.27 confirms that a decrease in the saliency ratio leads to a decrease in the CPSR.
Furthermore, applying 5-step skew leads to a reduction in the machine's torque and power across all
speeds.
(a) (b)
(c) (d)
Figure 5.27. CPSR performance for the symmetric and asymmetric design for maximum current density
of 10A/mm2. (a) torque vs. speed for the symmetric design; (b) power vs. speed for the symmetric design; (c)
torque vs. speed for the asymmetric design; (d) power vs. speed for the asymmetric design.
77
To assess the performance strength of the studied designs, an efficiency map for each design
was created using ANSYS Electronic Desktop's Machine-Toolkit. The machine was supplied with
DC bus voltage of 600 V and operated for MTPA at a maximum current density of 10 A/mm2. The
efficiency maps are presented in Figure 5.28 for the analyzed rotor structures for counter-clockwise
rotation.
Figure 5.28 shows that the efficiency is also reduced when skewing is applied. This occurs
because the machine produces less power at the shaft for a given input power, as a result of cross-
coupling between the various modules and leakage between each slice. Both designs exhibit a similar
behavior, reaching 96% efficiency for the skewless machine and 95% for the 5-step skew machine at
the rated speed.
(a) (b)
(c) (d)
Figure 5.28. Efficiency map when the machine is operated a MTPA for a maximum current density of
10A/mm2. (a) skewless symmetric design; (b) skewless asymmetric design; (c) 5-step skew symmetric design;
(d) 5-step skew asymmetric design.
5.6. Manufactured prototype

Figure 5.29 presents a view of the motor's components. The stator and rotor cores are
constructed from 0.35 mm non-oriented silicon steel M350-35A laminations and feature 36 slots
housing a three-phase single-layer distributed winding. The winding is configured with 6 turns, with
each turn consisting of 10 parallel wires, resulting in a fill factor of ~0.45.
To maintain safe operating temperatures for the motor, the housing incorporates a spiral water
jacket, ensuring efficient heat dissipation. The housing itself is constructed from aluminum, which
not only enhances heat transfer capabilities but also reduces the overall weight of the motor. This
combination of materials and design elements contributes to the motor's optimal performance and
life-time.
78
(a) (b)
(c) (d)
Figure 5.29. Exploded view of the motor assembly. (a) asymmetric rotor; (b) asymmetric rotor; (c) stator
lamination and winding assembled in the housing (jacket water); (d) shaft.
5.7. Summary
This chapter described the design tools used to create symmetrical and asymmetrical rotor
models for a SynRM using the same stator. An optimization was performed using MOGA and FEM
simulations to obtain the designs, and discrete skew was applied in 5-steps to reduce torque ripple.
The contour maps of the main performance indices for both designs show that the asymmetric design
is superior in terms of average torque and torque ripple for specific operating points. The optimal
designs were evaluated mechanically to ensure the rotor's mechanical integrity and conform to
standards for a maximum speed of 10,000 rpm. The magnetic flux density was illustrated in both
designs and showed that saturation levels occur in the end-point parts of the barriers and ducts. The
influence of applying skew on the performance of symmetric and asymmetric models was analyzed.
79
It is possible to observe that most of the performance indexes decrease when skew is applied on the
rotor structure, but also a reduction of more than 50% in the torque ripple is achieved. Both machines
had similar efficiency, but the asymmetric design generated more torque under the torque-speed
curve. Finally, a first view of the prototype is presented, highlighting its different parts and materials.
80
6. Conclusions and future work
6.1. Conclusion
The main aim of this thesis is to offer a comprehensive understanding of Synchronous
Reluctance Machines (SynRMs), addressing both the fundamental principles of operation and the
development of various design techniques. The research conducted in this thesis has successfully
contributed to the enhancement of SynRMs' performance by developing design techniques and
guidelines that primarily focus on increasing rotor anisotropy. Through the optimization of rotor
design, the potential for significant improvements in machine efficiency and overall performance
was demonstrated.
The incorporation of an analytical model in the electric machine design process is an effective
means to simplifying the design and save valuable time. A significant achievement of this study is
the development and validation of a precise analytical model for SynRMs. This model combines two
approaches: the calculation of air-gap flux density and average torque employs the magnetic
potential of both the rotor and stator, while torque ripple is determined by assessing the energy
stored in the air-gap. This model is also adaptable for machines featuring multiple flux barriers.
Comparisons with Finite Element Analysis (FEA) have produced promising results,
particularly in terms of air-gap flux density and electromagnetic torque. However, a harmonic
analysis has revealed that the analytical model tends to overestimate air-gap flux density and torque
due to certain underlying assumptions made during its development. Nevertheless, the model offers
valuable functionalities, including the capacity to extract machine parameters within the d-q
reference frame, facilitating preliminary control strategy analysis. This model equips designers and
researchers with a valuable tool for predicting the machine's performance under various operational
conditions, thereby speeding up the design and analysis process.
The findings of this work demonstrate that an asymmetrical rotor structure in a SynRM
provides better performance than a symmetrical rotor structure. The advantage of the asymmetric
rotor structure with respect to machine behavior varies depending on the design technique
employed. By selecting different machine parameters during the optimization phase, it is possible to
enhance various performance indices. The position and opening of the flux barrier in the air-gap
govern torque ripple, while the thickness of the flux barriers along the q-axis regulates mean torque.
Combining all these parameters with the current angle improves the power factor when the
maximum torque is achieved at a larger current angle. This analysis was conducted in the context of
two different SynRM topologies. The first one featured a 48-slot, 2-pole pairs configuration with 2
flux barriers per pole, and the other was a 36-slot, 3-pole pairs configuration with 3 flux barriers per
pole.
In both cases, the asymmetric design managed torque ripple more effectively due to its ability
to cancel air-gap harmonics, while the mean torque remained relatively stable. However, in the 48-
slot topology, there was only a slight improvement in the power factor since it was selected as the
objective function during the optimization stage. By quantifying the impact of these asymmetries in
each topology, a deeper understanding of the machine's behavior was gained, enabling more precise
design decisions to be made.
The imperative of achieving less than 15% torque ripple in electromobility applications
necessitated the implementation of skew in the design of the two rotor prototypes. To address this
challenge, a study on the discrete-skew methodology was conducted, and a method was proposed
81
to gain a deeper understanding of the impact of skewing angle and its determination during the
SynRM design phase. This method takes into account the influence on all torque harmonic
components and offers a means to visualize the reduction of specific undesired harmonic content. A
skew reduction factor was introduced, which theoretically inform the potential of the method to
completely eliminate a desired harmonic component. However, in practical application, complete
elimination is not feasible, and the reduction of the selected harmonic component is usually around
80%.
By applying this technique during the design phase of the prototypes, it becomes possible to
ensure that torque ripple remains below 15% for nearly all the operation points analyzed.
Nevertheless, a significant reduction in mean torque is observed when skewing is applied. The
reduction is not limited to the desired harmonic component; rather, it extends to the other remaining
components. Furthermore, the cross-coupling effect between the q-axis and d-axis is amplified, and
a new phenomenon of cross-coupling effect emerges between the different sections into which the
rotor is divided for skew application. For both rotor structures, the application of skewing techniques
results in an approximate 20% reduction in mean torque compared to the skewless machine. The
saliency ratio and the power factor are also reduced by applying skew, but the efficiency doesn’t
show a remarkable change. An important fact in applying this technique is that it is not necessary to
use 3D FEA simulation for its application, 2D FEA is effective in predicting the optimal step skewing
which allows a considerable time-saving in the design stage.
6.2. Future work
A critical aspect of this thesis involved analyzing and comparing the performance of SynRM
with enhanced rotor anisotropy, incorporating asymmetries, against traditional designs across
various operating conditions. Our findings have clearly demonstrated the superior performance of
the optimized designs, highlighting the advantages of rotor asymmetries in enhancing machine
efficiency.
The practical feasibility of the optimized SynRM designs, including rotor asymmetries, was
rigorously validated through FEA and the next step is the experimental testing of the prototypes.
Two rotor structures sharing a common stator were built for the experimental tests. This validation
will confirm that the proposed design techniques are not only theoretically sound but also real-world
applicable, thus offering engineers and designers reliable solutions for improving machine
performance.
To validate the analytical model, a unique rotor prototype without bridges at the flux barriers
was fabricated and installed on an induction machine stator. The absence of the required tools for
measuring torque ripple has delayed the validation of the model through experimental testing.
However, I anticipate being able to conduct this experiment in the near future. Another essential
feature that is needed for improved results, particularly in the context of SynRMs, is the
incorporation of saturation into the analytical model.
The incorporation of modulation techniques to eliminate combinations of harmonic
components and reduce torque ripple without impacting the mean torque can be integrated as an
additional feature into the skew method developed in the thesis.
82
References
[1] M. U. Naseer, A. Kallaste, B. Asad, T. Vaimann, and A. Rassõlkin, “Analytical modelling of synchronous
reluctance motor including non-linear magnetic condition,” IET Electr Power Appl, vol. 16, no. 4, pp.
511–524, Apr. 2022, doi: 10.1049/ELP2.12172.
[2] M. N. F. Ibrahim, A. S. Abdel-Khalik, E. M. Rashad, and P. Sergeant, “An Improved Torque Density
Synchronous Reluctance Machine With a Combined Star-Delta Winding Layout,” IEEE Transactions on
Energy Conversion, vol. 33, no. 3, pp. 1015–1024, Sep. 2018, doi: 10.1109/TEC.2017.2782777.
[3] K. B. Tawfiq, M. N. Ibrahim, E. E. El-Kholy, and P. Sergeant, “Performance Improvement of

Synchronous Reluctance Machines—A Review Research,” IEEE Trans Magn, vol. 57, no. 10, pp. 1–11,
Oct. 2021, doi: 10.1109/TMAG.2021.3108634.
[4] M. Murataliyev, M. Degano, M. Di Nardo, N. Bianchi, and C. Gerada, “Synchronous Reluctance

Machines: A Comprehensive Review and Technology Comparison,” Proceedings of the IEEE, vol. 110,
no. 3, pp. 382–399, Mar. 2022, doi: 10.1109/JPROC.2022.3145662.
[5] D. Gerada, A. Mebarki, N. L. Brown, C. Gerada, A. Cavagnino, and A. Boglietti, “High-speed electrical
machines: Technologies, trends, and developments,” IEEE Transactions on Industrial Electronics, vol. 61,
no. 6, pp. 2946–2959, 2014, doi: 10.1109/TIE.2013.2286777.
[6] C. Babetto, G. Bacco, and N. Bianchi, “Analytical Power Limits Curves of High-Speed Synchronous
Reluctance Machines,” IEEE Trans Ind Appl, vol. 55, no. 2, pp. 1342–1350, Mar. 2019, doi:
10.1109/TIA.2018.2875663.
[7] B. Wang, J. Wang, B. Sen, A. Griffo, Z. Sun, and E. Chong, “A Fault-Tolerant Machine Drive Based on
Permanent Magnet-Assisted Synchronous Reluctance Machine,” IEEE Trans Ind Appl, vol. 54, no. 2, pp.
1349–1359, Mar. 2018, doi: 10.1109/TIA.2017.2781201.
[8] C. Babetto and N. Bianchi, “Synchronous Reluctance Motor with Dual Three-Phase Winding for Fault-
Tolerant Applications,” Proceedings - 2018 23rd International Conference on Electrical Machines, ICEM 2018,
pp. 2297–2303, Oct. 2018, doi: 10.1109/ICELMACH.2018.8506771.
[9] H. Heidari et al., “A Review of Synchronous Reluctance Motor-Drive Advancements,” Sustainability

2021, Vol. 13, Page 729, vol. 13, no. 2, p. 729, Jan. 2021, doi: 10.3390/SU13020729.
[10] M. Villani, G. Fabri, A. Credo, L. Di Leonardo, and F. Parasiliti Collazzo, “Line-Start Synchronous
Reluctance Motor: A Reduced Manufacturing Cost Avenue to Achieve IE4 Efficiency Class,” IEEE
Access, vol. 10, pp. 100094–100103, 2022, doi: 10.1109/ACCESS.2022.3208154.
[11] N. G. Ozcelik, U. E. Dogru, M. Imeryuz, and L. T. Ergene, “Synchronous Reluctance Motor vs. Induction
Motor at Low-Power Industrial Applications: Design and Comparison,” Energies 2019, Vol. 12, Page
2190, vol. 12, no. 11, p. 2190, Jun. 2019, doi: 10.3390/EN12112190.
[12] H. Heidari et al., “A Review of Synchronous Reluctance Motor-Drive Advancements,” Sustainability,

vol. 13, no. 2, p. 729, Jan. 2021, doi: 10.3390/su13020729.
[13] A. Fratta, G. P. Toglia, A. Vagati, and F. Villata, “Ripple evaluation of high-performance synchronous
reluctance machines,” IEEE Industry Applications Magazine, vol. 1, no. 4, pp. 14–22, 1995, doi:
10.1109/2943.392459.
83
[14] J. Liang, Y. Dong, H. Sun, R. Liu, and G. Zhu, “Flux-Barrier Design and Torque Performance Analysis
of Synchronous Reluctance Motor with Low Torque Ripple,” Applied Sciences 2022, Vol. 12, Page 3958,
vol. 12, no. 8, p. 3958, Apr. 2022, doi: 10.3390/APP12083958.
[15] W. Zhao, Y. Sun, J. Ji, Z. Ren, and X. Song, “Phase Shift Technique to Improve Torque of Synchronous
Reluctance Machines with Dual M-Phase Windings,” IEEE Transactions on Industrial Electronics, vol. 69,
no. 1, pp. 5–17, Jan. 2022, doi: 10.1109/TIE.2021.3050359.
[16] M. N. F. Ibrahim, A. S. Abdel-Khalik, E. M. Rashad, and P. Sergeant, “An Improved Torque Density
Synchronous Reluctance Machine With a Combined Star-Delta Winding Layout,” IEEE Transactions on
Energy Conversion, vol. 33, no. 3, pp. 1015–1024, Sep. 2018, doi: 10.1109/TEC.2017.2782777.
[17] N. Bianchi, S. Bolognani, D. Bon, and M. Dai Pré, “Rotor flux-barrier design for torque ripple reduction
in synchronous reluctance motors,” Conference Record - IAS Annual Meeting (IEEE Industry Applications
Society), vol. 3, pp. 1193–1200, 2006, doi: 10.1109/IAS.2006.256683.
[18] C. Gallardo, J. A. Tapia, M. Degano, H. Mahmoud, and A. E. Hoffer, “Rotor Asymmetry Impact on
Synchronous Reluctance Machines Performance,” 2022 International Conference on Electrical Machines,
ICEM 2022, pp. 848–854, 2022, doi: 10.1109/ICEM51905.2022.9910665.
[19] J. K. Kostko, “Polyphase reaction synchronous motors,” Journal of the American Institute of Electrical
Engineers, vol. 42, no. 11, pp. 1162–1168, Nov. 2013, doi: 10.1109/joaiee.1923.6591529.
[20] A. Vagati, “Synchronous reluctance solution: a new alternative in a.c. drives,” in IECON Proceedings
(Industrial Electronics Conference), IEEE, 1994, pp. 1–13. doi: 10.1109/iecon.1994.397741.
[21] L. Xu and J. Yao, “A compensated vector control scheme of a synchronous reluctance motor including
saturation and iron losses,” in Conference Record of the 1991 IEEE Industry Applications Society Annual
Meeting, IEEE, 1992, pp. 298–304. doi: 10.1109/ias.1991.178170.
[22] A. Vagati, M. Pastorelli, and G. Franceschini, “High-performance control of synchronous reluctance

motors,” IEEE Trans Ind Appl, vol. 33, no. 4, pp. 983–991, 1997, doi: 10.1109/28.605740.
[23] R. E. Betz, R. Lagerquist, M. Jovanovic, T. J. E. Miller, and R. H. Middleton, “Control of synchronous

reluctance machines,” IEEE Trans Ind Appl, vol. 29, no. 6, pp. 1110–1122, 1993, doi: 10.1109/28.259721.
[24] N. G. Ozcelik, U. E. Dogru, M. Imeryuz, and L. T. Ergene, “Synchronous reluctance motor vs. Induction
motor at low-power industrial applications: Design and comparison,” Energies (Basel), vol. 12, no. 11, p.
2190, Jun. 2019, doi: 10.3390/en12112190.
[25] T. A. Lipo, “Synchronous reluctance machines-a viable alternative for ac drives?,” Electric Machines and
Power Systems, vol. 19, no. 6, pp. 659–671, 1991, doi: 10.1080/07313569108909556.
[26] X. D. Xue, K. W. E. Cheng, and N. C. Cheung, “Selection of electric motor drives for electric vehicles,”
2008 Australasian Universities Power Engineering Conference, AUPEC 2008, 2008, Accessed: Sep. 27, 2020.
[Online]. Available: https://ieeexplore.ieee.org/document/4813059
[27] R. R. Moghaddam, F. Magnussen, C. Sadarangani, and H. Lendenmann, “New theoretical approach to

the synchronous reluctance machine behavior and performance,” in Proceedings of the 2008 International
Conference on Electrical Machines, ICEM’08, IEEE, Sep. 2008, pp. 1–6. doi:
10.1109/ICELMACH.2008.4799845.
84
[28] D. A. Staton, T. J. E. Miller, and S. E. Wood, “Maximising the saliency ratio of the synchronous
reluctance motor,” IEE Proceedings B: Electric Power Applications, vol. 140, no. 4, pp. 249–259, Jul. 1993,
doi: 10.1049/ip-b.1993.0031.
[29] M. J. Kamper and A. F. Volschenk, “Effect of rotor dimensions and cross magnetisation on Ld and Lq
inductances of reluctance synchronous machine with cageless flux barrier rotor,” IEE Proceedings:
Electric Power Applications, vol. 141, no. 4, pp. 213–220, Jul. 1994, doi: 10.1049/IP-EPA:19941261.
[30] T. Matsuo and T. A. Lipo, “Rotor Design Optimization of Synchronous Reluctance Machine,” IEEE
Transactions on Energy Conversion, vol. 9, no. 2, pp. 359–365, Jun. 1994, doi: 10.1109/60.300136.
[31] J. Ahn et al., “Field weakening control of synchronous reluctance motor for electric power steering,” IET
Electr Power Appl, vol. 1, no. 4, p. 565, 2007, doi: 10.1049/iet-epa:20060212.
[32] S. M. Ferdous, P. Garcia, M. A. M. Oninda, and A. Hoque, “MTPA and field weakening control of
synchronous reluctance motor,” Proceedings of 9th International Conference on Electrical and Computer
Engineering, ICECE 2016, pp. 598–601, 2017, doi: 10.1109/ICECE.2016.7853991.
[33] A. E. Hoffer, R. H. Moncada, B. J. Pavez, J. A. Tapia, and L. Laurila, “A high efficiency control strategy
for synchronous reluctance generator including saturation,” Proceedings - 2016 22nd International
Conference on Electrical Machines, ICEM 2016, pp. 39–45, 2016, doi: 10.1109/ICELMACH.2016.7732503.
[34] V. Manzolini, D. Da Ru, and S. Bolognani, “An Effective Flux Weakening Control of a SyRM Drive
Including MTPV Operation,” IEEE Trans Ind Appl, vol. 55, no. 3, pp. 2700–2709, 2019, doi:
10.1109/TIA.2018.2886328.
[35] Y. Zhao, L. Ren, Z. Liao, and G. Lin, “A novel model predictive direct torque control method for
improving steady-state performance of the synchronous reluctance motor,” Energies (Basel), vol. 14, no.
8, 2021, doi: 10.3390/en14082256.
[36] E. M. Rashad, T. S. Radwan, and M. A. Rahman, “A maximum torque per ampere vector control
strategy for synchronous reluctance motors considering saturation and iron losses,” Conference Record -
IAS Annual Meeting (IEEE Industry Applications Society), vol. 4, pp. 2411–2417, 2004, doi:
10.1109/ias.2004.1348813.
[37] R. Morales-Caporal and M. Pacas, “A predictive torque control for the synchronous reluctance machine
taking into account the magnetic cross saturation,” IEEE Transactions on Industrial Electronics, vol. 54,
no. 2, pp. 1161–1167, Apr. 2007, doi: 10.1109/TIE.2007.891783.
[38] D. Mingardi, M. Morandin, S. Bolognani, and N. Bianchi, “On the Proprieties of the Differential Cross-
Saturation Inductance in Synchronous Machines,” IEEE Trans Ind Appl, vol. 53, no. 2, pp. 991–1000, 2017,
doi: 10.1109/TIA.2016.2622220.
[39] S. Tahi and R. Ibtiouen, “Finite element calculation of the dq-axes inductances and torque of
synchronous reluctance motor,” 2014 International Conference on Electrical Sciences and Technologies in
Maghreb, CISTEM 2014, pp. 1–5, 2014, doi: 10.1109/CISTEM.2014.7076979.
[40] A. E. Hoffer, R. H. Moncada, B. J. Pavez, and J. A. Tapia, “A Novel Method for Finite-Element Modeling
of a Commercial Synchronous Reluctance Machine,” IEEE Latin America Transactions, vol. 16, no. 3, pp.
806–812, 2018, doi: 10.1109/TLA.2018.8358659.
85
[41] Jorma. Haataja, “A comparative performance study of four-pole induction motors and synchronous
reluctance motors in variable speed drives,” 2003.
[42] by Reza Rajabi Moghaddam, F. Magnussen, and C. Sadarangani, “Synchronous Reluctance

Machine (SynRM) Design,” 2007, Accessed: Feb. 08, 2023. [Online]. Available:
http://urn.kb.se/resolve?urn=urn:nbn:se:kth:diva-153663
[43] H. Shao, S. Li, and T. G. Habetler, “Analytical Calculation of the Air-gap Flux Density and Magnetizing
Inductance of Synchronous Reluctance Machines,” 2018 IEEE Energy Conversion Congress and Exposition,
ECCE 2018, pp. 5408–5413, 2018, doi: 10.1109/ECCE.2018.8558160.
[44] M. Pohl and D. Gerling, “Analytical Model of Synchronous Reluctance Machines with Zhukovski
Barriers,” Proceedings - 2018 23rd International Conference on Electrical Machines, ICEM 2018, pp. 91–96,
2018, doi: 10.1109/ICELMACH.2018.8506737.
[45] S. Cai, H. Hao, M. J. Jin, and J. X. Shen, “A Simplified Method to Analyze Synchronous Reluctance
Machine,” 2016 IEEE Vehicle Power and Propulsion Conference, VPPC 2016 - Proceedings, Dec. 2016, doi:
10.1109/VPPC.2016.7791597.
[46] N. Bianchi, S. Bolognani, D. Bon, and M. D. Pré, “Torque harmonic compensation in a synchronous
reluctance motor,” IEEE Transactions on Energy Conversion, vol. 23, no. 2, pp. 466–473, Jun. 2008, doi:
10.1109/TEC.2007.914357.
[47] H. Mahmoud, N. Bianchi, G. Bacco, and N. Chiodetto, “Nonlinear Analytical Computation of the
Magnetic Field in Reluctance Synchronous Machines,” IEEE Trans Ind Appl, vol. 53, no. 6, pp. 5373–
5382, Nov. 2017, doi: 10.1109/TIA.2017.2746560.
[48] M. Barcaro and N. Bianchi, “Air-gap flux density distortion and iron losses in anisotropic synchronous
motors,” IEEE Trans Magn, vol. 46, no. 1, pp. 121–126, Jan. 2010, doi: 10.1109/TMAG.2009.2030675.
[49] G. Bacco and N. Bianchi, “Choice of flux-barriers position in synchronous reluctance machines,” 2017
IEEE Energy Conversion Congress and Exposition, ECCE 2017, vol. 2017-Janua, pp. 1872–1879, 2017, doi:
10.1109/ECCE.2017.8096023.
[50] H. Mahmoud and N. Bianchi, “Eccentricity in Synchronous Reluctance Motors—Part II: Different Rotor
Geometry and Stator Windings,” IEEE Transactions on Energy Conversion, vol. 30, no. 2, pp. 754–760, Jun.
2015, doi: 10.1109/TEC.2014.2384534.
[51] M. Degano, H. Mahmoud, N. Bianchi, and C. Gerada, “Synchronous reluctance machine analytical
model optimization and validation through finite element analysis,” Proceedings - 2016 22nd
International Conference on Electrical Machines, ICEM 2016, pp. 585–591, 2016, doi:
10.1109/ICELMACH.2016.7732585.
[52] X. Li, Y. Wang, and R. Qu, “Design of Synchronous Reluctance Motors with Asymmetrical Flux Barriers
for Torque Ripple Reduction,” in 2021 IEEE 4th Student Conference on Electric Machines and Systems
(SCEMS), IEEE, Dec. 2021, pp. 1–6. doi: 10.1109/scems52239.2021.9646121.
[53] C. Gallardo, J. A. Tapia, M. Degano, and H. Mahmoud, “Accurate Analytical Model for Synchronous
Reluctance Machine With Multiple Flux Barriers Considering the Slotting Effect,” IEEE Trans Magn, vol.
58, no. 9, Sep. 2022, doi: 10.1109/TMAG.2022.3189483.
86
[54] K. C. Kim, “A novel method for minimization of cogging torque and torque ripple for interior
permanent magnet synchronous motor,” IEEE Trans Magn, vol. 50, no. 2, pp. 793–796, 2014, doi:
10.1109/TMAG.2013.2285234.
[55] A. Credo, M. Villani, M. Popescu, and N. Riviere, “Application of Epoxy Resin in Synchronous
Reluctance motors with fluid-shaped barriers for e mobility,” IEEE Trans Ind Appl, vol. PP, no. c, pp. 1–
1, 2021, doi: 10.1109/tia.2021.3103826.
[56] X. Zeng, L. Quan, X. Zhu, L. Xu, and F. Liu, “Investigation of an asymmetrical rotor hybrid permanent
magnet motor for approaching maximum output torque,” IEEE Transactions on Applied
Superconductivity, vol. 29, no. 2, pp. 1–4, 2019, doi: 10.1109/TASC.2019.2893708.
[57] Y. Xiao, Z. Q. Zhu, J. T. Chen, D. Wu, and L. M. Gong, “A Novel Spoke-type Asymmetric Rotor Interior
PM Machine,” ECCE 2020 - IEEE Energy Conversion Congress and Exposition, pp. 4050–4057, 2020, doi:
10.1109/ECCE44975.2020.9236321.
[58] I. Petrov, P. Ponomarev, and J. Pyrhönen, “Asymmetrical geometries in electrical machines,”

International Review of Electrical Engineering, vol. 11, no. 1, pp. 20–27, 2016, doi: 10.15866/iree.v11i1.7739.
[59] A. E. Hoffer, I. Petrov, J. J. Pyrhonen, J. A. Tapia, and G. Bramerdorfer, “Analysis of a tooth-coil winding
permanent-magnet synchronous machine with an unequal teeth width,” IEEE Access, vol. 8, pp. 71512–
71524, 2020, doi: 10.1109/ACCESS.2020.2987872.
[60] X. Sun, G. Sizov, and M. Melfi, “Asymmetrical design in electrical machines,” 2019 IEEE Energy
Conversion Congress and Exposition, ECCE 2019, pp. 3786–3792, 2019, doi: 10.1109/ECCE.2019.8913231.
[61] D. M. Ionel, “Interior Permanent Magnet Motor Including Rotor With Flux Barriers,” US8102091B2,
2008 [Online]. Available: http://www.freepatentsonline.com/y2008/0224558.html
[62] K. M. Rahman, S. Jurkovic, E. L. Kaiser, and P. J. Savagian, “Interior permanent magnet machine with
pole-to-pole asymmetry of rotor slot placement,” US8933606B2, 2015 [Online]. Available:
https://patents.google.com/patent/US8933606B2/en
[63] D. Takizawa, “Electromagnetic steel sheet formed body, rotor core, rotor, rotating electrical machine,
and vehicle,” US8860276B2, 2014 [Online]. Available:
[64] T. Suzuki, A. Maemura, Y. Kawazoe, Y. Fukuma, and T. Inoue, “Rotor, rotating electric machine,
vehicle, elevator, fluid machine, and processing machine,” US8227953B2, 2012 [Online]. Available:
[65] W. Zhao, F. Xing, X. Wang, T. A. Lipo, and B. il Kwon, “Design and Analysis of a Novel PM-Assisted
Synchronous Reluctance Machine with Axially Integrated Magnets by the Finite-Element Method,”
IEEE Trans Magn, vol. 53, no. 6, Jun. 2017, doi: 10.1109/TMAG.2017.2662717.
[66] F. Xing, W. Zhao, and B. il Kwon, “Design and optimisation of a novel asymmetric rotor structure for a
PM-assisted synchronous reluctance machine,” IET Electr Power Appl, vol. 13, no. 5, pp. 573–580, 2019,
doi: 10.1049/iet-epa.2018.0184.
87
[67] A. J. Pina and L. Xu, “Modeling of synchronous reluctance motors aided by permanent magnets with
asymmetric rotor poles,” Proceedings - 2015 IEEE International Electric Machines and Drives Conference,
IEMDC 2015, pp. 412–418, 2016, doi: 10.1109/IEMDC.2015.7409092.
[68] G. Dajaku and D. Gerling, “New methods for reducing the cogging torque and torque ripples of
PMSM,” 2014 4th International Electric Drives Production Conference, EDPC 2014 - Proceedings, pp. 1–7,
2014, doi: 10.1109/EDPC.2014.6984396.
[69] W. Zhao, T. A. Lipo, and B. il Kwon, “Optimal design of a novel asymmetrical rotor structure to obtain
torque and efficiency improvement in surface inset PM motors,” IEEE Trans Magn, vol. 51, no. 3, pp. 3–
6, 2015, doi: 10.1109/TMAG.2014.2362146.
[70] W. Zhao, F. Zhao, T. A. Lipo, and B. il Kwon, “Optimal design of a novel v-type interior permanent
magnet motor with assisted barriers for the improvement of torque characteristics,” IEEE Trans Magn,
vol. 50, no. 11, pp. 18–21, 2014, doi: 10.1109/TMAG.2014.2330339.
[71] W. Zhao, D. Chen, T. A. Lipo, and B. il Kwon, “Performance Improvement of Ferrite-Assisted

Synchronous Reluctance Machines Using Asymmetrical Rotor Configurations,” IEEE Trans Magn, vol.
51, no. 11, Nov. 2015, doi: 10.1109/TMAG.2015.2436414.
[72] Y. Fan, C. Tan, S. Chen, and M. Cheng, “Design and analysis of a new interior permanent magnet motor
for EVs,” in 2016 IEEE 8th International Power Electronics and Motion Control Conference, IPEMC-ECCE
Asia 2016, IEEE, May 2016, pp. 1357–1361. doi: 10.1109/IPEMC.2016.7512487.
[73] E. Howard, M. J. Kamper, and S. Gerber, “Asymmetric Flux Barrier and Skew Design Optimization of
Reluctance Synchronous Machines,” IEEE Trans Ind Appl, vol. 51, no. 5, pp. 3751–3760, 2015, doi:
10.1109/TIA.2015.2429649.
[74] M. Davoli, C. Bianchini, A. Torreggiani, and F. Immovilli, “A design method to reduce pulsating torque
in PM assisted synchronous reluctance machines with asymmetry of rotor barriers,” IECON Proceedings
(Industrial Electronics Conference), pp. 1566–1571, 2016, doi: 10.1109/IECON.2016.7793919.
[75] M. Sanada, K. Hiramoto, S. Morimoto, and Y. Takeda, “Torque ripple improvement for synchronous
reluctance motor using an asymmetric flux barrier arrangement,” IEEE Trans Ind Appl, vol. 40, no. 4, pp.
1076–1082, 2004, doi: 10.1109/TIA.2004.830745.
[76] S. Ferrari, E. Armando, and G. Pellegrino, “Torque Ripple Minimization of PM-assisted Synchronous
Reluctance Machines via Asymmetric Rotor Poles,” 2019 IEEE Energy Conversion Congress and Exposition,
ECCE 2019, pp. 4895–4902, Sep. 2019, doi: 10.1109/ECCE.2019.8912470.
[77] A. Vagati, A. Canova, M. Chiampi, M. Pastorelli, and M. Repetto, “Design refinement of synchronous
reluctance motors through finite-element analysis,” IEEE Trans Ind Appl, vol. 36, no. 4, pp. 1094–1102,
2000, doi: 10.1109/28.855965.
[78] X. B. Bomela and M. J. Kamper, “Effect of stator chording and rotor skewing on performance of
reluctance synchronous machine,” IEEE Trans Ind Appl, vol. 38, no. 1, pp. 91–100, Jan. 2002, doi:
10.1109/28.980362.
[79] T. Hamiti, T. Lubin, and A. Rezzoug, “A simple and efficient tool for design analysis of synchronous
reluctance motor,” IEEE Trans Magn, vol. 44, no. 12, pp. 4648–4652, 2008, doi:
10.1109/TMAG.2008.2004536.
88
[80] Y. Wang, D. M. Ionel, V. Rallabandi, M. Jiang, and S. J. Stretz, “Large-Scale Optimization of

Synchronous Reluctance Machines Using CE-FEA and Differential Evolution,” IEEE Trans Ind Appl, vol.
52, no. 6, pp. 4699–4709, Nov. 2016, doi: 10.1109/TIA.2016.2591498.
[81] J. Juergens, A. Fricasse, L. Marengo, J. Gragger, M. de Gennaro, and B. Ponick, “Innovative design of
an air cooled ferrite permanent magnet assisted synchronous reluctance machine for automotive
traction application,” Proceedings - 2016 22nd International Conference on Electrical Machines, ICEM 2016,
pp. 803–810, Nov. 2016, doi: 10.1109/ICELMACH.2016.7732618.
[82] B. Ban, S. Stipetic, and T. Jercic, “Minimum Set of Rotor Parameters for Synchronous Reluctance
Machine and Improved Optimization Convergence via Forced Rotor Barrier Feasibility,” Energies 2021,
Vol. 14, Page 2744, vol. 14, no. 10, p. 2744, May 2021, doi: 10.3390/EN14102744.
[83] N. Bernard, L. Dang, L. Moreau, and S. Bourguet, “A Pre-Sizing Method for Salient Pole Synchronous
Reluctance Machines with Loss Minimization Control for a Small Urban Electrical Vehicle Considering
the Driving Cycle,” Energies 2022, Vol. 15, Page 9110, vol. 15, no. 23, p. 9110, Dec. 2022, doi:
10.3390/EN15239110.
[84] B. Ban and S. Stipetic, “Systematic Metamodel-Based Optimization Study of Synchronous Reluctance
Machine Rotor Barrier Topologies,” Machines 2022, Vol. 10, Page 712, vol. 10, no. 8, p. 712, Aug. 2022,
doi: 10.3390/MACHINES10080712.
[85] N. Bianchi, S. Bolognani, D. Bon, and M. D. Pré, “Rotor flux-barrier design for torque ripple reduction
in synchronous reluctance and PM-assisted synchronous reluctance motors,” IEEE Trans Ind Appl, vol.
45, no. 3, pp. 921–928, 2009, doi: 10.1109/TIA.2009.2018960.
[86] N. Bianchi, E. Fornasiero, E. Carraro, S. Bolognani, and M. Castiello, “Electric vehicle traction based on
a PM assisted synchronous reluctance motor,” 2014 IEEE International Electric Vehicle Conference, IEVC
2014, 2014, doi: 10.1109/IEVC.2014.7056146.
[87] P. Lazari, J. Wang, and B. Sen, “3-D Effects of Rotor Step-Skews in Permanent Magnet-Assisted
Synchronous Reluctance Machines,” IEEE Trans Magn, vol. 51, no. 11, Nov. 2015, doi:
10.1109/TMAG.2015.2446511.
[88] M. Hofer and M. Schroedl, “Comparison of a flux barrier and a salient pole synchronous reluctance
machine for high rotational speeds in electric traction applications,” 2017 20th International Conference
on Electrical Machines and Systems, ICEMS 2017, Oct. 2017, doi: 10.1109/ICEMS.2017.8056108.
[89] T. H. Lee, J. H. Lee, K. P. Yi, and D. K. Lim, “Optimal Design of a Synchronous Reluctance Motor Using
a Genetic Topology Algorithm,” Processes 2021, Vol. 9, Page 1778, vol. 9, no. 10, p. 1778, Oct. 2021, doi:
10.3390/PR9101778.
[90] T. H. Lee, D. K. Lim, K. Y. Moon, and K. W. Jeon, “Topology Optimization Combined with a Parametric
Algorithm for Industrial Synchronous Reluctance Motor Design,” Processes 2022, Vol. 10, Page 746, vol.
10, no. 4, p. 746, Apr. 2022, doi: 10.3390/PR10040746.
[91] J. C. Baziruwiha, M. J. Kamper, and S. Botha, “High Pole Number Epoxy-Casted Rotor Reluctance
Synchronous Wind Generator,” 2022 IEEE Energy Conversion Congress and Exposition, ECCE 2022, 2022,
doi: 10.1109/ECCE50734.2022.9948139.
89
[92] T. Hubert, M. Reinlein, A. Kremser, and H. G. Herzog, “Torque ripple minimization of reluctance
synchronous machines by continuous and discrete rotor skewing,” 2015 5th International Conference on
Electric Drives Production, EDPC 2015 - Proceedings, Nov. 2015, doi: 10.1109/EDPC.2015.7323229.
[93] O. Ocak and M. Aydin, “An Innovative Semi-FEA Based, Variable Magnet-Step-Skew to Minimize
Cogging Torque and Torque Pulsations in Permanent Magnet Synchronous Motors,” IEEE Access, vol.
8, pp. 210775–210783, 2020, doi: 10.1109/ACCESS.2020.3038340.
[94] O. Korman, M. Degano, M. di Nardo, and C. Gerada, “A Novel Flux Barrier Parametrization for
Synchronous Reluctance Machines,” IEEE Transactions on Energy Conversion, vol. 8969, no. c, pp. 1–11,
2021, doi: 10.1109/TEC.2021.3099628.
[95] M. S. Islam, A. Shrestha, and M. Islam, “Effect of Step Skew in Synchronous Reluctance Machines for
High Performance Applications,” 2022 IEEE Energy Conversion Congress and Exposition, ECCE 2022, 2022,
doi: 10.1109/ECCE50734.2022.9947850.
[96] C. Gallardo, C. Madariaga, J. A. Tapia, and M. Degano, “A Method to Determine the Torque Ripple
Harmonic Reduction in Skewed Synchronous Reluctance Machines,” Applied Sciences 2023, Vol. 13, Page
2949, vol. 13, no. 5, p. 2949, Feb. 2023, doi: 10.3390/APP13052949.
[97] A. Tessarolo, M. Degano, and N. Bianchi, “On the analytical estimation of the airgap field in
synchronous reluctance machine,” in Proceedings - 2014 International Conference on Electrical Machines,
ICEM 2014, 2014, pp. 239–244. doi: 10.1109/ICELMACH.2014.6960187.
[98] A. Tessarolo, M. Degano, and N. Bianchi, “On the analytical estimation of the airgap field in
synchronous reluctance machine,” Proceedings - 2014 International Conference on Electrical Machines,
ICEM 2014, pp. 239–244, 2014, doi: 10.1109/ICELMACH.2014.6960187.
[99] C. J. Fraser, “Electrical Machines,” in Systems, Controls, Embedded Systems, Energy, and Machines, CRC
Press, 2017, pp. 217–260. doi: 10.1201/9781420037043-13.
[100] A. Tessarolo, “Modeling and analysis of synchronous reluctance machines with circular flux barriers
through conformal mapping,” IEEE Trans Magn, vol. 51, no. 4, pp. 1–11, Apr. 2015, doi:
10.1109/TMAG.2014.2363434.
[101] B. Gaussens, E. Hoang, O. De La Barrière, J. Saint-Michel, M. Lecrivain, and M. Gabsi, “Analytical

approach for air-gap modeling of field-excited flux-switching machine: No-load operation,” IEEE Trans
Magn, vol. 48, no. 9, pp. 2505–2517, 2012, doi: 10.1109/TMAG.2012.2196706.
[102] N. Bianchi, M. Degano, and E. Fornasiero, “Sensitivity analysis of torque ripple reduction of
synchronous reluctance and interior PM motors,” IEEE Trans Ind Appl, vol. 51, no. 1, pp. 187–195, 2015,
doi: 10.1109/TIA.2014.2327143.
[103] A. Fratta, G. P. Troglia, A. Vagati, and F. Villata, “Evaluation of torque ripple in high performance
synchronous reluctance machines,” in Conference Record - IAS Annual Meeting (IEEE Industry Applications
Society), Publ by IEEE, 1993, pp. 163–170. doi: 10.1109/ias.1993.298919.
[104] S. Taghavi and P. Pillay, “A Novel Grain-Oriented Lamination Rotor Core Assembly for a Synchronous
Reluctance Traction Motor with a Reduced Torque Ripple Algorithm,” IEEE Trans Ind Appl, vol. 52, no.
5, pp. 3729–3738, Sep. 2016, doi: 10.1109/TIA.2016.2558162.
90
[105] T. Mohanarajah, J. Rizk, A. Hellany, M. Nagrial, and A. Klyavlin, “Torque Ripple Improvement in
Synchronous Reluctance Machines,” in 2018 2nd International Conference On Electrical Engineering, EECon
2018, IEEE, 2018, pp. 44–50. doi: 10.1109/EECon.2018.8541021.
[106] M. N. F. Ibrahim, P. Sergeant, and E. Rashad, “Simple design approach for low torque ripple and high
output torque synchronous reluctance motors,” Energies (Basel), vol. 9, no. 11, p. 942, Nov. 2016, doi:
10.3390/en9110942.
[107] E. S. Obe, “Calculation of inductances and torque of an axially laminated synchronous reluctance
motor,” IET Electr Power Appl, vol. 4, no. 9, pp. 783–792, 2010, doi: 10.1049/iet-epa.2009.0197.
[108] S. Taghavi and P. Pillay, “A Sizing Methodology of the Synchronous Reluctance Motor for Traction
Applications,” IEEE J Emerg Sel Top Power Electron, vol. 2, no. 2, pp. 329–340, Jun. 2014, doi:
10.1109/JESTPE.2014.2299235.
[109] G. Bacco and N. Bianchi, “Design Criteria of Flux-Barriers in Synchronous Reluctance Machines,” IEEE
Trans Ind Appl, vol. 55, no. 3, pp. 2490–2498, 2019, doi: 10.1109/TIA.2018.2886778.
[110] G. Pellegrino, F. Cupertino, and C. Gerada, “Barriers shapes and minimum set of rotor parameters in
the automated design of Synchronous Reluctance machines,” Proceedings of the 2013 IEEE International
Electric Machines and Drives Conference, IEMDC 2013, pp. 1204–1210, 2013, doi:
10.1109/IEMDC.2013.6556286.
[111] F. Cupertino, G. Pellegrino, and C. Gerada, “Design of synchronous reluctance motors with
multiobjective optimization algorithms,” IEEE Trans Ind Appl, vol. 50, no. 6, pp. 3617–3627, Nov. 2014,
doi: 10.1109/TIA.2014.2312540.
[112] X. Liu, H. Chen, J. Zhao, and A. Belahcen, “Research on the Performances and Parameters of Interior
PMSM Used for Electric Vehicles,” IEEE Transactions on Industrial Electronics, vol. 63, no. 6, pp. 3533–
3545, Jun. 2016, doi: 10.1109/TIE.2016.2524415.
[113] K. Wang, Z. Q. Zhu, G. Ombach, M. Koch, S. Zhang, and J. Xu, “Optimal slot/pole and flux-barrier layer
number combinations for synchronous reluctance machines,” in 2013 8th International Conference and
Exhibition on Ecological Vehicles and Renewable Energies, EVER 2013, IEEE, Mar. 2013, pp. 1–8. doi:
10.1109/EVER.2013.6521583.
[114] M. Fitouri, Y. Bensalem, and M. N. Abdelkrim, “Comparison between 2D and 3D Modeling of

Permanent Magnet Synchronous Motor Using FEM Simulations,” Proceedings of the 17th International
Multi-Conference on Systems, Signals and Devices, SSD 2020, pp. 681–685, Jul. 2020, doi:
10.1109/SSD49366.2020.9364256.
[115] H. Mahmoud, M. Degano, G. Bacco, N. Bianchi, and C. Gerada, “Synchronous Reluctance Motor Iron
Losses: Analytical Model and Optimization,” 2018 IEEE Energy Conversion Congress and Exposition,
ECCE 2018, pp. 1640–1647, Dec. 2018, doi: 10.1109/ECCE.2018.8558292.
[116] K. Wang, Z. Q. Zhu, G. Ombach, M. Koch, S. Zhang, and J. Xu, “Optimal slot/pole and flux-barrier layer
number combinations for synchronous reluctance machines,” 2013 8th International Conference and
Exhibition on Ecological Vehicles and Renewable Energies, EVER 2013, 2013, doi:
10.1109/EVER.2013.6521583.
91
[117] M. Degano et al., “Optimised Design of Permanent Magnet Assisted Synchronous Reluctance Machines
for Household Appliances,” IEEE Transactions on Energy Conversion, 2021, doi:
10.1109/TEC.2021.3076675.
[118] M. Barcaro, G. Meneghetti, and N. Bianchi, “Structural analysis of the interior PM rotor considering
both static and fatigue loading,” IEEE Trans Ind Appl, vol. 50, no. 1, pp. 253–260, 2014, doi:
10.1109/TIA.2013.2268048.

Gallardo Sánchez - César Tesis

Uploaded by

Copyright:

Available Formats

Gallardo Sánchez - César Tesis

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Gallardo Sánchez - César Tesis

Uploaded by

Copyright:

Available Formats

University of Concepción

Department of Electrical Engineering

César Gallardo Sánchez

DEVELOPMENT OF DESIGN TECHNIQUES TO

Thesis for the degree of Doctor of Science at the

Supervisor: Professor Juan A. Tapia Ladino

Professor Michele Degano

César Gallardo Sánchez

Keywords: analytical model, asymmetric rotor, design guidelines, electromagnetic torque,

ACKNOWLEDGEMENTS .......................................................................................................................................... III

LIST OF FIGURES ........................................................................................................................................................ VI

NOMENCLATURE AND ACRONYMS ................................................................................................................... XI

CHAPTER I: SYNCHRONOUS RELUCTANCE MACHINE: AN OVERVIEW. .................................................. 5

1.1. OPERATING PRINCIPLE ....................................................................................................................................... 5

1.2. ELECTROMAGNETIC TORQUE .......................................................................................................................... 6

1.3. INTERNAL POWER FACTOR AND POWER FACTOR ................................................................................... 8

1.4. CONTROL STRATEGY ........................................................................................................................................... 9

1.5. INFLUENCE OF SATURATION AND CROSS-COUPLING EFFECT ......................................................... 11

1.6. EFFICIENCY ............................................................................................................................................................ 13

1.7. DESIGN TECHNIQUES. MACHINE PERFORMANCE IMPROVEMENT ................................................ 15

1.7.1. Analytical models vs. Finite elements methods ............................................................................................... 15

1.8. SUMMARY .............................................................................................................................................................. 19

CHAPTER II: ACCURATE ANALYTICAL MODEL FOR SYNCHRONOUS RELUCTANCE MACHINE .. 20

2.1. MAIN ASSUMPTIONS ......................................................................................................................................... 20

2.2. AIR-GAP FLUX DENSITY CALCULATION ..................................................................................................... 21

2.3. TORQUE CALCULATION .................................................................................................................................... 24

2.5. PARAMETERS ON THE D-Q REFERENCE FRAME....................................................................................... 32

2.6. SUMMARY .............................................................................................................................................................. 34

CHAPTER III: ROTOR ASYMMETRIC IMPACT ON SYNCHRONOUS RELUCTANCE MACHINE

3.1.1. Sizing method .................................................................................................................................................. 35

3.2. RESULTS ANALYSIS ............................................................................................................................................. 39

3.2.1. Saliency ratio ................................................................................................................................................... 39

3.3. SUMMARY .............................................................................................................................................................. 45

CHAPTER IV: A METHOD TO DETERMINE THE TORQUE RIPPLE HARMONIC REDUCTION IN

4.1. SELECTED MACHINES ........................................................................................................................................ 46

4.2. ANALYTICAL METHOD DERIVATION FOR DISCRETE SKEWING ....................................................... 47

4.3. FINITE ELEMENT VALIDATION: RESULTS AND DISCUSSION ............................................................. 49

4.3.1. FEA evaluation original designs (skewless machines) .................................................................................... 50

4.4. COMPARISON, ANALYSIS, AND RECOMMENDATIONS ........................................................................ 56

4.5. SUMMARY .............................................................................................................................................................. 58

CHAPTER V: PROPOSED DESIGN OF A SYNCHRONOUS RELUCTANCE MACHINE ............................ 59

5.1. SET-UP OF THE OPTIMIZATION PROCESS .................................................................................................. 59

5.2. RESULTS FROM THE OPTIMIZATION PROCESS ....................................................................................... 61

5.3. STRUCTURAL ANALYSIS ................................................................................................................................... 63

5.4. OPTIMAL DESIGN ADJUSTMENT................................................................................................................... 64

5.5. IMPACT OF ROTOR SKEWING ON THE PERFORMANCE OF THE MACHINE .................................. 75

5.6. MANUFACTURED PROTOTYPE ....................................................................................................................... 77

5.7. SUMMARY .............................................................................................................................................................. 78

6. CONCLUSIONS AND FUTURE WORK ............................................................................................................... 80

6.1. CONCLUSION ........................................................................................................................................................ 80

6.2. FUTURE WORK ...................................................................................................................................................... 81

Nomenclature and acronyms

Interest in Synchronous Reluctance Machines (SynRMs) is growing due to their many

III. Outline of the doctoral dissertation

IV. Contributions of the doctoral dissertation

The specific contributions of this Thesis are summarized as follows:

V. List of publications and fundings

This doctoral dissertation contains material from the following papers.

❖ C. Gallardo, J. A. Tapia, M. Degano, H. Mahmoud and A. E. Hoffer, "Rotor Asymmetry