Design of Transverse Flux Permanent Magnet Machines for ...

Design of Transverse Flux Permanent Magnet Machines for Large Direct-Drive

Wind Turbines

Design of Transverse Flux Permanent Magnet Machines for Large Direct-Drive

Wind Turbines

PROEFSCHRIFT

ter verkrijging van de graad van doctor aan de Technische Universiteit Delft,

op gezag van de Rector Magnificus prof.ir. K.C.A.M. Luyben, voorzitter van het College voor Promoties,

in het openbaar te verdedigen op woensdag 20 oktober 2010 om 10.00 uur door

Deok-je BANG

Master of Engineering, Pukyong National University

geboren te Busan, Korea

Dit proefschrift is goedgekeurd door de promotor: Prof.dr. J.A. Ferreira Copromotor: Dr.ir. H. Polinder Samenstelling promotiecommissie: Rector Magnificus, voorzitter Prof.dr. J.A. Ferreira, Technische Universiteit Delft, promotor Dr.ir. H. Polinder, Technische Universiteit Delft, copromotor Prof.dr. B. Mecrow, Newcastle University, UK Prof.dr. G.J.W. van Bussel, Technische Universiteit Delft Prof.dr. Z. Chen, Aalborg University, Denmark Prof.dr.ir. G. Lodewijks, Technische Universiteit Delft Prof.ir. L. van der Sluis, Technische Universiteit Delft ISBN 978-90-5335-336-3 Printed by Ridderprint B.V. Pottenbakkerstraat 15-17 2984 AX Ridderkerk The Netherlands Cover design by Cha-joong Kim Cover pictures: Deok-je Bang Copyright © 2010 by Deok-je Bang All rights reserved. No part of the material protected by this copyright notice may be reproduced or utilised in any form or by any means, electronic or mechanical, including photocopying, recording or by any information storage and retrieval system without permission from the publisher or author.

to my wife, Nam-yeon for her support and love

i

Acknowledgments In April 2006 I started my Ph.D research at the Electrical Power Processing group, and now I am ending up my research. The time spent for Ph.D. research will be never forgotten because I learned a lot of things about study, research and life. I would hereby like to express my gratitude for the people who have supported me during my Ph.D. research period. First and foremost I would like to thank to my co-promoter Dr. Henk Polinder, who provided right guidance when I had vague ideas and when I was struggling with the TF (transverse flux or terrible flux) machines. Dr. Polinder also guided me how to represent my research results more explicitly. Besides his guidance on my research he also made me think about life and family. I would also like to express my gratitude to Prof. Dr. Jan Abraham Ferreira, who was my promoter. Prof. Ferreira’s openness of mind and positive thinking way inspired and motivated me to find new ideas to get round a bottleneck in my research issues. Without the supervision of both promoters, this thesis would not have been possible. I would also like to thank Dr. Bas Gravendeel, the external evaluator of my first year research evaluation, for his comments, suggestions and encouragement. The UpWind project, a European project funded under the EU’s Sixth Framework Programme (FP6), is gratefully acknowledged for their financial support to my research. I would also like to thank my dear friend Bong-Jun Kim, Mr. Dong-Sang Park and Wintech Co., Ltd. in Korea for their financial and technical support in building the experimental setup. Special appreciation is given to the members of my thesis committee, Professors B. Mecrow, G.J.W. van Bussel, Z. Chen, G. Lodewijks and L. van der Sluis for their time and constructive comments on my thesis. I would like to thank Cha-joong Kim for designing the cover of the thesis. It looks great, thank you. I would also like to thank Frank van der Pijl who translated the propositions and summary of my thesis into Dutch. Bridget (Mrs. Saywell) and Veronika (Ms. Pisorn, now Mrs. Borisavljevic) are acknowledged for checking the English language. Thanks a lot for your kind comments and edition.

ii

My gratitude goes to friends and colleagues in Electrical Power Processing group. Especially, I would like to thank my roommates Frank van der Pijl, Aleksandar Borisavljevic and Balazs Czech, and ex-roommate Mattia Scuotto for sharing the office with me and for helping me in all possible ways. I would also like to thank Ghanshaym Shrestha and Zhihui Yuan – we had a lot of discussion on our Ph.D. research and life. Others at the group whom I would like to acknowledge for their help and friendship are; Anoop Jassal, Dalibor Cvoric, Yi Wang, Hung Vu Xuan, Sam Ani, Ivan Josifovic, Jinku Hu, Marcelo Gutierrez Alcaraz, Johan Wolmarans, Milos Acanski, Xun Gong and Rodrigo Teixeira Pinto. I would also like to thank ir. S.W.H. de Haan, Dr. Paul Bauer, Dr. M.J. Hoeijmakers, Dr. Jelena Popovic, Mr. Bart Roodenburg and Mr. Robertus Schoevaars. I would also like to thank Dr. Yi Zhou, Dr. Dongsheng Zhao and Milan Hajder, ex-colleagues in the group. I would also like to thank Mrs. Suzy Sirks-Bong, secretary of Electrical Sustainable Energy department, who gave me the great help on whatever campus life and personal difficulties. In this regard, I would also like to mention Mrs. Laura Bruns who is secretary of Electrical Power Processing group. Thank you for all your help and support and for being such wonderful colleagues. I would like to express my gratitude to my family, family in law and friends. They have given me their love, support and encouragement. Especially, I would like to thank my father for being my pride. Most of all, I would like to thank my wife Nam-yeon for her support, sacrifice, patience and love during last four and half years. Lastly, I would like to thank my children Bo-hyun and Min-woo for their understanding when I had to work many hours for my Ph.D. research and thesis writing. From now I will do my best to repay you guys for your favors and love.

Table of Contents

iii

Table of Contents Acknowledgments.......................................................................................................................i Table of Contents......................................................................................................................iii Chapter 1: Introduction ...........................................................................................................1

1.1 Background ........................................................................................................................1 1.2 Problem Statement..............................................................................................................5 1.3 Research Objective and Approach ......................................................................................6 1.4 Outline of thesis .................................................................................................................7

Chapter 2: Overview of Generator Systems for Wind Turbines .............................................9

2.1 Introduction........................................................................................................................9 2.2 Wind turbine technology.....................................................................................................9

2.2.1 Fixed speed concept ...................................................................................................10 A. Stall control ...............................................................................................................12 B. Active stall control .....................................................................................................12 C. Pitch control..............................................................................................................12

2.2.2 Limited variable speed concept...................................................................................13 2.2.3 Variable speed concept with gearbox ..........................................................................14

A. DFIG system with three-stage gearbox ......................................................................15 B. SCIG system with three-stage gearbox .......................................................................16 C. PMSG system with three-stage gearbox .....................................................................17 D. PMSG system with single-stage gearbox ...................................................................17 E. EESG with mechanical and hydraulic gearboxes .......................................................19

2.2.4 Variable speed direct-drive concept ............................................................................22 A. Direct-drive EESG system..........................................................................................23 B. Direct-drive PMSG system.........................................................................................24

2.2.5 Generator systems on the market ................................................................................26 2.2.6 Other potential concepts .............................................................................................28

2.3 Direct-drive and geared generator systems........................................................................30 2.4 Conclusions......................................................................................................................33

Chapter 3: Direct-Drive PM Generators for Large Wind Turbines .....................................35

3.1 Introduction......................................................................................................................35 3.2 Criteria to assess direct-drive generators for wind turbines ...............................................35

Table of Contents

iv

3.3 PM machines for direct-drive ...........................................................................................36 3.3.1 RFPM machine ..........................................................................................................38 3.3.2 AFPM machine ..........................................................................................................47 3.3.3 TFPM machine...........................................................................................................57

3.4 Conclusions......................................................................................................................78 3.4.1 Advantages and disadvantages of PM machines..........................................................78 3.4.2 Active mass-competitiveness of PM machines............................................................79 3.4.3 Selection of PM machine configurations for large direct-drive wind generators..........81

Chapter 4: Mechanical Structure of Direct-Drive Wind Generators ...................................83

4.1 Introduction......................................................................................................................83 4.2 Large direct-drive wind generators ...................................................................................83

A. Conventional structure....................................................................................................85 B. Lightweight structure......................................................................................................87 C. Large direct-drive wind turbines .....................................................................................88 D. Total mass of different large direct-drive wind generators...............................................88

4.3 Total mass estimation of geared and direct-drive generators for wind turbines ..................94 A. Geared generator ............................................................................................................ 94 B. Mass estimation of geared and direct-drive generators ....................................................95

4.4 Conclusions......................................................................................................................96 Chapter 5: Modeling of PM Machines ...................................................................................99

5.1 Introduction......................................................................................................................99 5.2 Main dimensions of PM machines ....................................................................................99

5.2.1 RFPM machine ........................................................................................................ 100 5.2.2 TFPM machines ....................................................................................................... 102

5.3 Generalization of magnetic circuit of PM machines ........................................................ 109 5.3.1 Linear model ............................................................................................................ 109 5.3.2 Nonlinear model....................................................................................................... 112 5.3.3 Magnetic circuit modeling of PM machines.............................................................. 116

5.4 Verification of magnetic circuit analysis model............................................................... 122 5.4.1 Verification of no-load case ...................................................................................... 122 5.4.2 Verification in the case with a load ........................................................................... 131

5.5 Conclusions.................................................................................................................... 134 Chapter 6: Comparative Design of PM Generators for Large Direct-Drive Wind Turbines ................................................................................................................................................ 135

6.1 Introduction.................................................................................................................... 135 6.2 Selection of generator types for large direct-drive wind turbines ..................................... 135

Table of Contents

v

6.3 Analytical design of PM generators for direct-drive wind turbines .................................. 139 6.3.1 RFPM generator ....................................................................................................... 141 6.3.2 TFPM generators...................................................................................................... 143 6.3.3 Comparison of PM generators .................................................................................. 150

6.4 Conclusions.................................................................................................................... 156 Chapter 7: TFPM Machine with Multiple-Modules for Large Direct-Drive ..................... 159

7.1 Introduction.................................................................................................................... 159 7.2 TFPM machine with modular structure........................................................................... 160 7.3 Analytical modelling of TFPM generator with multiple-modules .................................... 164 7.4 Verification of magnetic circuit analysis model............................................................... 168

7.4.1 Verification of no-load case ...................................................................................... 173 7.4.2 Verification in the case with a load ........................................................................... 175

7.5 Design of TFPM generators with multiple-modules for large direct-drive wind turbines . 177 7.6 Conclusions.................................................................................................................... 185

Chapter 8: Challenges of Minimizing the Structural Mass of Large Direct-Drive Wind

Generators .......................................................................................................... 187 8.1 Introduction.................................................................................................................... 187 8.2 Ring-shaped generator concept with multi-sets ............................................................... 189 8.3 New supporting and guiding concepts............................................................................. 192

8.3.1 Buoyant rotating (moving) structure ......................................................................... 193 8.3.2 Bearingless drive ...................................................................................................... 197 8.3.3 Hydraulic bearing..................................................................................................... 205

8.4 Mass estimation of new direct-drive generator for large wind turbines............................ 206 8.4.1 Electromagnetic part................................................................................................. 206 8.4.2 Structural part........................................................................................................... 209 8.4.3 Total mass comparison ............................................................................................. 217

8.5 Conclusions.................................................................................................................... 220 Chapter 9: Conclusions and Recommendations................................................................... 221

9.1 Conclusions.................................................................................................................... 221 9.2 Recomendations for further research............................................................................... 226

References .............................................................................................................................. 227 Appendix 1............................................................................................................................. 241 Appendix 2............................................................................................................................. 245 Appendix 3............................................................................................................................. 249

Table of Contents

vi

List of Publications ................................................................................................................ 253 Summary................................................................................................................................ 255 Samenvatting ......................................................................................................................... 259 Curriculum Vitae................................................................................................................... 263

Introduction

1

Chapter 1

Introduction

1.1 Background The limited amount of fossil fuel that is available and the negative of global warming have made it necessary to harvest renewable energy. Therefore the use of renewable energy sources such as wind energy, solar energy, ocean energy and hydro power has been increasing. Wind energy has high potential because wind is everywhere on the earth. Wind energy produces cheaper electricity than other renewable energy sources. Therefore, wind energy has achieved the fastest growth. The average annual growth rate of wind energy has been about 30% since 1995. [Che 2006][Flo 2006][EWE 2009] In 2008, the world total installed wind power capacity reached 121 GW, which is twenty-five times more than in 1995 when the installed capacity was 4.8 GW as illustrated in Fig. 1-1 [Han 2007]. By the end of 2020, it is expected that the installed wind power capacity will increase to 1,260 GW, which will be about 12% of the world’s electricity consumption. The European Wind Energy Association (EWEA) has set a target to supply 23% of European electricity consumption through wind energy by 2030. [Flo 2006] In order to make wind energy more attractive and competitive, technological developments have been conducted on wind turbine concepts and generator system. Up-scaling and offshore wind turbines are a key driver in technological developments.

Wind Turbine Concepts and Generator Systems In order to maximize the energy harnessed, to minimize the cost, to improve the power quality and to ensure safety together with the growth of the size, various wind turbine concepts have been developed during last three decades [Man 2002]. Until the late 1990’s, most wind turbine manufacturers built fixed-speed stall-controlled turbines with a squirrel cage induction generator system and a three-stage gearbox (SCIG 3G). Due to the high thrust load in the rotor blades of

Chapter 1

2

the fixed-speed turbines and other reasons, most wind turbines with a power rating over 1.5 MW have changed to the variable speed pitch-controlled concept since the late 1990’s. In addition, direct-drive wind turbines have been built to increase energy yield, reduce gearbox failures, lower maintenance problems and to obtain better power quality to the grid since 1991. A hybrid system, namely the Multibrid system, has been built and discussed as an interesting alternative for wind turbines. The Multibrid system is equipped with a medium speed permanent magnet synchronous generator and a single-stage gearbox (PMSG 1G). An electrically excited synchronous generator system (EESG) with mechanical and hydraulic gearboxes has also appeared on the market. The principles of these different wind turbine concepts and generator systems will be described in the next chapter. Generator systems used in variable speed pitch-controlled wind turbine concepts can be classified as: • a doubly-fed induction generator system with a three-stage gearbox (DFIG 3G) • a squirrel cage induction generator system with a three-stage gearbox (SCIG 3G) • a permanent magnet synchronous generator system with a three-stage gearbox (PMSG 3G) • a permanent magnet synchronous generator system with a single-stage gearbox (PMSG 1G) • a direct-drive electrically excited synchronous generator system (EESG DD) • a direct-drive permanent magnet synchronous generator system (PMSG DD)

Fig. 1-1: World cumulative and annual wind power capacity

Introduction

3

Up-scaling wind turbines In reducing the cost of energy production from wind, the scaling up of both the size and the power rating of wind turbines has been a key driver. The average power of new wind turbines installed in the EU was about 105 kW in 1990, and the average power of those installed in 2007 increased to 1.7 MW. [EWE 2009] Large wind turbines are continually being researched and developed by manufacturers and research programmes. Fig. 1-2 illustrates the growth of the size and the power rating of wind turbines in the past, present and future [GAR 2008]. The world’s largest wind turbine is currently the prototype of Clipper’s Britannia model, rated at 7.5 MW, to be installed offshore near the UK. A German wind turbine manufacturer, Enercon GmbH also built a large turbine prototype that is the E-126 model with 126 m rotor diameter and 6 MW power rating. Fig. 1-3 shows the E-126 model wind turbine of Enercon GmbH. [HUF 2009][Tre 2009] Looking at the future of the design of very large offshore wind turbines, the UpWind project which is a European project funded under the EU's Sixth Framework Progamme (FP6) was created to research very large wind turbines, rated at 10 MW and higher.

Fig. 1-2: Development of power and size of wind turbines. Source: Garrad Hassan

Chapter 1

4

Offshore wind turbines The location of wind turbines is moving offshore because of the following reasons: • There are higher wind speeds with less turbulence. • The space to install the turbines on land and onshore is limited in some regions of Europe. The investment cost of offshore wind farms is considerably higher than the cost of onshore wind farms. Thus offshore wind power capacity currently accounts for a small amount (approximately 1 %) of the total installed wind power capacity in the world. 1.5 GW of capacity was located offshore by the end of 2008. However, the costs of offshore wind turbines are partly compensated by a higher total electricity production from the turbines, due to higher wind speeds and less

Fig. 1-3: E-126 model of Enercon GmbH in Germany. [HUF 2009]

Introduction

5

turbulence as described above. The energy production indicator described in [EWE 2009] is normally around 2,000-2,500 full load hours per year for an onshore installation. This figure goes up to 4,000 full load hours per year for a typical offshore installation.

1.2 Problem Statement Different generator systems such as geared and direct-drive generator systems have been discussed and compared by a number of authors. [Pol 2006][Wid 2006][Byw 2004][Gra 1996][Dub 2004][Lam 2004][Han 2001][Che 2005][Dub 2000][Böh 1997][Sie 1998][Böh 2003][Poo 2003][Han 2007][Car 1994][Soe 2005][Ann 1996] In the next chapter, the comparison results of the different generator systems will be discussed in detail. The features of such generator systems are shortly described in this section. Based on energy yield, reliability and maintenance problems, direct-drive generator systems have been regarded as a better concept than geared generator systems. Among different generator types, the permanent magnet synchronous generator (PMSG) is superior in overall efficiency. The PMSG has also been discussed as a concept with higher force density than the electrically excited (EE) generators. PMSG can thus be small and lightweight compared to induction generators and direct-drive electrically excited synchronous generators (EESG DD). Therefore, this thesis focuses on the research of the direct-drive permanent magnet synchronous generator system (PMSG DD) for large wind turbines. In this section, problems of the PMSG DD for large wind turbines are defined. Based on the problem definition, research questions are formulated.

Low speed and high torque rating The rotor of a direct-drive generator for a wind turbine is directly connected to the rotor hub. Thus the direct-drive generator operates at low speed. When a wind turbine is scaled up, the rotational speed of the rotor decreases because the tip speed of rotor blades is kept constant. The torque rating is thus increased inversely proportional to the decrease of the rotational speed. Thus a high torque generator that has to handle a high tangential force and that has a large air gap diameter is required for large direct-drive wind turbines. A large air gap diameter of the generator results in a large mass of the construction. This large mass of the direct-drive generator for large wind turbines increases costs. Therefore, large direct-drive generators have disadvantages such as large diameter, large mass and high cost in order to get high torque rating comparable to geared generator systems. When scaling up the power of direct-drive generators for wind turbines, the inactive part (structural part) becomes a dominant part of the total mass. [McD 2006] Therefore, it seems that

Chapter 1

6

only electromagnetic design to reduce the active material (electromagnetic material) is not sufficient to make the PMSG DD more attractive for large wind turbines.

Various PM machine configurations A number of PM machine configurations have been proposed and built for direct-drive applications. According to the flux path direction and the electromagnetic construction, such PM machines can be classified as follows. (1) Longitudinal flux and transverse flux PM machine configurations (2) Radial flux and axial flux PM machine configurations (3) Slotted and slotless PM machine configurations (4) Surface-mounted and flux-concentrating PM machine configurations (5) Iron-cored and coreless PM machine configurations Out of these different PM machine configurations, axial flux (AF) PM machines have advantages such as short axial length and higher torque over volume ratio, but disadvantages such as lower torque over mass ratio and structural instability. In a slotted AFPM machine configuration with large diameter, it is difficult to maintain the air gap and to produce the stator core. Radial flux (RF) machines are structurally stable, thus most of low speed generators use the RF machine configuration. Transverse flux (TF) PM machines have advantages such as higher force density, considerably lower copper losses and simpler windings than longitudinal flux (LF) PM machines. Thus it may become difficult to conclude which PM machine configuration is the most suitable for large direct-drive wind generators. Based on the problem statement defined above, research questions of the thesis are formulated as: • How can the mass and cost of PMSG DD be reduced to make it competitive with geared

generator systems? • Which PM machine configuration is the most suitable for large direct-drive wind turbines? • How do we design and assess LFPM and TFPM machines for large direct-drive wind turbines? • How can we minimize both the mass and cost of structural parts of large direct-drive

generator?

1.3 Research Objective and Approach The main objective of the thesis is to find a design solution for large direct-drive wind generators with highest possible energy yield and lowest possible cost. In order to achieve this objective, it is necessary to increase the force density and to significantly reduce the mass of both the active

Introduction

7

and inactive parts of the generator. Based on this, the main objective of the thesis can be revised as: Develop a PM machine with (1) higher force density and (2) the electromagnetic and structural material minimized for large direct-drive wind turbines. In order to achieve this main objective, the following issues are covered in the thesis. • Advantages and disadvantages of different generator systems for wind turbines are addressed. • Mass-competitiveness of direct-drive generators is identified to find a suitable generator

configuration for large direct-drive wind turbines. • Electromagnetic analysis models for longitudinal flux and transverse flux PM machines are

developed. • Different PM generators for large direct-drive wind turbines are compared. • A suitable TFPM machine configuration for large direct-drive wind turbines is proposed. • A new direct-drive generator construction is proposed to minimize the structural mass and the

cost.

1.4 Outline of thesis

Chapter 2 To gain an understanding of the advantages and the disadvantages of different generator systems for wind turbines, an overview of contemporary technologies of the systems is given. A quantitative comparison between direct-drive and geared generator systems is made based on energy yield, mass and cost.

Chapter 3 The aim of the chapter is to find permanent magnet (PM) machines with high active mass-competitiveness for large direct-drive wind turbines. To list the features of PM machines with various configurations, an overview of PM machines such as the axial flux PM (AFPM) machine, the radial flux PM (RFPM) machine and the transverse flux PM (TFPM) machine is given. To identify the active mass-competitiveness of those machines for large direct-drive applications, active masses of PM machines with various configurations discussed in the scientific literature are investigated as a function of the torque rating.

Chapter 4 The chapter identifies the total mass-competitiveness and estimates the total mass of direct-drive generators for large wind turbines. To identify the total mass-competitiveness of direct-drive generators compared to geared generators, an overview of different generators discussed in

Chapter 1

8

scientific literature is given based on the structural mass, the total mass, the size and the torque rating. To estimate the total mass of different generators in scaling up the power rating up to 20 MW, it is assumed that the value of the generator total mass to torque ratio is kept constant.

Chapter 5 The aim of the chapter is to develop electromagnetic analysis models for various configurations of PM machines. RFPM machine and TFPM machine are chosen for direct-drive wind generators in the chapter. The analysis models and relevant equations for the machines are developed. The analysis models are verified by the experiments and the finite element analyses.

Chapter 6 The chapter gives a comparative design of different PM machines in order to find a suitable configuration for large direct-drive wind turbines. A RFPM machine and four TFPM machines are selected and designed electromagnetically for 5 MW and 10 MW direct-drive applications. For the comparative design, the analytical models developed in chapter 5 are used. Those PM machines are assessed based on the criteria of mass, cost, efficiency and force density.

Chapter 7 The aim of this chapter is to derive new configurations of large direct-drive wind generators that would enable active mass reduction, and facilitate manufacture and maintenance. A flux-concentrating TFPM generator with multiple-modules of rotor and stator segmented is proposed in the chapter. An analytical design model of the proposed TFPM generator is developed, and the model is verified by the experiments of a downscaled generator. The proposed generator is designed for 5 MW and 10 MW direct-drive wind turbines. The generators designed are assessed based on active mass, cost, loss, efficiency and force density.

Chapter 8 Some challenges and solutions of large low-speed direct-drive generators for wind turbines are discussed in the chapter. A ring-shaped generator, a buoyant generator rotor and bearings with high flexibility are proposed as a solution to minimize the structural (inactive) mass. A rough design of 5 MW, 10 MW and 20 MW direct-drive PM generators with new configurations is given to show possibilities to overcome the limits of conventional large direct-drive generators which are heavy and expensive.

Chapter 9 The chapter concludes the thesis and gives recommendations for further research.

Overview of Generator Systmes for Wind Turbines

9

Chapter 2

Overview of Generator Systems for Wind Turbines

2.1 Introduction The aim of this chapter is to gain an understanding of the advantages and the disadvantages of contemporary technologies of generator systems for wind turbines. As discussed in the previous chapter, various wind turbine concepts have been developed in order to maximize the energy harnessed, to minimize the cost, to improve the power quality and to ensure safety during the last three decades. This chapter starts with an overview of existing wind turbine technologies to gain an understanding of the advantages and the disadvantages of different generator systems for wind turbines. Generator systems of wind turbines over 2 MW power rating are also investigated to understand the technology trend for large wind turbines. Some potential generator concepts for wind turbines are also discussed. Secondly, a comparison between direct-drive and geared generator systems is discussed to identify the advantages and the disadvantages of direct-drive generator systems. In the comparison, the energy yield, the cost and the mass of generator systems are considered as criteria.

2.2 Wind turbine technology Different wind turbine concepts have been discussed by a number of authors. In this section, the features, the advantages and the disadvantages of different wind turbine concepts discussed in [Lar 2000][Han 2001][Han 2007][Vih 2002][Hof 2002] are described. The wind turbine concepts can be classified according to the rotational speed, the power regulation method and the drive train construction. Considering the rotational speed, wind turbine concepts can be classified into the fixed speed concept, the limited variable speed concept and the variable speed concept. Considering the power regulation method, the concepts can be classified

Chapter 2

10

into the stall control and the pitch control concept. When the drive train is considered, the concepts can be classified into the geared drive and the direct-drive. Considering the generator type, different configurations of induction machines and synchronous machines have been used for wind turbines. For an overview of wind turbine technologies in this section, different wind turbine concepts are categorized as Table 2-2-1. Table 2-2-1: Different wind turbine concepts

2.2.1 Fixed speed concept The rotor speed of fixed speed wind turbine is fixed and determined by the frequency of the supply grid, the gear ratio and the generator design. The generator system of the fixed speed wind turbine is equipped with a squirrel cage induction generator and a three-stage gearbox (SCIG 3G). The SCIG is directly connected to the grid through a transformer. The SCIG draws reactive

Power regulation method

Drive train Gearbox Generator Converter Sub-section

Stall GD 3 stage SCIG - 2.2.1 A

Active stall GD 3 stage SCIG - 2.2.1 B

Fixed Speed Concept

Pitch GD 3 stage SCIG - 2.2.1 C

Limited Variable Speed Concept

Pitch GD 3 stage WRIG Variable resistance

2.2.2

DFIG Partial scale 2.2.3 A

SCIG Full-scale 2.2.3 B 3 stage

PMSG Full-scale 2.2.3 C

1 stage PMSG Full-scale 2.2.3 D

GD

2 stage + Hydro

EESG - 2.2.3 E

- EESG Full-scale 2.2.4 A

Variable Speed Concept

Pitch

DD - PMSG Full-scale 2.2.4 B


11

power from the grid, and therefore it is necessary to include a capacitor bank for reactive power compensation. The fixed speed wind turbine was extended to use a soft-starter in order to achieve a smoother grid connection. The grid connection schemes of the fixed speed wind turbine are illustrated in Fig. 2-2-1.

Grid

CapacitorSCIG

Gearbox

(a)

(b) Fig. 2-2-1: Fixed speed wind turbine concepts with SCIG 3G system [Han 2007] The advantages of the fixed speed wind turbine can be summarized as follows. • Simplicity • Robustness • Higher reliability • Low cost of electrical parts The fixed speed wind turbine concept was the most commonly used concept during the 1980’s and 1990’s. However, the fixed speed wind turbines have some disadvantages: • Higher mechanical stress • Limited power quality control To overcome the disadvantages of the fixed speed wind turbine concept, a two speed wind turbine concept has been developed. In the two speed concept, a pole changeable SCIG, which leads two rotation speeds, is used. The pole changeable SCIG is useful to increase the efficiency of rotor blades and to reduce the audible noise at low wind speeds. The pole changeable SCIG has been used for both the fixed speed (active) stall control concept and the fixed speed pitch control concept. [Lar 2000][Han 2001][Han 2007]

Chapter 2

12

This sub-section describes three different fixed speed wind turbine concepts, namely the stall control concept, the active stall control concept and the pitch control concept.

A. Stall control The rotor blades of fixed speed stall control wind turbine are directly fixed on the hub. When wind speeds are higher than the rated wind speed, the stall effect on the rotor blades limits the power. This fixed speed stall control wind turbine is completely passive while the wind causes the power regulation by itself. This concept is thus called a passive stall control concept or shortly the stall control concept. This concept is also known as the “Danish concept”. The angle of the rotor blades of fixed speed stall control wind turbine are adjusted only once when the turbine is installed. This means that the performance of the rotor blades is optimal only at one wind speed. When the rotor blades are in deeper stall at high wind speeds, the power output drops. Therefore, the efficiency of the blades of the turbine is not constant over wide range of wind speeds. The fixed speed stall control concept has been used by manufacturers such as Neg-Micon (currently Vestas), Made and Nordex.

B. Active stall control The fixed speed active stall control wind turbine has stalling rotor blades together with a blade pitch system. Stall is caused or delayed by turning the rotor blades. When wind speeds are higher than the rated wind speed, the attack angle of the rotor blades is increased. Thus the stall effect is increased to react to the output power. This fixed speed active stall control concept can avoid the overshoot of the output power at the beginning of a wind gust. Consequently, the fixed speed active stall control wind turbine can supply power to the grid with better quality than the fixed speed stall control wind turbine. In order to turn the rotor blades, the pitch system uses electric motors or hydraulic actuators. Slip rings are also used to transmit energy to the electric motors or the hydraulic actuators. Therefore, the pitch system of the fixed speed active stall control wind turbine requires more maintenance. The fixed speed active stall control concept has been used by manufacturers such as Bonus (currently Siemens, 1.3 and 2.3 MW) and Neg-Micon (currently Vestas, 1.65 MW).

C. Pitch control


13

The fixed speed pitch control wind turbine has rotor blades together with a blade pitch system. The difference between the fixed speed pitch control wind turbine and the fixed speed active stall control wind turbine is the turning direction of the rotor blades. The fixed speed pitch control wind turbine turns the rotor blades to decrease the attack angle of the blades in regulating the output power. The fixed speed pitch control wind turbine can make power control, controlled startup and emergency stopping possible. However, the fixed speed pitch control concept needs more powerful mechanical pitch system than the fixed speed active stall control concept. The fixed speed pitch control concept is not fast enough to regulate the output power, so that power fluctuations can occur even in small fluctuations of wind speeds. This concept has been used by manufacturers such as Suzlon (1.25 MW) and Mitsubishi (1 MW) on the market. [Han 2007]

2.2.2 Limited variable speed concept The limited variable speed wind turbine also has rotor blades together with a blade pitch system. This limited variable speed wind turbine concept is known as the OptiSlip concept, which has been promoted by Vestas since the middle of the 1990’s. The generator system of this concept is equipped with a wound rotor induction generator and a three-stage gearbox (WRIG 3G). The WRIG 3G has a variable external rotor resistance. The external rotor resistance is changed by a power electronic converter which is controlled optically and mounted on the generator rotor. Thus, the total rotor resistance of the generator rotor can be changed, and it is possible to control the slip. Consequently, the power output of the generator system can be controlled, and the output power fluctuations can be reduced. By using optical coupling, slip rings and brushes requiring maintenance are not necessary. Fig. 2-2-2 gives the grid connection scheme of this concept. The most suitable torque-speed characteristic can be chosen to obtain the optimal speed at the operating point. Only speeds over the synchronous speed are possible for generator operation, and the rotor power is not fed back into the grid. The rotational speed range of the generator can be controlled between 0 to 10 % above synchronous speed. [Han 2001] Manufacturers such as Vestas (V66-1.65 MW) and Suzlon (2 MW) use this concept on the current market.

Grid Gearbox

WRIG

Converter

Fig. 2-2-2: Limited variable speed concept with WRIG 3G system (OptiSlip) [Han 2007]

Chapter 2

14

2.2.3 Variable speed concept with gearbox The variable speed wind turbine concept has been developed to achieve maximum aerodynamic efficiency over a wide range of wind speeds, to dampen power and torque peaks caused by wind gust, to allow the turbine to accelerate, and to store energy during wind gusts [Han 2001][Vih 2002]. Thus, since the late 1990’s most wind turbines over a 1.5 MW power rating have been changed to a variable speed concept. The variable speed wind turbine has the following advantages compared to the fixed speed wind turbine [Dub 2004][Lam 2000]. • Improved output power quality • Increased energy capture • Reduced acoustic noise • Reduced mechanical stress of the drive train Considering the power regulation method, the variable speed wind turbine concept can be classified into the variable speed stall control concept and the variable speed pitch control concept. On the market the variable speed pitch control concept has been commonly used, although the variable speed stall control concept has been discussed in research and development programmes such as [Pol 2007][Ban 2007]. Therefore, the variable speed pitch control wind turbine is a major focus of this thesis. The variable speed pitch control wind turbine also has rotor blades together with a blade pitch system to regulate the output power. The variable speed pitch control concept has two operating modes depending on wind speed. Between the cut-in wind speed and the rated wind speed, the variable speed pitch control wind turbine operates at a fixed pitch with variable rotor blade speeds in order to maintain an optimal tip speed ratio. The tip speed ratio means the tip speed of rotor blades divided by the wind speed. In variable rotational speed mode, the electrical frequency of the generator varies, and is different from the grid frequency. Thus, a power electronic converter is needed to decouple the generator from the grid. This variable speed pitch control wind turbine concept has been successfully applied to large wind turbines even though the cost of the converter is higher than the fixed speed concept. When the power has reached the required power, the generator torque controls the electrical power output, while the pitch control is used to maintain the rotor speed within acceptable limits. During gusts the generator power can be maintained at a constant level, while the rotor blade speed increases. The increased energy in wind is stored in the kinetic energy of the rotor blades. If the wind speed decreases the reduced aerodynamic torque results in a deceleration of the rotor blades while the generator power is kept constant. If the wind speed remains high, the rotor blade pitch angle can be changed to reduce the efficiency of rotor blades and torque, once again reducing the rotor speed [Man 2002]. Considering the generator system, the variable speed pitch control wind turbine can be classified into the geared generator type and the direct-drive generator type. In this sub-section, 2.2.3, the variable speed pitch control wind turbine with a different geared generator system is discussed.


15

The variable speed pitch control wind turbine with direct-drive generator system will be discussed in the next sub-section, 2.2.4. The advantages of geared generator systems compared with the direct-drive generator system can be summarized as follows. • Small diameter • Small mass • Low cost However, geared generator systems have the following disadvantages. • Loss and heat dissipation from friction of gearbox • Regular maintenance of gearbox • Audible noise from gearbox As categorized in Table 2-2-1, the following generator systems are discussed for the variable speed pitch control wind turbine with gearbox in this sub-section. • Doubly-fed induction generator system with three-stage gearbox (DFIG 3G) • Squirrel cage induction generator system with three-stage gearbox (SCIG 3G) • Permanent magnet synchronous generator system with three-stage gearbox (PMSG 3G) • Permanent magnet synchronous generator system with single-stage gearbox (PMSG 1G) • Electrically excited synchronous generator system with mechanical and hydraulic gearboxes

A. DFIG system with three-stage gearbox The doubly-fed induction generator system with a three-stage gearbox (DFIG 3G) is the most popular type on the current wind turbine market. The stator of the DFIG is constructed in the same way with the squirrel cage induction generator (SCIG) used in the fixed speed wind turbines. The rotor of the DFIG is equipped with a three-phase winding. The grid connection scheme of the DFIG 3G is illustrated in Fig. 2-2-3. The stator is directly connected to the grid, and the rotor is connected to the grid through a partial scale power electronic converter. The power rating of the partial scale power electronic converter is roughly 30% of the power rating of the turbine. The converter of the DFIG consists of a rotor side converter and a grid side converter. These two converters are controlled independently. Thus, the converter of the DFIG enables reactive power compensation and smooth grid connection. The range of dynamic speed control of the DFIG depends on the power rating of the power electronic converter. Thus, the rotational speed of the DFIG varies from around ± 30% of the synchronous speed by controlling the rotor active power flow direction. It is an advantage of a DFIG that the speed is variable within a sufficient range with limited converter cost. The active and reactive power of the stator can be controlled independently by

Chapter 2

16

controlling the rotor currents with the converter. Furthermore, the grid-side converter can control its reactive power, independently of the generator operation. This allows the performance of voltage support towards the grid. However, the DFIG 3G has the following disadvantages. • High torque peaks in the machine and large stator peak currents under grid fault conditions

depending on control • Regular maintenance of the brush-slip ring set to bring power to the rotor • External synchronization circuit required between the stator and the grid to limit the start-up

current • In the case of grid disturbances, ride-through capability of DFIG is required so that the

control strategies may be very complex. Detailed transient models and good knowledge of the DFIG parameters are required to make a correct estimate of occurring torques and speeds [Che 2005].

Grid Gearbox

Converter

DFIG

Fig. 2-2-3: Variable speed wind turbine concept with DFIG 3G system [Han 2007]

B. SCIG system with three-stage gearbox In order to achieve the variable speed operation with a squirrel cage induction generator with a three-stage gearbox (SCIG 3G), the capacitor bank and soft-starter used in the fixed speed concept are replaced by a full-scale power electronic converter (approximately 120% of nominal generator power). Fig. 2-2-4 illustrates the grid connection scheme of the SCIG 3G with a full-scale converter. The full-scale converter enables variable speed operation at all wind speeds. Thus, it is not necessary to keep the generator frequency the same as the grid frequency. Compared with the SCIG system used in the fixed speed wind turbine, a disadvantage of the SCIG 3G for variable speed wind turbine is the high cost of the power electronic converter. Siemens uses this concept for the model Bonus 107-3.6 MW on the market.

Grid Gearbox Converter

SCIG Fig. 2-2-4: Variable speed wind turbine concept with SCIG 3G system and full-scale converter [Han 2007]


17

C. PMSG system with three-stage gearbox An alternative generator system in geared wind turbines is a permanent magnet synchronous generator system with a three-stage gearbox (PMSG 3G). A full-scale power electronic converter is used in the PMSG 3G. Compared with the DFIG 3G, the PMSG 3G has the following advantages: • the generator has better efficiency, • the generator may be slightly cheaper, • the generator is brushless, • it can be used both in 50 Hz and 60 Hz grids, • the grid-fault ride through capability is less complex, and the following disadvantages: • larger and more expensive converter (100% of rated power instead of 30%), • larger converter losses compared to DFIG. On the wind turbine market, this system has been used by Made for the AE-5x 800kW models, by GE Energy for the multi-megawatt series [GE 2010]. A manufacturer, Clipper Liberty turbine uses four sets of 660 kW PMSG with a gearbox for 2.5 MW model. Vestas also use this concept in the V112 3MW model.

Fig. 2-2-5: Variable speed wind turbine concept with PMSG 3G system [Han 2007]

D. PMSG system with single-stage gearbox This generator system, PMSG 1G, is equipped with a PMSG, a single-stage planetary gearbox and a full-scale converter. This system has been introduced as the Multibrid concept. The single-stage gearbox increases the speed by approximately a factor of 10. Therefore, this system has gained attention because it has advantages such as both a higher speed than the direct-drive concept and a lower number components than the multiple-stage gearbox system. The grid connection scheme

Chapter 2

18

of this concept is shown in Fig. 2-2-6. Fig. 2-2-7 shows the WinWind wind turbine with PMSG 1G. Wind turbine manufacturers such as Mulibrid (see Fig. 2-2-8) and WinWind use this concept on the current market.

Fig. 2-2-6: Variable speed wind turbine concept with PMSG 1G [Han 2007]

Fig. 2-2-7: WinWind wind turbine with PMSG 1G [Win 2006]


19

Fig. 2-2-8: Multibrid M5000, 5MW [Mul 2006]

E. EESG with mechanical and hydraulic gearboxes The generator system of this concept is equipped with a two-stage gearbox, a hydraulic gearbox, and an electrically excited synchronous generator (EESG). This generator is a fixed-speed 13.8 kV 4-pole synchronous generator, and is directly connected to the grid. Therefore, a frequency converter is not necessary in this system. A German company, DeWind has used this concept for the D8.2 model-2 MW. Fig. 2-2-9 shows the D8.2 wind turbine of DeWind. For the hydraulic gearbox, DeWind has used the WinDrive technology of Voith Turbo. The WinDrive gearbox hydraulics uncouple the mechanical gearbox output shaft from the generator input shaft. Thus drive train vibrations are dampened. Shocks and peak loads are substantially reduced. The DeWind expects this technology to achieve mass reduction of the drive train by up to 20% and the nacelle by up to 10% in the future [Eiz 2007]. Fig. 2-2-10(a) gives the schematic diagram of the generator system of the D8.2 DeWind [DeW 2006]. A downscaled model of the WinDrive with a two-stage gearbox is given in Fig. 2-2-10(b).

Chapter 2

20

Fig. 2-2-9: Dewind D8-2, 2MW [DeW 2006]

(a) Schematic diagram of the D8.2 model [DeW 2006]

(b) Photograph of WinDrive downscaled (taken at EWEC2007) Fig. 2-2-10: Schematic diagram of D8.2 model of DeWind and WinDrive of Voith Turbo


21

A wind turbine manufacturer in New Zealand, Windflow, has also introduced a hydraulic variable speed system, which is a patented Torque Limiting Gearbox (TLG) system. The TLG’s hydraulic system effectively eliminates inertial torques by enabling the generator speed to remain constant while the wind turbine speed varies. This also avoids the use of power electronics. Windflow expects the TLG system to be lighter than the generator systems of conventional variable speed wind turbines. Fig. 2-2-11(a) gives a comparison of the top mass of wind turbines from different manufacturers. The TLG system eliminates overloads on the gearbox by providing the torque smoothing even in very gusty conditions. Thus, Windflow also expects the power quality using the TLG system to be better than the power quality of conventional gearbox systems. Fig. 2-2-11(b) and (c) give power qualities of wind turbines without and with the application of the TLG system, respectively. [Win 2007]

(a) Top mass comparison of different wind turbines (kg/kW installed)

(b) (c)

Fig. 2-2-11: Comparison of wind turbine top mass, and comparison of power quality before and after applying Windflow’s concept [Win 2009]

Chapter 2

22

2.2.4 Variable speed direct-drive concept The rotor of a direct-drive generator is directly connected to the hub of the rotor blades, so that the generator rotates at low speed. This low speed makes it necessary to produce a high torque, which means the generator must have a large diameter. In order to increase the efficiency, to reduce the active mass, and to keep the end winding losses small, a direct-drive generator is usually designed with a large diameter and small pole pitches. When considering energy yield and reliability, direct-drive systems seem to be better than geared systems. The advantages of the direct-drive generator system compared to the geared generator system can be summarized as follows: • simplified drive train by omitting the gearbox • higher efficiency in partial load • high reliability and high availability • low noise of the drive train • flexible control of the entire system by the converter However direct-drive generators have the following disadvantages [Pol 2006][Wid 2006][Gra 1996][Dub 2004][Lam 2000][Böh 2003]: • large diameter • large mass • high cost • specific design to supply high electrical torques at low speeds

Fig. 2-2-12: Direct-drive synchronous generator


23

A conventional structure of a direct-drive synchronous generator is presented in Fig. 2-2-12. The direct-drive generators available on the market can be classified into two concepts, namely the electrically excited synchronous generator (EESG) and the permanent magnet synchronous generator (PMSG). In this section, direct-drive EESG and PMSG system are described.

A. Direct-drive EESG system The electrically excited synchronous generator (EESG) is usually built with a rotor carrying the field system provided by a DC excitation. For the DC excitation, slip rings and brushes, or a brushless exciter employing a rotating rectifier are used. The grid connection scheme of an EESG for a direct-drive wind turbine is illustrated in Fig. 2-2-13. The stator of an EESG carries a multi-phase winding that is similar to the winding of the induction generator. The rotor of an EESG can be constructed with salient or cylindrical poles. Between these two types, salient poles are mostly used for low speed machines, and are the most useful for wind turbine generator applications. In order to arrange space for the excitation windings, the pole pitch of the EESG must be large enough. All the generator power is processed through a power electronic converter. At the generator side of the converter, the amplitude and frequency of the voltage can be fully controlled independent of the grid. The generator speed is fully controllable over a wide range, even at very low speeds. The active and reactive power can be also fully controlled in the case of normal and disturbed grid conditions. [Wid 2006] The EESG can control the flux, thus it can minimize the loss in different power ranges. Furthermore, the EESG does not require permanent magnets (PMs), which would represent a large fraction of the generator costs. Therefore, the EESG is mostly used for large direct-drive applications on the wind turbine market [Soe 2005]. Enercon, a German manufacture currently uses this concept on the market. MTorres also uses this concept. Fig. 2-2-14 gives the E-112 4.5 MW model of Enercon with large generator diameter (nearly 12 m) [DeV 2004].

Grid EESG

Converter

Converter

Fig. 2-2-13: Scheme of a direct-drive EESG system [Han 2007]

Chapter 2

24

Fig. 2-2-14: Enercon E-112, 4.5MW [Ene 2006]

B. Direct-drive PMSG system The permanent magnet synchronous generator (PMSG) is excited by a permanent magnet. The PMSG is connected to the grid through a full-scale power electronic converter. The grid connection scheme of the PMSG for direct-drive wind turbines is shown in Fig. 2-2-15. The advantages of PMSG system compared to the EESG system can be summarized as follows: • higher efficiency and energy yield • no additional power supply for the magnetic field excitation • improved thermal characteristics due to the absence of field losses • higher reliability due to the absence of mechanical components such as slip rings • lightweight, hence higher power to mass ratio However, the PMSG systems have the following disadvantages: • high cost of PM • high cost of the converter • difficulties to handle in manufacture • demagnetization of PM at high temperature


25

In recent years, the performance of PMs has improved and the cost of PM has decreased. Thus the use of PMs seems more attractive than before. Additionally, the cost of power electronics is decreasing. This figure may make the direct-drive PMSG system (PMSG DD) with a full-scale power converter attractive for large wind generators in the future. Fig. 2-2-16 shows a 1.5(2) MW direct-drive wind turbine system of Zephyros (currently STX Windpower). EWT (Emergya Wind Technologies, 2 MW), STX Windpower, Scanwind (3.5 MW), Leitwind (1.5 MW), Vensys (2.5 MW), Goldwind (1.5 MW), and Unison (750 kW) use this concept. Jeumont (750 kW) also used this concept with axial flux machine, but it is not currently available.

Grid

PMSG

Converter

Fig. 2-2-15: Scheme of a direct-drive PMSG system [Han 2007]

Fig. 2-2-16: Zephyros Z72-1.5(2)MW. Source Zephyros [Pol 2006]

Chapter 2

26

2.2.5 Generator systems on the market Wind turbines with a power rating of 2 MW or higher are reviewed in order to understand the generator technologies for large wind turbines on the market. Clipper Liberty (USA) and Enercon (Germany) have built 7.5 MW and 6 MW prototype wind turbines, respectively. The type of generator system, power rating and manufacturer of large wind turbines are summarized in Ttable 2-2-2. [Ene 2006][Win 2006][STX 2010][Ves 2007][Sie 2006][Rep 2006][Nor 2006][Mul 2006][GEE 2006][Gam 2006][Eco 2006][Sin 2010][EWT 2010] Most manufactures use a gearbox in the generator system. The most commonly used generator type is still the induction generator and the doubly-fed induction generator system with a three-stage gearbox (DFIG 3G) is the most dominant. Fig. 2-2-17 depicts the market share of the different generator concepts based on the wind turbine market data over a 10 year period (1995–2005) [Han 2007]. As shown in Fig. 2-2-17, the market share of the fixed speed wind turbine concept (SCIG) has decreased to about one third in last 10 years, from almost 70% in 1995 to almost 19% in 2005. The market share of the OptiSlip concept (WRIG) has declined since 1997 in favor of the more attractive variable speed concept (DFIG). The trend depicted in Fig. 2-2-17 clearly indicates that the WRIG type is being phased out of the market. Incorporation of the DFIG wind turbine concept has increased from 0% to almost 62% of yearly installed wind power during the initial 10 years, and clearly became the most dominant concept by the end of 2005. Market penetration of the synchronous generator concept (EESG or PMSG) has altered little over 10 years, with no dramatic changes observed for SCIG, WRIG and DFIG. During the 10 years, the EESG and PMSG with full-scale converters have ranked third or fourth (Fig. 2-2-17) [BTM 2005] [Han 2007].


27

Table 2-2-2: Large wind turbines on the market with a power rating over 2 MW

Drive train Generator Power / Rotor diameter / Speed Manufacturer

4.5 MW / 120 m / 14.9 rpm Vestas (DK)

3.6 MW / 104 m / 15.3 rpm GE (US)

2 MW / 90 m / 19 rpm Gamesa (ES)

3 MW / 113 m / rpm Sinovel (CN)

3 MW / 109 m / 13.2 rpm Acciona (ES)

5 MW / 126 m / 12.1 rpm Repower (DE)

2.5 MW / 90 m / 14.85 rpm Nordex (DE)

3 MW / 100 m / 14.25 rpm Ecotecnia (ES)

2 MW / 90 m / 20.7 rpm DeWind (DE)

DFIG

2 MW / 90.6 m / 18.1 rpm Hyosung (KR)

SCIG 3.6 MW / 107 m / 13 rpm Siemens (DK)

WRIG 2.1 MW / 88 m / 17.6 rpm Suzlon (IN)

3 MW / 112 m / 12.8 rpm Vestas (DK)

2.x MW / 88 m / 16.5 rpm GE (US)

Multiple-stage gearbox

PMSG

2 MW / 88 m / Unison (KR)

5 MW / 116 m / 14.8 rpm Multibrid (DE) Single-stage gearbox PMSG

3 MW / 90 m / 16 rpm Winwind (FI) Hydro-controlled Multi-stage gearbox

EESG 2 MW / 90 m / 20.7 rpm DeWind (DE)

EESG 4.5 MW / 114 m / 13 rpm Enercon (DE)

EESG 1.65 MW / 70 m / 20 rpm MTorres (ES)

PMSG 2 MW / 82.7 m / 18.5 rpm STX (NL)

PMSG 2 MW / 90.5 m / 15.8 rpm EWT (NL)

PMSG 3.5 MW / / 19 rpm Scanwind (NO)

PMSG 2.5 MW / m / 14.5(16) rpm Vensys (DE)

PMSG 1.5 MW / 70 m / 19 rpm Goldwind (CN)

Direct-drive

PMSG 2 MW / 83.3 m / 19 rpm JSW (JP)

Chapter 2

28

2.2.6 Other potential concepts Other potential concepts for wind generators are the direct-drive induction generator for direct grid connection [Gri 1991][Del 1992], the switched reluctance generator [deH 1994][Tor 1995][Tor 2002][Tor 1993][Mue 2005], the brushless doubly-fed induction generator (BDFIG) [Bol 2006][Soe 2005][Run 2004][Car 2006], the transverse flux permanent magnet (TFPM) generator, and the high temperature superconducting generator (HTSG). Direct-drive induction generator The direct-drive induction generator has the axial flux configuration and a segmented stator that is fixed to the wind turbine tower. The stator winding is directly connected to the grid, thus this generator is simple. [Gri 1991][Del 1992] It was noted that the damping of the generator with a high slip is no problem, even though the generator is directly connected to the grid. A direct-drive induction generator rated at 500 kW with about 9 m diameter and 40 rpm [Gri 1991].

Fig. 2-2-17: World market share of yearly installed power for different wind turbine generator types, Data source:[Han 2007]


29

Brushless doubly-fed induction generator The BDFIG consists of a dual-cage in the rotor and two multi-phase windings, namely the power winding and the control winding, with different pole pairs in the stator. The power winding is directly connected to the grid, and the control winding is connected to the grid through an AC/DC/AC bi-directional power electronic converter. The BDFIG has similar properties to a DFIG with respect to speed range, active and reactive power control. The BDFIG does not have an electrical connection between the rotor and the stator. This is an advantage of the BDFIG. However, the operation principle and the assembly of the BDFIG are complex. In addition, a number of authors in [Bol 2006][Soe 2005] noted that slip rings and brushes are not the most critical components in the DFIG system. Therefore, the BDFIG is not likely to be a commercial success within the years to come [Soe 2005]. Transverse flux permanent magnet generator The transverse flux permanent magnet (TFPM) generator has a magnetic flux path perpendicular to the direction of the rotor rotation. Compared with the conventional PM generators, the TFPM generator has been discussed as a generator with high force density. Thus, the TFPM generators have more potential to reduce both the volume and the mass than the longitudinal flux permanent magnet (LFPM) generators. However, Dubois in [Dub 2004] concluded that LFPM generators are better than the TFPM generator with toothed rotor based on the mass and the cost of active part, if the air gap length is larger than 1.5 mm. Therefore, it is needed to find TFPM generators which are lighter and cheaper than LFPM generators for large direct-drive wind turbines. High temperature superconducting generator HTS (High Temperature Superconducting) generator concepts are also discussed for direct-drive wind turbine applications. In order to investigate the feasibility of the HTS generator for very large direct-drive wind turbines, a research and development on a 10 MW HTS direct-drive wind generator was initiated by NREL (National Renewable Eneragy Laboratory) and AMSC (American Superconductor) in the US. The estimated mass of the HTS generator is about 120 tonnes, which is 30 to 40% of the mass of conventional direct-drive generators. The voltage and rotational speed in design of the 10 MW HTS generator are 6 kV and 11 rpm, respectively. [Com 2009]

Chapter 2

30

2.3 Direct-drive and geared generator systems In order to identify the features of direct-drive generator system, comparisons of different generator systems in references are reviewed in this section. Comparisons of different generator systems for wind turbines have been discussed by a number of authors. [Pol 2006][Wid 2006][Byw 2004][Gra 1996][Dub 2004][Lam 2004][Han 2001][Che 2005][Dub 2000][Böh 1997][Sie 1998][Böh 2003][Poo 2003][Han 2007][Car 1994][Soe 2005][Ann 1996] A 1.5 MW wind turbine with a direct-drive electrically excited synchronous generator system (EESG DD) has been compared to a doubly-fed induction generator system with a three-stage gearbox (DFIG 3G) by Böhmeke and Siegfriedsen et al. [Böh 1997][Sie 1998] They conclude that the EESG DD is heavier and more expensive than the DFIG 3G. The top masses of different wind turbines rated up to 5 MW have been also compared in [Böh 2003]. The top mass of the Multibrid concept (a permanent magnet synchronous generator with one-stage gearbox: PMSG 1G) was estimated at the same mass with the concept of DFIG 3G, which is the lightweight concept. Some authors such as Polinder, Bywaters, Grauers, Dubois, Lampola and Poore et al. have discussed direct-drive PM synchronous generator systems (PMSG DD) for wind turbines. The elimination of the excitation losses and the reduction of the mass of active material have been indicated as advantages of the PMSG DD in [Byw 2004][Gra 1996][Dub 2004][Lam 2000][Poo 2003] and [Pol 2006]. A comparison between the PMSG DD and a squirrel cage induction generator system with a three-stage gearbox (SCIG 3G) has been drawn by Grauers [Gra 1996]. The power rating of wind turbines investigated in [Gra 1996] is from 30kW up to 3MW. A SCIG 3G rated at 500 kW has also been compared to PMSG DD in [Ann 1996]. Lampola has compared the total mass of the 500 kW PMSG DD and the geared drive induction generator system [Lam 2000]. Recently, Polinder et al have compared 3 MW direct-drive and geared generator systems based on their energy yield, cost, and mass [Pol 2006]. From a review of different generator systems described in the above references, the advantages (+) and the disadvantages (-) of the generator systems are summarized in Table 2-3-1 based on different criteria. Geared generator systems have been discussed as a better system than direct-drive generator systems in terms of generator diameter, mass, cost and tower head mass compared to. Direct-drive generator systems are better than geared generator systems based on axial length, efficiency, loss and energy yield. In order to identify the advantages and the disadvantages of direct-drive generator systems quantitatively, the comparison results of different 3 MW generator systems in [Pol 2006], namely DFIG 3G, EESG DD, PMSG DD and PMSG 1G are discussed. Fig. 2-3-1 gives both annual losses and annual energy yield of the different generator systems. The annual losses of such generator systems include the losses in gearbox, generator and converter. A cost comparison of such generator systems is given in Fig. 2-3-2.


31

Table 2-3-1: Advantages and disadvantages of different generator systems SCIG 3G DFIG 3G PMSG

1G EESG DD

PMSG DD

0.5 & 3 MW [Gra 1996]

•Diameter •Axial length •Mass •Efficiency

+ - + -

- + - +

1.5 MW [Böh 1997] [Böh 1998]

•Cost •Mass

+ +

- -

0.3 - 5 MW [Böh 2003]

•Tower head mass + + -

0.5 MW [Ann 1996]

•Annual energy yield - +

3 MW [Pol 2006]

•Mass •Cost •Loss •Annual energy yield

+ + + - - - -

+ + + -

- - - - - +

- -

+ + + +

DFIG 3G EESG DD PMSG DD PMSG 1G

Annu

al lo

sses

[MW

h]

0

200

400

600

800

1000

1200

Annu

al E

nerg

y yi

eld

[GW

h]

0

2

4

6

8

10

12

Converter loss Generator loss Gearbox loss Annual Energy yield

Fig. 2-3-1: Annual losses and energy yield of different generator systems

Chapter 2

32

From Fig. 2-3-1 and Fig. 2-3-2, it is concluded that: The energy yield of direct-drive synchronous generator systems is higher than the energy yield of geared generator systems. In terms of losses and energy yield, the PMSG is superior compared to the EESG. The DFIG 3G has a higher loss and lower energy yield than other generator systems. The converter loss and cost of DFIG 3G are the lowest because the generator system uses a partial-scale power converter (about 30 % of power rating). The mass and cost of the DFIG 3G are the lowest in different generator systems. The EESG DD is the heaviest and the most expensive generator system, and the PMSG DD is addressed as the second expensive generator system. Fig. 2-3-2 shows that the dominant part to determine the cost-competitiveness between the EESG DD and the PMSG DD is the active part. Among active materials used in generators, PMs are the most expensive material. Permanent magnet material in 3 MW PMSG DD accounts for 7 % of active mass and 26 % of active material cost. Therefore, minimizing PM material is an important issue to make the PMSG DD more attractive in terms of the cost.

DFIG 3G EESG DD PMSG DD PMSG 1G

Cos

t of G

ener

ator

sys

tem

[kEu

ro]

0

200

400

600

800

ConverterActive partInactive partGearbox

Fig. 2-3-2: Cost of different generator systems


33

2.4 Conclusions In this chapter, different generator systems for wind turbines have been reviewed to understand contemporary technologies of the generator systems. Different wind turbine concepts such as the fixed speed concept, the limited variable speed concept and the variable speed concept have been investigated in section 2.2. In the section, generator systems of large wind turbines with a power rating of 2 MW or higher were also investigated, and some potential generator concepts for wind turbines were also discussed. From the investigation of different wind turbine concepts, it is concluded that: • Among different wind turbine concepts, the variable speed wind turbine concept is the most

used concept because of its improved output power quality, increased energy capture, reduced acoustic noise and reduced mechanical stress of drive train.

• Based on the wind turbine market data, the most commonly used generator concept is still the induction generator, and the doubly-fed induction generator system with a three-stage gearbox (DFIG 3G) is the most dominant.

• The size and the power rating of wind turbines are being scaled up. • Generator systems with a full-scale power converter are attractive to deal with grid problems. • Among potential generator concepts for wind turbines, the transverse flux permanent magnet

(TFPM) generator and the high temperature superconducting (HTS) generator have more potential to reduce the mass than conventional generators. However, the cost-competitiveness of those generators must be more strengthened to use widely.

To identify the advantages and the disadvantages of direct-drive generator systems for large wind turbines, a quantitative comparison between direct-drive and geared generator systems was made based on energy yield, mass and cost in section 2.3. From the comparison, it is concluded that: • Geared generators are lighter and cheaper than direct-drive generators. • The DFIG 3G is a solution with low mass and low cost. • Converter loss and cost of the DFIG 3G are the lowest since the generator system uses a

partial-scale power converter with about 30 % of power rating. • Direct-drive generators have a high energy yield compared to geared generators. • The direct-drive permanent magent synchronous generator system (PMSG DD) is a solution

with low loss and high energy yield. • The DFIG 3G has the highest ratio of the annual energy yield to the generator system cost,

although DFIG 3G’s energy yield is the minimum among different four generator systems. Therefore, the DFIG 3G has been dominantly used on the wind turbine market.

• The direct-drive electrically excited synchronous generator system (EESG DD) is the heaviest and the most expensive generator system.

• Among the active materials of generators, PMs are the most expensive material. If the price

Chapter 2

34

of permanent magnet material is highly increased, then the cost-competitiveness of the PMSG DD against the EESG DD will be weakened.

• In order to maximize the ratio of energy yield to generator system cost, a generator system with the maximum energy yield, the minimum cost and the minimum mass is necessary for wind turbines.

Direct-Drive PM Generators for Large Wind Turbines

35

Chapter 3


3.1 Introduction The aim of this chapter is to identify the active mass-competitiveness of permanent magent (PM) machines for large direct-drive wind turbines. As discussed in the previous chapter, the direct-drive permanent magnet synchronous generator system (PMSG DD) is a solution with high energy yield, high reliability and fewer maintenance problems. Considering the mass and cost, the doubly-fed induction generator system with a three-stage gearbox (DFIG 3G) is a lightweight and low-cost solution. Therefore, there are still different opinions about which generator system is the most suitable among existing wind turbine generator technologies. A generator system with both the maximum energy yield and the minimum cost can be defined as the most suitable generator system for wind turbines. If it is possible to reduce the cost of a PMSG DD without diminishing its performance to the cost of the DFIG 3G or below, then the PMSG DD would be the most suitable generator system. The material cost of the generators mainly depends on the mass. Therefore, this chapter focuses on identifying the mass-competitiveness of PM machines with various configurations. The chapter starts with the definition of criteria to assess different generator configurations. Next, an overview of PM machines with different configurations is given in order to list the features and mass-competitiveness of those machines. In the overview, such machines are classified as the radial flux (RF), the axial flux (AF) and the transverse flux (TF) configurations.

3.2 Criteria to assess direct-drive generators for wind turbines This section discusses the criteria needed to assess direct-drive generators for large wind turbines. Various design criteria for electric machines have been discussed in scientific literature [Bar 2009][Bia 2006][Dub 2004][Kle 1984][Kri 1987]. The criteria discussed in the literature are

Chapter 3

36

efficiency, force density, power/volume, torque/volume, power/mass, torque/mass, power/cost, torque/cost etc. Electric machines are generally designed to maximize the value of one or more of these criteria. A better choice for the machine design generally results from a compromise which increases a value of one parameter without diminishing the other ones [Car 1995a]. If it is possible to minimize the cost of an electric machine while yielding acceptable performance, then the machine would be a suitable machine. The capital cost of electric machines is determined by both the material cost and the manufacturing cost. The manufacturing cost depends on the manufacturing processes for the variants of a design. The material cost mainly depends on the mass of the machines. In the thesis, the manufacturing cost is not considered, thus the mass is chosen as a parameter to assess different PM machines. When scaling up the electric machines at the same rotational speed, we can say that the mass of the machines depends on the power rating or the torque rating. However, in the case of wind turbines, the rotational speeds are further decreased in scaling up the power rating. The mass of machines depends on their torque rating. Therefore, in the thesis the torque rating of the direct-drive wind generator is chosen as a parameter rather than the power rating. In order to identify the mass-competitiveness of different PM machines, the mass to torque ratio is used as a criterion to assess the machines. Two main parameters and a main criterion used to assess direct-drive generators for wind turbines in the chapter are repeatedly written as follows. • Parameters: mass and torque • Criterion: mass/torque

3.3 PM machines for direct-drive This section lists the features and mass-competitiveness of different PM machines as discussed in the scientific literature. The section focuses on characterizing the mass-competitiveness of different PM machines as a function of the torque. Direct-drive PM machines have advantages, which were described in the previous chapter, such as lower incidence of failure, increased energy yield and higher reliability compared to both electrically excited (EE) machines and geared machines. Additionally, the performance of PMs has been improving and the cost of PMs and power electronics has been decreasing in recent years. Therefore, direct-drive PM machines can be more attractive when used in large wind turbines. Such PM machines can be classified by both the direction of the magnetic flux and the structure as follows. • the longitudinal flux or transverse flux machine • the radial flux or axial flux machine • the iron-cored or coreless machine • the slotted or slotless machine


37

• the surface-mounted PM or flux-concentrating PM machine In the configuration of longitudinal flux machine, the direction of motion is parallel with the direction of magnetic flux. The direction of magnetic flux of transverse flux machine is perpendicular to the direction of motion. Linear models of a longitudinal PM machine and a transverse flux PM machine are illustrated in Fig. 3-3-1.

(a) a slotted surface-mounted PM machine

(b) a flux-concentrating TFPM machine Fig. 3-3-1: Longitudinal flux machine and Transverse flux machine In the thesis, PM machines are mainly categorized into three concepts as follows. • the radial flux PM (RFPM) machine • the axial flux PM (AFPM) machine • the transverse flux PM (TFPM) machine The RFPM machine produces a magnet flux in the radial direction in the air gap with longitudinal flux configuration. The AFPM machine is a machine producing the magnetic flux in the axial direction with longitudinal flux configuration. The TFPM machine is a PM machine producing the magnetic flux that is perpendicular to the moving direction.

Chapter 3

38

3.3.1 RFPM machine In order to identify the possibility and potential of radial flux permanent magnet (RFPM) machines for large direct-drive wind turbines, the machines discussed in the references are surveyed. Different RFPM machines have been discussed in a number of references [Gra 1996][Dub 2004][Lam 2000][Spo 1996a][Che 1995a][Spo 1996b][Gen 1995][Wes 1996][Spo 1996c][Spo 1998][Spo 2005][Tav 2006][Wu 2000][Che 2005b][Che 2000][Kim 2005][McD 2006][Lam 1996][Pol 1998][Sle 1992][Lib 2004][Wan 2005a][Wan 2005b]. In this thesis the RFPM machines are categorized as follows based on the electromagnetic configurations: • the slotted surface-mounted RFPM machine (Fig. 3-3-2) • the slotted flux-concentrating RFPM machine (Fig. 3-3-3) • the slotless RFPM machine (Fig. 3-3-4) • the ironless RFPM machine (Fig. 3-3-5)

Fig. 3-3-2: A slotted surface-mounted RFPM machine [Dub 2004]

Fig. 3-3-3: A slotted flux-concentrating RFPM machine [Spo 1996a]


39

Fig. 3-3-4: A slotless RFPM machine configuration

Fig. 3-3-5: Cross-section of a lightweight ironless RFPM generator for wind turbines [Spo 2005]

Chapter 3

40

A. Slotted surface-mounted RFPM machine The slotted surface-mounted RFPM machine consists of the stator with slotted iron cores and windings, and the rotor with yoke and surface-mounted permanent magnets shown in Fig. 3-3-2. The slotted surface-mounted RFPM machine has been discussed as a better choice for direct-drive generators of large wind turbines [Gra 1996][Lam 2000]. In [Lam 1996], a slotted surface-mounted RFPM machine for direct-drive wind turbine of 500 kW power rating was designed, and its electrical performance was calculated using finite element methods. In [Lam 2000], two different stator windings and five different rotor constructions were discussed in order to find suitable generator types for direct-drive wind turbines. A slotted surface-mounted RFPM machine for a 1.5 MW wind turbine was designed using magnetic equivalent circuit methods and analyzed using finite element methods in [Kim 2005]. Different direct-drive RFPM machines of 5 kW power rating at 50 rpm were investigated and compared in [Lib 2004][Lib 2005], where a surface-mounted PM type and an inset surface-mounted PM type were included for the slotted surface-mounted RFPM machine. A design and control for a slotted surface-mounted RFPM generator for a direct-drive wind turbine of 20 kW power rating at 110 rpm were discussed in [Wan 2005a][Wan 2005b]. The mass minimization of slotted surface-mounted RFPM generators for large direct-drive wind turbines of 2, 3 and 5 MW power ratings is discussed in [McD 2006]. In the reference, the optimum ratios of the axial length to the air gap diameter of the generators with conventional mechanical construction were chosen in such a way that minimizes the total mass of the generators. Slotted surface-mounted RFPM machines with outer rotor construction were discussed by other authors [Wu 2000][Che 2000][Lib 2004][Lib 2005]. A slotted surface-mounted RFPM machine with an outer rotor is represented in Fig. 3-3-6. The optimization techniques of a 20 kW slotted surface-mounted RFPM machine with an outer rotor were discussed in order to reduce cogging torque and to improve performance using finite element methods in [Wu 2000]. In this research, design optimizations for different numbers of poles and different types of lamination and magnet materials were performed. A design and finite element analysis of a slotted surface-mounted RFPM generator with an outer rotor for direct-drive wind turbines were presented in [Che 2000]. In [Dub 2004], the slotted surface-mounted RFPM generator was used as a reference for the comparison of different generator topologies. In order to make a comparison with different generators, an analytical model for the generator was developed. Analytical models of the slotted flux-concentrating RFPM generator and the slotted surface-mounted AFPM generator were also developed. The analytical models were used for the optimization procedure to have the lowest cost of active material and the lowest torque/mass of the generators with different rotor diameters of about 1 m up to 4 m.


41

Fig. 3-3-6: A slotted surface-mounted RFPM machine with an outer rotor [Wu 2000] B. Slotted flux-concentrating RFPM machine The slotted flux-concentrating RFPM machine consists of a stator with slotted iron cores and windings, and a rotor with flux-concentrating permanent magnets and iron cores as shown in Fig. 3-3-3. The slotted flux-concentrating RFPM machine has been discussed as a machine type that allows an air gap flux density higher than the remanent flux density of permanent magnets [Spo 1996a][Dub 2004]. In [Spo 1996a], a type of slotted flux-concentrating RFPM generator using ferrite permanent magnets for 400 kW direct-drive wind generator was discussed. A slotted flux-concentrating RFPM generator with modular structure for direct-drive wind turbines was discussed in [Che 1995][Spo 1996b][Spo 1996c]. The stator module of the generator is divided circumferentially into a large number of E-cores, each carrying a coil which is wound on a bobbin and fitted prior to assembly of the generator as shown in Fig. 3-3-7. The rotor consists of a large number of blocks with steel pole sides and ferrite permanent magnets tangentially magnetized. In [Spo 1996c], this modular generator was compared with the generator discussed in [Spo 1996a] based on electromagnetic parameters, active mass, efficiency and so on. The cancellation of noise and vibration as well as an alternative damping system for this modular generator were discussed in [Gen 1995][Wes 1996]. A mathematical model of the modular generator was described and its predictions were compared with measured losses in [Spo 1998]. A slotted flux-concentrating RFPM machine with toroidal windings was discussed in [Han 2005][Han 2006][Wid 2006]. The windings in the stator were placed in flat slots, and the magnets were fixed on the rotor support structure made of non-ferromagnetic material. Soft magnetic cores were attached on both PM poles to make the flux path easy. Fig. 3-3-8 depicts this machine.

Chapter 3

42

In [Dub 2004] analytical design and optimization of the slotted flux-concentrating RFPM machines with different rotor diameters of about 1 m up to 4 m were discussed including the active mass estimation.

Fig. 3-3-7: Modular-shaped flux-concentrating PM machine with ferrite magnets [Che 1995][Spo 1996b][Gen 1995][Wes 1996][Spo 1996c]

Fig. 3-3-8: Schematic diagram of a RFPM machine with flux-concentrating arrangement [Han 2005]


43

C. Slotless RFPM machine The slotless RFPM machine with two rotors and one stator shown in Fig. 3-3-4 was discussed in [Kor 2004]. In the reference, design and optimization of the generator were carried out using finite element methods. It was concluded that this generator concept makes end-windings shorter, thus reducing the mass of the generator, increasing its efficiency and reducing the active material cost. However, the data designed for the generator was not enough to determine the trends of active mass in scaling up the torque rating. Tavner et al. described how the large number of pole pairs and air gap diameter affect the design of large low-speed direct-drive machines in [Tav 2006]. They also discussed and compared the output torque with the ratio of structural mass to active mass of different machines, such as induction machines, wound rotor synchronous machines, PM synchronous machines, slotless armature PM machines and ironless machines. The slotless armature PM machine was discussed as a type which offered the best prospect for constructing large, low speed machines with light structures. A single bearing structure was discussed as a design to save on bearing cost. The structure consists of cantilevered supports for both the rotor and the stator as in Fig. 3-3-9.

Fig. 3-3-9: A single bearing generator structure with cantilevered supports [Tav 2006]

Chapter 3

44

D. Ironless RFPM machine A new ironless RFPM machine for large direct-drive wind turbines was proposed by Spooner et al. [Spo 2005]. The proposed machine consists of a pair of spoke wheels which carry the rotor and the stator as shown in Fig. 3-3-5. The rotor consists of a steel rim with surface-mounted rare-earth permanent magnets, and the stator consists of air-gap windings with a non-metallic support structure. The absence of the stator iron core avoids the radial attractive force, thus the lightweight spoke structures are sufficiently stiff. The active mass and total mass of the proposed ironless generator rated at 5 MW and 13.7 rpm were compared to slotted and slotless RFPM generators in terms of shear stress and efficiency. The proposed generator was lighter than other types of RFPM generator and even lighter than a geared generator. The proposed generator was also discussed as an example of a type with less maintenance demands and with the prospect of creating a fault-tolerant electrical system. E. Active mass of RFPM machines In order to determine the trends of active masses of RFPM machines during the scaling up of the torque rating, active masses of different RFPM machines discussed in the references are addressed as a function of the torque. Among different RFPM machines, not enough references could be found to determine general characteristics of torque and acctive mass of the slotless RFPM machines and the ironless RFPM machines. Therefore, the following slotted RFPM machines are chosen to characterize active mass of the machines in scaling up the torque rating. • R1: Slotted surface-mounted RFPM machine with inner rotor and rare earth magnets (RF-

SM-ST-IR-NdFeB) • R2: Slotted surface-mounted RFPM machine with outer rotor and rare earth magnets (RF-

SM-ST-OR-NdFeB) • R3: Slotted flux-concentrating RFPM machine with inner rotor and ferrite magnets (RF-FC-

ST-IR-Ferrite) • R4: Slotted flux-concentrating RFPM machine with inner rotor and rare earth magnets (RF-

FC-ST-IR-NdFeB) • R5: Slotted flux-concentrating RFPM machine with outer rotor and rare earth magnets (RF-

FC-ST-OR-NdFeB) Fig 3-3-10 depicts the active masses of RFPM machines for different configurations. Fig. 3-3-11 depicts the ratio of active mass to torque rating of different RFPM machines. Considering the active mass characteristics of the RFPM machines addressed in those figures, the following trends are found.


45

• There is no clear difference between slotted surface-mounted RFPM machines with inner rotor concept (R1: RF-SM-ST-IR-NdFeB) and outer rotor concept (R2: RF-SM-ST-OR-NdFeB). Therefore, it cannot be concluded that one concept is lighter than another concept for large torque machines.

• It seems that the slotted flux-concentrating RFPM machines with ferrite magnets (R3: RF-FC-ST-IR-Ferrite) are heavier than the machines with rare earth magnets (R4: RF-FC-ST-IR-NdFeB).

• It is not clear to draw the conclusion that the slotted flux-concentrating RFPM machine with outer rotor (R5: RF-FC-ST-OR-NdFeB) is lighter than the machine with inner rotor (R4: RF-FC-ST-IR-NdFeB) for large torque machines. However, the RFPM machine with R5 concept was addressed as lighter machine than the machines with R1, R2 and R3 concepts at low torque rating, about 1 kNm.

• The RFPM machines with R1 concept and the RFPM machines with R4 concept seem lighter than other RFPM machines. The RFPM machines with R1 concept showed the lightest active mass characteristics.

Chapter 3

46

Torque rating [kNm]

10-2 10-1 100 101 102 103 104

Act

ive

mas

s [k

g]

1

10

100

1000

10000

100000

R1

R1R1

R1R1

R1 R1R1

R1R1 R1

R1

R1

R1

R1

R1

R1R1

R1R1

R1

R1

R1

R1

R1R1R1

R1R1 R1R1R1R1R1

R1R1R1R1

R1R1R1R1

R1R1

R1

R1R1R1R1

R1R1R1

R1R1R1

R1R1R1

R2R2

R2R2

R2R2R2R2

R3R3R3

R3

R4

R4R4

R4

R4R4

R4

R4R4R4R5

RF-SM-ST-IR-NdFeBR1RF-SM-ST-OR-NdFeBR2RF-FC-ST-IR-FerriteR3RF-FC-ST-IR-NdFeBR4RF-FC-ST-OR-NdFeBR5

40 ton / 3820 kNm

Fig. 3-3-10: Active masses of RFPM machines as a function of the torque ratings

Torque rating [kNm]

10-2 10-1 100 101 102 103 104

Act

ive

mas

s / T

orqu

e ra

ting

[kg/

kNm

]

1

10

100

1000

10000

R1

R1

R1R1

R1

R1

R1R1

R1R1

R1R1

R1

R1R1

R1

R1R1

R1R1

R1

R1

R1

R1

R1R1

R1

R1R1

R1R1R1R1R1

R1R1R1R1

R1R1

R1R1

R1 R1R1

R1R1R1R1

R1R1R1R1R1R1 R1R1R1

R2R2

R2R2R2R2R2

R2

R3R3R3

R3

R4

R4

R4

R4

R4R4

R4

R4R4R4R5

RF-SM-ST-IR-NdFeBR1RF-SM-ST-OR-NdFeBR2RF-FC-ST-IR-FerriteR3RF-FC-ST-IR-NdFeBR4RF-FC-ST-OR-NdFeBR5

Fig. 3-3-11: Ratios of active mass to torque of RFPM machines as a function of the torque ratings


47

3.3.2 AFPM machine To determine the potential of axial flux permanent magnet (AFPM) machines for use in direct-drive wind turbines, a survey is made of the machines discussed in the scientific literature. Different AFPM machines have been discussed in a number of references [Söd 1996][Söd 1997][Bum 2005][Oka 2006][Boc 2006][Car 1994][Cha 1999][Eas 2002][Che 2005a][Che 2005b][Par 2005a][Par 2005b][Spo 1992][Wu 1995][Wu 1995][Mue 2005b][Sah 2001][Viz 2005][Viz 2006][Car 1995][Car 1999]. In this thesis the AFPM machines are categorized as follows based on the electromagnetic configurations: • the slotted surface-mounted AFPM machine (Fig. 3-3-12) • the slotless AFPM machine with toroidal stator (TORUS) (Fig. 3-3-17) • the coreless AFPM machine (Fig. 3-3-21) A. Slotted surface-mounted AFPM machine The slotted surface-mounted axial flux permanent magnet (AFPM) machine consists of a stator with slotted iron cores and windings, and a rotor with yoke and surface-mounted magnets as represented in Fig. 3-3-12.

Fig. 3-3-12: A slotted surface-mounted AFPM machine [Dub 2004]

Chapter 3

48

The slotted surface-mounted AFPM machine has been discussed as a better concept for some applications where compactness is required [Sit 2000][Dub 2004]. Sahin described a high speed AFPM machine with slotted, double-stator and internal rotor configurations to combine with a flywheel as shown in Fig. 3-3-13 [Sah 2001].

Fig. 3-3-13: A slotted surface-mounted AFPM machine combined with a flywheel [Sah 2001] The potential of soft magnetic composite (SMC) material was discussed for application in low-speed, direct-drive, axial flux PM wind generators by Chen et al. Comparative design studies were conducted on the PM generators with different configurations with both laminated core and SMC core [Che 2005a]. Fig. 3-3-14 depicts the generator with one stator and two rotors. A slotted surface-mounted AFPM machine with concentrated windings for a small wind turbine was discussed by Parviainen et al. [Par 2005a]. The power rating of the prototype was 1.6 kW at 250 rpm. Parviainen discussed the use of an analytical method to perform the preliminary design of the slotted surface-mounted AFPM machines with one-rotor-two-stators configuration as shown in Fig. 3-3-15. The performance and construction of the low-speed RFPM and AFPM machines were compared in the power range from 10 kW to 500kW at 150-600 rpm [Par 2005b].


49

Fig. 3-3-14: A slotted surface-mounted AFPM machine with one stator and two rotors [Che 2005a]

Fig. 3-3-15: A slotted surface-mounted AFPM machine with one-rotor-two-stators [Par 2005b] Vizireanu et al. discussed a 9-phase, 2.7MW slotted AFPM generator for direct-drive wind turbines [Viz 2005]. The generator consists of two PM outer rotors with 90 poles and one interior 9-phase stator with 81 slots with distributed windings. Vizireanu and Brisset et al. also discussed 5 MW 9-phase concentrated winding AFPM wind generator with 6 different configurations, which are characterized by the combination of the number of poles and the number of slots for one magnetic periodicity. The configurations for a comparative study are 8 poles and 9 slots

Chapter 3

50

(8p9s), 10p9s, 14p9s, 16p18s, 22p18s and 26p18s [Viz 2006][Bri 2008]. Fig. 3-3-16 illustrates the external view of 10p9s prototype. In the references, it was noted that the total active mass of the 8p9s configuration was lower than other configurations.

(a) Half generator (b) Ons stator (c) Half rotor Fig. 3-3-16: A 10 pole, 9 slot, 9-phase AFPM machine prototype [Viz 2006][Bri 2008] B. Slotless AFPM machine with a toroidal stator (TORUS machine) The slotless axial flux permanent magnet (AFPM) machine with a toroidal stator (TORUS machine) consists of a stator with a laminated steel tape (toroidal stator) and wound coils, and the rotor with yoke discs and surface-mounted magnets as shown in Fig. 3-3-17.

Fig. 3-3-17: A slotless AFPM machine with toroidal-stator (TORUS machine) [Boc 2006]


51

The slotless AFPM machine has been discussed as a better type of machine for lightness, compactness, short axial length and suitable integration [Spo 1992][Wu 1995a][Wu 1995b]. Caricchi et al. proposed slotless AFPM machines for direct-drive wind turbine applications [Car 1994b]. Both mechanical and electromagnetic designs of a 100 kW slotless AFPM machine for direct-drive wind turbines were discussed in [Söd 1996][ Söd 1997]. Chalmers et al. also presented a slotless AFPM generator for direct-drive variable speed wind turbine application. The power rating of the generator built is 5 kW at 200 rpm [Cha 1999]. Boccaletti et al. analyzed the algorithms and the relevant numerical optimization techniques for the design of slotless AFPM machines [Boc 2006]. They described some numerical optimization algorithms and their applications to the design process. Objective functions such as mean torque, power losses and total mass were discussed. A slotless AFPM machine with two rotors for ship propulsion was discussed in [Car 1995a]. A new machine topology, having the counter-rotation of two rotors, was proposed in order to recover energy from the rotational flow of the main propeller slip stream, as shown in Fig. 3-3-18. A design of the AFPM machine with power ratings from 1 MW to 20 MW was made. In order to validate the operating principle of the new machine, a down-scaled prototype with 510 W power rating at 195 rpm was built.

Fig. 3-3-18: Disassembled view of a slotless AFPM machine with two rotors: a) main and counter-rotating propellers, b) radial bearing, c) outer shaft, d) PM rotor, e) motor bearing, f) assembly ring, g) machine stator, h) assembly ring, i) motor bearing, l) PM rotor, and m) inner shaft [Car 1995a]

Chapter 3

52

In [Car 1999], a slotless AFPM machine with a multi-stage arrangement was discussed for ship propulsion drives as shown in Fig. 3-3-19. When the available space of the machine is small in diameter, the required torque can be achieved by means of stacking the AFPM machines. The machines discussed in [Car 1999] have the design characteristics of 4-stages with power ratings of 7.6 MW and 14 MW at 195 rpm. The active mass of the machines estimated are 7.3 tonnes and 13.6 tonnes, respectively.

Fig. 3-3-19: A slotless AFPM machine with 4-stages [Car 1999] Mueller et al. have described the features of three different configurations of the slotless AFPM machine, namely a single rotor-single stator arrangement, a double rotor-single stator arrangement and a multistage arrangement. The layouts of these configurations are represented in Fig. 3-3-20. Analytical expressions about the theory of circular plate, elastic beam and cylindrical shell were introduced in the structural design of the machine for direct-drive wind turbines. Mechanical finite element analysis models were used in order to verify the analytical expressions introduced. In the structural analysis of the slotless AFPM machine, they focused on the double rotor with single stator configuration. The power rating of the machine was a range of typical wind turbine ratings from 100 kW at 60 rpm up to 5 MW at 11.5 rpm. The active mass and the structural mass of the machine was analyzed and discussed in [Mue 2005b].


53

Fig. 3-3-20: Different configurations of slotless AFPM machines [Mue 2005b]: i) stator support structure, ii) stator iron, iii) copper winding, iv) permanent magnets, v) rotor iron, vi) rotor disc (external), vii) rotor disc (intermediate), viii) bearing, ix) rotor shaft C. Coreless AFPM machine The coreless axial flux permanent magnet (AFPM) machine consists of a stator with a non-ferromagnetic non-conducting structure and wound coils, and a rotor with yoke discs and surface-mounted magnets as shown in Fig. 3-3-21. The coreless AFPM machine has been discussed as a better type of machine for compactness, lightness and efficiency [Bum 2005][Car 1995b][Eas 2002]. Bumby et al. described the design and development of a coreless AFPM generator for direct-drive small wind turbines. The generator consists of two rotor discs with magnets located around its periphery. The stator is made of non-ferromagnetic non-conducting material and has a number of bobbin-wound armature coils located around its periphery. The generator produces 1 kW at 300 rpm or 2 kW at 500 rpm with an electrical efficiency substantially greater than 90% [Bum 2005]. The active mass of the generator is estimated at about 32 kg. A coreless AFPM machine that consists of multi-stage structure and water-cooled ironless stator was proposed for direct-drive wheel motors [Car 1995b]. In the proposed machine, the toroidal iron core is eliminated, and the stator winding consists of diamond-shaped coils to reduce end-winding losses. Fig. 3-3-22 depicts a cross-section of the two-stage coreless AFPM machine with a diamond-shaped coil. A prototype of a two stage axial flux PM machine rated at 200 Nm and 1100 rpm was designed and built. The mass of the machine was about 26.8 kg.

Chapter 3

54

Fig. 3-3-21: A coreless AFPM machine [Bum 2005]

Fig. 3-3-22: A coreless AFPM machine with two-stage [Car 1995b]


55

D. Active mass of AFPM machines In order to determine the trends of active masses of AFPM machines during the scaling up of the torque rating, active masses of different AFPM machines discussed in the references are addressed as a function of the torque. The AFPM machines are categorized as follows. • A1: Slotted surface-mounted AFPM machine with single-side (AF-SM-SS-ST) • A2: Slotted surface-mounted AFPM machine with double-sides and single-rotor (AF-SM-DS-

ST-SR) • A3: Slotted surface-mounted AFPM machine with double-sides and double-rotors (AF-SM-

DS-ST-DR) • A4: Slotless surface-mounted AFPM machine with double-sides (AF-SM-DS-SL: Torus

machine) • A5: Coreless surface-mounted AFPM machine (AF-SM-CL) Fig. 3-3-23 depicts the active masses of AFPM machines with different configurations. Fig. 3-3-24 depicts the ratio of active mass to torque rating of the different AFPM machines. Considering the active mass characteristics of the AFPM machines addressed in those figures, the following trends are found. • There is no clear difference that would allow the conclusion that one machine type is lighter

than other machine types in torque ratings lower than 10 kNm. • The data designed for the coreless AFPM machines (A5: AF-SM-CL) was not enough to

determine the trends of active masses during the scaling up of the torque rating. • In torque ratings higher than 10 kNm, the slotless surface-mounted AFPM machines with

toroidal-stator (A4: AF-SM-DS-SL) seem lighter than the slotted AFPM machines (A1: AF-SM-SS-ST, A2: AF-SM-DS-ST-SR, A3: AF-SM-DS-ST-DR).

• Among slotted surface-mounted AFPM machines, the double-sided machine type with double-rotors (A3) seems lighter than the other two types of slotted surface-mounted AFPM machines (A1, A2).

Chapter 3

56

Torque rating [kNm]

10-2 10-1 100 101 102 103 104

Act

ive

mas

s [k

g]

1

10

100

1000

10000

100000

A1A1A1

A1

A1A1

A1

A1A1

A1

A1A1A1

A2

A2A2

A2A2A2

A2A2A2

A2A2A2

A2A2A2

A3

A3A3

A3

A3

A4A4

A4

A4

A4

A4

A4

A4

A4

A4A4

A4

A4

A4

A4

A4

A5 A5

AF-SM-SS-STA1AF-SM-DS-ST-SRA2AF-SM-DS-ST-DRA3AF-SM-DS-SL (TORUS)A4AF-SM-CLA5

28 ton / 4152 kNm

Fig. 3-3-23: Active masses of AFPM machines as a function of the torque ratings

Torque rating [kNm]

10-2 10-1 100 101 102 103 104

Act

ive

mas

s / T

orqu

e ra

ting

[kg/

kNm

]

1

10

100

1000

10000

A1A1A1

A1

A1

A1

A1

A1

A1

A1

A1A1A1

A2A2A2

A2A2A2

A2A2A2 A2A2A2

A2A2A2

A3

A3A3A3

A3

A4

A4A4

A4 A4

A4A4

A4

A4

A4A4

A4A4

A4A4

A4

A5

A5

AF-SM-SS-STA1AF-SM-DS-ST-SRA2AF-SM-DS-ST-DRA3AF-SM-DS-SL (TORUS)A4AF-SM-CLA5

Fig. 3-3-24: Ratios of active mass to torque of AFPM machines as a function of the torque ratings


57

3.3.3 TFPM machine Different TFPM machines discussed in the references are surveyed in order to identify the possibility and potential of the machines for large direct-drive wind turbines. The features, configurations, advantages and disadvantages of the TFPM machines have been discussed in a number of references [Weh 1986][Weh 1988a][Weh 1988b][Weh 1991][Har 1993a][Har 1993b][Weh 1994][Mit 1995][Weh 1995][Har 1996][Mec 1996][Har 1997b][Har 1997c][Hen 1997][Hua 1997a][Hua 1997b][Mit 1997][Paj 1997][Har 1998][Mad 1998][Mad 1999][Mul 1999] [Has 1999][Bli 2000][Dub 2000][Has 2000][Lam 2000][Lan 2000] [Voy 2000][Ars 2001][Han 2001][Ars 2002][Dic 2002][Dub 2002][Kas 2002a][Kas 2002b][Pay 2002][Ran 2002][Guo 2003][Hus 2003][Lu 2003a][Lu 2003b][Lu 2003c][Lu 2003d][Nje 2003][Shi 2003][Bao 2004][Dub 2004][Mas 2004][Rah 2004][Bao 2005a][Bao 2005b][Gie 2005][Sch 2005][Wer 2005][Zha 2005][Guo 2006][Sve 2006a][Sve 2006b]. The major difference between TFPM machine and RFPM and AFPM machines is that the TFPM machine allows an increase of the space for the windings without decreasing the available space for the main flux. The TFPM machines can be also made with a very small pole pitch compared with other machines. This feature of the machine results in higher force density than RFPM and AFPM machines. The copper winding of the TFPM machine is simple, and the non-active copper winding is considerably smaller than the copper windings of other machines. Thus, the active mass of the TFPM machine needed to produce the required torque can be smaller than the active mass of other machines. In other words, low values of the active mass to torque ratio can be achieved by the TFPM machine. However, the TFPM machine has disadvantages such as a complicated construction, a low power factor and difficulties in manufacturing. The TFPM machine seems to be a suitable machine type for direct-drive applications because of its high specific torque, even though it has a large number of individual parts and special methods of manufacture and assembly [Han 2001]. Based on the electromagnetic configurations, the TFPM machines can be classified as follows. • surface-mounted PM type or flux-concentrating PM type • single-sided type or double-sided type • single winding type or double winding type • type with stator bridge core, C-core, U-core, E-core, or claw pole core • inner rotor type or outer rotor type In this thesis, TFPM machines are categorized into two types: the surface-mounted TFPM machine type and the flux-concentrating TFPM machine type.

Chapter 3

58

A. Surface-mounted TFPM machine Different configurations of surface-mounted TFPM machines were discussed by a number of authors. In this thesis the surface-mounted TFPM machines are mainly categorized as double-sided and single-sided. Double-sided surface-mounted TFPM machines were discussed in [Weh 1986][Mul 1999][Kas 2002a]. Single-sided surface-mounted TFPM machines were discussed in [Zha 2005][Weh 1988a][Har 1993a][Har 1996][Har 1997b][Dub 2000][Dub 2004][Lu 2003a][Lu 2003b][Lu 2003c][Lu 2003d][Hen 1997][Kas 2002b][Gie 2005][Ars 2002][Dic 2002][Guo 2003][Guo 2006][Ars 2001][Sve 2006a][Sve 2006b][Mad 1999]. Weh et al. discussed the achievable force densities of different PM machines such as a flux-concentrating RFPM machine with laminated claw poles, a surface-mounted RFPM machine with concentrated windings, and a double-sided surface-mounted TFPM machine with double windings (see Fig. 3-3-25(a)) [Weh 1986]. Another double-sided, surface-mounted TFPM machine with reduced flux leakage was also discussed as shown in Fig. 3-3-25(b). It was stated that both current loading and flux density were the determining factors for force density. The current loading of the TFPM machine was higher than seven times the current loading of the RFPM machines with the same current density.

(a) (b) Fig. 3-3-25: Double-sided surface-mounted TFPM machines [Weh 1986] A double-sided TFPM machine combined with modular structure, axial flux configuration and double windings for wind turbine generators was proposed by Muljadi et al. as shown in Fig. 3-3-26 [Mul 1999]. They showed that the ring stator winding could be easily assembled and automatically produced and the winding is exposed to open air, which improves cooling. A


59

single-phase machine with 650 W power rating at 667 rpm was discussed. The machine consists of stator and rotor cores made on a per-pole basis with 18 pole PMs. Torque ripple reduction of a double-sided axial type TFPM machine was discussed by Kastinger et al. [Kas 2002a].

Fig. 3-3-26: A double-sided TFPM machine with axial air gap [Mul 1999] A single-sided surface-mounted TFPM machine with single winding shown in Fig. 3-3-27 was proposed by Weh [Weh 1988a]. An outer rotor type of the machine of Fig. 3-3-27 was discussed by Harris et al. as shown in Fig. 3-3-28 [Har 1993a][Har 1996]. In [Har 1993a] torque densities of a single-sided surface-mounted TFPM machine with outer rotor were compared with the torque density of an induction machine. The TFPM machine of which the torque rating is 69.1 Nm showed the torque density is 4.9 times higher by volume and 7 times higher by mass than the values of the induction machine.

Fig. 3-3-27: A single-sided surface-mounted TFPM machine with single windings [Weh 1988a]

Chapter 3

60

Lu et al. discussed a new torque equation, the estimation of inductance, an analytical model and the power factor of single-sided surface-mounted TFPM machines with single windings and an outer rotor [Lu 2003a][Lu 2003b][Lu 2003c][Lu 2003d].

Fig. 3-3-28: A single-sided surface-mounted TFPM machine with single windings and an outer rotor [Har 1993a][Har 1996] The design, performance analysis, and experimental results of the surface-mounted transverse flux PM machine using soft magnetic composites (SMC) have been discussed in [Guo 2003][Guo 2006]. Fig. 3-3-29 represents the surface-mounted TFPM machine with outer rotor and SMC cores. The measured performance of the SMC TFPM machine was compared with three different motors, namely an induction motor, a radial flux PM DC motor and an SMC claw pole motor [Guo 2006]. The output power rating of the SMC TFPM motor was 640 W at 1800 rpm. The SMC TFPM motor had the highest values of torque to volume ratio compared to the other motors. The torque to volume ratio of the SMC TFPM motor was almost 4.5 times higher than the torque to volume ratio of the induction motor.

Fig. 3-3-29: Outer rotor surface-mounted TFPM machine with SMC cores [Guo 2003][Guo 2006]


61

An analytical approach to dimension and to analyze the performance of the TFPM machine was proposed by Arshad et al. [Ars 2001]. In the reference, they focused on a single-sided surface-mounted TFPM machine with an outer rotor. The design results of the machine were compared with the results of both a double-sided flux-concentrating TFPM machine with single windings [Anp 2001] and a surface-mounted TFPM machine [Paj 1997]. In terms of the ratios of torque to active mass and torque to volume, the single-sided surface-mounted TFPM machine with an outer rotor discussed in [Ars 2001] had higher values than the other two machines. The double-sided flux-concentrating TFPM in [Anp 2001] had the highest power factor, and the second was the surface-mounted TFPM machine with an outer rotor. A novel single-sided surface-mounted TFPM machine with single windings for direct-drive wind generators was proposed and analyzed in [Sve 2006a][Sve 2006b]. The power ratings of the generators are 3, 5, 7 and 10 MW at 18.6 rpm, 13.2 rpm, 10.6 rpm and 8.3 rpm, respectively. Fig. 3-3-30 illustrates the novel TFPM machine that consists of a hollow rotor with the surface-mounted PMs embraced by the laminated stacks with the windings placed in the slots.

(a) (b)

(c)

Fig. 3-3-30: A novel TFPM topology for a direct-drive wind turbine [Sve 2006a][Sve 2006b]

Chapter 3

62

A single-sided surface-mounted TFPM machine with single windings and stator bridges shown in Fig. 3-3-31 was discussed by Henneberger et al. [Hen 1997].

Fig. 3-3-31: Surface-mounted TFPM machine with stator bridges [Hen 1997] Kastinger et al. proposed a new concept of a single-sided TFPM machine with single windings and stator bridges. The stator of the machine consists of laminated U-cores with stator bridges and ring-shaped windings [Kas 2002b]. The rotor of the machine consists of laminated iron cores and an axially magnetized PM ring as shown in Fig. 3-3-32.

Fig. 3-3-32: Single-sided TFPM machine with stator bridges and axially magnetized magnets [Kas 2002b]


63

Gieras discussed the analysis and performance characteristics of a single-sided surface-mounted TFPM machine with single windings, stator bridges and an outer rotor [Gie 2005]. The stator of the machine consists of U-shaped and I-shaped laminated iron cores and ring windings. The rotor consists of a back iron core with surface-mounted PMs. Fig. 3-3-33 represents the machine discussed in [Gie 2005]. The power rating of the machine was 6 kW at 594 rpm.

Fig. 3-3-33: A surface-mounted single-sided TFPM machine with stator bridges [Gie 2005] A single-sided surface-mounted TFPM machine with intermediate poles was proposed by Zweygbergks in 1992. The machine called the Z-machine was discussed and compared with a conventional surface-mounted TFPM machine by Arshad et al. [Ars 2002]. Fig. 3-3-34 illustrates the Z-machine. The output of the Z-machine was twice the output of the conventional surface-mounted TFPM machine with the same volume. Iron losses of the Z-machine were twice as high as the losses of the conventional TFPM machine. It was also stated that the normalised cogging torque of the Z-machine was slightly lower than the other machine. The cogging frequency of the Z-machine would be twice of the cogging frequency of the other machine.

Fig. 3-3-34: Z-machine proposed by Zweygbergks [Ars 2002]

Chapter 3

64

Dickinson et al. have improved the performance of the single-sided surface-mounted TFPM machine with single windings and claw poles [Dic 2002]. To build the claw pole structure of the machine soft magnetic composites (SMC) were used. In order to reduce the leakage flux of the machine, they optimised the shape of claw pole using both finite element analyses and an equivalent reluctance model. Fig. 3-3-35(a) illustrates the main flux path of a claw pole surface-mounted TFPM machine without leakage flux. The claw pole structure before and after optimisation is shown in Fig. 3-3-35(b) and (c) respectively. The average torque of the machine was 25.6 Nm, which corresponds to 9.3 Nm/kg at the thermally limited current.

(a) (b) (c) Fig. 3-3-35: Claw pole surface-mounted TFPM machine [Dic 2002] Power factor analyses of different TFPM machines were discussed by Zhao et al. [Zha 2005]. The machines include the single-sided surface-mounted TFPM machine with single windings and an outer rotor, the single-sided surface-mounted TFPM machine with single windings and stator bridges, and the double-sided flux-concentrating TFPM machine with single windings. They compared the power factor as a function of the pole number, the electrical loading and the electromagnetic torque. They also noted that a high magnetomotive force or a high electrical loading causes a low power factor. B. Flux-concentrating TFPM machine Compared to the surface-mounted TFPM machines, the flux-concentrating TFPM machines have been discussed as a superior machine type regarding their specific torque and power factor. However, the disadvantages of their complicated construction and lower robustness were also discussed.


65

In the thesis the flux-concentrating TFPM machines are also categorized into the double-sided type and the single-sided type. Double-sided flux-concentrating TFPM machines were discussed in [Weh 1988a][Weh 1988b][Weh 1995][Lan 2000][Mec 1996][Mit 1995][Mit 1997][Voy 2000][Hua 1997a][Hua 1997b][Sch 2005][Has 2000][Dub 2000][Dub 2004][Mad 1999][Ran 2002][Mas 2001][Pay 2002][Hus 2003][Bao 2004][Bao 2005a][Bao 2005b][Weh 1995][Has 1999][Har 1997b]. The single-sided flux-concentrating TFPM machines were discussed in [Hen 1997][Dub 2000][Dub 2004][Mad 1999][Nje 2003b][Har 1997b]. A double-sided flux-concentrating TFPM machine with double windings and U-core arrangement shown in Fig. 3-3-36 was proposed by Weh [Weh 1988b][Weh 1995]. The specific sizing and power density (power to volume ratio) equations of the TFPM machine were discussed by Huang et al. [Hua 1997a][Hua 1997b]. Power densities of three different 75 kW machines, namely the 4-pole induction machine, the 16-pole TFPM machine with ferrite magnets, and the 16-pole TFPM machine with rare earth magnets, were compared using the sizing and power density equations. In the comparison between the induction machine and the two TFPM machines, the TFPM machine with rare earth magnets had the highest power density over the whole speed range. The induction machine had the lowest power density. A double-sided flux-concentrating TFPM machine with a single winding and U-core arrangement was proposed in [Weh 1995]. Fig. 3-3-37 illustrates the machine. A cogging torque analysis of a double-sided flux-concentrating TFPM machine was discussed by Schmidt [Sch 2005]. The machine configuration is the same as the machine discussed in [Weh 1995] as shown in Fig. 3-3-38.

Fig. 3-3-36: TFPM machine with double-sided with double winding and U-core arrangement [Weh 1995]

Chapter 3

66

Fig. 3-3-37: A double-sided flux-concentrating TFPM machine with single windings and U-core arrangement [Weh 1995]

Fig. 3-3-38: A double-sided flux-concentrating TFPM machine with single windings and U-core arrangement [Sch 2005] Hasubek et al. discussed an analysis of the design limitations of a double-sided flux-concentrating TFPM machine with a passive rotor and single windings as shown in Fig. 3-3-39


67

[Has 2000]. In this reference, optimum dimensions of pole pitches, PM widths, and rotor pole widths were discussed.

Fig. 3-3-39: A single-sided flux-concentrating TFPM machine with a passive rotor and single windings [Has 2000] A double-sided flux-concentrating TFPM machine with single windings and C-core arrangement was proposed by Weh in [Weh 1995], where the stator core consists of laminated steel as shown in Fig. 3-3-40.

Fig. 3-3-40: A double-sided TFPM machine with single windings and C-core stator made of laminated steel [Weh 1995]

Chapter 3

68

A modified design of the transverse flux machine, which offers a high torque density, was proposed for the electric propulsion of ships by Mitcham et al. in [Mit 1995][Mit 1997] as shown in Fig. 3-3-40. The machine is a type of double-sided flux-concentrating TFPM machine with single windings and C-core arrangement proposed in [Weh 1995]. In [Mit 1997], it was also concluded that the TFPM machine offers a high mechanical integrity, a low noise and vibration.

Fig. 3-3-41: A double-sided flux-concentrating TFPM machine with single windings [Mit 1997] Rang et al. discussed two configurations of TFPM machines, namely a double-sided flux-concentrating TFPM machine with double windings and a double-sided flux-concentrating TFPM machine with single windings and C-core arrangement [Ran 2002]. They also discussed an analytical design approach used for the second configuration, which is the same as the machine type discussed in [Mit 1995][Mit 1997]. Fig. 3-3-41 illustrates the double-sided flux-concentrating TFPM machine with single windings and C-cores discussed by Mitcham et al. A theoretical analysis and some experimental results of a double-sided flux-concentrating TFPM machine with single windings and C-core arrangement were discussed for the ship propulsion by Payne et al. [Pay 2002]. A condition monitoring technique based on the flux analysis of the machine was also discussed in the reference. Husband et al. introduced a demonstrator of a 2 MW TFPM machine rated at 308 rpm for ship propulsion [Hus 2003]. The TFPM machine was a double-sided flux-concentrating type with single windings and C-core stator arrangement as shown in Fig. 3-3-42. The mass of the machine was estimated at 13 tonnes with the size of 1.475 x 1.5 x 1.55 metres.


69

Fig. 3-3-42: A double-sided flux-concentrating type with single windings and C-core stator arrangement [Hus 2003] A double-sided flux-concentrating TFPM machine with single windings and a C-core stator made of the soft magnetic composite (SMC) was proposed by Maddison [Mad 1999]. Fig. 3-3-43 illustrates the TFPM machine discussed in the reference.

Fig. 3-3-43: Double-sided flux-concentrating TFPM machine with single windings and a C-core stator made of SMC [Mad 1999] Masmoudi et al. discussed an analytical method to assess the cogging torque of a double-sided flux-concentrating TFPM machine with single windings [Mas 2001].

Chapter 3

70

Bao et al. discussed cogging torque reduction and power loss reduction of a double-sided TFPM machine with single winding and a C-core stator [Bao 2004][Bao 2005a][Bao 2005b]. Fig. 3-3-44 illustrates the TFPM machine discussed by Bao et al.

Fig. 3-3-44: A a double-sided TFPM machine with single winding and C-core stator [Bao 2004][Bao 2005a][Bao 2005b] Torque ripples, normal force fluctuation, low power factor and complicated construction have been discussed as the disadvantages of TFPM machines compared to conventional (longitudinal flux) machines. In order to reduce the torque ripples and the normal force fluctuation, the optimizations of electromagnetic dimensions and current wave forms can provide some solutions. Wer et al. proposed the cogging torque reduction of the TFPM machine by an optimal current control [Wer 2005]. A flux-concentrating TFPM machine with double windings and an E-core was discussed in [Weh 1995]. The stator of the TFPM machine consists of E-cores with laminated steel and two windings. The rotor consists of two sets of flux-concentrating magnets and iron cores as shown in Fig. 3-3-45.


71

Fig. 3-3-45: A flux-concentrating TFPM machine with an E-core configuration [Weh 1995]

Fig. 3-3-46: A double-sided flux-concentrating TFPM machine with single windings and a passive rotor [Has 1999] Hasubek et al. discussed a double-sided flux-concentrating TFPM machine with single windings and a passive rotor as shown in Fig. 3-3-46. The machine consists of a stator with flux-concentrating magnets, iron cores and windings, and a rotor with iron cores. The force density of the TFPM machine was compared to that of the longitudinal flux machines [Has 1999]. Harris et al. compared the relative advantages and disadvantages of three different topologies of TFPM machines, namely a single-sided surface-mounted TFPM machine with single windings and an outer rotor, a single-sided surface-mounted TFPM machine with single windings and

Chapter 3

72

stator bridges, and a double-sided flux-concentrating TFPM machine with single windings [Har 1997b]. They also discussed the power factor of the transverse flux machine in [Har 1997c]. A single-sided flux-concentrating TFPM machine with single windings and stator bridges was proposed by Henneberger et al. as Fig. 3-3-47 [Hen 1997]. This is a single-sided type of the double-sided flux-concentrating TFPM machine with U-cores shown in Fig. 3-3-37.

Fig. 3-3-47: A single-sided flux-concentrating TFPM machine with single windings and stator bridges [Hen 1997] Dubois discussed different topologies of TFPM machines such as a single-sided surface-mounted TFPM machine with single windings and stator bridges, a double-sided flux-concentrating TFPM machine with double windings, and double-sided flux-concentrating TFPM machines with single windings [Dub 2000][Dub 2004]. A new single-sided flux-concentrating TFPM machine with single windings, stator bridges and a toothed outer rotor was proposed as shown in Fig. 3-3-48 [Dub 2002][Dub 2004]. The active mass of a prototype was 68 kg per phase, and the nominal torque rating was 1 kNm per phase at 1 mm air gap length and 100 rpm. In [Dub 2004], the proposed TFPM machine with toothed rotor was compared with conventional PM synchronous machines based on the cost and the active mass. In the comparison, the proposed TFPM machine showed better cost and mass characteristics for diameters of 0.5 m and 1.0 m. However, the conventional PM synchronous machines showed better cost-competitiveness and mass-competitiveness for diameters larger than 1.0 m. In order to identify the mass-competitiveness of those different PM machines discussed in [Dub 2004], the ratios of active mass to torque rating are addressed as Fig. 3-3-49. In the figure, the conventional PM machines show that the value of the active mass to torque ratio decreases when the torque rating is increasing. The TFPM machine with toothed rotor (T6: TF-FC-SS-SW-SB in Fig. 3-3-49) proposed in [Dub 2004] shows the


73

same characteristics with the conventional (longitudinal flux) machines in low torque ratings. However, the active mass to torque ratio of the proposed TFPM machine with toothed rotor (T6) increases if the torque rating is higher than 5 kNm.

Fig. 3-3-48: TFPM machine with toothed rotor [Dub 2004][Dub 2002]

Torque rating [kNm]

10-1 100 101 102 103

Act

ive

mas

s / T

orqu

e ra

ting

[kg/

kNm

]

10

100

1000

R1

R1

R1

R1

R1

R1

R1

R1R1

R1

R1

R1

R1

R1R1

R1

R4

R4

R4

R4

R4

R4

A1

A1

A1

A1

A1

A1

T6

T6

T6

T6T6T6

T6

T6T6

RF-SM-ST-IR-NdFeBR1RF-FC-ST-IR-NdFeBR4AF-SM-SS-STA1TF-FC-SS-SW-SBT6

Fig. 3-3-49: Active masses of flux-concentrating TFPM machines as a function of the torque ratings

Chapter 3

74

Some authors [Mad 1998][Bli 2000] have discussed the single-sided flux-concentrating TFPM machine with single windings and claw poles. The effect of the stator pole overlap for the TFPM machine was discussed by Maddison et al. in [Mad 1998]. This results in a maximum 6% increase in torque output at 30% overlap of stator pole. Maddison also discussed about different TFPM machines, namely (a) a single-sided surface-mounted TFPM machine with single windings, (b) a single-sided surface-mounted TFPM machine with single windings and stator bridges, (c) a single-sided surface-mounted TFPM machine with single windings and claw poles, (d) a single-sided flux-concentrating TFPM machine with single windings and claw poles, (e) a single-sided flux-concentrating TFPM machine with single windings and stator bridges, and (f) a double-sided flux-concentrating TFPM machine with single windings [Mad 1999]. The single-sided flux-concentrating TFPM machine with single windings and claw poles discussed in [Mad 1999] is illustrated in Fig. 3-3-50.

Fig. 3-3-50: Single-sided flux-concentrating TFPM machine with single windings and claw poles [Mad 1999] Njeh et al. investigated the cogging torque of a single-sided flux-concentrating TFPM machine with single windings and claw poles based on three-dimensional (3D) finite element analyses [Nje 2003a]. They selected this type of TFPM machine because of its simplicity and high power density. A standard converter fed three-phase TFPM machine arranged axially was considered to reduce the drive cost. The rotor of the machine consists of soft magnetic composite (SMC) cores and flux-concentrating permanent magnets. The stator consists of SMC cores with claw poles and ring-shaped windings.


75

The cogging torque reduction of about 25 % by skewing 90 degree-electrical angle of the rotor elements was stated in [Nje 2003b]. An optimisation of a single-sided flux-concentrating TFPM machine with single windings and claw poles was discussed, considering the stator pole overlap as shown in Fig. 3-3-51 in [Mas 2004]. The maximization of the output torque and the minimization of the cogging torque were discussed while considering the stator pole overlap ratios. The maximum torque is produced at an overlap of 30 %.

Fig. 3-3-51: A single-sided flux-concentrating TFPM machine with different overlap of claw poles [Mas 2004] C. Active mass of TFPM machines In order to determine the trends of active masses of TFPM machines during the scaling up of the torque rating, active masses of different TFPM machines discussed in the references are addressed as a function of the torque. The TFPM machines are categorized as follows. • T1: Surface-mounted single-sided TFPM machine with single windings, stator bridges and an

outer rotor (TF-SM-SS-SW-SB-OR) • T2: Surface-mounted single-sided TFPM machine with single windings and claw poles (TF-

SM-SS-SW-CP) • T3: Surface-mounted single-sided TFPM machine with single windings (TF-SM-SS-SW) • T4: Surface-mounted single-sided TFPM machine with single windings and an outer rotor

(TF-SM-SS-SW-OR) • T5: Surface-mounted double-sided TFPM machine with double windings (TF-SM-DS-DW)

Chapter 3

76

• T6: Flux-concentrating single-sided TFPM machine with single windings and stator bridges (TF-FC-SS-SW-SB, TFPM machine with toothed rotor in [Dub 2004])

• T7: Flux-concentrating single-sided TFPM machine with single windings and claw poles (TF-FC-SS-SW-CP)

• T8: Flux-concentrating double-sided TFPM machine with single windings and U-core arrangement (TF-FC-DS-SW-Ucore)

• T9: Flux-concentrating double-sided TFPM machine with single windings and C-core arrangement (TF-FC-DS-SW-Ccore)

• T10: Flux-concentrating double-sided TFPM machine with double windings (TF-FC-DS-DW) Fig. 3-3-52 depicts the active masses of surface-mounted TFPM machines and flux-concentrating TFPM machines as a function of the torque. Fig. 3-3-53 depicts the ratio of active mass to torque rating of different TFPM machines as a function of the torque. Considering the active masses addressed in those figures, the following trends are found. • Flux-concentrating TFPM machines seem lighter than surface-mounted TFPM machines in

torque ratings lower than 10 kNm. • The flux-concentrating single-sided TFPM machine with single windings and claw poles (T7:

TF-FC-SS-SW-CP) seems the lightest, even though the machine was discussed for lower torque ratings than 10 kNm.

• The ratio of active mass to torque of TFPM machines decreases when the torque increases. However, the flux-concentrating single-sided TFPM machines with single windings and stator bridges (T6: TF-FC-SS-SW-SB) discussed in [Dub 2002][Dub 2004] represent different trends with other TFPM machines.

• The surface-mounted single-sided TFPM machine with single windings (T3: TF-SM-SS-SW) discussed in [Sve 2006a][Sve 2006b] seem not enough to determine trends of active mass in low torque ratings.

• In torque ratings lower than 0.2 kNm, the surface-mounted single-sided TFPM machine with single windings and claw poles (T2: TF-SM-SS-SW-CP) seems lighter than the surface-mounted single-sided TFPM machine with single windings, stator bridges and an outer rotor (T1: TF-SM-SS-SW-SB-OR), the surface-mounted single-sided TFPM machine with single windings and an outer rotor (T4: TF-SM-SS-SW-OR) and the surface-mounted double-sided TFPM machine with double windings (T5: TF-SM-DS-DW).


77

Torque rating [kNm]

10-3 10-2 10-1 100 101 102 103 104

Act

ive

mas

s [k

g]

1

10

100

1000

10000

100000

T1 T1

T2

T2

T3

T3T3

T4

T4

T5

T6

T6

T6

T6

T6

T6

T6

T6

T6

T6

T7

T7T7 T7T7

T8

T8

T9

T9T9T10

TF-SM-SS-SW-SB-ORT1TF-SM-SS-SW-CPT2TF-SM-SS-SWT3TF-SM-SS-SW-ORT4TF-SM-DS-DWT5TF-FC-SS-SW-SBT6TF-FC-SS-SW-CPT7TF-FC-DS-SW-UcoreT8TF-FC-DS-SW-CcoreT9TF-FC-DS-DWT10

36 ton / 3617 kNm

Fig. 3-3-52: Active masses of surface-mounted TFPM machines as a function of torque ratings

Torque rating [kNm]

10-3 10-2 10-1 100 101 102 103 104

Act

ive

mas

s / T

orqu

e ra

ting

[kg/

kNm

]

1

10

100

1000

10000

T1

T1T2 T2

T3T3 T3

T4

T4

T5

T6T6

T6T6 T6T6

T6T6 T6

T6T7T7

T7

T7T7

T8

T8

T9T9T9

T10

TF-SM-SS-SW-SB-ORT1TF-SM-SS-SW-CPT2TF-SM-SS-SWT3TF-SM-SS-SW-ORT4TF-SM-DS-DWT5TF-FC-SS-SW-SBT6TF-FC-SS-SW-CPT7TF-FC-DS-SW-UcoreT8TF-FC-DS-SW-CcoreT9

TF-FC-DS-DWT10

Fig. 3-3-53: Ratios of active mass to torque of TFPM machines as a function of torque ratings

Chapter 3

78

3.4 Conclusions The topic covered in this chapter was to identify the active mass-competitiveness of permanent magnet (PM) machines for large direct-drive wind turbines. In section 3-2, the active mass and torque rating were chosen as the main parameters to assess different PM machines. To identify the active mass-competitiveness of PM machines, the ratios of the mass to torque rating of PM machines were investigated as a function of the torque rating. Section 3.3 gave an overview of radial flux PM (RFPM) machines, axial flux PM (AFPM) machines and transverse flux PM (TFPM) machines. Based on the electromagnetic configuration, the RFPM machines were classified into small groups of the slotted surface-mounted RFPM machine, the slotted flux-concentrating RFPM machine, the slotless RFPM machine and the ironless RFPM machine. The AFPM machines were categorized as the slotted surface-mounted AFPM machine, the slotless surface-mounted AFPM machine with toroidal-stator (TORUS machine) and the coreless AFPM machine. The TFPM machines were categorized into two types: the surface-mounted TFPM machine and the flux-concentrating TFPM machine. From the overview of different PM machines, the advantages and disadvantages of those machines are listed in sub-section 3.4.1. The active mass-competitiveness of PM machines is discussed in subsection 3.4.2.

3.4.1 Advantages and disadvantages of PM machines The following advantages of the radial flux (RF) machines can be listed. • Structural stability • Easy production (compared to the slotted axial flux machine and transverse flux machine) Because of these advantages, the RF machines have mostly been used for the low-speed megawatt wind generators. The following advantages of the axial flux permanent magnet (AFPM) machines compared to the radial flux permanent magnet (RFPM) machines have been discussed in the references. • Simple winding (in a slotless machine) • Low cogging torque and noise (in a slotless machine) • Short axial length • Higher torque/volume ratio However, the AFPM machines have the following disadvantages compared to the RFPM machines. • Lower torque/mass ratio • Larger outer diameter and large mass of PM (in a slotless machine)


79

• Structural instability • Difficulty in maintaining the air gap in large diameter (in a slotted machine) • Difficult production of the stator core (in a slotted machine) • Non-uniform flux density distribution in teeth due to non-uniform cross-section of teeth (in a

slotted machine) In order to apply AFPM machines in large direct-drive wind generators, these disadvantages must be overcome since they results in high costs and difficult manufacturing. The main advantages of transverse flux permanent magnet (TFPM) machines, compared to the longitudinal flux machines (RFPM and AFPM machines), can be summarized as follows. • Higher force density • Considerably lower copper losses • Simple winding TFPM machines have the following disadvantages compared to RFPM and AFPM machines. • Complicated construction • Low power factor • Low cost-advantage and low mass-competitiveness in higher torque ratings

3.4.2 Active mass-competitiveness of PM machines The ratios of the mass to torque rating of PM machines discussed in this section are represented in Fig. 3-4-1 as a function of the torque ratings. In the figure PM machines are categorized into the RFPM machine (“R” in the figure), the AFPM machine (“A” in the figure) and the TFPM machine (“T” in the figure). The three lines represent the linear regressions of those PM machines. From the comparison of the ratios of active mass to torque rating (active mass/torque ratio) of different PM machines, the following characteristics are drawn. • TFPM machines seem lighter than RFPM machines in low torque ratings (lower than 50

kNm). • RFPM machines show the lightest characteristics in large torque ratings (larger than 100

kNm). • According to the review of the active mass and the torque rating of different PM machines, it

is concluded that the active mass/torque ratios of those machines decrease when the torque rating increases. However, the TFPM machine with toothed rotor, (T6: TF-FC-SS-SW-SB) proposed in [Dub 2004], showed different characteristics of the active mass/torque ratio. The value of the ratio in [Dub 2004] was increasing when the torque rating is larger than 5 kNm.

• Slotted flux-concentrating RFPM machines with inner rotor and rare earth magnets (R4: RF-FC-ST-IR-NdFeB) and slotted surface-mounted RFPM machines with inner rotor and rare

Chapter 3

80

earth magnets (R1: RF-SM-ST-IR-NdFeB) seem lighter than other RFPM machines. The R1 (RF-SM-ST-IR-NdFeB) machine is the lightest machine among RFPM machines.

• Slotless surface-mounted AFPM machine with double-sides (A4: AF-SM-DS-SL, TORUS machine) seem lighter than other AFPM machines in high torque ratings. However, the Torus machine (A4) requires large outer diameter, large amount of PM material. This results in the increase in cost. Therefore, the A4 machine is not considered as a machine configuration for the design of large direct-drive PM wind generators in the next chapters.

• Flux-concentrating single-sided TFPM machine with single windings and claw poles (T7: TF-FC-SS-SW-CP) seems the lightest machine in low torque ratings (lower than 10 kNm). However, not enough references could be found to determine the trends of active masses of the T7 machine in scaling up the torque rating.

• The active mass of the flux-concentrating TFPM machines seems smaller than the mass of other machines. However, the designs and studies of the TFPM machines with high torque ratings were rarely discussed in references. Therefore, it is necessary to further investigate the mass characteristics of the flux-concentrating TFPM machines with high torque ratings.

Fig. 3-4-1: Ratios of active mass to torque of PM machines as a function of the torque ratings


81

3.4.3 Selection of PM machine configurations for large direct-drive wind generators From the overview of different PM machines and the comparison of the active mass/torque ratio of those machines, the following PM machine configurations are selected for the design of large direct-drive PM wind generators in the next chapters. • RFPM machine: Slotted surface-mounted RFPM machine with inner rotor and rare earth

magnets (R1: RF-SM-ST-IR-NdFeB) is selected among different RFPM machines. • AFPM machine: AFPM machines discussed in this chapter are not considered for the design

in the next chapters. • TFPM machine: Flux-concentrating TFPM machines with single windings are selected.

Chapter 3

82

Mechanical Structure of Direct-Drive Wind Generators

83

Chapter 4


4.1 Introduction Scaling up the power of direct-drive wind generators, the structural part of the generators becomes a dominant part of the total mass. Therefore, the active mass minimization of the generators is not sufficient to make the direct-drive generators more attractive for large wind turbines. This chapter thus identifies the total mass-competitiveness of direct-drive generators for large wind turbines. In order to identify the total mass-competitiveness of direct-drive generators compared to geared generators, this chapter starts with a short review of different wind generators discussed in the references. In the review, the mass, size and torque rating of the generators are addressed. The power ratings of the generators are from 1.5 MW to 5 MW [McD 2006][ENE 2006][Eng 2007][STX 2010]. Next, the total mass of geared and direct-drive generators discussed in the references is addressed as a function of the torque rating. The ratio of the total mass per torque rating ( m T ) is inducted as a criterion to evaluate the total mass-competitiveness of geared and direct-drive generators. To estimate the total mass of the different generators in scaling up the wind turbines up to 20 MW, the values of the m T ratios of the different generators are assumed to be kept constant.

4.2 Large direct-drive wind generators The rotor of a direct-drive generator for a wind turbine is directly connected to the rotor blades. Thus, the direct-drive generator operates at low speed. When the wind turbine is scaled up, the rotational speed is decreased further because the tip speed is kept constant. In order to scale up the power of the direct-drive generator P, the increase in torque T must be thus inversely proportional to the decrease of the mechanical angular speed ωm as given in (4-2-1).

Chapter 4

84

mP Tω= (4-2-1) The torque of the generator T is proportional to the tangential force F and the air gap diameter Dg. The tangential force F can be defined as the product of the tangential force density Fd and the air gap area Ag. Thus, the generator power P can be also defined as (4-2-2).

2

2 d g s mP F D lπ ω= (4-2-2)

where ls is the axial length of the generator. The torque is proportional to the air gap diameter squared, thus the direct-drive generator is generally built with a larger diameter to produce higher torque. Larger air gap diameter results in an increase of materials to construct the generator against the normal stress due to the flux density between the rotor and stator. Therefore, the direct-drive generator that operates at low speed has a high torque rating, a large diameter, a large mass and a high cost. The most common mechanical structure for electric machines supports the rotor shaft simply between two bearings housed in the stator frame as illustrated in Fig. 4-2-1 [Tav 2006].

Fig. 4-2-1: The traditional mechanical structure of electric machines [Tav 2006] Direct-drive wind generators have meanwhile used different mechanical structures as Fig. 4-2-2. An inner rotor with a double bearing concept, the conventional concept for wind turbines, is illustrated in Fig. 4-2-2(a). Fig. 4-2-2(b) depicts a concept with an inner rotor with a single bearing, which is often used to save on bearing costs. A concept with an outer rotor with a double bearing is shown in Fig. 4-2-2(c).


85

(a) Inner rotor with double

bearings (b) Inner rotor with single

bearing (c) Outer rotor with double

bearing Fig. 4-2-2: Different mechanical structures of direct-drive wind generators

A. Conventional structure Traditionally the rotor of generator is connected to a shaft mounted on bearings that enable the rotation in the stator as shown in Fig. 4-2-1. The structure of Fig. 4-2-2(a) is widely used on the wind turbine market by Enercon GmbH, whose world market share was about 10 % in 2008. Fig. 4-2-3 depicts the generator structure of Enercon’s E-112 model. The generator is an electrically excited direct-drive synchronous generator, whose total mass and diameter are about 220 tonnes and 12 m, respectively [Eng 2007].

Fig. 4-2-3: Structure of 4.5 MW electrically excited direct-drive synchronous generator. Source: Enercon GmbH [ENE 2006]

Chapter 4

86

In [McD 2006] the mass minimization of the permanent magnet excited direct-drive generator with traditional mechanical structure was discussed for different power ratings: 2, 3 and 5 MW. In order to minimize the total mass of the generator, the ratio of axial length to air gap diameter Krad has been optimized. Fig. 4-2-4 depicts the structure of the rotor and stator discussed in [McD 2006]. Fig. 4-2-5 depicts the total mass of the generator as a function of the ratio Krad. The Krad of 2, 3 and 5 MW generators chosen as the optimum value are 0.2, 0.22, and 0.27 respectively. According to [McD 2006], it is also shown that the structural mass of large direct-drive wind generator is dominant in the total mass when scaling up the power rating.

Fig. 4-2-4: Structure of the rotor and stator for structural optimization [McD 2006]

Fig. 4-2-5: Total mass of 2, 3 and 5 MW PMSG DD as a function of the ratio, Krad [McD 2006]


87

B. Lightweight structure A single-bearing structure of Fig. 4-2-2(b) has been used by Zephyros B.V., currently STX Windpower B.V. [STX 2010]. A cone-shaped hollow structure with a single bearing is used for the generator instead of a traditional main shaft with two bearings as illustrated in Fig. 4-2-6. The diameter of the generator is smaller than the diameter of the conventional electrically excited synchronous generator. In this structure, the mechanical load path seems shorter than the traditional structure with a main shaft. Therefore, the structural mass of this structure can be smaller than the mass of the traditional structure.

Fig. 4-2-6: Structure of 1.5 MW permanent magnet direct-drive synchronous generator with a single bearing [STX 2010] A new direct-drive generator for wind turbines has been proposed in [Eng 2007]. The fundamental idea of the machine - the NewGen (see Fig. 4-2-7) is to reduce the stiffness demand by removing the load path from the rotor, the shaft and the stator by putting the bearings close to the air gap.

(a) (b) Fig. 4-2-7: New-Gen generator [Eng 2007]

Chapter 4

88

C. Large direct-drive wind turbines Large direct-drive wind turbines of different manufacturers are listed in Table 4-2-1. According to the table, the following trends can be observed. • Large direct-drive wind turbines use radial flux machines more than other types of machines. • The axial flux machine has not been used at power levels above 1 MW. • The transverse flux machine has not been used for large direct-drive wind generators. Table 4-2-1: Large direct-drive wind turbines of different manufacturers

Generator Type Power / Speed Manufacturer

EESG (RF) 4.5 MW / 13 rpm Enercon

PMSG (RF) 3.5 MW / 19 rpm Scanwind

PMSG (RF) 2.5 MW / 14.5(16) rpm Vensys

PMSG (RF) 2 MW / 18.5 rpm STX Windpower

PMSG (RF) 2 MW / 15.8 rpm EWT (Emergya Wind Technologies)

PMSG (RF) 2 MW / 19 rpm JSW (Japan Steel Works)

EESG (RF) 1.65 MW / 20 rpm MTorres

PMSG (RF) 1.5 MW / 23 rpm Leitwind

PMSG (RF) 1.5 MW / 19 rpm Goldwind

PMSG (AF) 0.75 MW / 25 rpm Jeumont (not available)

D. Total mass of different large direct-drive wind generators Different direct-drive generators discussed in the previous sub-section are compared based on the mass as a function of the torque rating. Table 4-3-1 gives the parameters and the generator masses of 1.5 MW Zephyros, 4 MW NewGen and 4.5 MW Enercon wind turbines, respectively. The design parameters and results of the 2, 3 and 5 MW direct-drive permanent magnet generators in [McD 2006] are given in Table 4-3-2.


89

Table 4-3-1: Parameters and total masses of 1.5, 4 and 4.5 MW direct-drive generators

Power 1.5 MW

(Zephyros) 4 MW

(NewGen) 4.5 MW

(Enercon)

Generator type PMSG PMSG EESG

Rotor speed [rpm] 18 19 13

Torque rating [kNm] 862 2010 3306

Diameter [m] 4 9 12

Total mass [ton] 40 36.9 220

Mass/torque [kg/kNm] 46.4 18.4 66.5

Remarks Market available 140 kW prototype Market available

Table 4-3-2: Parameters and Masses of 2, 3 and 5 MW PMSG DD

Rated power 2 MW 3 MW 5 MW

Rotor speed [rpm] 19.5 16 12.5

Torque rating [kNm] 979 1790 3820

Air gap diameter [m] 4.3 5.1 6.1

Krad [-] 0.2 0.22 0.27

Active mass [ton] 14.6 22.4 39.9

Inactive mass [ton] 10.4 19.6 50.1

Total mass [ton] 25 42 90

Mass/torque [kg/kNm] 25.53 23.46 23.56

Air gap [mm] Air gap diameter / 1000

The generator masses discussed above are illustrated in Fig. 4-3-1 as a function of the torque rating. The ratios of total mass per torque rating mtot/T for the 1.5 MW (Zhphyros) and the 4.5 MW (Enercon) generators are higher than the ratios of theoretically optimized 2, 3 and 5 MW generators in [McD 2006]. In the practical design, the total masses of the generators will be larger than the theoretical design because detailed parts for manufacturing were not included in the theoretical design.

Chapter 4

90

The NewGen concept seems to be the lightest direct-drive generator concept among different direct-drive generators currently existing. The total mass of the NewGen generator seems to be competitive with a doubly-fed induction generator with a three-stage gearbox (DFIG 3G).

Torque [kNm]

979 1790 3820

Mas

s [to

n]

10

100

Zephyros 1.5 MW(m / T = 46.4 kg/kNm)

Enercon 4.5 MW(m / T = 66.5 kg/kNm)

2 MW

3 MW

5 MW

DFIG 3G 4 MW(m / T = 17.4 kg/kNm)

NewGen 4 MW(m / T = 18.4 kg/kNm)

minactive_RFPMG

mactive_RFPMG

mtotal_RFPMG

Fig. 4-3-1: Mass comparison of different generator concepts The structural masses of the slotted surface-mounted radial flux permanent magnet (RFPM) machines and the slotless axial flux permanent magnet (AFPM) machines (TORUS machines) discussed in the included references are illustrated in Fig. 4-3-2. The ratio of structural mass per torque rating of the PM machines is illustrated in Fig. 4-3-3. Considering the structural mass characteristics of those PM machines, the following characteristics are drawn. • Slotted surface-mounted RFPM machines (R1 machine in chapter 3, about 50 tonnes at 5

MW) seem lighter than slotless AFPM machines (A4 machine in chapter 3, about 240 tonnes at 5 MW) in structural mass.

• The values of structural mass per torque ratios of the slotless AFPM machines (A4) seem to remain constant as a function of the torque rating.

• The values of structural mass per torque ratios of surface-mounted RFPM machines (R1) seem to increase as a function of the torque rating. However, not enough references were found that showed stuructral mass and torque characteristics. Therefore, it is necessary to further investigate the mass characteristics of the machine.


91

Torque rating [kNm]

10 100 1000 10000

Stru

ctur

al m

ass

[kg]

100

1000

10000

100000

1000000

Surface mounted RFPM machines [McD 2006]Slotless AFPM machines (TORUS)

Fig. 4-3-2: Structural masses of surface-mounted RFPM machines (R1) and slotless AFPM machines (A4) as a function of torque ratings

Torque rating [kNm]

10 100 1000 10000

Stru

ctur

al m

ass

/ Tor

que

ratin

g [k

g/kN

m]

1

10

100

1000

Surface mounted RFPM machines [McD 2006]Slotless AFPM machines (TORUS)

Fig. 4-3-3: Ratios of structural mass to torque rating of surface-mounted RFPM machines (R1) and slotless AFPM machines (A4) as a function of torque ratings

Chapter 4

92

Total masses of different direct-drive machines discussed in the references are illustrated in Fig. 4-3-4 as a function of the torque rating. Fig. 4-3-4 also includes total masses of both a flux-concentrating double-sided TFPM machine with single windings and C-core arrangement (T9: TF-FC-DS-SW-Ccore in chapter 3 [Weh 1995] and a high temperature superconducting (HTS) machine estimated by NREL and AMSC in the US [Com 2009]. Fig. 4-3-5 depicts the ratio of total mass to torque rating of different direct-drive machines. According to the total mass comparison in these figures, the following trends can be expected. • Slotless surface-mounted AFPM machines with toroidal stator (A4: TORUS machines in

chapter 3) and an electrically excited synchronous generator (EESG) with traditional mechanical structure seem heavier than other direct-drive machines.

• The flux-concentrating double-sided TFPM machines with C-core arrangements (T9) seem lighter than the TORUS machines. However, there is not enough data for higher torque ratings to allow the conclusion that the TFPM machine is lighter than other PM machines.

• The high temperature superconducting generator (HTSG) seems the lightest generator concept.

• Considering the ratios of total mass to torque rating of different direct-drive machines, the mass-competitiveness of those machines can be ordered as: 1st(Best)-HTSG, 2nd-R1 machine with air gap bearings (NewGen concept), 3rd-R1 machine by theoretical design, 4th-T9 machine, 5th-R1 machine with a single-bearing concept, 6th-A4 (TORUS) machine & EESG with a traditional mechanical structure.


93

Torque rating [kNm]

0.001 0.01 0.1 1 10 100 1000 10000

Tota

l mas

s [k

g]

1

10

100

1000

10000

100000

1000000

PMSG (theoretical design)Slotless AFPM machines (TORUS)FCTFPMG with C-coreHTSG (estimation by NREL)PMSG (Zephyros)PMSG (NewGen)EESG (Enercon)

Fig. 4-3-4: Total mass of large direct-drive generators for wind turbines as a function of torque rating

Torque rating [kNm]

0.001 0.01 0.1 1 10 100 1000 10000

Tota

l mas

s / T

orqu

e ra

ting

[kg/

kNm

]

10

100

1000

PMSG (theoretical design)Slotless AFPM machines (TORUS)FCTFPMG with C-coreHTSG (estimation by NREL)PMSG (Zephyros)PMSG (NewGen)EESG (Enercon)

Fig. 4-3-5: Total mass/Torque rating of large direct-drive generators for wind turbines as a function of torque rating

Chapter 4

94

4.3 Total mass estimation of geared and direct-drive generators for wind turbines

A. Geared generator In order to identify the interrelationship between the mass and torque of the gearboxes, the mass of the gearboxes manufactured by a Germany based company, Bosch Rexroth AG [Bos 2007] is investigated as a function of the torque. The torque of those gearboxes is between 230 kNm and 2070 kNm, which are for 660 kW and 3300 kW wind turbines. Fig. 4-4-1 depicts that the mass to torque ratio m T of the gearboxes decreases roughly from 14 kg/kNm to 12.5 kg/kNm when the torque increases. In Fig. 4-4-1, the mass of the generator was not included, so that the total mass of geared generators must be larger than the mass of the gearboxes. In this section, the total mass of the doubly-fed induction generator with a three-stage gearbox (DFIG 3G) described in [Eng 2007] is taken to define the ratio of total mass to torque m T of geared generators. For the estimation of the total mass of geared generators in scaling up, the value of m T is assumed as constant with the following value.

A geared generator concept (DFIG 3G) with m/T=17.4 kg/kNm

0

5000

10000

15000

20000

25000

0 500 1000 1500 2000 2500

Torque [kNm]

Mas

s [k

g]

Fig. 4-4-1: Mass of gearboxes manufactured by Bosch Rexroth AG. Data source: [Bos 2007]


95

B. Mass estimation of geared and direct-drive generators The total mass of different direct-drive generators discussed in the previous section is estimated as a function of the torque. To identify the total mass-competitiveness of those direct-drive generators, the mass of the doubly-fed induction generator with a three-stage gearbox (DFIG 3G) is taken as a reference. The direct-drive generator concepts used to estimate the total mass are described as follows. • New-Gen direct-drive generator concept with m/T=18.4 kg/kNm • A traditional direct-drive PM generator concept with m/T=25 kg/kNm • Zephyros direct-drive PM generator concept with m/T=46.4 kg/kNm • Enercon direct-drive EE generator concept with m/T=66.5 kg/kNm Fig. 4-4-2 represents the total mass of the four different direct-drive generators and a DFIG 3G as a function of power rating.

Fig. 4-4-2: Total mass estimation of different generators for wind turbines

Chapter 4

96

4.4 Conclusions This chapter identified the structural mass-competitiveness and the total mass-competitiveness of direct-drive wind generators. For the identification, an overview of different generators discussed in the scientific literature was given based on the structural mass, the total mass and the torque rating in section 4.2. The following generators were included in the overview. • Slotted surface-mounted radial flux permanent magnet (RFPM) machines (R1: RF-SM-ST-

IR-NdFeB in chapter 3) • Slotless surface-mounted axial flux permanent magnet (AFPM) machines with toroidal-stator

(A4: TORUS machines in chapter 3) • Electrically excited synchronous generator (EESG) • Flux-concentrating double-sided TFPM machine with C-core arrangements (T9: TF-FC-DS-

SW-Ccore in chapter 3) • High temperature superconducting generator (HTSG) From the overview, it is concluded that: • Scaling up the power ratings of direct-drive wind generators, the structural part of the

generators becomes dominant in the total mass. • The structural mass of the R1 machine was smaller than the structural mass of the A4

machines. • The A4 machine and EESG with traditional mechanical structure were heavier than other

direct-drive machines in the total mass. • The total mass of the T9 machne was smaller than the total mass of the A4 machines.

However, the data designed for the machine was not enough to determine the trends of the mass in higher torque ratings. Therefore, it is needed to investigate the mass-competitiveness of T9 machines in high torque ratings. The active mass-competitiveness of the T9 machine will be compared with different PM machines in chatper 6.

• Total mass of six different direct-drive machines were investigated. The total mass-competitiveness of the six machines are ordered as: 1st(Best)-HTSG, 2nd-R1 machine with air gap bearings (the NewGen concept), 3rd-R1 machine by theoretical design, 4th-T9 machine, 5th-R1 machine with a single-bearing concept (the Zephyros concept), 6th-A4 machine and EESG with a traditional mechanical structure.

• The HTSG requires a special material and a special cooling system, thus the electromagnetic design and the mass estimations of the HTSG seem not enough to idenfify its own


97

competitiveness. Therefore, the HTSG is not considered for large direct-drive wind turbines in the thesis, even though the HTSG was addressed as the lightweight solution.

In section 4.3, the total mass of currently used direct-drive and geared wind generators was estimated as a function of torque rating to evaluate the total mass-competitiveness of those generators. For the estimation, the total mass/torque ratio of each generator concept was assumed to be constant in scaling up the power rating of wind turbines. The doubly-fed induction generator with a three-stage gearbox (DFIG 3G) was taken as a reference generator. The total mass of four different direct-drive wind generators currently used was estimated up to 20 MW power ratings. From the total mass estimation of different wind generators, the following results are obtained: • The direct-drive electrically excited synchronous generator (EESG DD, Enercon concept)

was addressed as the heaviest generator. • The R1 machine with air gap bearings (the NewGen concept) was constructed without a main

shaft, torque arms and mechanical structure to maintain the air gap. Thus the machine with NewGen concept was addressed as the lightest generator among direct-drive generators and was comparable to the DFIG 3G in terms of the total mass-competitiveness. The total mass/torque ratio of the NewGen direct-drive generator (18.4 kg/kNm) was 6 % larger than the ratio of the DFIG 3G (17.4 kg/kNm). However, the machine of NewGen concept requires high accuracy and allows for small tolerances which increase the cost of the machine.

From the above results, it is concluded that: • To significantly reduce structural mass of a direct-drive generator for large wind turbines, a

generator construction without a shaft and torque arms would be a solution. • To strengthen the cost-competitiveness of the generator without a shaft and torque arms, low

accuracy and large tolerances must be acceptable for the generator in constructing.

Chapter 4

98

Modeling of PM Machines

99

Chapter 5


5.1 Introduction The aim of this chapter is to create a generalized analytical model for various topologies of permanent magnet (PM) machines. In the review of PM machines in Chapter 3, it was discussed that surface-mounted radial flux permanent magnet (RFPM) machines have been mostly used for direct-drive applications. The flux-concentrating transverse flux permanent magnet (TFPM) machine has been discussed as a machine with high force density which results in reduction of volume and mass. Therefore, surface-mounted RFPM and flux-concentrating TFPM machines are chosen in the modelling of PM machines for direct-drive applications in this chapter. A number of TFPM machine topologies have been proposed and discussed with a derivation of analytical models as discussed in Chapter 3. However, it seems rather difficult to apply those analytical models for various topologies of TFPM machines because the models are limited to specified topologies. Therefore, a generalized analytical model for various topologies of PM machines is discussed in this chapter with the following outline. Firstly, a determination of geometric parameters of both a surface-mounted RFPM machine and a flux-concentrating TFPM machine is discussed. Secondly, generalized electromagnetic circuit analysis models of both machines are developed. The models include nonlinear B-H curve characteristics of iron cores. Next, the proposed analysis model is verified through a comparison with finite element analysis and experimental results.

5.2 Main dimensions of PM machines This section concentrates on determining the main dimensions of both a surface-mounted radial flux permanent magnet (RFPM) machine and a flux-concentrating transverse flux permanent

Chapter 5

100

magnet (TFPM) machine. The main dimensions determined in the section will be used for the analytical design of PM generators for wind turbines in the next chapters.

5.2.1 RFPM machine Fig. 5-2-1 depicts a linearized cross-section of two poles of a surface-mounted radial flux permanent magnet (RFPM) machine with full pitch windings. In the figure, gl is the air gap

length, sτ is the slot pitch, sb is the stator slot width, tb is the stator tooth width, syh is the stator

yoke height, sh is the stator slot height, ml is the magnet length, ryh is the rotor yoke height, pb

is the magnet width and pτ is the pole pitch.

sb

mlpb ryh

syh

sh

tb

gl

pτ

sτ

Fig. 5-2-1: A linearized cross-section of a surface-mounted RFPM machine In [Gra 1996][Cat 2001][Pol 2006][McD 2006], the following dimensional parameters were discussed for direct-drive electric machines. • Air gap length gl of direct-drive wind generator was defined as a function of the air-gap

diameter gD , 1000g gl D= in [Gra 1996].

• Slot pitch sτ of electric machine was discussed to choose between 19 mm and 38 mm in [Cat 2001].

• Stator slot width sb , tooth width tb , magnet width pb and slot pitch sτ were defined as a

function of the pole pitch pτ in [Pol 2006][Cat 2001].

• The stator slot height sh was discussed to choose between 3 and 7 times of the slot width sb .

• Considering the flux paths and the continuity of flux, the stator yoke height syh and the rotor

yoke height ryh need to be larger than 1.5 times the tooth width tb .


101

• In order to minimize total mass of direct-drive RFPM generators for 2, 3 and 5 MW wind turbines, the optimum ratios of the axial length to the air-gap diameter rad s gK l D= were

chosen as 0.2, 0.23 and 0.27, respectively. [McD 2006] Considering the above defined parameters, the dimensions and parameters of the RFPM machine are determined in Table 5-2-1. Table 5-2-1: Parameters and dimensions of RFPM machines

Aspect ratio of generator g

srad D

lK = [-]

Force density 40dF = [kN/m2]

Air gap diameter (Generator rotor diameter) 32

grad d

TDK Fπ

= [m]

Axial length of generator 2

2 gennoms

g d m

Pl

D Fπ ω= [m]

Air gap length 1000Dl g

g = [m]

Magnet height 2.5m gl l= [m]

Stator diameter 2s g gD D l= + [m]

Number of phases 3m = [-] Stator slot pitch 0.033sτ = [m]

Number of slots per pole per phase 1q = [-]

Pole pitch p smqτ τ= [m]

Number of pole pairs p

g2

Dp τπ

= [-]

Rotor pole width 0.8p pb τ= [m]

Stator slot width per slot pitch 0.45s

s

bτ = [-]

Stator slot width ss s

s

bb ττ= × [m]

Stator tooth width t s sb bτ= − [m]

Stator slot height 5.3s sh b= [m]

Stator yoke height 2.22sy th b= [m]

Rotor yoke height ry syh h= [m]

Air gap area g g sA D lπ= [m2]

Chapter 5

102

5.2.2 TFPM machines Flux-concentrating double-sided transverse flux permanent magnet (TFPM) machines with single windings and C-core arrangement were discussed in chapter 3 as the machines which offer a high torque density. Therefore, a flux-concentrating double-sided TFPM machine is chosen in the modelling of a TFPM machine. Fig. 5-2-2 depicts a linearized model and dimensional parameters of a double-sided single winding flux-concentrating TFPM machine with C-cores.

pτ Fig. 5-2-2: A linearized double-sided single winding flux-concentrating TFPM machine with C-cores In Fig. 5-2-2, gl is the air gap length, sb is the stator slot width, spl is the stator pole length, syh is

the stator yoke height, sh is the stator slot height, ml is the magnet length, rh is the rotor height,

spb is the stator pole width, rpb is the rotor pole width and pτ is the pole pitch.

In order to determine the optimum electromagnetic dimensions and parameters of TFPM machines, two-dimensional (2D) static analyses of electromagnetic fields are done using finite element analysis (FEA) software, the Flux2D. The static analysis model of electromagnetic fields is built as Fig. 5-2-3. Dirichlet boundary condition with zero vector potential (A=0) is applied on the outer border lines of the model.


103

Stator core

magnetic shield

Rotor core

Stator coremagnetic shield

lm

pτ brp

lg

bsp

p2τ

bs -2lg

Winding

6r 10μ =

PM

Fig. 5-2-3: Two-dimensional static analysis model of electromagnetic fields of TFPM machines The following processes and conditions are used to find optimum electromagnetic dimensions and parameters of TFPM machines by FEA. (1) The remanent flux density and the relative recoil permeability of permanent magnets used in

the model are rmB =1.2 [T] and rmμ =1.05 [-], respectively. (2) The flux density and the magnetic field intensity curve of the iron core used in the analysis

model are given in Fig. 5-2-4. The iron core is 35PN380, an electrical steel model produced by POSCO in Korea.

(3) The force density dF of the machine is defined as an objective function to maximize under the electromagnetic dimensions and parameters given.

(4) In order to determine the electromagnetic dimensions of the TFPM machine achieving the maximum force density dF , the magnet length ml , the stator pole width spb and the pole pitch

pτ are used as variables in the FEA.

(5) The analyses to find the optimum ratios of the magnet length to the pole pitch ( m pl τ ) and

the stator pole width to the pole pitch ( sp pb τ ) are done. In the analyses, the magneto-motive

Chapter 5

104

force mmf by current, the air-gap length gl and the pole pitch pτ are fixed to 3000 [AT], 1

[mm] and 20 [mm], respectively. Force densities dF are calculated as a function of the

magnet length ml and the stator pole width spb . After investigating dF as a function of ml and

spb , the optimum ratios of m pl τ and sp pb τ are calculated. Fig. 5-2-5 depicts the force

density dF investigated as a function of the stator pole width spb and the magnet length ml .

Maximum force densities dF at different m pl τ and sp pb τ ratios are marked with circles in

Fig. 5-2-5. (6) The magneto-motive force mmf by current and the air-gap length gl are fixed to 3000 [AT]

and 1 [mm], respectively. The magnet length ml and the stator pole width spb are fixed using

the optimum ratios of m pl τ and sp pb τ determined in the previous progress. Force densities

dF of the machine are calculated as a function of the pole pitch pτ in order to determine the

optimum size of pτ . Fig. 5-2-6 depicts the force density dF investigated as a function of the

pole pitch pτ .

(7) Force densities dF of the machine are calculated as a function of the magneto-motive force

mmf by current in order to determine the optimum value of mmf . Fig. 5-2-7 depicts the force density dF as a function of the magneto-motive force mmf for a TFPM machine with

5gl = [mm], 50pτ = [mm]. Considering the saturation effect of the machine, it seems a better

choice to determine the mmf at 10,000 [Ampere-turn] that is multiplication of 62 10× [Ampere/m] and air-gap length gl [m].


105

bsp/τp

0.5 0.6 0.7 0.8 0.9 1.0

Forc

e de

nsity

, Fd [

kN/m

2 ]

20

25

30

35

40

45

lm/τp=0.5

lm/τp=0.4

lm/τp=0.3

lm/τp=0.2

Fig. 5-2-5: Force density dF as a function of the ratios of PM length ml to pole pitch pτ ,

and stator pole width spb to pole pitch pτ

Magnetic intensity, H [A/m]

10 100 1000 10000 100000

Flux

den

sity

, B [T

]

0.0

0.5

1.0

1.5

2.0

2.5

B-H curve of 35PN380 core

Fig. 5-2-4: Flux density and magnetic field intensity (B-H) curve of the iron core used in the analysis model

Chapter 5

106

Magnetomotive force/Air gap length, mmf/lg [AT/m]

0 1x106 2x106 3x106 4x106 5x106 6x106

Forc

e de

nsity

, Fd

[kN

/m2 ]

0

20

40

60

80

lg=5mm, τp=50mm, lm/τp=0.4, bsp/τp=0.8

Fig. 5-2-7: Force density of a TFPM machine with 5gl = [mm] and 50pτ = [mm] dF as a

function of the mmf by current

τp/lg [ - ]

0 10 20 30 40

Forc

e de

nsity

, Fd [

kN/m

2 ]

0

10

20

30

40

50

60

lm/τp=0.4, bsp/τp=0.8

Fig. 5-2-6: Force density dF as a function of pole pitch pτ


107

According to the FEA results represented in Fig. 5-2-5, Fig. 5-2-6 and Fig. 5-2-7, the following is taken: • The air-gap length is a main parameter in the electromagnetic design of TFPM machines. • The optimum pole pitch is determined by 10p glτ = .

• The optimum magnet length is determined by 0.4m pl τ= .

• The optimum stator pole width is determined by 0.8sp pb τ= .

• The optimum magneto-motive force by current mmf is determined by [ ] 2000000[ / ] [ ]s s gmmf AT N I AT m l m= = × .

In order to achieve the maximum force density of TFPM machines under the limited design condition, electromagnetic dimensions and parameters of TFPM machines are determined in Table 5-2-2.

Chapter 5

108

Table 5-2-2: Parameters and dimensions of TFPM machines Air gap diameter (Generator rotor diameter) . .g TFPMG g RFPMGD D= [m]

Air gap length 1000

gg

Dl = [m]

Pole pitch 10p glτ = [m]

Ratio of magnet height to pole pitch 0.4m

p

lτ = [-]

Magnet height mm p

p

ll ττ= × [m]

Ratio of stator pole width to pole pitch 0.8p

p

bτ = [-]

Stator pole width pp p

p

bb ττ= × [m]

Rotor pole width pr p mb lτ= − [m]

Magneto-motive force mmf by current 2000000cslot s gmmf N I l= = × [AT]

Current density 3sJ = [A/mm2]

Number of conductors per slot cslot scslot

s

N IN I= [-]

Cross-section area of conductors per slot 1 610cslot s

Cu phs

N IAJ

=⋅

[m2]

Slot filling factor 0.65sfillk = [-]

Cross-section area of slot 1Cu phs

sfill

AA k= [m2]

Stator slot width s sb A= [m]

Stator slot height s sh b= [m]

Number of pole pairs 2

g

p

Dp

πτ

= [-]

Stator pole length .max

max

psp

e cslot p p

el

p N B bω= [m]

Stator height S sy s Rh h h h= + + [m]

Stator yoke height sy sth h= [m]

Rotor height R sph l= [m]

Air gap area g g sA D lπ= [m2] Here, ( )2s sp sl m l b= + [m2]


109

5.3 Generalization of magnetic circuit of PM machines

5.3.1 Linear model In electric machines it is generally necessary to establish flux between a stationary part and a moving part. This involves causing flux to cross an air gap. To produce a strong magnetic field in an air gap, permanent magnet (PM) materials are being used for electric machines [Sle 1992]. In the research of this thesis, PM machines are thus considered for use in generators of large direct-drive wind turbines. In order to show how to model the magnetic circuit and how to calculate the flux of PM machines, a generalized configuration of the magnetic circuit is given as Fig. 5-3-1.

Fig. 5-3-1: Cross-section of a simplified magnetic circuit of a PM machine In Fig. 5-3-1 syh is the height of yoke of the stationary part, sth is the height of the tooth of the

stationary part, gl is the length of air gap, ml is the height of the magnets and ryh is the height of

the yoke of the moving part. A flux path caused by the magnets is represented by a bold line with arrows. This path is used as a contour of the magnetic circuit to apply Ampere’s circuit law. The contour is used to calculate the flux density due to the permanent magnets in the air gap and the iron cores. An equation for the B-H characteristic of the magnets in the second quadrant is given as

Chapter 5

110

0m rm m rmB H Bμ μ= + (5.3.1)

where, mB is the flux density of the magnets [T], 0μ is the permeability of free space [H/m], rmμ

is the recoil permeability of the magnets [ - ], mH is the magnetic field intensity of the magnets

[A/m] and rmB is the remanent flux density of the magnets [T]. Equations for the B-H characteristic of the air gap and iron cores are written as (5.3.2), (5.3.3), (5.3.4) and (5.3.5), respectively.

0g gB Hμ= (5.3.2)

0 .ry r ry ryB Hμ μ= (5.3.3)

0 .st r st stB Hμ μ= (5.3.4)

0 .sy r sy syB Hμ μ= (5.3.5)

B is the flux density [T], H is the magnetic field intensity [A/m] and rμ is the relative

permeability [ - ]. The subscripts of g , ry , st and sy represent the air gap, the yoke of the moving part, the tooth of the stationary part and the yoke of the stationary part, respectively. Using Ampere’s circuit law, the contour in Fig. 5-3-1 is expressed as 2 2 2 0g g m m ry ry st st sy syH l H l H l H l H l+ + + + = (5.3.6)

where l is the magnetic field length. At low values of the magnetic field intensity, the flux density in iron cores increases almost linearly. The permeability of iron cores with low magnetic intensity is much larger than the permeability of the air and the magnets. Therefore, (5.3.6) can be simplified by neglecting terms of the magnetic field intensity of iron cores for the range of low magnetic field intensity as 2 2 0g g m mH l H l+ = (5.3.7).

From the continuity of flux, the flux φ is written as

g g m m ry ry st st sy syB A B A B A B A B Aφ = = = = = (5.3.8)

where A is the area. In order to calculate the flux density in the air gap, (5.3.7) is reformulated as (5.3.9) using (5.3.1), (5.3.2) and (5.3.8).

0 0

2 2 0

gg rm

g mg m

rm

AB BB Al l

μ μ μ

−+ = (5.3.9)


111

Therefore, the air gap flux density is determined by (5.3.10).

0

0 0

2 12 2

rm mg

g g mrm

rm m

B lB l A lA

μ μμ μ μ

= ×+

(5.3.10)

Assuming that the flux φ varies sinusoidally with time, the flux φ [Wb], the flux linkage λ [Wb-turns] and the no-load phase voltage e [V] thus are expressed as (5.3.11), (5.3.12) and (5.3.13).

max( ) sin et tφ φ ω= (5.3.11)

where axmφ is the amplitude of the flux [Wb], eω is the angular frequency (= 2 fπ , here f is the

frequency) [rad/s] and t is time [s]. From the equations expressed above and Faraday’s law, the flux linkage λ and the no-load voltage e are given as

sNλ φ= (5.3.12)

dedtλ

= (5.3.13)

where sN is the total number of windings per phase [turns]. The no-load phase voltage induced in the sN turns coil with time is written as

( ) sde t Ndtφ

= (5.3.14).

(5.3.14) is reformulated by substituting (5.3.11) into (5.3.14) as

max( ) cos coss e e ee t N t e tφ ω ω ω∧

= = (5.3.15). The amplitude of the no-load phase voltage is given as

max maxs e ee N φ ω λ ω∧

= = (5.3.16). Generator power is formulated as a function of the number of phases m [ - ], the terminal voltage

tV [V], the nominal current snomI [A] and the power factor cosφ [ - ] as

cost snomP mV I φ= (5-3-17).

Chapter 5

112

This chapter focuses on the electromagnetic design of generators, and the power factor cosφ is assumed as 1 (one) in the analytical modelling. The equation of generator power is re-formulated as

snomP mEI= (5-3-18). Here, E is the root-mean-square (RMS) value of the no-load phase voltage. The minimum nominal current snomI necessary to produce the power of a three-phase electric machine is written by

3snomPIE

= (5.3.19)

The no-load phase voltage of a surface-mounted RFPM machine is formulated as

.max22 w cslot m g s gE k N p r l Bω= (5.3.20)

where wk is the winding factor, cslotN is the number of conductors per a slot, p is the number of

pole pair, mω is the mechanical angular velocity, gr is the radius of generator rotor, sl is the stack

length in axial direction, and .maxgB is the amplitude of air-gap flux density.

Electromagnetic analysis models and relevant expressions on TFPM machines have been discussed and formulated by a number of authors as discussed in [Ban 2008b]. However, most models and equations are not sufficient for use with various topologies of TFPM machines because of restrictiveness of the models and equations. Therefore, in this thesis a generalized formulation of no-load phase voltage of TFPM machines is developed and proposed as

.max2p

cslot m g sp pp

bE N p r l B

πω

τ

⎛ ⎞= ⎜ ⎟⎜ ⎟⎝ ⎠

(5.3.21)

where pb is the pole width, pτ is the pole pitch, spl is the pole length in axial direction, and

.maxpB is the amplitude of flux density in the pole.

5.3.2 Nonlinear model When the magnetic field intensity H in iron cores is increased, their flux density increases nonlinearly, which leads to a decrease in the permeability of the iron cores. Further increasing the magnetic field’s intensity results in the iron cores being saturated and an increase in the reluctances of the magnetic path. Therefore, it is necessary to consider the nonlinear


113

characteristic of the flux density of the iron cores when designing electric machines meant to operate in the region of higher flux density. The flux caused by the magnets was calculated as (5.3.8) from the continuity of flux. The flux caused by the magnets is also calculated as a function of the magnetic field intensity of the magnets mH , the magnetic field length of the magnets ml and the total reluctance in the

equivalent circuit tR as (5.3.22).

m m

t

H lR

φ = (5.3.22)

The equivalent circuit of the magnetic reluctance of the model shown in Fig. 5-3-1 could be simplified as Fig. 5-3-2. In the figure, the white rectangles represent the iron core reluctances, the white rectangles with bold lines represent the air gap reluctances. The rectangles hatched represent the magnet reluctances.

: Iron core reluctance

: Air gap reluctance

: PM reluctance

mF mF

Fig. 5-3-2: A simplified equivalent reluctance model of Fig. 5-3-1 The total magnetic reluctance of the magnetic circuit of the model is given by

t g m ry st syR R R R R R= + + + + (5.3.23)

where gR is the reluctance of the air gap, mR is the reluctance of the magnets, ryR is the

reluctance of the yoke of the moving part, stR is the reluctance of the tooth of the stationary part

and syR is the reluctance of the yoke of the stationary part.

Chapter 5

114

General expressions of the magnetic reluctances are given as a function of the magnetic field length, its permeability and area.

0

2 gg

g

lR

Aμ= (5.3.24)

2 mm

m m

lRAμ

= (5.3.25)

ryry

ry ry

lR

Aμ= (5.3.26)

2st stst

st st st st

l hRA Aμ μ

= = (5.3.27)

sysy

sy sy

lR

Aμ= (5.3.28)

Considering the nonlinear B-H characteristic of the iron cores and the above expressions, the flux density in the air gap gB is reformulated as

m mg

t g

H lBR A

= (5.3.29)

From the continuity of flux, it is possible to calculate the flux density in the magnets mB , the

yoke of the moving part ryB , the tooth of the stationary part stB and the yoke of the stationary

part syB . After determining the flux density in the air gap, the procedure and the expressions to

calculate the flux φ , the flux linkage λ and the no-load voltage e are the same with as in section 5.3.1. The procedure to determine the flux density, the flux, the flux linkage and the no-load induced voltage of a magnetic circuit including nonlinear characteristic is made as the following steps. Fig. 5-3-3 depicts a flow chart of the procedure. (Step1) Input predefined parameters of PM machines (Step2) Determine the initial values of the flux densities of the air gap gB , the magnets mB , the

yoke of the moving part ryB , the tooth of the stationary part stB and the yoke of the

stationary part syB using Ampere’s circuit law and the continuity of flux


115

(Step3) Assume the initial value of the permeability of the tooth of the stationary part stμ . In the

thesis we assume 0300stμ μ= as an initial value.

(Step4) Calculate the magnetic field intensity of the tooth of the stationary part stH by

( 2)

( 3)

st Stepst

st Step

BH

μ= .

(Step5) Re-calculate the flux density of the tooth of the stationary part stB with the B-H curve of

the iron core ( , )st stB f H BH curve= − . Re-calculate the flux densities of the yoke of the

moving part ryB and the yoke of stationary part syB from the continuity of flux.

(Step6) Re-calculate the permeability of the tooth of the stationary part stμ by stst

st

BH

μ = .

Calculate the permeability of the yoke of the moving part ryμ and the yoke of the

stationary part syμ .

(Step7) Calculate the total reluctance of the magnetic circuit tR by t g m ry st syR R R R R R= + + + + .

(Step8) Calculate the flux φ by m m

t

H lR

φ = .

(Step9) Re-calculate the flux density of the tooth of the stationary part stB by stst

BAφ

= using φ

calculated in (Step8). (Step10) Re-calculate the magnetic field intensity of the tooth of the stationary part stH by

( 9)

( 6)

st Stepst

st Step

BH

μ= .

(Step11) Re-calculate the flux density of the tooth of the stationary part stB with the B-H

curve of the iron core ( 10)( , )st st StepB f H BH curve= − .

(Step12) Compare stB calculated in (Step9) with stB calculated in (Step11). If

( 11) ( 9)st Step st StepB B accuracy− ≤ , then proceed to the next step. If not, go back to (Step 6).

(Step13) Re-calculate the flux densities of the air gap gB , the magnets mB , the yoke of the

moving part ryB and the yoke of the stationary part syB using the continuity of flux.

(Step14) Re-calculate the flux φ , and calculate the flux linkage λ and the no-load induced voltage e .

Chapter 5

116

(Step 1) Input predefined parameters

(Step 2) Determine initial values of flux densities from Ampere’s law and flux continuity,

(Step 3) Assume initial value of permeability of iron core,

(Step 4) Calculate the magnetic intensities of iron cores by

(Step 5) Re-calculate flux densities of iron cores by BH-curve and flux continuity

(Step 6) Re-calculate permeability of iron cores by

(Step 7) Calculate the total reluctance of magnetic circuit

(Step 8) Calculate the flux using the mmf by the magnets and the total reluctance

(Step 9) Re-calculate the flux density of the tooth of stationary part using the flux calculated in (Step 8)

(Step 10) Re-calculate the magnetic field intensity of the tooth of stationary part by

(Step 11) Re-calculate the flux density of the tooth of stationary part by BH-curve

(Step 13) Re-calculate the flux densities of the air gap, the magnets and the iron cores

(Step 12)( 11) ( 9)st Step st StepB B accuracy− ≤

(Step 14) Re-calculate the flux, and calculate the flux linkage and the no-load voltage

YesNo

stB

stμ

( 2)st st Step stH B μ=

st st stB Hμ =

( 9) ( 6)st st Step st StepH B μ=

Fig. 5-3-3: Flow chart showing the process to determine the flux density, the flux, the flux linkage and the no-load induced voltage of the magnetic circuit of a PM machine

5.3.3 Magnetic circuit modeling of PM machines A. RFPM machine In this section, a contour is sketched to which Ampere’s circuit law is applied as shown in Fig. 5-3-4. The figure depicts two poles of the magnetic circuit of a surface-mounted RFPM machine, which consists of a stator with slotted iron cores and full pitch windings, and a rotor with iron cores and magnets.


117

Fig. 5-3-4: The contour of a surface-mounted RFPM machine to apply Ampere’s circuit law In the modelling of the magnetic circuit of the RFPM machine, it is assumed that • the flux density crosses the magnets and the air gap perpendicularly • there is no leakage flux in the circuit • the fluxes in the air gap, the magnets, the rotor yoke, the stator teeth and the stator yoke are

equal. In order to fulfill the above assumptions, the shape of the magnets is modified from Fig. 5-3-4 to Fig. 5-3-5. In Fig. 5-3-5, the flux paths in a contour are represented with lines and arrows. A simplified equivalent circuit of the magnetic reluctance of the RFPM machine illustrated in Fig. 5-3-4 and Fig. 5-3-5 is the same with the equivalent circuit shown in Fig. 5-3-2.

Fig. 5-3-5: A linearized RFPM machine with the modified magnets and the flux paths

Chapter 5

118

B. TFPM machine Fig. 5-3-6 depicts a flux-concentrating TFPM machine with five poles and a contour to which Ampere’s circuit law is applied. The machine consists of a stator with iron cores and windings, a rotor with flux-concentrating iron cores and the magnets. A simplified equivalent circuit model of the magnetic reluctances of the TFPM machines is illustrated in Fig. 5-3-7.

Fig. 5-3-6: The contour of a flux-concentrating TFPM machine to apply Ampere’s circuit law


119

mF



: PM reluctance

Fig. 5-3-7: A simplified equivalent circuit of magnetic reluctances of flux-concentrating TFPM machines In the equivalent circuit of the TFPM machine, the leakage fluxes of TFPM machines are much larger than the leakage fluxes of the RFPM machines with full pitch windings. Therefore, the leakage fluxes of TFPM machines are included in the equivalent circuit in the thesis. Electromagnetic characteristics of the TFPM machines are the same and repetitive in every one pole pair. Therefore, the electromagnetic equivalent circuit of one pole pair is considered for the analytical model. The equivalent circuit including the leakage fluxes is illustrated in Fig. 5-3-8, which is made by cutting the stator in the middle and spreading both the stator and the rotor. The white rectangles are the iron core reluctances and the white rectangles with bold lines are the air gap reluctances. The blue rectangles hatched are the PM reluctances and the red rectangles dotted are the leakage flux reluctances. In the modelling of the magnetic circuit of the TFPM machine, it is assumed that • the flux density crosses the magnets perpendicularly • the flux density crosses the air gap on iron cores perpendicularly • the leakage fluxes are modelled as shown in Fig. 5-3-8.

Chapter 5

120

The equivalent circuit is modified as Fig. 5-3-9 including the fluxes and the magneto-motive force due to the magnets. In order to determine the fluxes Aφ , Bφ , Cφ , Dφ , Eφ , Fφ , Gφ , Hφ , Iφ ,

Jφ , Kφ and Lφ in Fig. 5-3-9 (a), Kirchhoff’s voltage law is applied to the fluxes 1Φ , 2Φ , 3Φ ,

4Φ and 5Φ in Fig. 5-3-9 (b), (c), (d) and (e). The procedure and equations to calculate the magnetic reluctances and the fluxes in the equivalent circuits in Fig. 5-3-9 are described in Appendix 1.

Fig. 5-3-8: Equivalent circuit of magnetic reluctances of flux-concentrating TFPM machines including leakage flux reluctances


121

11R

12R

13R

14R 15R

9R 8R

5R

4R

3R

1R 2R

18R

19R

28R

27R

21R

22R

29R 20R

AφHφ

KφIφ

GφJφ

Cφ

DφEφ

Fφ

Lφ

17R

16RmF

7R 6R10R26R

25R

24R

23R

Bφ

(a)

11R

12R

13R

14R 15R

9R 8R

5R

4R

3R

1R 2R

18R

19R

28R

27R

21R

22R

29R 20R 17R

16RmF

7R 6R10R26R

25R

24R

23R

1Φ2Φ

11R

12R

13R

14R 15R

9R 8R

5R

4R

3R

1R 2R

18R

19R

28R

27R

21R

22R

29R 20R 17R

16RmF

7R 6R10R26R

25R

24R

23R

3Φ

(b) (c)

11R

12R

13R

14R 15R

9R 8R

5R

4R

3R

1R 2R

18R

19R

28R

27R

21R

22R

29R 20R 17R

16RmF

7R 6R10R26R

25R

24R

23R

4Φ

11R

12R

13R

14R 15R

9R 8R

5R

4R

3R

1R 2R

18R

19R

28R

27R

21R

22R

29R 20R 17R

16RmF

7R 6R10R26R

25R

24R

23R

5Φ

(d) (e)

Fig. 5-3-9: Equivalent circuit of magnetic reluctances of flux-concentrating TFPM machines including leakage flux reluctance and magneto-motive force caused by magnets

Chapter 5

122

5.4 Verification of magnetic circuit analysis model This section discusses the verification of the proposed analysis model of TFPM machines for two cases, a no-load case and a case with a load. In order to verify the analytical model for the no-load case, the no-load induced voltage obtained by the analytical model is compared with the no-load voltage obtained through three-dimensional finite element analyses (3D FEA) and the measurement. To validate the proposed analysis model for the case with a load, the force obtained through the analysis model is compared with the force obtained through the FEA and the static force measurement. The TFPM machine considered for the verification is a single-sided, single winding flux-concentrating TFPM machine with a U-core and passive rotor as illustrated in Fig. 5-4-1. Solid iron cores were chosen to build the TFPM machine for easier manufacturing.

Fig. 5-4-1: A linearized single-sided single winding flux-concentrating TFPM machine with U-core and passive rotor

5.4.1 Verification of no-load case A. Analysis model Analysis models and formulations in section 5.3 are used for the analysis of a single-sided, single winding flux-concentrating TFPM machine with a U-core and passive rotor. The equivalent


123

circuit of magnetic reluctances of the TFPM machine in Fig. 5-4-1 is represented by cutting the rotor in the middle and spreading both the stator and the rotor as Fig. 5-4-2. The parameters and dimensions of the single-sided, single winding flux-concentrating TFPM machine with a U-core and passive rotor is given in Table 5-4-1. The parameters and dimensions in Table 5-4-1 are determined by Table 5-2-2 in sub-section 5-2-2. Material characteristics of the TFPM machine are given in Table 5-4-2. Fig. 5-4-3 depicts the characteristics of the flux density and the magnetic intensity of the solid iron cores used for the TFPM machine. The amplitudes of the flux density in the stator core between magnets, and the flux, the flux linkage and the no-load induced voltage per pole pair obtained by the analysis are given in Table 5-4-3.

: Iron core reluctance : Air-gap reluctance

: Magnet reluctance : Leakage flux reluctance

1

2

3

4

5

6a

7 8

9

10b

11

12

13

14

15

16

17

18

19

20b

21

22

26

23

24

25

27

28

29a

6b

10a

20a

29b

Fig. 5-4-2: The equivalent circuit of magnetic reluctances of a single-sided single winding flux-concentrating TFPM machine with a U-core and passive rotor

Chapter 5

124

Table 5-4-1: Parameters and dimensions of the TFPM machine Air gap length 2gl = [mm]

Pole pitch 20pτ = [mm]

Magnet height 10ml = [mm]

Stator pole width 14pb = [mm]

Rotor pole width 10prb = [mm]

Number of conductors per slot 288cslotN = [Turn]

Stator slot width 50sb = [mm]

Stator slot height 35sh = [mm]

Number of pole pairs 7.5p = [-]

Stator pole length 40spl = [mm]

Stator height 75Sh = [mm]

Stator yoke height 40syh = [mm]

Rotor height 40Rh = [mm]

Table 5-4-2: Material characteristics of the TFPM machine

Iron core type Solid core (S20c)

B-H curve: see Fig. 5-4-3

Resistivity of copper, Cuρ 0.025 [µΩm]

Remanent flux density of permanent magnets, rmB 1.3 [T]

Relative recoil permeability of permanent magnets, rmμ 1.06 [-]

Permeability of free space, 0μ 4π×10-7 [H/m]

Iron core, Feρ 7800 [kg/m3]

Permanent magnet, pmρ 7600 [kg/m3]Density

Copper, Cumassρ 8900 [kg/m3]


125

(a) B-H curve measurements (b) B-H curve used in the analysis Fig. 5-4-3: Flux density and magnetic intensity (B-H) curve of the iron core (S20c) Table 5-4-3: Amplitudes of flux density, flux, flux linkage and no-load induced voltage of a single-sided single winding flux-concentrating TFPM machine with a U-core and passive rotor by analytical model

Flux density in the stator core, B∧

1.46 [T]

Flux in the stator core per pole pair, _pole pairφ∧

0.000586 [Wb]

Flux linkage in the stator per pole pair, _pole pairλ∧

0.168665 [Wb]

No-load induced voltage per pole pair, _pole paire∧

26.49 [V]

B. Finite element analysis Fig. 5-4-4 depicts two poles of the single-sided, single winding flux-concentrating TFPM machine with a U-core and passive rotor for three-dimensional finite element analysis (3D FEA). Table 5-4-4 gives the amplitudes of the flux density, the flux, the flux linkage and the no-load induced voltage obtained by the 3D FEA. Comparing the no-load voltage obtained from the 3D FEA with the voltage obtained by the analytical model, the following result is taken: • The no-load voltage obtained by the analytical model is 2 % higher than the voltage obtained

from the 3D FEA.

Chapter 5

126

Fig. 5-4-4: 3D FEA model for a single-sided single winding flux-concentrating TFPM machine with a U-core and passive rotor Table 5-4-4: Amplitudes of flux density, flux, flux linkage and no-load induced voltage of single-sided single winding flux-concentrating TFPM machine with U-core and passive rotor by 3D FEA


1.35 [T]


0.000540 [Wb]


0.155416 [Wb]


24.41 [V]

C. Measurement Fig. 5-4-5 (a) and (b) depict the sketches of the linearized single-sided, single winding flux-concentrating TFPM machine with U-cores and passive iron cores. Fig. 5-4-5 (c) depicts a method of assembling the iron cores and the permanent magnets using bolts [Ban 2007c]. The TFPM machine was built using the parameters, dimensions and material characteristics described in Table 5-4-1 and Table 5-4-2.


127

(a) A set of iron cores (grey and dark grey color), permanent magnets (dark blue color), and

assembly components

(b) A set of passive iron cores (grey color) and assembly components

(c) A method of assembling iron cores (grey and dark grey color) and permanent magnets (dark

blue color) using bolts [Ban 2007c] Fig. 5-4-5: Sketches of a linearized single-sided single winding flux-concentrating TFPM machine with U-cores and passive iron cores

Chapter 5

128

In Fig. 5-4-6, (a) depicts the U-cores with holes for bolting with the next U-core and with projecting parts for assembling. The cores were mechanically designed to guide permanent magnets in bonding on the core and to avoid the detachment of the magnets. Fig. 5-4-6 (b) shows a permanent magnet before assembling with cores, and (c) illustrates a set of U-core and the magnets bonded on the core.

(a) U-cores with holes for bolting and with projecting parts for assembling

(b) a permanent magnet before assembling with cores

(c) a set of a U-core and permanent magnets bonded on the core

Fig. 5-4-6: Photographs of U-cores, a permanent magnet and a set of the core and the magnets


129

Fig. 5-4-7 shows the assembly progress of the linearized TFPM machine. In Fig. 5-4-7, (a) illustrates the progress to assemble the U-cores and the magnets using a jig. After assembling them on the jig, a set of the cores and the magnets is fixed on the base frame using setscrews. Fig. 5-4-7(b) illustrates a winding machine with a winding jig, a winding with tapping progress and a winding formed with a thermo-couple. The progress to assemble the copper windings is shown in Fig. 5-4-7(c).

(a) Assembly progress of the sets of U-cores and magnets

(b) Winding machine and progress of copper winding

(c) Progress to assemble the copper windings with a set of the U-cores and the magents

Fig. 5-4-7: Photographs of assembly progress of the linearized TFPM machine

Chapter 5

130

A set of iron cores, magnets and copper windings is equipped on a vehicle guided by wheels, and a set of passive iron cores is equipped on the stationary track as shown in Fig. 5-4-8. The vehicle is driven by the external motor drives which are coupled with the wheels. Therefore, it is possible to measure the no-load induced voltage of the TFPM machine by driving the external motors with the wheels. It is also possible to change and control the speed of the vehicle by the motor drives.

(a) A set of iron cores, permanent magnets, copper windings and assembly components

(b) A set of passive iron cores

(c) TFPM machine installed with guide wheels

Fig. 5-4-8: Photographs of the linearized TFPM machine built Fig. 5-4-9 depicts no-load voltages measured at 1 m/s vehicle speeds (air-gap speeds). Table 5-4-5 gives the flux density, the flux, the flux linkage calculated by the no-load voltage and the amplitude of the no-load voltage. Comparing the no-load voltage obtained from the measurement with the voltage obtained by the analytical model, the following result is taken:


131

• The no-load voltage obtained by the analytical model is 8 % higher than the voltage obtained from the measurement.

Fig. 5-4-9: No-load induced voltages (y-axis: 6.67 V/div/pole pair) measured at 1 m/s air gap speeds (phase angle difference: 45 o) Table 5-4-5: Amplitudes of flux density, flux, flux linkage and no-load induced voltage of single-sided single winding flux-concentrating TFPM machine with a U-core and passive rotor as measured


1.49 [T]


0.000595 [Wb]


0.171463 [Wb]


26.93 [V]

5.4.2 Verification in the case with a load To validate the proposed analysis model of the TFPM machine for the case with a load, the force obtained in the analysis model is compared with the force obtained through the 3D FEA and the static force measurement. Using (5.3.18) and (5.3.21), the Lorenz force of a phase of the TFPM machine for the analysis model is formulated as

.max .max2 2sp spsnom snom

cslot p sp g cslot p sp snomg gp p

b bI IF E pN B l v pN B l Iv v

π πτ τ

⎛ ⎞ ⎛ ⎞= = =⎜ ⎟ ⎜ ⎟⎜ ⎟ ⎜ ⎟

⎝ ⎠ ⎝ ⎠ (5.4.1)

Chapter 5

132

The thrust force per pole pair of the TFPM machine obtained through the 3D FEA and the static force measurement are represented in Fig. 5-4-10 as a function of the rotor displacement. Due to the effect of the cogging force and the reluctance force, the sinusoidal distribution of the thrust force of electric machines is distorted. The effect of the cogging force and the reluctance force in TFPM machines is noticeably high compared to longitudinal flux permanent magnet (LFPM) machines. Thus, the force distribution of the TFPM machine is not sinusoidal as shown in Fig. 5-4-10.

Displacement

0.0 po

le pit

ch

0.2 po

le pit

ch

0.4 po

le pit

ch

0.6 po

le pit

ch

0.8 po

le pit

ch

1.0 po

le pit

ch

Thru

st fo

rce

per p

ole

pair

[N]

-100

0

100

200

300

400

500

FEAMeasurement

0 [AT/m]

mmf / lg = 2.5x106[AT/m]

1.25x106[AT/m]

Fig. 5-4-10: Thrust force per pole pair obtained through 3D FEA and static force measurement The analytical model and the formulation (5.4.1) do not include the effect of the cogging and reluctance force in calculating the thrust force. Therefore, the following assumptions are also made to verify the analytical model of the TFPM machine in the case with a load. • The force distribution obtained by the analytical model is sinusoidal. • The force at the rotor displacement of 0.5 pole pitch is the maximum. • In the force distribution obtained through the 3D FEA and the static force measurement, the

cogging effect is omitted. Thus the force at the rotor displacement of 0.5 pole pitch is compared with the force obtained by both the analytical model and (5.4.1).


133

The peak thrust force obtained through the 3D FEA, the static force measurement, the analytical model and the above assumptions is given as a function of the magnetomotive force per air gap length, mmf/lg in Table 5-4-6. The peak force by the analytical model is compared with the peak force by the 3D FEA and the measurement, and the comparison results are represented in the Fig. 5-4-11. At the rated magnetomotive force per air gap length, mmf/lg=2x106 [AT/m], the peak force obtained by the analytical model is 10 % and 7 % larger than the force obtained by the 3D FEA and by the measurement, respectively. Increasing the mmf/lg, it becomes bigger that the differences between the force by the analytical model and the force by the 3D FEA and the measurement, because the analytical model did not include the saturation effect by the stator current. Table 5-4-6: Thrust force per a pole pair obtained through the analytical model

Magnetomotive force/Air gap length, mmf / lg [AT/m]

0 0.63x106 1.25x106 1.88x106 2.5x106 3.1x106

peakF by 3D FEA [N] 0.0 113.9 221.5 313.2 386.4 438.0

peakF by measurement [N] 0.0 134.0 230.7 321.8 391.5 441.8

peakF by analytical model [N] 0.0 114.3 229.3 344.1 458.6 573.2


0.0 1.0x106 2.0x106 3.0x106 4.0x106

F peak

.mea

sure

men

t / F

peak

.ana

lysi

s [-]

F peak

.FE

A /

F peak

.ana

lysi

s [-]

0.0

0.2

0.4

0.6

0.8

1.0

1.2

1.4

Fpeak.measurement / Fpeak.analsysis

Fpeak.FEA / Fpeak.analsysis

Fig. 5-4-11: Comparion of thrust force obtained by the 3D FEA, the measurement and the analytical model

Chapter 5

134

5.5 Conclusions The main topic of discussion in this chapter was the development of a generalized analytical model for various topologies of permanent magnet (PM) machines. A slotted surface-mounted radial flux permanent magnet (RFPM) machine and flux-concentrating transverse flux permanent magnet (TFPM) machines were chosen in the modelling of these machines for direct-drive wind turbines. Firstly, a determination of geometric parameters of RFPM machine and TFPM machine was discussed. For the RFPM machine, the geometric parameters were determined by an investigation of the parameters of RFPM machines discussed in scientific literature. To determine the geometric parameters of TFPM machines, two-dimensional finite element analyses (2D FEA) of electromagnetic fields were done. In the analyses it was focused on finding the optimum geometrical parameters to maximize the force density. The parameters to optimize were the magnet length, the stator pole width, the pole pitch and the magnetomotive force by current. Next, generalized electromagnetic circuit analysis models were created for both the RFPM and TFPM machines. Nonlinear B-H characteristics of iron cores were included in the models. Leakage fluxes of TFPM machines are much larger than the RFPM machine with full pitch windings, thus the leakage fluxes were included in the models of TFPM machines. Furthermore, to verify the analytical model of TFPM machines in no-load case, the no-load induced voltage obtained through the analytical model was compared with the no-load voltages obtained through the three-dimensional finite element analyses (3D FEA) and the measurement. The no-load voltage obtained by the analytical model at 1 m/s air gap speed was 8 % higher than the voltage obtained by the 3D FEA and 2 % lower than the voltage obtained by the measurement. To validate the analytical model in the case with a load, the force obtained by the analytical model was compared with the force obtained through 3D FEA and the static force measurement as a function of the displacement and the stator current. At the nominal stator current, the force obtained by the analytical model is 10 % and 7 % larger than the force obtained by the 3D FEA and by the static force measurement, respectively. However, increasing the stator current larger than the nominal current, the differences between the force by the analytical model and the force by the 3D FEA and the measurement become bigger. This was caused by no consideration of the saturation effect by the stator current in the analytical model. The analytical models developed in this chapter will be used for a comparative design of PM wind generators in the next chapter. In the design the stator current will be limited to the nominal current.

Comparative Design of PM Generators for Large Direct-Drive Wind Turbines

135

Chapter 6


6.1 Introduction The aim of this chapter is to assess different topologies of permanent magnet (PM) generators for large direct-drive wind turbines. For the assessment, a comparative design of different PM generators for 5 MW and 10 MW direct-drive wind turbines is represented using the analysis models derived in Chapter 5. This chapter consists of the following outline. First, a selection of types of PM generators for large direct-drive wind turbines is discussed. A surface-mounted radial flux permanent magnet (RFPM) machine and four different flux-concentrating transverse flux permanent magnet (TFPM) machines are chosen for a comparative design. Next, the electromagnetic aspects of the chosen PM generators are designed, taking into consideration of the parameters of 5 MW and 10 MW direct-drive wind turbines. These generators are assessed based on the criteria of mass, loss, cost, efficiency and force density.

6.2 Selection of generator types for large direct-drive wind turbines Different topologies of permanent magnet (PM) machines have been discussed in a number of references as discussed in Chapter 3. Out of these different PM machines, surface-mounted radial flux permanent magnet (RFPM) machines have been discussed as a better choice for large direct-drive wind turbines in references. Considering the force density of electric machines, flux-concentrating TFPM machines have been discussed as potentially having higher force density than surface-mounted TFPM machine topologies. Single winding topologies of TFPM machines have been discussed as a suitable type for simpler construction and lower copper losses.

Chapter 6

136

Therefore, the following five different types of PM generators have been selected for the comparative design in this chapter. 1) RFPMG: a slotted surface-mounted RFPM generator with full pitch windings and inner rotor

(R1 machine, RF-SM-ST-IR-NdFeB in chapter 3) 2) TFPMG-1: a single-sided, single winding flux-concentrating TFPM generator with U-core 3) TFPMG-2: a double-sided, single winding flux-concentrating TFPM generator with C-core 4) TFPMG-3: a single-sided, single winding flux-concentrating TFPM generator with U-core

and passive rotor 5) TFPMG-4: a double-sided, single winding flux-concentrating TFPM generator with C-core

and passive rotor Linearized structures of the selected five different PM generators are illustrated in Figs. 6-2-1 to 6-2-5. Fig. 6-2-1 depicts a surface-mounted RFPM generator with PMs and a back yoke in the rotor, and windings, slots and a back yoke in the stator (RFPMG). Fig. 6-2-2 depicts the single-sided air gap TFPM generator which consists of flux-concentrating cores with PMs in the rotor, and U-cores with single winding in the stator (TFPMG-1). The double-sided air gap TFPM generator with flux-concentrating cores and PMs in the rotor, and with C-cores and single windings in the stator (TFPMG-2) is depicted in Fig. 6-2-3. Fig. 6-2-4 depicts the single-sided air gap TFPM generator which consists of flux-concentrating U-cores, PMs and single windings in the stator with a passive rotor (TFPMG-3). The double-sided air gap TFPM generator with flux-concentrating C-cores, PMs and single windings in the stator with a passive rotor (TFPMG-4) is depicted in Fig. 6-2-5.

Fig. 6-2-1: Surface-mounted RFPM generator (RFPMG)


137

Fig. 6-2-2: Single-sided single winding flux-concentrating TFPM generator with a U-core (TFPMG-1)

Fig. 6-2-3: Double-sided single winding flux-concentrating TFPM generator with a C-core (TFPMG-2)

Chapter 6

138

Fig. 6-2-4: Single-sided single winding flux-concentrating TFPM generator with a U-core (TFPMG-3)

Fig. 6-2-5: Double-sided single winding flux-concentrating TFPM generator with a C-core (TFPMG-4)


139

6.3 Analytical design of PM generators for direct-drive wind turbines Using the formulations and analytical models of PM machines discussed in chapter 5, a surface-mounted RFPM generator (RFPMG) and four different TFPM generators (TFPMG-1, TFPMG-2, TFPMG-3 and TFPMG-4) selected in section 6.2 are designed for 5 MW and 10 MW direct-drive wind turbines in this section. Table 6-3-1 gives the parameters of the wind turbines and the requirements for the generators. Cost models and material characteristics [Pol 2006] of the generators are given in Table 6-3-2 and Table 6-3-3, respectively. Table 6-3-1: Wind turbine parameters and generator requirements

Wind turbine parameters

Rated grid power, P 5 [MW] 10 [MW]

Rotor blades diameter, rD 126 [m] 178 [m]

Rotor blades tip speed, tipv 80 [m/s] 80 [m/s]

Rated rotor speed, N 12.1 [rpm] 8.6 [rpm]

Generator requirements

Nominal power, gennomP 5.56 [MW] 11.12 [MW]

Nominal torque, gennomT 4.38 [MNm] 12.38 [MNm]

Table 6-3-2: Generator cost models

Cost models

Iron core cost, Fek 3 [€/kg]

Copper cost, Cusk 15 [€/kg]

Permanent magnet cost, pmk 25 [€/kg]

Chapter 6

140

Table 6-3-3: Generator material characteristics

Material characteristics

Laminated electrical steel core

4 [W/kg] at 1.5 [T] and 50 [Hz] Specific hysteresis losses of iron cores

SMC core (Somaloy 700)

7.93 [W/kg] at 1.3 [T] and 50 [Hz] in [Lee 2009] - Sample core size:

Do: 57.2 [mm], Di: 26.4 [mm], H: 5.6 [mm] Laminated electrical steel core

1 [W/kg] at 1.5 [T] and 50 [Hz] Specific eddy current losses of iron cores,


0.17 [W/kg] at 1.3 [T] and 50 [Hz] in [Lee 2009] - Sample core size:

Do: 57.2 [mm], Di: 26.4 [mm], H: 5.6 [mm]





Iron core, Feρ Lamination steel: 7700 [kg/m3] SMC core: 7440 [kg/m3]

Permanent magnet, pmρ 7600 [kg/m3] Density



141

6.3.1 RFPM generator Fig. 6-3-1 depicts the external shape, the dimensional parameters and the flux paths of the surface-mounted RFPM generator with full pitch windings (RFPMG). In the figure, dotted lines with arrows represent the flux paths. Table 6-3-4 gives the design results with dimensions, flux density, current and no-load voltage of RFPMG for 5 MW and 10 MW direct-drive wind turbines. The parameters and dimensions of the RFPMG in Table 6-3-4 were determined by the dimensions and parameters discussed in Table 5-2-1 in chapter 5. Table 6-3-5 gives the design results with mass, cost and losses of RFPMG for 5 MW and 10 MW direct-drive wind turbines.

sb

ml

pτ ryh

syh

sh

tb

(a) (b) Fig. 6-3-1: Dimensional parameters and flux paths of RFPMG

Chapter 6

142

Table 6-3-4: Dimensions, current, air-gap flux density and no-load voltage of RFPMG for 5 MW and 10 MW direct-drive wind turbines

5 [MW] 10 [MW]

Aspect ratio of generator, srad

g

lKD

= [-] 0.27 0.3

Air gap diameter (Generator rotor diameter), gD [m] 6.36 8.68

Axial length of generator, sl [m] 1.72 2.61

Air gap length, gl [mm] 6.36 8.68

Magnet height, ml [mm] 15.9 21.7

Stator diameter, sD [m] 6.37 8.699

Number of phases, m [-] 3 3

Stator slot pitch, sτ [mm] 33 33

Number of slots per pole per phase, q [-] 1 1

Pole pitch, pτ [mm] 100 99.7

Number of pole pairs, p [-] 100 137

Rotor pole width, pb [mm] 80 80

Stator slot width, sb [mm] 15 14.96

Stator tooth width, tb [mm] 18 18

Stator slot height, sh [mm] 80 80

Stator yoke height, syh [mm] 40 40

Rotor yoke height, ryh [mm] 40 40

Air gap area, gA ( )g sD lπ= [m2] 34.29 71.04

Nominal current, snomI [A] 606.2 551.9

Number of conductors per slot, cslotN [turns] 3.35 3.35

Peak flux density in the air-gap, gB∧

(= pmB∧

) [T] 0.97 1.07

RMS value of no-load voltage, E [V] 3057.3 6716.1

Force density, dF [kN/m2] 40.16 40.16


143

Table 6-3-5: Active mass, cost and loss of the surface-mounted RFPM generator for 5 MW and 10 MW direct-drive wind turbines

Mass of active material 5 [MW] 10 [MW]

Copper mass, CusM 7,491 14,961

Stator core mass, FesM 22,530 46,542

Permanent magnets mass, pmM 3,314 9,374

Rotor core mass, FerM 10,443 21,669

Mass of active material [kg]

Generator mass, genM 43,805 92,546

Copper cost, CusK 112,365 224,421

Stator core cost, FesK 67,671 139,625

Permanent magnets cost, pmK 82,857 234,350

Rotor core cost, FerK 31,329 65,006

Cost of active material [€]

Generator cost, genK 294,222 663,403

Copper loss, CusP 162 270.1

Stator core loss, FesP 25.3 60.3Loss [kW]

Generator loss, genP 187.3 330.4

Efficiency, nomη [%] 96.6 97

6.3.2 TFPM generators Fig. 6-3-2, Fig. 6-3-3, Fig. 6-3-4 and Fig. 6-3-5 depict the external shapes, the dimensional parameters and the flux paths of the four different flux-concentrating TFPM generators (TFPMG-1, TFPMG-2, TFPMG-3 and TFPMG-4). In the figures, dotted lines with arrows represent the flux paths simplified.

Chapter 6

144

δ

Rh

phsl 1.

syhsb

sh

Spl

Spb

pτml Rpb

Sh

sth

(a) (b) Fig. 6-3-2: Dimensional parameters and flux paths of TFPMG-1

δphsl 1.

Sh

Spl

δ

Spb

pτRpb ml

syh

sth

sb

sh

Rh



145

pτ

Stator

Core

PM

phsl 1.

sb

sh

Spl

Winding

Core

Rotor

Spb ml

Rpbδ

Rh

syh

Sh

sth

(a) (b) Fig. 6-3-4: Dimensional parameter and flux paths of TFPMG-3

pτ

Spb ml

δphsl 1.

Spl

δRpb

Stator

Rotor

Core

PM

Winding

CoreSh

syh

sth

sb

sh

Rh


Chapter 6

146

As discussed in chapter 5, the leakage fluxes of TFPM machines are much larger than the leakage fluxes of RFPM machines with full pitch windings. Therefore, the leakage fluxes are included in the equivalent circuits of magnetic reluctances of TFPM generators as illustrated in Fig. 6-3-6 and Fig. 6-3-7. Fig. 6-3-6 depicts the equivalent circuits of the magnetic reluctances of TFPMG-1 and TFPMG-2. Fig. 6-3-7 depicts the equivalent circuits of the magnetic reluctances of TFPMG-3 and TFPMG-4.

Fig. 6-3-6: Equivalent circuits of magnetic reluctances of TFPMG-1 and TFPMG-2


147

Fig. 6-3-7: Equivalent circuits of magnetic reluctances of TFPMG-3 and TFPMG-4 Table 6-3-6 gives the design results with dimensions, flux density, current and no-load voltage of the four different TFPM generators for 5 MW and 10 MW direct-drive wind turbines. In the design, the dimensions and parameters of the TFPM generators were determined by the dimensional parameters discussed in Table 5-2-2 in chapter 5. Table 6-3-7 gives the design results with mass, cost and losses of the four different TFPM generators for 5 MW and 10 MW direct-drive wind turbines.

Chapter 6

148

Table 6-3-6: Dimensions, current, air-gap flux density and no-load voltage of TFPMG-1, TFPMG-2, TFPMG-3 and TFPMG-4 for 5 MW and 10 MW direct-drive wind turbines

TFPMG-1 TFPMG-2 TFPMG-3 TFPMG-4

5 MW 10 MW 5 MW 10 MW 5 MW 10 MW 5 MW 10 MW

Air gap diameter, gD [m] 6.36 8.68 6.36 8.68 6.36 8.68 6.36 8.68

Axial length of generator, sl [m] 1.03 1.464 1.03 1.464 1.02 1.468 1.06 1.515

Air gap length, gl [mm] 6.36 8.68 6.36 8.68 6.36 8.68 6.36 8.68

Magnet height, ml [mm] 25.4 34.7 25.4 34.7 25.4 34.7 25.4 34.7

Stator diameter, sD [m] 6.37 8.699 6.37 8.699 6.37 8.699 6.37 8.699

Number of phase, m [-] 3 3 3 3 3 3 3 3

Pole pitch, pτ [mm] 63.6 86.9 63.6 86.9 63.6 86.9 63.6 86.9

Number of pole pairs, p [-] 157 157 157 157 157 157 157 157

Pole width, spb [mm] 50.9 69.5 50.9 69.5 50.9 69.5 50.9 69.5

Stator slot width, sb [mm] 80.8 94.4 80.8 94.4 80.8 94.4 80.8 94.4

Stator pole length, spl [mm]

(=Stator tooth width, stb ) 130.4 196.8 130.4 196.8 130.1 197.4 135.9 205.3

Stator slot height, sh [mm] 80.8 94.4 80.8 94.4 80.8 94.4 80.8 94.4

Stator yoke height, syh [mm] 130.4 196.8 130.4 196.8 130.1 197.4 135.9 205.3

Rotor height, Rh [mm] 130.4 196.8 291.6 528.1 130.1 197.4 135.9 205.3

Air gap area, gA ( )g sD lπ= [m2] 20.47 39.93 20.47 39.93 20.43 40.03 21.12 41.31

Nominal current, snomI [A] 606.2 551.9 606.2 551.9 606.2 551.9 606.2 551.9

Number of conductors per slot,

cslotN [turn] (= Number of conductors per phase)

20.98 31.46 20.98 31.46 20.98 31.46 20.98 31.46

Peak flux density in air-gap, gB∧

[T]

1.59 1.6 0.76 0.65 1.66 1.65 1.53 1.54

RMS value of no-load voltage, E [V]

3057 6,716 3058 6,716 3059 6,716 3057 6,716

Force density, dF [kN/m2] 67.3 71.5 67.3 71.5 67.4 71.3 65.2 69.1


149

Table 6-3-7: Active mass, cost and loss of TFPMG-1, TFPMG-2, TFPMG-3 and TFPMG-4 for 5 MW and 10 MW direct-drive wind turbines

TFPMG-1 TFPMG-2 TFPMG-3 TFPMG-4

5 MW 10 MW 5 MW 10 MW 5 MW 10 MW 5 MW 10 MW

Mass of active material [kg]

Copper, CusM 2,294 4,269 2,294 4,269 2,294 4,269 2,294 4,269

Stator core,

FesM 11,699 32,429 25,263 83,044 17,487 48,888 28,552 82,778

Permanent magnets, pmM 8,114 23,888 3,615 10,114 11,909 33,293 19,444 56,372

Rotor core,

FerM 11,915 35,077 5,308 14,852 7,910 23,519 1,648 3,849

Generator, genM 34,021 95,662 36,480 112,281 39,601 109,969 51,938 147,268

Cost of active material [€]

Copper, CusK 34,410 64,035 34,410 64,035 34,410 64,035 34,410 64,035

Stator core,

FesK 35,097 97,286 75,790 249,133 52,462 146,664 85,656 248,334

Permanent magnets, pmK 202,846 597,188 90,371 252,872 297,722 832,320 486,099 1,409,302

Rotor core,

FerK 35,744 105,230 15,924 44,558 23,732 70,557 4,945 11,546

Generator, genK 308,097 863,741 216,495 610,599 408,326 1,113,576 611,100 1,733,217

Loss [kW]

Copper, CusP 58.5 108.8 58.5 108.8 58.5 108.8 58.5 108.8

Stator core, FesP 34.8 68.6 75.1 175.6 92.9 182.7 138.9 286.1

Rotor core, FerP 60.4 126.4 11.2 18.4 24.5 51.2 4.8 7.9

Generator, genP 153.6 303.8 144.8 302.9 175.9 342.7 202.2 402.9

Efficiency, nomη [%]

97.2 97.3 97.4 97.3 96.8 96.9 96.4 96.4

Chapter 6

150

6.3.3 Comparison of PM generators Fig. 6-3-8 depicts the active mass of five different PM generators which are RFPMG, TFPMG-1, TFPMG-2, TFPMG-3 and TFPMG-4 for 5 MW direct-drive wind turbines. Among these five generators, TFPMG-1 seems the lightest generator and TFPMG-4 seems the heaviest generator. The copper mass of RFPMG is larger than TFPM generators. Fig. 6-3-9 depicts the losses of the five different generators. TFPMG-4 has the largest loss and TFPMG-2 has the smallest loss among the five different generators. The flux density in the stator cores of TFPMG-3 is higher than the flux density in the stator cores of RFPMG. Therefore, the stator core loss of TFPMG-3 is larger than the stator core loss of RFPMG, even though stator core mass of TFPMG-3 is smaller than the stator core mass of RFPMG. Fig. 6-3-10 depicts the cost of the five different generators. It shows that TFPMG-2 is the cheapest generator, RFPMG is the 2nd cheapest generator and TFPMG-4 is the most expensive generator.

Generator type_5MW

RFPMG TFPMG-1 TFPMG-2 TFPMG-3 TFPMG-4

Act

ive

mas

s [to

n]

0

10

20

30

40

50

60Copper massStator core massPM massRotor core massGenerator mass

Fig. 6-3-8: Active mass comparison of different 5 MW PM generators


151

Generator type_5MW


Loss

[kW

]

0

50

100

150

200

250Copper lossStator core lossRotor lossGenerator loss

Fig. 6-3-9: Loss comparison of different 5 MW PM generators

Generator type_5MW


Cos

t [kE

uro]

0

100

200

300

400

500

600

700

Copper costStator core costPM costRotor core costGenerator cost

Fig. 6-3-10: Cost comparison of different 5 MW PM generators

Chapter 6

152

Fig. 6-3-11 depicts the active mass of the five different PM generators for 10 MW direct-drive wind turbines. Among these five generators, RFPMG seems the lightest generator, TFPMG-1 seems the 2nd lightest generator and TFPMG-4 seems the heaviest generator. The copper mass of RFPMG is larger than TFPM generators. Fig. 6-3-12 depicts the losses of the five different generators. TFPMG-4 has the largest loss, and TFPMG-1 and TFPMG-2 has the smallest loss among the five different generators. Fig. 6-3-13 depicts the cost of the five different generators. It shows that TFPMG-2 is the cheapest generator, RFPMG is the 2nd cheapest generator and TFPMG-4 is the most expensive generator.

Generator type_10MW


Act

ive

mas

s [to

n]

0

20

40

60

80

100

120

140

160Copper massStator core massPM massRotor core massGenerator mass

Fig. 6-3-11: Active mass comparison of different 10 MW PM generators


153

Generator type_10MW


Loss

[kW

]

0

100

200

300

400

500

Copper lossStator core lossRotor lossGenerator loss

Fig. 6-3-12: Loss comparison of different 10 MW PM generators

Generator type_10MW


Cos

t [kE

uro]

0

500

1000

1500

2000

Copper costStator core costPM costRotor core costGenerator cost

Fig. 6-3-13: Cost comparison of different 10 MW PM generators

Chapter 6

154

Fig. 6-3-14 depicts the efficiency and the force density of the five different generators for 5 MW and 10 MW direct-drive wind turbines. The efficiencies of those generators obtained through the analytical design are between 96.4 % and 97.4 %. The differences of efficiency among those generators are not large. Force densities of TFPM generators are between 65.2 and 71.5 [kN/m2], which are higher than the force density of RFPMG, 40.16 [kN/m2]. Fig. 6-3-15 depicts the cost/torque ratio and the mass/torque ratio of the five different generators. TFPMG-2 shows the highest cost-competitiveness among different generators for both 5 MW and 10 MW wind turbines. For 5 MW, the cost/torque ratio of TFPMG-2 is 49.4 [Euro/kNm], the ratio of RFPMG is 67.2 [Euro/kNm] and the ratio of TFPMG-1 is 70.3 [Euro/kNm]. The active mass/torque ratio of TFPMG-1 that is the lightest generator is 7.77 [kg/kNm], and the ratio of RFPMG that is the heaviest generator is 10 [kg/kNm]. For 10 MW, the cost/torque ratio of TFPMG-2 is 49.3 [Euro/kNm], the ratio of RFPMG is 53.6 [Euro/kNm] and the ratio of TFPMG-1 is 69.8 [Euro/kNm]. RFPMG is also addressed as the 2nd cheapest generator, but the difference between the ratios of RFPMG and TFPMG-1 is larger than the difference at 5 MW. The active mass/torque ratios of RFPMG, TFPMG-1, TFPMG-2, TFPMG-3 and TFPMG-4 are 7.48, 7.73, 9.07, 8.88 and 11.9 [kg/kNm], respectively.

Generator type


Effic

ienc

y [%

],Fo

rce

dens

ity [k

N/m

2 ]

0

20

40

60

80

100

120

140 Efficiency- 5 MW Efficiency-10MW Force density- 5MW Force density-10MW

Fig. 6-3-14: Comparison of efficiency and force density of different 5 MW and 10 MW PM generators


155

Generator type


Cos

t/Tor

que

[Eur

o/kN

m]

0

20

40

60

80

100

120

140

160

Mas

s/To

urqu

e [k

g/kN

m]

0

2

4

6

8

10

12

14

16Cost/Torque_5MW Cost/Torque_10MW Mass/Torque_5MW Mass/Torque_10MW

Fig. 6-3-15: Comparison of cost/torque and mass/torque of different 5 MW and 10 MW PM generators Table 6-3-8 gives an overview of comparison results of the five PM generators based on the criteria of active mass, cost, efficiency and force density. In the table, the strengths of the five generators are indicated with following marks. • ++ : very strong • + : strong • Δ : middle • - : weak • -- : very weak

Chapter 6

156

Table 6-3-8: Comparison of the five different PM generators for 5 MW and 10 MW direct-drive wind turbines RFPMG TFPMG-1 TFPMG-2 TFPMG-3 TFPMG-4

5 MW - ++ + Δ -- Active mass

10 MW ++ + - Δ --

5 MW + Δ ++ - -- Cost

10 MW + Δ ++ - --

5 MW - (96.6%)

+ (97.2%)

++ (97.4%)

Δ (96.8%)

-- (96.4%) Efficiency

10 MW Δ (97%)

++ (97.3%)

++ (97.3%)

- (96.9%)

-- (96.4%)

5 MW -- ++ ++ ++ + Force density

10 MW -- ++ ++ + -

6.4 Conclusions In order to assess different topologies of permanent magnet (PM) generators for large direct-drive wind turbines, a comparative design of different PM generators for 5 MW and 10 MW direct-drive wind turbines was discussed in this chapter. From the overview of different PM machines and the identification of the active mass-competitiveness of those machines in chapter 3, the following PM generators were selected for the comparative design in this chapter. • RFPMG: A slotted surface-mounted radial flux permanent magnet generator with full pitch

windings, inner rotor and rare earth magnets • TFPMG-1: A single-sided, single winding flux-concentrating transverse flux permanent

magnet generator with U-core • TFPMG-2: A double-sided, single winding flux-concentrating transverse flux permanent

magnet generator with C-core • TFPMG-3: A single-sided, single winding flux-concentrating transverse flux permanent

magnet generator with U-core and passive rotor • TFPMG-4: A double-sided, single winding flux-concentrating transverse flux permanent

magnet generator with C-core and passive rotor


157

Using the formulations and the analytical models developed in chapter 5, the selected five PM generators were electromagnetically designed for 5 MW and 10 MW direct-drive wind turbines. In the design, the electromagnetic dimensions and parameters of the generators were determined by the dimensions and parameters discussed in chapter 5. These five generators were assessed based on the criteria of active mass, loss, cost, efficiency and force density. From the comparative design, the following results were obtained: • TFPMG-1 was addressed as the lightest generator whose active mass is 78 [%] of the mass of

RFPMG for 5 MW wind turbines. • In the design of the generators for 10 MW wind turbines, RFPMG was addressed as the

lightest generator. TFPMG-1 was addressed as the second lightest generator whose active mass is 3.3 [%] larger than the mass of RFPMG.

• TFPMG-2 had the smallest loss and the lowest cost compared to the other generators for both 5 MW and 10 MW turbines.

• TFPMG-4 was addressed as the generator with the largest mass, the highest cost and the largest loss among the five different generators for both 5 MW and 10 MW turbines.

• TFPMG-3 and TFPMG-4 were more expensive than the other generators, since both generators need large mass of permanent magnets which are the most expensive active material.

• TFPMG-2 and TFPMG-4 were more complicated than the others to construct because these two generators have double-sided air gaps.

• Therefore, TFPMG-1 is selected as a suitable generator for large direct-drive wind turbines. In the next chapter, a new configuration of TFPMG-1 with multiple-modules will be discussed for large direct-drive wind turbines.

• In [Dub 2004] it was concluded that the TFPM machine with toothed rotor (T6 machine in chapter 3) was a valuable option in terms of the active mass and cost, if the air gap length can be kept below 1.5 mm. However, the design results in this chapter indicated that the conclusion in [Dub 2004] is not valid for all configurations of flux-concentrating TFPM machines.

Chapter 6

158

TFPM Machine with Multiple-Modules for Large Direct-Drive

159

Chapter 7


7.1 Introduction The aim of this chapter is to develop new configurations of large direct-drive wind generators that would enable active mass reduction and facilitate manufacture and maintenance. In the last chapter, five different permanent magnet generators for 5 MW and 10 MW direct-drive wind turbines were desined electromagnetically and compared based on active mass, loss, cost, efficiency and force density. Among the five different generators, the flux-concentrating transverse flux permanent magnet generator with single-sided, single winding and U-core configuration (TFPMG-1) was addressed as the lightest generator for 5 MW. The surface-mounted PM generator with full pitch windings (RFPMG) was address as the lightest generator for 10 MW. The flux-concentrating TFPM generator with double-sided, single winding and C-core configuration (TFPMG-2) had the smallest loss and the lowest cost of active material compared to other generators. However, the TFPMG-2 was more complicated than the others to construct because this generator has double-sided air gaps. This constructive difficulties result in the increase of manufacturing cost. Therefore, the TFPMG-1 is selected as a suitable generator for large direct-drive wind turbines, and the RFPMG is considered as a reference generator in the design. To make the TFPMG-1 more competitive in terms of the active mass, cost, efficiency and force density, new configurations of the TFPMG-1 are developed, and the generators are designed for 5 MW and 10 MW direct-drive wind turbines in this chapter. This chapter begins with a description of the new configuration of TFPMG-1 for large direct-drive wind turbines. The proposed TFPM generator consists of multiple-modules of rotor and stator. Secondly, an analytical design model of the proposed TFPM generator is developed, and the model is verified by the experiments of a downscaled TFPM generator. Next, the proposed TFPM generator is designed for 5 MW and 10 MW direct-drive wind turbines. In the design, the number of slots per phase is taken as a variable. The proposed TFPM generators with various numbers of slots are assessed based on active mass, cost, loss, efficiency and force density. The

Chapter 7

160

designed generators are also compared with the RFPMG and the TFPMG-1 discussed in chapter 5.

7.2 TFPM machine with modular structure Various configurations of transverse flux permanent magnet (TFPM) machines have been proposed and discussed in a number of references. TFPM machines with flux-concentrating configurations have higher force density which results in volume reduction and consequently mass reduction. Thus, the TFPM machine with flux-concentrating configuration is considered for large direct-drive wind generators. This section starts with a description that lists unsuitable configurations of flux-concentrating TFPM machines for large direct-drive wind turbines. Next, suitable configurations of flux-concentrating TFPM machines for large direct-drive wind turbines are listed. Furthermore, a new configuration of a flux-concentrating TFPM machine with multiple-modules is proposed for large direct-drive wind turbines. Conventional flux-concentrating TFPM machines have the following disadvantages: • TFPM machines with double-sided air gaps and double windings are complicated to construct. • Considering the winding structure of TF machines, mostly ring-shaped windings have been

used because they lead to lower copper losses and simpler construction. However, the ring-shaped windings with a large diameter are difficult to manufacture and repair.

• When enlarging PM machines, the electromagnetic dimensions of the machines are increased together with an increase in magnet size. Large size of magnets thus makes manufacture more difficult and increases the cost of the machines.

• In order to fix the magnets on the iron cores in the conventional configuration, bonding is widely used. However, when bonding magnets to affix iron cores, the magnets can detach as shown in Fig. 7-2-1(a). In order to avoid the detachment of the magnets, mechanical stacking with bolting discussed in chapter 5 [Ban 2007c] can be an alternative method of affixing the magnets. However, this mechanical stacking and bolting method seems unsuitable for a rotational machine because it is difficult to limit mechanical tolerance accumulated in tangential stacking.

• To increase the volume of magnets with maintaining the length of pole pitch, the height of magnets is increased together with the increase of the height of iron cores in a conventional flux-concentrating TFPM machine configuration as shown in Fig. 7-2-1(b). The increase in magnet volume results in an increase in the volume of iron cores, and consequently the mass and cost of the cores are also increased.


161

• TF machines have three-dimensional flux paths, thus their construction and manufacture are more complicated than that of longitudinal flux (LF) machines. Due to these disadvantages of TF machines, it would be difficult to achieve their mass production and cost-competitiveness compared to LF machines.

Unsuitable configurations of flux-concentrating TFPM machines for large direct-drive wind turbines described above are summarized as follows: • double-sided air gap • double windings • ring-shaped windings with a large diameter • large size of iron cores and magnets • using the bonding method to affix magnets • assembling magnets and rotor cores in tangential stacking • difficulties in mass production In order to overcome the disadvantages of the flux-concentrating TFPM machines with unsuitable configurations described above, the following configurations of the machines are proposed as suitable configurations for large direct-drive wind turbines: • flux-concentrating TFPM machine with single-sided and single winding configuration • a multiple-module configuration of TFPM machine with multiple-slots per phase to reduce

the active material by shortening flux paths instead of one-module configuration [Ban 2008a]: Electric machines with shorter flux paths enable to reduce the active material, since shorter flux paths result in material reduction by decreasing slot pitch and slot height as illustrated in Fig. 7-2-2.

• racetrack-shaped windings instead of ring-shaped windings: A poly phase transverse flux motor with racetrack-shaped windings was also proposed in [Gla 2002]. However, the end-winding length of the motor is longer than the winding length in the slot, thus its end-winding

(a) (b)

Fig. 7-2-1: Conventional PMs and iron cores configuration of the flux-concentrating TFPM machine

Chapter 7

162

loss is large. Therefore, a racetrack-shaped winding with short end-winding length is needed for large direct-drive TFPM machines.

• a claw pole configuration of a TFPM machine with an increased iron core area [Ban 2008a][Dub 2004] to produce higher induced voltage, which results in higher force density and lower mass/torque ratio as discussed in chapter 3. A generator with an increased iron core area is suitable for increasing the no-load voltage because the voltage is proportional to the iron core area as given in (7-2-1).

2p cslot core coree N B A fπ= (7-2-1)

where pe is the no-load voltage, cslotN is the number of conductors per slot, coreB is the flux

density in iron cores, coreA is the area of iron cores to link the flux, and f is the frequency.

• a configuration with segmented iron cores and segmented magnets in order to facilitate manufacture for large direct-drive generators

• modular structures of the rotor and stator in order to facilitate manufacturing and maintenance

Fig. 7-2-3 depicts a sketch of the proposed flux-concentrating TFPM machine with the configuration of single-sided, single winding, racetrack-shaped windings, claw poles, multiple-modules and multiple-slots per phase. In Fig. 7-2-3 the claw pole cores with blue lines are showing the stator cores. In order to shorten flux paths, a multiple-module configuration with multiple-slots per phase [Ban 2008a] is used. The yellow racetrack-shaped structure represents the copper winding. The blue hexahedra with black arrows represent the permanent magnets (PMs), and the white hexahedra between the PMs represent the flux-concentrating cores in the rotor. In order to facilitate manufacturing of the rotor with magnets and iron cores, a new configuration of magnets and iron cores is proposed in Fig. 7-2-4. The parts with grey colour in Fig. 7-2-4 are non-ferromagnetic parts to assemble magnets and iron cores. The configuration in Fig. 7-2-3 is

(a) (b) Fig. 7-2-2: Configurations with one slot and two slots per a phase


163

modified to the configuration segmented as Fig. 7-2-4(a). The magnet and iron core segments are rearranged as in Fig. 7-2-4(b). This new configuration allows for an increase in the volume of magnets while maintaining the pole pitch length without increasing the height of the iron cores as shown in Fig. 7-2-4(c). In order to facilitate manufacture and assembly of magnets and iron cores, the configuration in Fig. 7-2-4(d) is proposed as an alternative assembling method, using bolting. The non-ferromagnetic parts in Fig. 7-2-4(d) can be made easily by the extrusion or the drawing method in manufacturing. Therefore, this configuration makes easier mass-production of flux-concentrating TFPM machines. The configuration proposed in Fig. 7-2-4 can also be used for longitudinal flux PM machines.

+

+=

Fig. 7-2-3: New TFPM machine with flux-concentrating configuration

Chapter 7

164

4

32

1

5 6

1

(a) (b)

(c) (d)

Fig. 7-2-4: New PMs and iron cores configuration of the flux-concentrating TFPM machine

7.3 Analytical modelling of TFPM generator with multiple-modules A sketch of the proposed TFPM generator with two slots per phase is illustrated in Fig. 7-3-1. Fig. 7-3-2 depicts the tangential and axial views of the generator with dimensional parameters. The dotted lines in Fig. 7-3-2 represent the main flux paths produced by the PM magneto-motive force. The electromagnetic dimensions and parameters of the proposed TFPM generator are determined by Table 5-2-2 in Chapter 5.


165

Fig. 7-3-1: 3D sketch of multiple-module TFPM generator with two slots per phase

pτ

ml

spb

rpb

Sh

sb

sh

Rh

gl

Spl

stb

(a) Cross-section in tangential view (b) Cross-section in axial view

Fig. 7-3-2: Cross-section view of multiple-module TFPM generator with two slots per a phase

Chapter 7

166

Electromagnetic reluctances in every pole pair are the same and repetitive. Electromagnetic reluctances in a pole are symmetrical with the reluctances in the next pole. Therefore, the equivalent circuit of electromagnetic reluctances in one pole is considered for the analytical model. Fig. 7-3-3 illustrates the equivalent circuits of the reluctance model of the TFPM generator. The white rectangles represent iron core reluctances, and the white rectangles with bold lines represent air gap reluctances. The blue rectangles hatched represent PM reluctances and the red rectangles dotted represent leakage flux reluctances. In order to formulate the flux equations of the equivalent circuit in Fig. 7-3-3, the equivalent circuit is modified as in Fig. 7-3-4. The flux densities, the flux, the flux linkages in the air gap, the PM and the iron cores are determined by the calculation procedure described in chapter 5.

.8mF

.10mF

7R6R

5bR

5aR

4R

3R

2R

1R

8R

9R

10R

11R12R

13aR

13bR

14R

15R

17R

16R

20R

21R

31R

30R

28R

29R

35R

34R27R

36R



: PM reluctance

: Leakage flux reluctance

Fig. 7-3-3: Equivalent circuit of magnetic reluctances of the proposed TFPM generator


167

.82

mF

82

R

7R

6R

5bR

5aR

4R

3R

2R 1R

.82

mF

82

R 9R

.102

mF

102

R

.102

mF

102

R

11R

12R

13aR

13bR

14R

15R

17R 16R

20R

21R

31R

30R

28R

29R

35R

34R

27R 36R

AφBφ

Kφ

Iφ

Cφ

Jφ

Gφ

Hφ

Dφ

Eφ

Fφ

Lφ

(a)

.82

mF

82

R

7R

6R

5bR

5aR

4R

3R

2R 1R

.82

mF

82

R 9R

.102

mF

102

R

.102

mF

102

R

11R

12R

13aR

13bR

14R

15R

17R 16R

20R

21R

31R

30R

28R

29R

35R

34R

27R 36R

1Φ 2Φ

.82

mF

82

R

7R

6R

5bR

5aR

4R

3R

2R 1R

.82

mF

82

R 9R

.102

mF

102

R

.102

mF

102

R

11R

12R

13aR

13bR

14R

15R

17R 16R

20R

21R

31R

30R

28R

29R

35R

34R

27R 36R3Φ

(b) (c)

.82

mF

82

R

7R

6R

5bR

5aR

4R

3R

2R 1R

.82

mF

82

R 9R

.102

mF

102

R

.102

mF

102

R

11R

12R

13aR

13bR

14R

15R

17R 16R

20R

21R

31R

30R

28R

29R

35R

34R

27R 36R4Φ

.82

mF

82

R

7R

6R

5bR

5aR

4R

3R

2R 1R

.82

mF

82

R 9R

.102

mF

102

R

.102

mF

102

R

11R

12R

13aR

13bR

14R

15R

17R 16R

20R

21R

31R

30R

28R

29R

35R

34R

27R 36R

5Φ

(d) (e)

Fig. 7-3-4: Modified equivalent circuit of magnetic reluctances of the proposed TFPM generator

Chapter 7

168

In order to determine the fluxes Aφ , Bφ , Cφ , Dφ , Eφ , Fφ , Gφ , Hφ , Iφ , Jφ , Kφ , and Lφ in Fig. 7-3-

4(a), Kirchhoff’s voltage law is applied to the fluxes 1Φ , 2Φ , 3Φ , 4Φ and 5Φ in Fig. 7-3-4(b), (c), (d) and (e). The procedure and equations to calculate the magnetic reluctances, the fluxes, the flux linkage and the no-load voltage in the equivalent circuits in Fig. 7-3-4 are described in Appendix 2.

7.4 Verification of magnetic circuit analysis model This section discusses the verification of the magnetic circuit analysis model discussed in the last section, 7.3. To verify the analysis model at no-load, the no-load induced voltage obtained through the analysis model is compared with the no-load voltage obtained through the measurement of a downscaled TFPM generator. To validate the analysis model at a load, the force of the generator obtained through the analysis model is compared with the force obtained through the static force measurement. The electromagnetic dimensions and parameters of the downscaled TFPM generator are given in Table 7-4-1. These dimensions and parameters were determined by Table 5-2-2 in chapter 5. Material characteristics of the TFPM generator are given in Table 7-4-2. Table 7-4-1: Electromagnetic dimensions and parameters of downscaled TFPM generator with multiple-modules Air gap length, gl 4 [mm]

Pole pitch, pτ 40 [mm]

Magnet height, ml 8 [mm]

Stator pole width, pb 32 [mm]

Rotor pole width, prb 24 [mm]

Number of conductors per slot, cslotN 576 [Turn]

Stator slot width, sb 30 [mm]

Stator slot height, sh 30 [mm] Number of pole pairs, p 40 [-] Stator pole length, spl 20 [mm]

Stator height, Sh 70 [mm]

Stator yoke height, syh 20 [mm]

Rotor height, Rh 20 [mm]


169

Table 7-4-2: Material characteristics of TFPM machine

Iron core type Solid core (S20c)





Iron core, Feρ 7800 [kg/m3]



The downscaled TFPM generator is built as in Fig. 7-4-1, Fig. 7-4-2 and Fig. 7-4-3. The generator consists of multiple-sets of the stator and rotor. Considering easier manufacturing of the generator, a solid iron core is used to construct both the stator and the rotor. Fig. 7-4-1 depicts segmented rotor cores and magnets, an assembly process of the cores and magnets, and a set of assembled rotor. Fig. 7-4-2 depicts a set of stator core, a racetrack-shaped winding, and a set of assembled stator. Fig. 7-4-3 depicts the TFPM generator with structural components, rotor and stator sets.

Chapter 7

170

(a) Segmented cores with an aluminium plate for assembly

(b) Assembly process of the magnets and the segmented cores (c) A segmented core

(d) A set of assembled rotor cores and magnets

(e) Assembly process of the rotor

Fig. 7-4-1: Rotor cores and magnets with segmented construction of the proposed TFPM generator


171

(a) Stator core (b) Racetrack-shaped copper winding

(c) A set of assembled stator

Fig. 7-4-2: The set of stator cores and racetrack-shaped winding of the proposed TFPM generator

Chapter 7

172

(a) Structural components with rollers

(b) Front view of the generator without stator

Stator

U

UV

WV

W

Rotor

Rollers

(c) Generator with rotor and stator Fig. 7-4-3: Proposed generator with sets of rotor and stator


173

7.4.1 Verification of no-load case A. Analytical results The peak flux density and the peak no-load voltage at 0.25 m/s air-gap speed by the analysis model of the TFPM generator are given in Table 7-4-3. Table 7-4-3: Peak flux density and peak no-load induced voltage of the downscaled TFPM generator with multiple-modules by analytical model

Flux density in the stator core, pB∧

1.06 [T] at 2 [mm] air gap 0.93 [T] at 4 [mm] air gap

No-load induced voltage per two pole pairs, 2 _pole paire

∧

28.21 [V] at 2 [mm] air gap18.96 [V] at 4 [mm] air gap

B. Experimental results Fig. 7-4-4 depicts the experimental setup of the downscaled generator. The TFPM generator is driven by a motor drive set integrated into a gearbox. The specifications of the experimental setup are given in Table 7-4-4. Table 7-4-4: Specifications of experimental setup

Driving motor

- 3 phase, AC machine - Nominal power: 14.3 [kW] - Nominal speed: 2,600 [rpm] - Nominal torque: 52.7 [Nm]

Pulley & belt - 1st pulley diameter: 152.4 [mm] - 2nd pulley diameter: 304.8 [mm]

Gearbox 43:1 gear ratio

Generator diameter - Outer diameter: 1.3 [m] - Inner diameter: 1 [m]

Air gap length 2 & 4 [mm]

Fig. 7-4-5 depicts the measured no-load voltages of three phases at 4 mm air gap and at 0.25 m/s air gap speed. Table 7-4-5 gives the peak values of no-load voltages measured at 2 mm and 4 mm air gap and 0.25 m/s air gap speed.

Chapter 7

174

Stator

Rotor

Rollers

Power analyzer

Torque sensor

Gearbox

Driving motor

Motor driving unit

Fig. 7-4-4: Proposed TFPM generator with experimental setup

Fig. 7-4-5: No-load induced voltages measured at 0.25 m/s


175

Table 7-4-5: Peak no-load induced voltage of the downscaled TFPM generator with multiple-modules by measurements

No-load induced voltage per two pole pairs, 2 _pole paire

∧

27.93 [V] at 2 [mm] air gap18.73 [V] at 4 [mm] air gap

The peak no-load voltages of the generator with 2 [mm] and 4 [mm] air gap length measured at 0.25 [m/s] air gap speed are 1 [%] and 1.2 [%] lower than the voltages obtained through the analysis model. Therefore, the analysis model of the proposed TFPM generator is used for the design of the generator for large direct-drive wind turbines in the next section.

7.4.2 Verification in the case with a load In order to validate the analysis model of the proposed TFPM generator with a load, the force obtained through the analysis model is compared with the force obtained from static force measurements. Using a force equation (5.4.1) in chapter 5, the force of the proposed TFPM generator is calculated. To measure the thrust force of the proposed TFPM generator, a linearized type of the generator is built and equipped on a test bench as shown in Fig. 7-4-6. Fig. 7-4-7 depicts the mearsued thrust force per pole pair of the generator as a function of the rotor displacement. Due to the effect of the cogging force and the reluctance force, the sinusoidal distribution of the thrust force was distorted as shown in Fig. 7-4-7. Fig. 7-4-8 depicts the differences between the thrust force obtained through the static force measurements and the force obtained through the analytical model. In Fig. 7-4-8, it is indicated that the peak force measured at 25 % and 50 % of the nominal current is 5 % and 11 % lower than that obtained through the analytical model. During the static force measurements, it was not able to increase the current more than 50 % of the nominal current because of the current capacity limitation of the power supply.

Fig. 7-4-6: Proposed TFPM generator equipped on a test bench to measure the static force

Chapter 7

176

Displacement

0.0

pole

pitc

h

0.2

pole

pitc

h

0.4

pole

pitc

h

0.6

pole

pitc

h

0.8

pole

pitc

h

1.0

pole

pitc

h

Thru

st fo

rce

per p

ole

pair

[N]

-100

0

100

200

300

400

500

600

7004000 AT at lg=2 mm 2000 AT at lg=2 mm 0 AT at lg=2 mm

4000 AT at lg=4 mm 2000 AT at lg=4 mm 0 AT at lg=4 mm

Fig. 7-4-7: Thrust force per a pole pair of the proposed TFPM generator by the measurement


0 1x106 2x106 3x106 4x106

F peak

.mea

sure

men

t / F

peak

.ana

lysi

s [-]

0.0

0.2

0.4

0.6

0.8

1.0

1.2

1.4

Fpeak.measurement/Fpeak.analysis in Chapter 5 Fpeak.FEA/Fpeak.analysis in Chapter 5 Fpeak.measurement/Fpeak.Analysis at lg=4mm

Fig. 7-4-8: Thrust force differences between the measurement and the analysis


177

7.5 Design of TFPM generators with multiple-modules for large direct-drive wind turbines Using the formulations and the analytical models derived in the last section, 7.4, the proposed TFPM generator with multiple-modules and multiple-slots per phase is designed for 5 MW and 10 MW direct-drive wind turbines in this section. In the design the number of slot per phase is variable as 1, 2, 4 and 8. The TFPM generators designed with various numbers of slots per phase are assessed based on active mass, cost, loss, efficiency and force density. Design results of the generators are also compared with a surface-mounted RFPM generator and a flux-concentrating TFPM generator, namely the RFPMG and the TFPMG-1 discussed in chapter 6. Wind turbine parameters and generator requirements for 5 MW and 10 MW wind turbines are given in Table 7-5-1. In the design of the TFPM generators with multiple-modules, material characteristics and cost models used for the generators are given in Table 7-5-2. Table 7-5-1: Wind turbine parameters and generator requirements

Wind turbine parameters

Rated grid power, P 5 [MW] 10 [MW]

Rotor blade diameter, rD 126 [m] 178 [m]

Rotor blade tip speed, tipv 80 [m/s] 80 [m/s]

Rated rotor speed, N 12.1 [rpm] 8.6 [rpm]


Nominal power, gennomP 5.56 [MW] 11.12 [MW]

Nominal torque, gennomT 4.38 [MNm] 12.38 [MNm]

Chapter 7

178

Table 7-5-2: Material characteristics and cost models for generator

Material characteristics

Specific hysteresis losses of iron cores


7.93 [W/kg] at 1.3 [T] and 50 [Hz]

Specific eddy current losses of iron cores,


0.17 [W/kg] at 1.3 [T] and 50 [Hz]



Relative recoil permeability of permanent magnets,

rmμ 1.05 [-]


Iron core, Feρ SMC core: 7440 [kg/m3]



Cost models

Iron core cost, Fek 3 [€/kg]

Copper cost, Cusk 15 [€/kg]

Permanent magnet cost, pmk 25 [€/kg]

In the analytical design for the proposed generator, the following parameters are used as input parameters. (1) nominal power, gennomP [MW]

(2) rotational speed, N [rpm] (3) number of phases, phm [ - ]

(4) power factor, cosφ [ - ] (5) diameter of rotor, gD [m]

(6) nominal current, sI [A]

(7) RMS value of no-load voltage, pe [V]

(8) current density, sJ [A/mm2]

(9) slot filling factor of stator conductors, sfillk [ - ]


179

(10) remanent flux density of permanent magnets, rmB [T]

(11) relative recoil permeability of permanent magnets, rmμ [ - ]

(12) permeability of free space, 0μ [H/m] (13) B-H curve data of iron cores (14) axial length per three phases, sl [mm]

(15) number of slots per phase, modulem [ - ]

(16) width of stator tooth, tb [mm]

(17) height of stator yoke, syh [mm]

(18) length of stator pole, spl [mm]

Among the input parameters listed above, the values of the parameters from (1) to (13) are the same with the TFPMG-1. Thus the parameters from (1) to (13) are kept constant in the design of the proposed TFPM generator. The parameter (14) axial length sl is determined by two cases of assumptions in the design: • Case-1: The axial length of the proposed TFPM generator is same as the axial length of the

TFPMG-1. • Case-2: The pole area of the proposed TFPM generator is same as the pole area of the

TFPMG-1. The parameter (15) number of slots per phase modulem is variable as 1, 2, 4 and 8. The parameters

(16) width of stator tooth stb and (17) height of stator yoke syh of the proposed TFPM generator

are obtained by dividing stb and syh of TFPMG-1 with the number of slots per phase modulem . The

parameter (18) length of stator pole spl is determined by 65 % [Mad 1998][Mas 2004]of the

width of a stator module. In the design procedure of the proposed generators under the limited design condition, the following geometric parameters are adjusted in order to obtain the required no-load induced voltage of the generators. 1. height of rotor, Rh [mm]

2. height of magnet, ml [mm] In the design of the proposed TFPM generator with multiple-modules and multiple-slots per phase, the following configurations of the generator are considered. • CP_1mod_ls: single-sided, single winding flux-concentrating TFPM generator with claw poles,

one slot per phase and limited axial length (Case-1) • CP_2mod_ls: single-sided, single winding flux-concentrating TFPM generator with claw poles,

two slots per phase and limited axial length (Case-1)

Chapter 7

180

• CP_4mod_ls: single-sided, single winding flux-concentrating TFPM generator with claw poles, four slots per phase and limited axial length (Case-1)

• CP_8mod_ls: single-sided, single winding flux-concentrating TFPM generator with claw poles, eight slots per phase and limited axial length (Case-1)

• CP_1mod_Ap: single-sided, single winding flux-concentrating TFPM generator with claw poles, one slot per phase and limited pole area (Case-2)

• CP_2mod_Ap: single-sided, single winding flux-concentrating TFPM generator with claw poles, two slots per phase and limited pole area (Case-2)

• CP_4mod_Ap: single-sided, single winding flux-concentrating TFPM generator with claw poles, four slots per phase and limited pole area (Case-2)

• CP_8mod_Ap: single-sided, single winding flux-concentrating TFPM generator with claw poles, eight slots per phase and limited pole area (Case-2)

In order to identify the competitiveness of the proposed TFPM generators listed above, the design results of the generators are compared with the results of the following PM generators discussed in chapter 6. • RFPMG: a surface-mounted TFPM generator with full pitch windings • TFPMG-1: a single-sided, single winding flux-concentrating TFPM generator with a U-core Table 7-5-3 and Table 7-5-4 give the design results of parameters (15), (16), (17), (18) and (19) of the proposed 5 MW TFPM generators with limited axial length (Case-1) and limited pole area (Case-2), respectively. Design results of the parameters of the proposed 10 MW TFPM generators are given in Table 7-5-5 and Table 7-5-6. The electromagnetic dimensions and parameters of these TFPM generators were determined by Table 5-2-2 in chapter 5. Table 7-5-3: Number of slots per phase, width of stator tooth, height of stator yoke, length of stator pole of 5 MW generators with limited axial length (Case-1)

number of modules per slot, modulem [-] 1 2 4 8

axial length per three phases, sl [m] 1.03 1.03 1.03 1.03

width of stator tooth, stb [mm] 130.4 65.2 32.6 16.3

height of stator yoke, syh [mm] 130.4 65.2 32.6 16.3

length of stator pole, spl [mm] 223.2 111.6 55.8 27.9

height of rotor, Rh [mm] 33.8 33.8 34.7 36.8


181

Table 7-5-4: Number of slots per phase, width of stator tooth, height of stator yoke, length of stator pole of 5 MW generators with limited pole area (Case-2)



width of stator tooth, tb [mm] 52.2 26.1 13.1 6.5




Table 7-5-5: Number of slots per phase, width of stator tooth, height of stator yoke, length of stator pole of 10 MW generators with limited axial length (Case-1)







Table 7-5-6: Number of slots per phase, width of stator tooth, height of stator yoke, length of stator pole of 10 MW generators with limited pole area (Case-2)







Chapter 7

182

Fig. 7-5-1 depicts the active mass of ten different PM generators for 5 MW direct-drive wind turbines. In the figure, copper mass, iron core mass, magnet mass and total active mass are represented. Among the ten different PM generators, CP_1mod_Ap seems to be the lightest generator and CP_8mod_ls seems to be the heaviest generator.

Generator type

RFP

MG

TFPM

G-1

CP

_1m

od_l

s

CP

_2m

od_l

s

CP

_4m

od_l

s

CP

_8m

od_l

s

CP

_1m

od_A

p

CP

_2m

od_A

p

CP

_4m

od_A

p

CP

_8m

od_A

p

Act

ive

mas

s [to

n]

0

20

40

60

80

100

Copper massStator core massPM massRotor core massGenerator mass

Fig. 7-5-1: Active mass of different PM generators for 5 MW direct-drive wind turbines Fig. 7-5-2 depicts the competitiveness of the different PM generators in terms of efficiency, force density, cost and active mass. From the results, it is taken that CP_1mod_ls has the highest efficiency (98.6 %) and CP_8mod_Ap has the lowest efficiency (90.2 %). The TFPM generators with claw poles and limited pole area have higher force density than the other generator. In terms of cost, CP_1mod_Ap seems cheaper than the other generators. CP_8mod_Ap seems to be the most expensive configuration. Fig. 7-5-3 depicts the active mass of the ten different PM generators for 10 MW direct-drive wind turbines. Among the ten different PM generators, CP_1mod_Ap and CP_2mod_Ap are addressed as the lightest generator and the second lightest generator, respectively. CP_8mod_ls seems to be the heaviest generator.


183

Generator type

RFP

MG

TFPM

G-1

CP_

1mod

_ls

CP_

2mod

_ls

CP_

4mod

_ls

CP_

8mod

_ls

CP_

1mod

_Ap

CP_

2mod

_Ap

CP_

4mod

_Ap

CP_

8mod

_Ap

Effic

ienc

y co

mpe

titiv

enes

s [-

], Fo

rce

dens

ity c

ompe

titiv

enes

s[-],

C

ost/T

orqu

e co

mpe

titiv

enes

s [-]

0.0

0.5

1.0

1.5

2.0

2.5

3.0

Mas

s/To

urqu

e co

mpe

titiv

enes

s [-]

0.0

0.5

1.0

1.5

2.0

2.5

3.0

Efficiency Force density Cost/Torque Mass/Torque

Fig. 7-5-2: Competitiveness of efficiency, force density, cost/torque ratio and mass/torque ratio of different PM generators for 5 MW direct-drive wind turbines

Generator type

RFP

MG

TFPM

G-1

CP_

1mod

_ls

CP_

2mod

_ls

CP_

4mod

_ls

CP_

8mod

_ls

CP_

1mod

_Ap

CP_

2mod

_Ap

CP_

4mod

_Ap

CP_

8mod

_Ap

Activ

e m

ass

[ton]

0

50

100

150

200

250

Copper massStator core massPM massRotor core massGenerator mass

Fig. 7-5-3: Active mass of different PM generators for 10 MW direct-drive wind turbines

Chapter 7

184

Fig. 7-5-4 depicts the competitiveness of different 10 MW PM generators in terms of efficiency, force density, cost and mass. According to the results, CP_1mod_ls seems to be the configuration with the highest efficiency and CP_8mod_Ap has the lowest efficiency. The TFPM generators with claw poles and limited pole area have higher force density than the other configurations. In terms of cost, CP_2mod_ls has the highest competitiveness. CP_8mod_Ap seems to be the most expensive generator configuration.

Generator type

RFP

MG

TFPM

G-1

CP

_1m

od_l

s

CP

_2m

od_l

s

CP

_4m

od_l

s

CP

_8m

od_l

s

CP

_1m

od_A

p

CP

_2m

od_A

p

CP

_4m

od_A

p

CP

_8m

od_A

p

Effic

ienc

y co

mpe

titiv

enes

s [-

], Fo

rce

dens

ity c

ompe

titiv

enes

s[-],

C

ost/T

orqu

e co

mpe

titiv

enes

s [-]

0.0

0.5

1.0

1.5

2.0

2.5

3.0

Mas

s/To

urqu

e co

mpe

titiv

enes

s [-]

0.0

0.5

1.0

1.5

2.0

2.5

3.0Efficiency Force density Cost/Torque Mass/Torque

Fig. 7-5-4: Competitiveness of efficiency, force density, cost/torque ratio and mass/torque ratio of different PM generators for 10 MW direct-drive wind turbines Table 7-6-1 gives an overview of competitiveness of different 5 MW and 10 MW PM generators based on the criteria of active mass, cost, efficiency and force density. In the table, the generator with the highest competitiveness is indicated with “9”, and the generator with the lowest competitiveness is indicated with “0”.


185

Table 7-6-1: Comparison of the eleven PM generators for 5 MW and 10 MW direct-drive wind turbines

RF

PMG

TFPM

G-1

CP_

1mod

_ls

CP_

2mod

_ls

CP_

4mod

_ls

CP_

8mod

_ls

CP_

1mod

_A

CP_

2mod

_A

CP_

4mod

_A

CP_

8mod

_A

Active mass 4 7 3 5 2 0 9 8 6 1

Cost 5 3 7 8 4 1 9 6 2 0

Efficiency 4 5 9 7 3 1 8 6 2 0 5 MW

Force density 0 5 2 3 1 4 9 9 9 9

Active mass 6 5 1 4 3 0 9 8 7 2

Cost 7 3 8 9 5 1 6 4 2 0

Efficiency 4 5 9 7 3 1 8 6 2 0 10 MW

Force density 0 5 4 1 3 3 9 9 9 6

7.6 Conclusions This chapter dealt with a new configuration of large direct-drive wind generators that would enable active mass reduction and facilitate manufacture and maintenance. A flux-concentrating transverse flux permanent magnet (TFPM) machine with multiple-modules segmented, multiple-slots per phase and racetrack-shaped copper windings was proposed to decrease the flux path length. An analytical design model of the proposed TFPM generator was developed, and a downscaled TFPM generator was built to verify the analytical model. The no-load voltages measured at two different air gap lengths (2 mm and 4 mm) were 1 % to 1.2 % lower than the voltage obtained by the analytical model. To validate the analytical model in the case with a load, the force obtained by the analytical model was compared with the force obtained through measurement. The peak force measured at 25 % and 50 % of the nominal current was 5 % and 11 % lower than that obtained through the analytical model. The analytical design model was used for the design of proposed TFPM generators for 5 MW and 10 MW direct-drive wind turbines. In the analytical design, the number of slots per phase was variable as 1, 2, 4 and 8. The TFPM generators designed with various numbers of slots per phase

Chapter 7

186

were assessed based on active mass, cost, loss, efficiency and force density. The following ten different PM generator configurations were discussed in the analytical design. • RFPMG: a surface-mounted TFPM generator with full pitch windings • TFPMG-1: a single-sided, single winding flux-concentrating TFPM generator with a U-core • CP_1mod_ls: single-sided, single winding flux-concentrating TFPM generator with claw poles,

one slot per phase and limited axial length • CP_2mod_ls: single-sided, single winding flux-concentrating TFPM generator with claw poles,

two slots per phase and limited axial length • CP_4mod_ls: single-sided, single winding flux-concentrating TFPM generator with claw poles,

four slots per phase and limited axial length • CP_8mod_ls: single-sided, single winding flux-concentrating TFPM generator with claw poles,

eight slots per phase and limited axial length • CP_1mod_Ap: single-sided, single winding flux-concentrating TFPM generator with claw

poles, one slot per phase and limited pole area • CP_2mod_Ap: single-sided, single winding flux-concentrating TFPM generator with claw

poles, two slots per phase and limited pole area • CP_4mod_Ap: single-sided, single winding flux-concentrating TFPM generator with claw

poles, four slots per phase and limited pole area • CP_8mod_Ap: single-sided, single winding flux-concentrating TFPM generator with claw

poles, eight slots per phase and limited pole area Among these different PM generators, CP_1mod_Ap was addressed as the lightest generator, CP_2mod_Ap was addressed as the second lightest generator and CP_8mod_ls was addressed as the heaviest generator for 5 MW and 10 MW wind turbines. From the design of the generators for 5 MW wind turbines, it was taken that the active mass of TFPMG-1, CP_2mod_ls, CP_1mod_Ap, CP_2mod_Ap and CP_4mod_Ap are smaller than that of RFPMG. The active mass of CP_1mod_Ap, CP_2mod_Ap and CP_4mod_Ap seems to be lighter than that of RFPMG for 10 MW wind turbines. In the chapter, the current density of the proposed TFPM generators was kept constant for various numbers of slots per phase. This represents the length of flux paths of the generators with large numbers of slots per phase was not shorter than the length of generators with small numbers of slots per phase. Therefore, the mass of the TFPM generator with eight slots per phase was larger than that of the gnerators with one slot per phase. To reduce the length of flux paths further more, increasing current density would be an alternative. To increase current density, cooling effect of the machine must be also increased.

Challenges of Minimizing the Structural Mass of Large Direct-Drive Wind Generators

187

Chapter 8


8.1 Introduction This chapter describes some solutions to the problems tied to large direct-drive wind generators, which are heavy and expensive, while they also have a higher energy yield than geared generators. The previous chapters in the thesis mainly discussed the electromagnetic design of large direct-drive wind generators to minimize the mass and cost of the active part. In case of large direct-drive wind generators, the structural mass is dominant in the total mass of the generators as discussed in chapter 4. Therefore, it is necessary to significantly reduce the mass of the structural part of large direct-drive wind generators in order to make those generators more attractive for industry use. The functions of the structural part of direct-drive wind generators could be defined as follows: • To maintain the air gap between the rotor and the stator • To transmit the torque from the rotor blades to the generator rotor If it is possible to perform the functions of the structural part with minimized mass and cost, this could be a solution to the problems with large direct-drive wind generators. In [Ban 2008a], the following generator concepts were discussed as promising use in large direct-drive wind turbines. • A generator without both a main shaft and torque arms • A generator with modular structures that are easy to manufacture, assemble, transport, install

and repair • A generator with flexible and very lightweight joint and guide systems to take the rotational

force from the rotor blades to the generator rotor

Chapter 8

188

• A multi-set generator of which each set can work individually: That means the generator can continue to produce electricity, even if a few parts fail.

(a)

(b)

Fig. 8-1: Lightweight and modular direct-drive generator concept for large wind turbine [Ban 2008a]


189

Fig. 8-1 illustrates the generator concept without a main shaft and torque arms discussed in [Ban 2008a]. The concept uses a large bearing to guide and support the rotor blades. Some auxiliary rollers are used to avoid further damage caused by the touchdown of the rotor on the stator. a detailed tangential view of the generator, including the auxiliary rollers and the large bearing, is illustrated in Fig. 8-1(b). As represented in Fig. 8-1, the generator is constructed without shaft and torque arms, thus it would be possible to significantly reduce its structural mass. If the generator concept with a large bearing and auxiliary rollers in Fig. 8-1 needs very small tolerance in constructing, the concept will not be attractive as far as cost is concerned even despite a lightweight structure. Therefore, the following two assumptions must be fulfilled to simultaneously minimize both the structural mass and the cost of large direct-drive wind generators: • A construction without a shaft and without torque arms • A constructon which does not require very small tolerance This chapter focuses on finding new concepts of large direct-drive wind generators which could fulfil the above assumptions. Firstly, a ring-shaped generator without a shaft and without torque arms is proposed. The generator consists of multi-sets of the rotor and the stator with multi-sets of the converter. Secondly, a buoyant generator rotor is proposed to reduce the structural mass in supporting the mass of the generator rotor. Next, a bearingless drive and hydraulic bearings, which do not require very small tolerance, are proposed to maintain the air gap between the rotor and the stator. Furthermore, the mass of the new direct-drive generators, integrated with ring-shaped construction and the buoyant rotor, is estimated for large wind turbines up to 20 MW. In order to identify the strengths of the proposed new generator, the estimated mass is compared with the mass of different generators currently in use.

8.2 Ring-shaped generator concept with multi-sets Conventional direct-drive generators have disadvantages such as large mass and difficulties in assembly, transport, installation and repair. These disadvantages raise their cost. When a component of a conventional generator fails, the generator systems cease operation. The failure of large scale generator systems is more serious than the failure of small scale generator systems. In order to overcome these disadvantages of large direct-drive generators, the following concepts are proposed in this section. • A ring-shaped machine without a shaft and without torque arms • Multi-sets of rotor and stator which are easy to assemble, transport, install and repair

Chapter 8

190

• Multi-sets of three-phases generators to continue electricity production in case of failure in a few components

Fig. 8-2 illustrates a conceptual drawing of the ring-shaped generator with multi-sets. The ring-shaped generator does not have a shaft and torque arms. In order to make the generator structure stable against the normal force between the rotor and the stator, the generator is constructed to have double-sided air gaps. Fig. 8-3 depicts the conceptual construction of a ring-shaped generator with multi-sets of 3-phase double-sided configurations. Multi-sets of the generator’s converter system are represented in Fig. 8-4, which shows that each stator set has its own converter set. The numbers of the rotor and stator sets can be changed depending on the applications.

Fig. 8-2: A ring-shaped direct-drive generator concept with multi-sets


191

Z

YX

Y

Z

XY

Z

YX

A A’

A A’

Cro

ss se

ctio

n vi

ew

A-A

’

1-1

1-2

2-1

2-2

3-1

3-2

4-1 4-

2

8-1 8-

2

7-1

7-2

6-1

6-2

5-1

5-2

1-1 1-

2

2-1

2-2

3-1

3-2

4-1

4-2

8-1

8-2

7-1

7-2

6-1

6-2

5-1 5-

2

Stat

or_1

Stat

or_2

Rot

or

Fi

g. 8

-3: R

ing-

shap

ed d

irect

-dri

ve g

ener

ator

con

stru

ctio

n

Chapter 8

192

8.3 New supporting and guiding concepts Mechanical bearings are mostly used to support and guide direct-drive wind generators. Conventional mechanical bearings have the following disadvantages.

Stator_1 Stator_2

1-1

1-2

1-1

1-2

Stator_1 Stator_2

8-1

8-2

8-1

8-2

2-1 & 2-2

3-1 & 3-2

4-1 & 4-2

5-1 & 5-2

6-1 & 6-2

7-1 & 7-2

2-1 & 2-2

3-1 & 3-2

4-1 & 4-2

5-1 & 5-2

6-1 & 6-2

7-1 & 7-2

GRID

Fig. 8-4: Power converter construction of the ring-shaped new bearingless PM generator with multi-sets


193

• Very small tolerance is required in their construction. The housing of conventional mechanical bearings is generally constructed with a one-body structure to reduce wear and to fulfil the tolerance requirements. Thus using conventional bearings for large-diameter applications increases difficulties in manufacturing, transport, installation and maintenance.

• Regular maintenance is required because of the wear on the solid surface from physical contact.

• The disadvantages summarized above increase the cost of the bearings. A supporting and guiding system without physical contact would provide higher flexibility, allowing the disadvantages of conventional bearing systems to be overcome. Therefore, this section focuses on finding new concepts of supporting and guiding systems with high flexibility and without physical contact. In order to provide support for a heavy structure without physical contact, this section proposes the use of buoyancy. To support the mass of the rotating part, the rotor of the generator is constructed with air chambers inside and the gap between the rotor and the stator is then filled with fluid. Using this buoyant rotor concept has the added advantage of not consuming energy to support the mass of the generator rotor. To guide the air gap of the generator without physical contact, the use of bearingless drive and/or hydraulic bearings, instead of the mechanical bearings, is proposed. In the proposed concept with a buoyant rotor, applying a bearingless drive and setting a hydraulic bearing are straightforward because there is no bearing force necessary to support the rotor mass and the fluid exists in the gap. When large direct-drive wind generators rotate at high speed, filling the air gap with fluid and using a hydraulic bearing is not be a satisfactory solution becaused of the involved relatively high friction and large eccentricity. However, large direct-drive wind generators are operated at very low speed. Considering this, the use of a buoyant rotor concept and relatively larger tolerances remain acceptable for large direct-drive wind generator applications. Therefore, new supporting and guiding systems with a buoyant rotating structure, a simplified bearingless drive and a hydraulic bearing are discussed in this section.

8.3.1 Buoyant rotating (moving) structure Using the buoyancy force, a very heavy structure floats easily in fluid. Therefore, the use of the buoyancy force for supporting the heavy rotating (moving) structure of large direct-drive wind generator has been introduced in this section. The advantages of a buoyant rotating structure could be summarized as follows. • easily supports a heavy structure without consuming energy

Chapter 8

194

• reduces structural material required for support • provides flexibility in supporting a heavy structure Fig. 8-5 depicts a simplified cross-sectional shape of a buoyant structure both in the stable state and in the unstable state.

For large direct-drive applications, the buoyant rotating part and the stationary part could be constructed as shown in Fig. 8-6. There is air in the chamber of the rotating part. Fluid fills the air gap between the buoyant rotating part and the stationary part. The iron cores and permanent magnets for the generator rotor are set on the rotating part. The iron cores and copper windings for the generator stator are set on the stationary part. In order to connect the generator rotor with the rotor blades of the wind turbine, the rotor has a protruding part on the outer side. The protruding part of the buoyant rotor can also be positioned on the inner diameter side or on the rotor blade side. In order to avoid leakage of fluid, some sealing systems must be included between the protruding part of the rotating part and the stationary part. Fig. 8-7 depicts a sketch of a sealing mechanism and a sketch of a wind turbine concept integrated with the proposed generator. A sealing mechanism using a flexible tube filled with compressed air could be a solution to the fluid leakage problem as shown in Fig. 8-7(a). Fig. 8-7(b) illustrates an integration of the proposed generator with a buoyant rotor and the rotor blade hub using flexible joints.

(a) Stable state (b) Unstable state Fig. 8-5: Buoyant structure in fluid


195

Fi

g. 8

-6: B

uoya

nt ro

tor c

once

pt fo

r lar

ge d

irect

-dri

ve

Chapter 8

196

(a) A sealing mechanism with a tube in which compressed air is filled

Flexible joints

Hub

Rotor blade

Bearings with small diameter

Hub

Rotor blade

Z

XYY

(b) Proposed generator with a buoyant rotor integrated with a rotor blade hub by flexible joints Fig. 8-7: Sketch of a sealing mechanism and an integration of the proposed generator with a buoyant rotor and a rotor blade hub


197

8.3.2 Bearingless drive To gain an understanding of the bearingless drive, the features of a conventional bearingless drive are described. Next, a new bearingless drive concept is proposed and described. A. Conventional bearingless drive Bearingless drives have been used to solve problems in certain special applications such as in outer space, harsh environments and in high speed machines. The various bearingless drives discussed and proposed by some authors [Chi 2005] can be classified as follows: (1) bearingless permanent magnet (PM) machine drives, (2) bearingless synchronous reluctance machine drives, (3) bearingless induction machine drives, (4) bearingless switched reluctance machine drives, and (5) bearingless homopolar, hybrid and consequent-pole machine drives. Of these bearingless drives, a bearingless permanent magnet (PM) machine drive has the following advantages: [Chi 2005]: (1) small size and lightweight, (2) high power factor and high efficiency, (3) suspension forces generated without excitation current in the main winding, (4) high inverter fault independence because magnetic suspension operates independently of the

machine winding current. Therefore, the bearingless permanent magnet machine drive concept is chosen for the research in the section. A significant feature of the bearingless drive compared to the electric machine with the magnetic bearing is that the bearing winding is integrated into the electric machine. Fig. 8-8(a) depicts the principle of the winding construction of the bearingless PM machine with four-poles. In order to achieve extensive decoupling between the generations of the radial bearing forces and the torque, the windings are designed with different numbers of poles. [Amr 2006] The winding arrangement of a primitive bearingless drive can be simplified as shown in Fig. 8-8(b), where the windings for four-poles are 4a and 4b and the windings for two-poles are 2a and 2b. Fig. 8-8(c) shows the rotor structure with inset-type permanent magnets. This structure consists of q-axis saliency and the PMs between the salient poles. Fig. 8-8(d) is a buried PM type rotor which improves radial suspension force generation because of the low PM magnetic reluctance [Chi 2005].

Chapter 8

198

As described above, bearingless drives need to control both the bearing force with bearing windings and the torque with torque windings. Therefore, bearingless drives are more complicated and expensive than the conventional electric machine drives. In the case of the large direct-drive machines, the mass of the rotating part is large. Thus it is expected that the power consumption of producing the bearing force, for supporting the rotating part against the gravity, will be large for large direct-drive wind generators. However, if a bearingless drive could contribute to significant mass reduction of the structural part, then the use of the bearingless drive could be acceptable for large direct-drive applications. Therefore, the following requirements must be fulfilled in order to use the bearingless drives in large direct-drive applications to be acceptable:

(a) Winding of a bearingless 4-pole PM

motor. Outside: 4-pole torque winding; inside: 2-pole bearing force winding [Wol 2006]

(b) 4-pole and 2-pole winding arrangements of a primitive bearingless drive [Chi 2005]

(c) Inset type PMs (d) Buried type PMs [Chi 2005] Fig. 8-8: Conventional bearingless drive concepts


199

• significant mass reduction of the structural part • minimized power consumption in process of producing the bearing force • simplified control and components B. New bearingless drive A new bearingless drive concept [Ban 2010a], which fulfils the requirements mentioned above is described. The new bearingless drive is a type of bearingless permanent magnet (PM) machine drive and is constructed with the double-sided axial flux configuration. The rotor of the bearingless machine consists of the PMs and the iron cores. The stator consists of the slotted iron cores and the windings. The bearingless machine can be constructed with the longitudinal flux (LF) configuration or the transverse flux (TF) configuration. Fig. 8-9 and Fig. 8-10 depict a cross-section of linearized bearingless permanent magnet machines with the longitudinal flux configuration and the transverse flux configuration, respectively. In these bearingless drive concepts, the mass of the rotating (moving) part is not supported by the bearing force. The mass is supported by the buoyancy force as discussed in the last sub-section, 8.3.1. Therefore, no power consumption is necessary to support the rotating part. The moving direction of the rotor is the Y-axis that is represented with a “ ” mark in Fig. 8-9 and Fig. 8-10. The flux paths of the TFPM machine are represented with dotted lines and arrows in Fig. 8-10. In the stable state, both air gap 1 and air gap 2 will be kept at the same air gap length. When air gap 1 is different from air gap 2, the difference can be relayed to the current controller by the gap sensor.

Chapter 8

200

Y

X

YZ

Z

X

moving direction

Stator 2

Stator 1

Rotor

air gap 1

air gap 2

Fig. 8-9: Bearingless PM machine with longitudinal flux configuration

Fig. 8-10: Bearingless PM machine with transverse flux configuration


201

Stator_1 Stator_2

Current controller

Rotor gap sensor

+ _

2i

Current controller

Rotor(moving direction)

g1–g2=0i i

air gap 1(g1)

air gap 2(g2)

Z

XY (a) Case of air gap 1 = air gap 2

(b) Case of air gap 1 > air gap 2

(c) Case with the movement in θY-direction Fig. 8-11: A new bearingless PM machine with different length of air gaps

Chapter 8

202

Fig. 8-11 depicts a control diagram of the new bearingless permanent magnet machine with current controllers for air gap control. The mark “ ” under the rotor means the rotor is supported against gravity, so there is no bearing force necessary to maintain the position of the rotor to Z-axis direction. The bearing force is only used to control the air gap length in the X-axis direction. The bearingless machines in the two cases (air gap 1 = air gap 2 and air gap 1 ≠ air gap 2) are represented in Fig. 8-11(a) and (b), respectively. In the first case, the current in the stator 1 windings is controlled to be equal to the current in the stator 2 windings, as shown in Fig. 8-11(a). When air gap 1 becomes larger than air gap 2, air gap 1 can be reduced by increasing the current of stator 1, as shown in Fig. 8-11(b). The changes of flux densities in different air gap lengths are roughly illustrated in Fig. 8-12, where the flux density B is represented as a function of the magnetic intensity H. Bair gap is the flux density in the stable case. Bsmall air gap and Blarge air gap represent the air gap flux densities in a small air gap and a large air gap, respectively. However, when the rotor moves to θY-direction as shown in Fig. 8-11(c), the bearingless drive machine concept seems insufficient to control both air gap 1 and air gap 2 equally. Therefore, the concept is modified as shown in Fig. 8-13 and Fig. 8-14. Using this modified concept, both air gaps 1 and 2 can be controlled even when the rotor is tilted in θY-direction. Fig. 8-13 depicts the bearingless machine in the case of air gap 1 = air gap 2. Fig. 8-14 depicts the bearingless machine with a rotor that has moved to θY-direction.

B

H

Blarge air gap

Bsmall air gap

Bair gap

Fig. 8-12: Flux densities in different air gap lengths


203

Fig. 8-13: New bearingless PM machine concept at air gap 1 = air gap 2

Chapter 8

204

Fig. 8-14: New bearingless PM machine concept with the rotor movement in θY-direction


205

8.3.3 Hydraulic bearing Using hydraulic bearings, it is possible to guide the rotating and stationary parts with flexibility and relatively large tolerance. In the bearingless drive generator system discussed in the previous section, the winding current of the bearingless generator may stop in some fault cases such as a control failure or an inverter fault. In such cases, the hydraulic bearings are sufficient, not only to guide the generator rotor together with the bearingless drive, but also to prevent further damage caused by the touchdown of the rotor on the stator. The advantages of using the hydraulic bearings are summarized as follows: • possible to provide flexibility in guiding a heavy and large structure • possible to prevent the touchdown of the rotor on the stator • possible to maintain the air gap when the bearingless drive fails • possible to reduce peaks in the power consumption of the bearingless drive In the unstable state of the buoyant structure, the hydraulic bearing can also work to make the structure stable, together with the bearingless drive as shown in Fig. 8-15. Pump systems for the hydraulic bearings can also be included on the outside of the generator.

Fig. 8-15: Buoyant rotating part with hydraulic bearings and stationary part

Chapter 8

206

8.4 Mass estimation of new direct-drive generator for large wind turbines The mass of the new direct-drive generator for large wind turbines with the power ratings from 5 MW to 20 MW is estimated in this section. In order to estimate the electromagnetic mass of the generator, a surface-mounted permanent magnet machine with double-sided air gap, axial flux configuration and full pitch windings is electromagnetically designed for 5 MW, 10 MW and 20 MW direct-drive wind turbines. In order to estimate the structural mass of the new direct-drive generator, three-dimensional finite element analysis (3D FEA) is done. The wind turbine specifications and generator requirements are given in Table 8-1.

8.4.1 Electromagnetic part The longitudinal flux permanent magnet generator with a double-sided air gap and axial flux configuration is designed for the estimation of electromagnetic mass. A cross-sectional view of the generator is presented in Fig. 8-16. The electromagnetic dimensions and mass of the generators obtained from the electromagnetic design is given in Table 8-2.

Table 8-1: Wind turbine specifications and generator requirements

Wind turbine specifications

Power [MW] 5 10 20

Speed [rpm] 12.1 8.6 6.1

Rotor blade diameter [m] 126 178 252

Rated wind speed [m/s] 12 12 12


Air gap diameter [m] 6.3 8.9 12.6

Air gap length [mm] 6.3 8.9 12.6

Force density [kN/m2] 40 40 40


207

b' a c' b a' c

b' a c' b a' c

moving direction

Y

X

YZ

Z

X

ssb

ml

pτ ryh

syh

ssh

stb

pb

Fig. 8-16: Cross-section view of a LFPM generator with double-side air gap configuration

Chapter 8

208

Table 8-2: Electromagnetic dimensions and mass of 5, 10 and 20 MW generators

Power [MW] 5 10 20

Dimensions

Outer radius, Rro [m] 3.57 5.05 7.14

Inner radius, Rri [m] 2.73 3.86 5.46

Number of pole pair, p [ - ] 99 140 198

Number of slots per pole per phase, q [ - ]

1 1 1

Air gap length, lg [mm] 6.3 8.9 12.6

Stator slot width, bss [mm] 15.0 15.0 15.0

Stator tooth width, bst [mm] 18.0 18.0 18.0

Stator slot height, hss [mm] 80.0 80.0 80.0

Stator yoke height, hsy [mm] 40.0 40.0 40.0

Rotor yoke height, hry [mm] 40.0 40.0 40.0

Magnet height, lm [mm] 17.0 23.0 31.5

Rotor pole width, bp [mm] 80.0 80.0 80.0

Mass

Copper [tonnes] 8.0 15.1 28.8

Stator iron [tonnes] 21.8 42.7 85.8

Rotor iron [tonnes] 10.1 20.9 40.2

PM [tonnes] 3.2 8.9 25.2

Total [tonnes] 43.1 87.6 180.0


209

8.4.2 Structural part The cross-section of the proposed direct-drive generator in tangential view is illustrated in Fig. 8-17 depicts. In the figure,

gl is the air gap length [mm],

ml is the magnet length [mm],

ryh is the rotor yoke height [mm],

ssh is the stator slot height [mm],

syh is the stator yoke height [mm],

rht is the rotor structure thickness in the radial direction [mm],

rbt is the rotor structure thickness in the axial direction [mm],

Fig. 8-17: Cross-section of the generator in tangential view

Chapter 8

210

sht is the stator structure thickness in the radial direction [mm],

sbt is the stator structure thickness in the axial direction [mm],

rhl is the rotor structure length in the radial direction [m],

rbl is the rotor structure length in the axial direction [m],

shl is the stator structure length in the radial direction [m],

sbl is the stator structure length in the axial direction [m],

soR is the outer radius of the stationary part [m],

roR is the outer radius of the rotating part [m],

riR is the inner radius of the rotating part [m], and

siR is the inner radius of the stationary part [m]. For the structural mass estimation of both the rotating and stationary part of the generator, the following assumptions are made. • Normal pressure acting on the rotor and the stator due to the flux density in the air gap is 320

[kN/m2] obtained by 202gB μ .

• Maximum air gap deformation is limited to 10% of the air gap length. • so roR R− and ri siR R− are defined as 10 times of the value of gl .

• Generally used structural steel is used for the structural part of the rotor and the stator. • Deformation due to gravity, buoyancy and thermal expansion is not considered in the

analyses. The dimensional parameters gl , ml , ryh , ssh , syh , rhl , shl , soR , roR , riR and siR in Fig. 8-17 are

determined from the electromagnetic design results and the above assumptions. Therefore, it is necessary to determine the dimensions of rht , rbt , sht , sbt , rbl , and sbl in Fig. 8-17 in the structural design of the generator. General expressions used to calculate structural deformation are given as the following equations [Tim 1968]:

4

max5

384wl

EIδ = (8-1)

maxl

Eσλ = (8-2)

where

maxδ is the maximum deformation by the bending force [m],

w is the intensity of distributed force [N/m],


211

l is the span of the structure [m], E is modulus of elasticity [N/m2], I is the area moment of inertia [m4],

maxλ is the maximum deformation by tension force [m], and

σ is the normal pressure (which is the normal force density in electrical machines) [N/m2]. To apply (8-1) and (8-2) in calculating the deformation caused by normal pressure acting on the rotor and the stator of the generator, the structure of the rotating part and the stationary part in Fig. 8-17 is modeled as Fig. 8-18. The structure before deformation is represented in Fig. 8-18(a), and the structure after deformation is represented in Fig. 8-18(b). Constructing the rotating part with multiple support structures in the air chamber enables a decrease in deformation caused by the bending force, r .maxδ . The number of support structures, rbn is chosen as 3, as modelled in Fig. 8-18. For the models of the rotating part and the stationary part in Fig. 8-18, equations (8-1) and (8-2) are reformulated as

( )( )

3

.max 4 312.8 1dn ro ri rh

rrb rh

F d d lE n t

δ−

=+

(8-3)

( ) 3

.max 312.8dn so si sh

ssh

F d d lEt

δ−

= (8-4)

( )( ).max 4 2

dn ro ri rbr

rb rb

F d d lE n t

λ−

=+

(8-5)

( ).max 8

dn so si sbs

sb

F d d lEt

λ−

= (8-6)

where subscripts r and s signify the rotor and the stator,

dnF is the normal force density in the air gap [m],

rbn is the number of support structure in the air chamber of the rotating structure [m],

rod is the outer diameter of the rotating part [m],

rid is the inner diameter of the rotating part [m].

sod is the outer diameter of the stationary part [m], and

sid is the inner diameter of the stationary part [m]. The procedure of deriving (8-3), (8-4), (8-5) and (8-6) from (8-1) and (8-2) is described in Appendix 3. Equations used to determine the structural mass and the air volume in the rotating part are also described in Appendix 3.

Chapter 8

212

Using the procedure in Appendix 3 and (8-3) to (8-6), the structural dimensions of rht , sht , rbt ,

sbt , rbl and sbl are determined. The dimensions and the structural mass obtained from the procedure and (8-3) to (8-6) for the models in Fig. 8-18 are given in Table A-3-1 and Table A-3-2 in Appendix 3. To further reduce the structural mass of the rotating part and the stationary part, the structural dimensions in Table A-3-1 are adjusted using three-dimensional finite element analyses (3D FEA). For the 3D FEA, the ANSYS version 11 is used. In the FEA, the structure of the stationary part is modified as Fig. 8-19 in order to reduce the maximum deformation at the center of the span. Fig. 8-20 depicts the full models of the rotating part and the stationary part. In order to save time for the analyses, partial models (1/60 model) of both the rotating part and the stationary part are built as shown in Fig. 8-21 and Fig. 8-22. Fig. 8-21 depicts the deformation results of the rotating part and Fig. 8-22 depicts the deformation results of the stationary part for the generator with 5 MW power rating.


213

(a) Models before deformation

sλ

sδ sδ

sλ

rδ rδ

rλ rλ

w w

(b) Models after deformation

Fig. 8-18: Simplified deformation models of rotating structure and stationary structure

w w

Fig. 8-19: Models of rotating part and stationary part for 3D FEA

Chapter 8

214

(a) Rotating part with cross-sectional views

(b) Stationary part with cross-sectional views Fig. 8-20: Deformation analysis results of 5 MW generator structure with full modelling


215

Fig. 8-21: Deformation analysis result of rotating part with partial modelling

Fig. 8-22: Deformation analysis result of stationary part with partial modelling

Chapter 8

216

Table 8-3 gives the structural dimensions of new 5 MW, 10 MW and 20 MW generators obtained by the 3D FEA. Table 8-4 gives the air gap deformation and the structural mass of the generators obtained by the 3D FEA including the mass of the fluid. Table 8-5 gives the volume of the air in the rotor structure and the mass of the rotor including both the electromagnetic and structural masses. In physics, the principle of buoyancy would be defined as: • A body wholly or partly immersed in a fluid is buoyed up by a buoyant force equal to the

mass of the fluid displaced by the body. This principle does not mean we need more fluid than the mass of the body in order to make the body afloat. Therefore, it is possible to make the rotor of proposed generator afloat with less mass of fluid than the mass of the rotor as represented in Table 8-4.

Table 8-3: Structural dimensions of new generators by 3D FEA

5 MW 10 MW 20 MW

Axial length, lrb [m] 1.19 1.33 1.49

Front thickness, Rear thickness, trh [mm]

10.3 17 21

Outer diameter, Dro [m] 7.14 10.07 14.26

Center diameter, Dr [m] 6.3 8.8 12.6

Inner diameter, Dri [m] 5.46 7.69 10.94

Rotor

Outer thickness, Center thickness, Inner thickness, trb [mm]

1.5 1.6 1.8

Air gap length [mm] 6.3 8.9 12.6

Axial length, lsb [m] 1.55 1.72 1.9

Front thickness, Rear thickness, tsh [mm] Diameter of curvature [m]

116

2.12

158 3.1

220

4.39

Outer diameter, Dso [m] 7.34 10.27 14.46

Center diameter, Ds [m] 6.3 8.8 12.6

Inner diameter, Dsi [m] 5.26 7.49 10.74

Stator

Outer thickness, Inner thickness, tsb [mm]

10.1 19.4 30.2


217

8.4.3 Total mass comparison In order to identify the mass-competitiveness of the new, proposed generator, the total mass is compared with the mass of different generators currently in use. For this mass comparison, the total mass of five different generators estimated in chapter 4 is used with the following assumption.

• The ratios of total mass per torque rating ( m T ) of each generator concepts remain constant while scaling up the turbines, when using the same materials and the same design technologies.

Table 8-4: Air gap deformation and structural mass of new generator by 3D FEA

5 MW 10 MW 20 MW

Air gap deformation

Rotor structure [mm] 0.311 0.427 0.6

Stator structure [mm] 0.315 0.444 0.63

Structural mass

Rotor structure [ton] 4.1 11.2 25.9

Stator structure [ton] 35.4 84.6 205.3

Fluid (water) [ton] 6.7 11.1 19.0

Total [ton] 46.2 106.9 250.2

Table 8-5: Air volume in rotor structure and total rotor mass

5 MW 10 MW 20 MW

Air volume in rotor [m3] 19.65 44.11 97.83

Total rotor mass [ton] (electromagnetic mass + structural mass)

17.36 40.35 91.35

Chapter 8

218

The generator concepts discussed in Chapter 4 are described as follows.

• Concept-1: EESG DD (Enercon concept, m/T=66.5 kg/kNm at 4.5 MW) • Concept-2: PMSG DD (Zephyros concept, m/T=46.4 kg/kNm at 1.5 MW) • Concept-3: PMSG DD (Theoretical design, m/T=25 kg/kNm at 2, 3 and 5 MW) • Concept-4: PMSG DD (NewGen concept, m/T=18.4 kg/kNm at 4 MW) • Concept-5: DFIG 3G (DFIG concept with m/T=17.4 kg/kNm at 3.5(4) MW where EESG DD signifies a direct-drive electrically-excited synchronous generator, PMSG DD signifies a direct-drive permanent magnet synchronous generator, and DFIG 3G signifies a doubly-fed induction generator with a three-stage gearbox.

The total mass estimation of the five generators is represented in Fig. 8-23 as a function of power rating. Table 8-6 and Fig. 8-24 show the total mass of generators with the power ratings of 5 MW, 10 MW and 20 MW. In Fig. 8-24, the total mass of the new direct-drive generator is represented with a bold line and crosses. According to this mass estimation and comparison, the proposed generator with buoyant rotor concept is heavier than Concept-4 (PMSG DD, NewGen concept) and Concept-5 (DFIG 3G concept) at 5 MW power rating. When scaling up larger than 9 MW, this mass estimation shows that the proposed generator becomes the lightest generator concept as shown in Fig. 8-24.

Power [MW]

0 5 10 15 20

Gen

erat

or to

tal m

ass

[ton]

0

500

1000

1500

2000

2500

EESG DD (Enercon concept) PMSG DD (Zephyros concept)PMSG DD (Theoretical design) PMSG DD (NewGen concept) DFIG 3G

Fig. 8-23: Total mass estimation of five different generator systems currently in use


219

Power [MW]

5 10 15 20

Gen

erat

or to

tal m

ass

[ton]

0

100

200

300

400

500

600

EESG DD (Enercon concept) PMSG DD (Zephyros concept)PMSG DD (Theoretical design) PMSG DD (NewGen concept) DFIG 3G PMSG DD with Buoyant rotor

Fig. 8-24: Generator total mass of the new PMSG DD with different generator systems

Table 8-6: Generator total mass

Total mass [tonnes] 5 MW 10 MW 20 MW

Concept-1 (Enercon structure, m/T=66.5 kg/kNm)

254.0 782.7 2207.1

Concept-2 (Zephyros structure, m/T=46.4 kg/kNm)

177.3 546.1 1540.1

Concept-3 (Theoretical design structure, m/T=25 kg/kNm)

95.5 294.3 829.8

Concept-4 (NewGen concept, m/T=18.4 kg/kNm)

70.3 216.6 610.7

Concept-5 (DFIG 3G, m/T=17.4 kg/kNm)

66.5 204.8 577.5

Proposed concept (m/T kg/kNm)

89.3 (22.64)

194.5 (17.52)

430.3 (13.74)

Chapter 8

220

8.5 Conclusions This chapter focused on finding new concepts of large direct-drive wind generators which enable the structural mass minimization. The following generator concepts have been proposed to overcome the disadvantages of large direct-drive generators: • a ring-shaped generator without a main shaft and without torque arms • multi-sets of generator and converter • a generator with a buoyant rotor • a generator with a simplified bearingless drive and/or hydraulic bearings to guide the air gap A rough design of the new direct-drive wind generators with 5 MW, 10 MW and 20 MW power ratings has been done. The total mass of the new generator has been compared with five different generator systems currently in use. From the rough design of the new generator and the total mass comparison, the following conclusions were obtained. • The proposed direct-drive generator with a buoyant rotor was lighter than the conventional

direct-drive PM generators, Concept-1 (EESG DD, Enercon concept), Concept-2 (PMSG DD, Zephyros concept) and Concept-3 (PMSG DD, Theoretical design) at all power ratings.

• When scaling up to larger than 9 MW, the proposed generator becomes lighter than Concept-4 (PMSG DD, NewGen concept) and Concept-5 (DFIG 3G concept). Therefore, the proposed direct-drive generator with a buoyant rotor could be a solution to minimizing the structural mass of large direct-drive wind generators.

As described above, this chapter focused on finding new concepts of large direct-drive wind generators to minimize the structural mass. Some issues which are needed to realize the proposed generator concenpt will be discussed in the next chapter, recommendations for further research.

Conclusions and Recommendations

221

Chapter 9

Conclusions and Recommendations This thesis dealt with finding a design solution for large direct-drive wind generator with highest possible energy yield and lowest possible cost, as described in chapter 1. Chapter 2 gave an overview of different generator systems for wind turbines in order to gain an understanding of the features of those generator systems. Chapter 3 investigated the active mass-competitiveness of permanent magnet (PM) machines with various configurations discussed in scientific literature. Chapter 4 identified the total-mass competitiveness and estimated the total mass of direct-drive and geared generators currently in use for wind turbines. Chapter 5 developed an analytical model for various configurations of PM machines. The model was verified by experiments and finite element analyses. Chapter 6 gave a comparative design of different PM machines for 5 MW and 10 MW direct-drive wind turbines using the analytical model developed in chapter 5. Chapter 7 derived a new configuration of large direct-drive PM wind generators that would enable active-mass reduction, and would facilitate their manufacture and maintenance. Chapter 8 proposed new concepts for a large direct-drive wind generator with a ring-shaped construction, buoyant rotor, hydraulic bearings and/or a simplified bearingless drive to significantly reduce the structural mass of the generator. In this chapter, conclusions are drawn and recommendations are given for further research.

9.1 Conclusions

Wind turbine generator systems To identify the advantages and disadvantages of direct-drive generator systems for large wind turbines, direct-drive and geared generator systems were compared based on energy yield, loss, mass and cost. In terms of mass and cost, geared generators are superior to direct-drive generators. A doubly-fed induction generator system with a three-stage gearbox (DFIG 3G) is a solution with small mass and low cost. In terms of energy yield, direct-drive generators have a higher energy yield than geared generators. The direct-drive permanent magnet synchronous

Chapter 9

222

generator system (PMSG DD) is a solution with low losses and high energy yield. The DFIG 3G has the highest ratio of annual energy yield to generator system cost, even though the PMSG DD has the highest energy yield overall. The direct-drive electrically excited synchronous generator (EESG DD) was the heaviest and most expensive generator system. Up-scaling and offshore wind turbines are being a key driver in technological developments. Therefore, a generator system which would enable to reduce the cost of energy production and which has higher energy yield is needed. In the thesis, the PMSG DD was chosen for large direct-drive wind generators because of its high energy yield. Therefore, the thesis focused on reducing its mass as much as possible in order to make the PMSG DD more cost-effective.

Permanent magnet machines for large direct-drive To identify the active mass-competitiveness of permanent magnet (PM) generators for large direct-drive applications, various configurations of PM machines such as the radial flux (RF), axial flux (AF) and transverse flux (TF) PM machines were surveyed. According to the survey of various PM machines, AFPM machines seem heavier than RFPM machines and TFPM machines. TFPM machines seem lighter than RFPM machines in low torque ratings. TFPM machines were rarely used for large torque applications. RFPM machines seem the lightest in large torque ratings. Slotted surface-mounted RFPM machines with inner rotors and flux-concentrating TFPM machines with single windings and claw poles showed higher mass-competitiveness than the other machines. Therefore, the following two PM machines were selected for the design of large direct-drive PM wind generators. • Slotted surface-mounted RFPM machine with inner rotor and rare earth magnets • Flux-concentrating TFPM machines with single windings

Mechanical structure of direct-drive wind generators To identify the structural mass-competitiveness and the total mass-competitiveness of direct-drive wind generators, an overview of different generators discussed in scientific literature was given based on the structural mass, the total mass and toque rating. Based on the ratio of the total mass/torque of different direct-drive wind generators, the mass-competitiveness of those generators were odered as: 1st(the lightest)-high temperature superconducting generator (HTSG), 2nd-PMSG with air gap bearings, 3rd-PMSG by theoretical design, 4th-flux-concentrating TFPMG with a C-core configuration, 5th-PMSG with a single-bearing concept, 6th-TORUS & EESG with a traditional mechanical structure.


223

In order to estimate the total mass of direct-drive and geared wind generators that are currently in use, the total mass/torque ratio of each generator concept was assumed to remain constant while scaling up the power rating of wind turbines. The DFIG 3G was taken as a reference generator. The total mass of different direct-drive wind generators currently in use was estimated up to 20 MW. The EESG DD was addressed as the heaviest generator. The PMSG DD with air gap bearings was addressed as the lightest generator among the other direct-drive generators. The total mass of the PMSG DD with air gap bearings was comparable with the mass of the DFIG 3G. From the above results, the following conclusions were obtained. • To significantly reduce structural mass of a direct-drive generator for large wind turbines, a

generator construction without a shaft and torque arms would be a solution. • To strengthen the cost-competitiveness of the generator without a shaft and torque arms, low

accuracy and large tolerances must be acceptable for the generator in constructing.

Electromagnetic modeling of PM machines A generalized electromagnetic design model was developed for both an RFPM machine and a TFPM machine. The procedure and relevant expressions for the RFPM and TFPM machines were discussed in order to determine the electromagnetic dimensions and parameters of the machines. To verify the analysis model of TFPM machines, no-load induced voltage obtained from the analysis model was compared with the no-load voltages obtained through three-dimensional finite element analyses (3D FEA) and measurement. The no-load voltage per pole pair in the analysis model was 8 % higher than the voltage obtained through 3D FEA, and 2 % lower than the measured voltage. To validate the analysis model at a load, the force obtained through the analysis model was compared with the force obtained in 3D FEA and static force measurement. At the nominal stator current, the force obtained by the analytical model is 10 % and 7 % larger than the force obtained by the 3D FEA and by the static force measurement, respectively. However, increasing the stator current larger than the nominal current, the differences between the force by the analytical model and the force by the 3D FEA and the measurement became bigger. This was caused by no consideration of the saturation effect by the stator current in the analytical model.

Comparison of PM generators for large direct-drive wind turbines A comparative design of PM generators for large direct-drive wind turbines was represented to assess different configurations of the generators based on the active mass, the losses and the cost. The following five configurations of PM generators were selected for the comparative design. • RFPMG: a surface-mounted RFPM generator with full pitch windings

Chapter 9

224

• TFPMG-1: a single-sided, single winding flux-concentrating TFPM generator with a U-core • TFPMG-2: a double-sided, single winding flux-concentrating TFPM generator with a C-core • TFPMG-3: a single-sided, single winding flux-concentrating TFPM generator with a U-core

and passive rotor • TFPMG-4: a double-sided, single winding flux-concentrating TFPM generator with a C-core

and passive rotor Among the five different generators, the TFPMG-1 was addressed as the lightest generator for 5 MW wind turbines. However, in the design of the generators for 10 MW wind turbines, the RFPMG was addressed as the lightest genertor. The TFPMG-2 had the smallest loss and the lowest cost compared to the other generators for both 5 MW and 10 MW turbines. The TFPMG-4 was addressed as the generator with the largest mass, the highest cost and the largest loss for both 5 MW and 10 MW turbines. The TFPMG-4 and TFPMG-3 were more expensive than the other generators, since both generators consist of large mass of permanent magnets which are the most expensive active material. The TFPMG-2 and TFPMG-4 were more complicated in constructing because these two genertors have double-sided air gaps. Therefore, the TFPMG-1 was selected as a suitable generator for large direct-drive wind turbines. In [Dub 2004] it was concluded that the TFPM machine with toothed rotor (T6 machine in chapter 3) was a valuable option in terms of the active mass and cost, if the air gap length can be kept below 1.5 mm. However, the design results in this chapter indicated that the conclusion in [Dub 2004] is not valid for all configurations of flux-concentrating TFPM machines.

New configuration of TFPM generator for large direct-drive wind turbines It was dealt to derive a new configuration of large direct-drive wind generators that would enable active mass reduction and facilitate manufacture and maintenance. A flux-concentrating transverse flux permanent magnet (TFPM) machine with multiple-modules segmented, multiple-slots per phase and racetrack-shaped copper windings was proposed to decrease the flux path length. An analytical design model of the proposed TFPM generator was derived, and the model was verified by the experiment of a downscaled TFPM generator. Using the analytical design model, the proposed TFPM generators for 5 MW and 10 MW direct-drive wind turbines were designed and compared based on active mass, cost, loss, efficiency and force density. The following ten different PM generator configurations were discussed in the analytical design. • RFPMG: a surface-mounted TFPM generator with full pitch windings • TFPMG-1: a single-sided, single winding flux-concentrating TFPM generator with a U-core • CP_1mod_ls: single-sided, single winding flux-concentrating TFPM generator with claw poles,

one slot per phase and limited axial length


225

• CP_2mod_ls: single-sided, single winding flux-concentrating TFPM generator with claw poles, two slots per phase and limited axial length

• CP_4mod_ls: single-sided, single winding flux-concentrating TFPM generator with claw poles, four slots per phase and limited axial length

• CP_8mod_ls: single-sided, single winding flux-concentrating TFPM generator with claw poles, eight slots per phase and limited axial length

• CP_1mod_Ap: single-sided, single winding flux-concentrating TFPM generator with claw poles, one slot per phase and limited pole area

• CP_2mod_Ap: single-sided, single winding flux-concentrating TFPM generator with claw poles, two slots per phase and limited pole area

• CP_4mod_Ap: single-sided, single winding flux-concentrating TFPM generator with claw poles, four slots per phase and limited pole area

• CP_8mod_Ap: single-sided, single winding flux-concentrating TFPM generator with claw poles, eight slots per phase and limited pole area

Among these different PM generators, CP_1mod_Ap was addressed as the lightest generator, CP_2mod_Ap was addressed as the second lightest generator and CP_8mod_ls was addressed as the heaviest generator for 5 MW and 10 MW wind turbines. From the design of the generators for 5 MW wind turbines, it was taken that the active mass of TFPMG-1, CP_2mod_ls, CP_1mod_Ap, CP_2mod_Ap and CP_4mod_Ap are smaller than that of RFPMG. The active mass of CP_1mod_Ap, CP_2mod_Ap and CP_4mod_Ap seems to be lighter than that of RFPMG for 10 MW wind turbines.

Minimizing the structural mass of large direct-drive wind generators The following generator concepts were proposed to minimize the structural mass of large direct-drive wind generators. • a ring-shaped generator without a main shaft and torque arms • multi-sets of generator and converter • a generator with a buoyant rotor • a generator with a simplified bearingless drive and/or hydraulic bearings to guide the air gap The proposed generator integrated with the concepts described above was designed for large direct-drive wind turbines with 5 MW, 10 MW and 20 MW power ratings. From the design, it was concluded that the proposed direct-drive generator was lighter than traditional direct-drive PM generators at all power ratings. When scaled up to larger than 9 MW, the proposed generator becomes lighter than a geared generator (DFIG 3G).

Chapter 9

226

9.2 Recomendations for further research

Criterion to evaluate transverse flux electrical machines To evaluate the mass-competitiveness of transverse flux permanent magnet (TFPM) machines, in the scientific literature mostly force density was used because higher force density results in the reduction of active material. However, in chapter 7, it was shown that a TFPM machine with higher force density had larger mass than a TFPM machine with lower force density. Therefore, it may be better to use the mass/torque ratio as the criterion to evaluate TFPM machines, instead of the force density.

Analytical model of TFPM machine A generalized analytical model to use for various configurarions of TFPM machines was proposed in chapter 5. To improve the the analytical model, the stator current may be included in further research.

Disadvantages of TFPM machine This thesis focused on finding a TFPM machine that would enable active mass reduction and facilitate manufacture and maintenance. Thus the disadvantages of the machine, low power factor and high cogging torque, were not discussed in the thesis. However, overcoming these disadvantages must be included in further research in order to strengthen the competitiveness of the machine.

Buoyant rotor construction and guiding system with high flexibility For the realization of the proposed direct-drive generator in chapter 8, the following issues must be included in further research. (1) sealing, (2) friction loss, (3) protection of the magnets, windings and iron cores from the corrosion, (4) protection of the generator structure from structural impact caused by the touchdown of the rotor on the stator, (5) structural deformation caused by gravity, buoyancy and thermal expansion, (6) design of the hydraulic bearing system and the simplified bearingless drive system, (7) verification by building a downscaled system

References

227

References [Amr 2006] W. Amrhein and S. Silber, “Magnetic bearing technology” in Handbook of Fractical-

Horsepower Drives”, Springer-Verlag Berlin Heidelberg 2008. [Ann 1996] Annon, “European wind turbine catalogue”, European commission, Brussels,

Belgium, pp. 64-67, 1996. [Anp 2001] P. Anpalaham, “Design of transverse flux machines using analytical calculations and

finite element analysis”, Tech. Licentiate Thesis, Royal Institute of Technology, Stockholm, 2001.

[Ars 2001] W.M. Arshad, T. Bäckström and C. Sadarangani, “Analytical design and analysis procedure for a transverse flux machine”, in Proc. 2001 IEEE Conf. Elec. Mach. and Drives, pp. 115-121.

[Ars 2002] W.M. Arshad, T. Bäckström and C. Sadarangani, “Investigating a transverse flux machine with intermediate poles”, in Proc. 2002 IEE Conf. Power Electronics, Mach. and Drives, pp. 325-328.

[Ban 2007a] D. Bang, H. Polinder, G. Shrestha, J.A. Ferreira, and R.P.J.O.M. van Rooij, “New active speed stall control compared to pitch control for direct-drive wind turbines”, in Proc. EWEC (European Wind Energy Conference & Exhibition, Milan, Italy, May 7-10 2007.

[Ban 2007b] D. Bang, H.G. Kim, and S.J. Cho, “Linear motor and control method”, Registration No.:10-0728078, 13 pages, Korea patent office (KR) granted patent publication on 13-Jun.-2007.

[Ban 2007c] D. Bang, Y.M. Yoon, I.H. Moon, and H.J. Kim, “Linear motor”, Registration No.:10-0753460, 12 pages, Korea patent office (KR) granted patent publication on 31-Aug.-2007.

[Ban 2008a] D. Bang, H. Polinder, G. Shrestha, and J.A. Ferreira, “Promising direct-drive generator system for large wind turbines”, EPE Journal, Vol. 13, pp. 7-13, September 2008.

[Ban 2008b] D. Bang, H. Polinder, G. Shrestha, and J.A. Ferreira, “Review of generator systems for direct-drive wind turbines”, in Proc. EWEC (European Wind Energy Conference & Exhibition), Brussels, Belgium, March 31 - April 3 2008.

[Ban 2008c] D. Bang, H. Polinder, G. Shrestha, and J.A. Ferreira, “Comparative design of radial and transverse flux PM generators for direct-drive wind turbines”, in Proc. 2008 IEEE International Conference on Electrical Machines, Vilamoura, Portugal, September 6-9 2008.

[Ban 2008d] D. Bang, H. Polinder, G. Shrestha, and J.A. Ferreira, “Design of a lightweight transverse flux permanent magnet machine for direct-drive wind turbines”, in Proc.

References

228

2008 IEEE Industry Applications Society Annual Meeting, Edmonton, Canada, October 5-9 2008.

[Ban 2009a] D. Bang, H. Polinder, G. Shrestha, and J.A. Ferreira, “Possible solutions to overcome drawbacks of direct-drive generator for large wind turbines”, in Proc. EWEC (European Wind Energy Conference & Exhibition), Marseille, France, March 16-19 2009.

[Ban 2009b] D. Bang, H. Polinder, G. Shrestha, and J.A. Ferreira, “Ring-shaped transverse flux PM generator for large direct-drive wind turbines”, in Proc. 2009 IEEE International conference on power electronics and drive systems”, Taipei, Taiwan, November 2-5 2009.

[Ban 2010a] D. Bang, H. Polinder, J.A. Ferreira and B.J. Kim, “Direct-drive electric machine”, Application No.:10-2009-88576, 52 pages, Korea patent office (KR). (accepted to grant)

[Bao 2004] G.Q. Bao, D. Zhang, J.H. Shi and J.Z. Jiang, “Optimal design for cogging torque reduction of transverse flux permanent motor using particle swarm optimization algorithm”, in Proc. 2004 IEEE Conf. Power Electronics and Motion Control(IPEMC), Vol. 1, pp. 260-263.

[Bao 2005a] G.Q. Bao and J.Z. Jiang, “A new transverse flux permanent motor for direct drive application”, in Proc. 2005 IEEE Conf. Elec. Mach. and Drives, pp. 1192-1195.

[Bao 2005b] G.Q. Bao, J.H. Shi and J.Z. Jiang, “Efficiency optimization of transverse flux permanent magnet machine using genetic algorithm”, in Proc. 2005 IEEE Conf. Elec. Mach. and Systems(ICEMS), Vol. 1, pp. 380-384.

[Bar 2009] P.D. Barba and M.E. Mognaschi, “Industrial design with multiple criteria: Shape optimization of a permanent-magnet generator”, IEEE Transactions on Magnetics, Vol. 45, No. 3, pp. 1482-1485, March 2009.

[Bia 2006] N. Bianchi, S. Bolognani and P. Frare, “Design criteria for high-efficiency SPM synchronous motors”, IEEE Transactions on Energy Conversion, Vol. 21, No. 2, pp. 396-404, June 2006.

[Bla 2004] Blaabjerg, F., Chen Z., Kjaer S.B., “Power Electronics as Efficient Interface in Dispersed Power Generation Systems”, IEEE Trans. on Power Electronics, Vol. 19, No. 5, Sept. 2004, pp. 1184-1194.

[Bli 2000] R. Blissenbach, G. Henneberger, U. Schäfer and W. Hackmann, “Development of a transverse flux traction motor in a direct drive system”, in Proc. 2000 Int. Conf. on Elec. Mach., pp. 1457-1460.

[Boc 2006] C. Boccaletti, S. Elia and E. Nisticò, “Deterministic and stochastic optimization algorithms in conventional design of axial flux PM machines”, International Symposium on Power Electronics, Electrical Drives, Automation and Motion, Speedam 2006, pp. 111-115, May 2006.

References

229

[Böh 1997] G. Böhmeke, R. Boldt and H. Beneke, “Geared drive intermediate solutions, comparisons of design features and operating economics”, in Proc. 1997 Europ. Wind Energy Conf., pp. 664-667.

[Böh 2003] G. Böhmeke, “Development and operational experience of the wind energy converter WWD-1”, in Proc. 2003 Europ. Wind Energy Conf.

[Bol 2006] I. Boldea, Variable speed generators, Taylor&Francis, 2006. [Bos 2007] http://www.boschrexroth.com/mobile-hydraulics-

catalog/Vornavigation/Vornavi.cfm?Language=EN&VHist=g54076,g54073,g61480&PageID=m3844, last accessed June 2010.

[BTM 2004] BTM Consults. “International Wind Energy Department—World Market Update 2004, Forecast 2005–2009”. A. Rasmussens, Ringkøbing, Denmark, 2005.

[Bum 2005] J. R. Bumby, R. Martin, “Axial-flux permanent-magnet air-cored generator for small-scale wind turbines”, IEE Proc.-Electr. Power Appl., Vol. 152, No. 5, pp. 1065-1075, September 2005.

[Byw 2004] G. Bywaters, V. John, J. Lynch, P. Mattila, G. Nortor, J. Stowell, M. Salata, O. Labath, A. Chertok and D. Hablanian, “Northern power systems windPACT drive train alternative design study report”, NREL, Golden, Colorado, report No. NREL/SR-500-35524, October 2004.

[Car 1994a] Carlson O., Grauers A., Svensson J., Larsson A., “A comparison of electrical systems for variable speed operation of wind turbines”, in Proc. European wind energy conf., pp.500-505, 1994.

[Car 1994b] F. Caricchi, F. Crescimbini, O. Hororati and R. Vivarelli, “Prototype of a wind-turbine directly-coupled axial-flux PM generator”, in Proc. European wind energy conf., pp.492-499, 1994.

[Car 1995a] F. Caricchi, F. Crescimbini and E. Santini, “Basic principle and design criteria of axial-flux PM machines having counter-rotating rotors”, IEEE Transactions on Industry Applications, Vol. 31, No. 5, pp. 1062-1068, September/October 1995.

[Car 1995b] F. Caricchi, F. Crescimbini, F. Mezzetti and E. Santini, “Multi-stage axial-flux PM machine for wheel direct drive”, in Proc. 1995 IEEE Conf. Ind. Appl., Vol. 1, pp. 679-684.

[Car 1999] F. Caricchi, F. Crescimbini and O. Honorati, “Modular axial-flux permanent-magnet motor for ship propulsion drives”, IEEE Transactions on Energy Conversion, Vol. 14, No. 3, pp. 673-679, September 1999.

[Cat 2001] J.J. Cathey, “Electric machines: Analysis and design applying MATLAB”, McGraw-Hill, 2001.

[Cha 1999] B. J. Chalmers, W. Wu and E. Spooner, “An axial-flux permanent-magnet generator for a gearless wind energy system”, IEEE Trans. Energy Conversion, Vol. 14, No. 2, pp. 251-257, June 1999.

References

230

[Che 1995] Z. Chen and E. Spooner, “A modular, permanent-magnet generator for variable speed wind turbines”, in Proc. 1995 IEE Conf. Elec. Mach. And Drives, pp. 453-457.

[Che 2000] J. Chen, C.V. Nayar and L. Xu, “Design and finite-element analysis of an outer-rotor permanent-magnet generator for directly coupled wind turbines”, IEEE Transactions on Magnetics, Vol. 36, No. 5, pp. 3802-3809, September, 2000.

[Che 2005a] Y. Chen and P. Pillay, “Axial-flux PM wind generator with a soft magnetic composite core”, in Proc. 2005 IEEE Conf. Ind. Appl., pp. 231-237.

[Che 2005b] Y. Chen, P. Pillay and A. Khan, “PM wind generator topologies”, IEEE Transactions on Ind. App. Vol. 41, No. 6, pp. 1619-1626, November/December 2005.

[Che 2006] Chen, Z., Blaabjerg, F. “WIND ENERGY-The World's Fastest Growing Energy Source”, IEEE Power Electronics Society Newsletter, Volume18, Number 3, 2006, ISSN 1054-7231.

[Chi 2005] A. Chiba, T. Fukao, O. Ichikawa, M. Oshima, M. Takemoto, and D.G. Dorrell, “Magnetic Bearings and Bearingless Drives”, London, U.K.: Newnes, 2005.

[Com 2009] http://www.compositesworld.com/news/nrel-to-validate-10-mw-superconductor-wind-turbine, last accessed July 2009.

[DeV 2004] E. De Vries, ‘Global wind technology – overview of developments 2003 - 2004,’ Renewable Energy World, July 2004.

[DeW 2009] DeWind, http://www.compositetechcorp.com/OurCompanies/tabid/145/Default.aspx, last accessed September 2006.

[Dic 2002] P. Dickinson, A.G. Jack and B.C. Mecrow, “Improved permanent magnet machines with claw pole armatures”, in Proc. 2002 Int. Conf. on Elec. Mach., paper 245.

[Dub 2000] M.R. Dubois, H. Polinder and J.A. Ferreira, “Comparison of generator topologies for direct-drive wind turbines”, in Proc. 2000 Nordic Countries Pow. and Indust. Elec., pp. 22-26.

[Dub 2002] M.R. Dubois, H. Polinder and J.A. Ferreira, “Transverse-flux permanent magnet (TFPM) machine with toothed rotor”, in Proc. 2002 IEE Conf. Power Electronics, Mach. and Drives, pp. 309-314.

[Dub 2004] M. Dubois, “Optimized permanent magnet generator topologies for direct drive wind turbines”, Ph.D. dissertation, Delft University of Technology, Delft, The Netherlands, 2004.

[Eas 2002] J. F. Eastham, F. Profumo, A. Tenconi, R. Hill-Cottingham, P. Coles and G. Gianolio, “Novel axial flux machine for aircraft drive: Design and modeling”, IEEE Trans. Magnetics, Vol. 38, No. 5, pp. 3003-3005, September 2002.

[Eco 2006] Ecotècnia, http://www.ecotecnia.com/index_ing.htm, last accessed September 2006. [Eiz 2007] Eize de Vries, “Drive and innovation: The DeWind D8.2 with Voith WinDrive”,

Renewable Energy World, Vol. 10, Issue 2, pp. 55-65, March/April 2007.

References

231

[ENE 2006] ENERCON GmbH, http://www.enercon.de/en/_home.htm, last accessed November 2006.

[Eng 2007] S. Engström and S. Lindgren, “Design of NewGen direct-drive generator for demonstration in a 3.5 MW wind turbine”, EWEC (European Wind Energy Conference & Exhibition, Milan, Italy, May 7-10 2007.

[Evo 2004] S. Evon and R. Schiferl, “Direct drive induction motors”, in Proc. 2004 IEEE the Pulp and Paper Industry Technical Conference, pp. 49-54.

[EWE 2009] European wind Energy Association, “The Economics of wind energy”, March 2009. [EWT 2010] Emergya Wind Technologies BV, http://www.ewtinternational.com/?id=49, last

accessed September 2010. [Fin 2006] L. Fingersh, M. Hand and A. Laxson, “Wind turbine design cost and scaling model”,

Technical Report, NREL/TP-500-40566, December 2006. [Flo 2006] J. Florence, “Global wind power expands in 2006”, http://www.earth-

policy.org/index.php?/indicators/C49/wind_power_2006, last accessed June 2009. [Gam 2006] Gamesa Elórica, http://www.gamesa.es/gamesa/index.html, last accessed September

2006. [GE 2010] GE Energy, http://www.gepower.com/businesses/ge_wind_energy/en/index.htm, last

accessed June 2010. [Gen 1995] G. Geng and E. Spooner, “Cancellation of noise and vibration in modular permanent-

magnet wind turbine generators”, in Proc. 1995 IEE Conf. Elec. Mach. and Drives, pp. 467-471.

[Gie 2005] J.F. Gieras, “Performance characteristics of a permanent magnet transverse flux generator”, in Proc. 2005 IEEE Conf. Elec. Mach. and Drives, pp. 1293-1299.

[Gla 2002] C.J.W. Glanni and J.H. Boyd, Jr., “Poly phase transverse flux motor”, Patent No.: US 6,492,758 B1, Dec. 2002.

[Gra 1996] A. Grauers, “Design of direct-driven permanent-magnet generators for wind turbines”, Ph.D. dissertation, Chalmers University of Technology, Göteborg, Sweden, 1996.

[Guo 2003] Y.G. Guo, J.G. Zhu, P.A. Watterson and W. Wu, “Design and analysis of a transverse flux machine with soft magnetic composite core”, in Proc. 2003 IEEE Conf. Elec. Mach. and Systems(ICEMS), Vol. 1, pp. 153-157.

[Guo 2006] Y.G. Guo, J.G. Zhu, P.A. Watterson and W. Wu, “Development of a PM transverse flux motor with soft magnetic composite core”, IEEE Transactions on Energy Conversion, Vol. 21, No. 2, pp. 426-434, June 2006.

[Had 2000] D. Hadjidj, A. Miraoui, and J.M. Kauffmann, “Modelling of a transverse Vernier hybrid reluctance motor”, The European Physical Journal Applied Physics, Vol. 11, pp. 183-188, 2000.

[Han 2001] L. H. Hansen, L. Helle, et al. “Conceptual survey of generators and power electronics for wind turbines”, Risø-R-1205(EN), 2001.

References

232

[Han 2005] R. Hanitsch and M. Widyan, “Design, construction and test of a permanent-magnet prototype machine for wind energy applications,” KOMEL Conf., Poland, May 2004, pp. 63-66.

[Han 2007] A.D. Hansen and L.H. Hansen, “Wind turbine concept market penetration over 10 years(1995-2004)”, Wind Energy, Vol.10, pp. 81-97, 2007.

[Har 1993a] M.R. Harris and B.C. Mecrow, “Variable reluctance permanent magnet motors for high specific output”, in Proc. 1993 IEE Conf. Elec. Mach. and Drives, pp. 437-442.

[Har 1993b] M.R. Harris, “Comparative electromagnetic parameters for alternative motor types”, IEE Colloquium on Motors and Drives for Battery Powered Propulsion, pp. 3/1-3/4, April 1993.

[Har 1996] M.R. Harris, G.H. Pajooman and S.M. Abu Sharkh, “Performance and design optimization of electric motors with heteropolar surface magnets and homopolar windings”, IEE Proc. Electr. Power Appl., Vol. 143, No. 6, pp. 429-436, Nov. 1996.

[Har 1997a] T. Hartkopf, M. Hofmann and S. Jöckel, “Direct-drive generators for megawatt wind turbines”, in Proc. 1997 Europ. Wind Energy Conf., pp. 668-671.

[Har 1997b] M.R. Harris, G.H. Pajooman and S.M.A. Sharkh, “Comparison of alternative topologies for VRPM(transverse-flux) electrical machines”, in Proc. 1997 IEE Colloquium on New Topologies for PM Machines, pp. 2/1-2/7.

[Har 1997c] M.R. Harris, G.H. Pajooman and S.M.A. Sharkh, “The problem of power factor in VRPM (transverse-flux) machines”, in Proc. 1997 IEE Conf. Elec. Mach. and Drives, pp. 386-390.

[Har 1998] M.R. Harris, G.H. Pajooman, S.M. Abu Sharkh and B.C. Mecrow, “Comparison of flux-concentrated and surface-magnet configurations of the VRPM (Transverse-Flux) machine”, in Proc. 1998 Int. Conf. on Elec. Mach., pp. 1119-1122.

[Has 1999] B.E. Hasubek and E.P. Nowicki, “Two dimensional finite element analysis of passive rotor transverse flux motors with slanted rotor design”, in Proc. 1999 IEEE Canadian Conf. Elec. and Computer Engineering, Vol. 2, pp. 1199-1204.

[Has 2000] B.E. Hasubek and E.P. Nowicki, “Design limitations of reduced magnet material passive rotor transverse flux motors investigated using 3D finite element analysis”, in Proc. 2000 IEEE Canadian Conf. Elec. and Computer Engineering, Vol. 1, pp. 365-369.

[Hen 1997] G. Henneberger and M. Bork, “Development of a new transverse flux motor”, in Proc. 1997 IEE Colloquium on New Topologies for PM Machines, pp. 1/1-1/6.

[Hof 2002] R. Hoffmann, “A comparison of control concepts for wind turbines in terms of energy capture”, Ph.D. dissertation Darmstadt University of Technology, Darmstadt, Germany, 2002.

[Hua 1997a] S. Huang, J. Luo and T.A. Lipo, “Analysis and evaluation of the transverse flux circumferential current machine”, in Proc. 1997 IEEE Conf. Ind. Appl. Soc., Vol. 1, pp. 378-384.

References

233

[Hua 1997b] S. Huang, J. Luo and T.A. Lipo, “Evaluation of the transverse flux circumferential current machine by the use of sizing equations”, in Proc. 1997 IEEE Conf. Elec. Mach. and Drives, pp. WB2/15.1-WB2/15.3.

[Huf 2009] www.huffingtonpost.com/michael-graham-richard/wind-power-success-storie_b_106975.html, last accessed June 2009.

[Hus 2003] S.M. Husband and C.G. Hodge, “The Rolls-Royce transverse flux motor development”, in Proc. 2003 IEEE Conf. Elec. Mach. and Drives, Vol. 3, pp. 1435-1440.

[Jos 2006] Joseph Florence, “Global Wind Power Expands in 2006”, June 28, 2006. http://www.earth-policy.org/Indicators/Wind/2006.htm

[Kan 2000] D.H. Kang, D. Bang, J.M. Kim, Y.H. Jeong, M.H. Kim, “A Study on the design of PM exited transverse flux linear motor for ropeless elevator”, Trans. KIEE (The Korean Institute of Electrical Engineers). Vol. 49B. No. 3, Mar, 2000 (in Korean).

[Kas 2002] G. Kastinger and A. Schumacher, “Reducing torque ripple of transverse flux machines by structural designs”, in Proc. 2002 IEE Conf. Power Electronics, Mach. and Drives, pp. 320-324.

[Kim 2005] K.C. Kim, S.B. Lim, K.B. Jang, S.G. Lee, J. Lee, Y.G. Son, Y.K. Yeo and S.H. Baek, “Analysis on the direct-driven high power permanent magnet generator for wind turbine”, in Proc. 2005 IEEE Conf. Elec. Mach. and Systems(ICEMS), Vol. 1, pp. 243-247.

[Kle 1984] F.N. Klein and M.E. Kenyon, “Permanent magnet DC motors design criteria and operation advantages”, IEEE Transactions on Industry Applications, Vol. IA-20, No. 6, November/December 1984.

[Kor 2004] G. Korouji, “Design, construction and test study of a wind energy generator with dual permanent magnet excitation”, Ph.D. dissertation, Berlin University of Technology, Berlin, Germany, 2004 (in German).

[Kri 1987] R. Krishnan, “Selection criteria for servo motor drives”, IEEE Transactions on Industry Applications, Vol. IA-23, No. 2, March/April 1987.

[Lam 1996] P. Lampola and J. Perho, “Electromagnetic analysis of a low-speed permanent-magnet wind generator”, International Conference on Opportunities and Advances in International Electric Power Generation, pp. 55-58, March 1996.

[Lam 2000] P. Lampola, “Directly driven, low-speed permanent-magnet generators for wind power applications”, Ph.D. dissertation, Helsinki University of Technology, Finland, 2000.

[Lan 2000] A. Lange, W.-R. Canders, F. Laube and H. Mosebach, “Comparison of different drive systems for a 75kW electrical vehicle drive”, in Proc. 2000 Int. Conf. on Elec. Mach., pp. 1308-1312.

[Lar 2000] Å. Larsson, “The power quality of wind turbines”, Ph.D. dissertation, Chalmers University of Technology, Göteborg, Sweden, 2000.

References

234

[Lee 2009] J.-Y. Lee, J.-W. Kim, S.-R. Moon, J.-H. Chang, S.-U. Chung, D.-H. Kang and J.-P. Hong, “Dynamic Characteristic Analysis Considering Core Losses in Transverse Flux Linear Machine with Solid Cores”, IEEE Trans. Magnetics, vol. 45, no. 3, pp. 1776-1779, 2009.

[Lib 2004] F. Libert and J. Soulard, “Design Study of Different Direct-Driven Permanent-Magnet Motors for a Low Speed Application”, in Proceedings of the Nordic Workshop on Power and Industrial Electronics (NORpie), Trondheim, Norway, June 2004.

[Lib 2005] F. Libert and J. Soulard, “Design Study of Low-Speed Direct-Driven Permanent-Magnet Motors with Concentrated Windings”, in Proceedings of the 6th International Symposium on Advanced Electromechanical Motion Systems (Electromotion), September 2005.

[Lu 2003] K.Y. Lu, E. Ritchie, P.O. Rasmussen and P. Sandholdt, “Modelling a single phase surface-mounted permanent magnet transverse flux machine based on Fourier Series Method”, in Proc. 2003 IEEE Conf. Elec. Mach. and Drives, Vol. 1, pp. 340-345.

[Lu 2003] K.Y. Lu, E. Ritchie, P.O. Rasmussen and P. Sandholdt, “General torque equation capable of including saturation effects for a single phase surface-mounted permanent magnet transverse flux machine”, in Proc. 2003 IEEE Conf. Ind. Appl. Soc., Vol. 2, pp. 1382-1388.

[Lu 2003] K.Y. Lu, E. Ritchie, P.O. Rasmussen and P. Sandholdt, “A simple method to estimate inductance profile of a surface-mounted permanent magnet transverse flux machine”, in Proc. 2003 IEEE Conf. Power Electronics and Drive Systems(PEDS), Vol. 2, pp. 1334-1338.

[Lu 2003] K.Y. Lu, E. Ritchie, P.O. Rasmussen and P. Sandholdt, “Modeling and power factor analysis of a single phase surface-mounted permanent magnet transverse flux machine”, in Proc. 2003 IEEE Conf. Power Electronics and Drive Systems(PEDS), Vol. 2, pp. 1609-1613.

[Mad 1998] C.P. Maddison, B.C. Mecrow and A.G. Jack, “Claw pole geometries for high performance transverse flux machines”, in Proc. 1998 Int. Conf. on Elec. Mach., pp. 340-345.

[Mad 1999] C. P. Maddison, “Transverse flux machines for high torque applications”, Ph.D. dissertation University of Newcastle upon Tyne, UK, 1999.

[Man 2002] J. Manwell, J. McGowan, A. Rogers, “Wind energy explained”, Wiley, 2002. [Mas 2001] A. Masmoudi and A. Elantably, “A simple assessment of the cogging torque in a

transverse flux permanent magnet machine”, in Proc. IEEE International Electric Machines and Drives Conference, 2001 (IEMDC 2001).

[Mas 2004] A. Masmoudi, A. Njeh, A. Mansouri, H. Trabelsi and A. Elantably, “Optimizing the overlap between the stator teeth of a claw pole transverse-flux permanent-magnet machine”, IEEE Transactions on Magnetics, Vol. 40, No. 3, pp. 1573-1578, May 2004.

References

235

[McD 2006] A.S. McDonald, M.A. Mueller and H. Polinder, “Comparison of generator topologies for direct-drive wind turbines including structural mass”, in Proc. of the International Conference on Electrical Machines(ICEM), pp. 360.1-7, September 2006.

[Mec 1996] B.C. Mecrow, A.G. Jack and C.P. Maddison, “Permanent magnet machines for high torque, low speed applications”, in Proc. 1996 Int. Conf. on Elec. Mach., pp. 461-466.

[Mit 1995] A.J. Mitcham and J.J.A. Cullen, “Motors and drives for surface ship propulsion: comparison of technologies”, in Proc. 1995 Electric Propulsion Conf., paper 4.

[Mit 1997] A.J. Mitcham, “Transverse flux motors for electric propulsion of ships”, in Proc. 1997 IEE Colloquium on New Topologies for PM Machines, pp. 3/1-3/6.

[Mue 2005a] M.A. Mueller, “Design and performance of a 20kW, 100rpm, switched reluctance generator for a direct drive wind energy converter”, in Proc. 2005 IEEE Conf. Elec. Mach. and Drives, pp. 56-63.

[Mue 2005b] M.A. Mueller, A.S. McDonald and D.E. Macpherson, “Structural analysis of low-speed axial-flux permanent-magnet machines”, IEE Proc.-Electr. Power Appl., Vol. 152, No. 6, pp. 1417-1426, November 2005.

[Mul 1999] E. Muljadi, C. P. Butterfield and Y. Wan, “Axial-flux modular permanent-magnet generator with a toroidal winding for wind-turbine applications”, IEEE Trans. Ind. Applicat., Vol. 35, No. 4, pp. 831-836, July/Aug. 1999.

[Mul 2006] Multibrid , http://www.multibrid.com/m5000/data.html, last accessed September 2006.

[Nje 2003a] A. Njeh, A. Masmoudi and A. Elantably, “3D FEA based investigation of the cogging torque of a claw pole transverse flux permanent magnet machine”, in Proc. 2003 IEEE Conf. Elec. Mach. and Drives, Vol. 1, pp. 319-324.

[Nje 2003b] A. Njeh, A. Masmoudi and A. Elantably, “Cogging torque minimization of a claw pole TF-motor: FEA in the Maxwell 3D environment”, in Proc. IEEE Conf. Signals, Systems, Decision, and Information Technology (SSD’03), Sousse, Tunisia, March 2003.

[Nor 2006] Nordex AG, http://www.nordex-online.com/en/products-services/wind-turbines.html, last accessed September 2006.

[Oka 2006] T. Okazaki, H. Sugimoto and T. Takeda, “Liquid nitrogen cooled HTS motor for ship propulsion”, in Proc. 2006 IEEE Con. Power Engineering Soc., pp. 1-6.

[Paj 1997] G.H. Pajooman, “Performance assessment and design optimization of VRPM (Transverse Flux) machines by finite element computation”, Ph.D. dissertation, Southampton Univ., UK, 1997.

[Par 2005a] A. Parviainen, J. Pyrhonen and P. Kontkanen, “Axial flux permanent magnet generator with concentrated winding for small wind power applications”, in Proc. 2005 IEEE International Conference on Electric Machines and Drives, pp. 1187-1191, May 2005.

References

236

[Par 2005b] A. Parviainen, “Design of axial-flux permanent-magnet low-speed machines and performance comparison between radial-flux and axial-flux machines”, Ph.D. dissertation Lappeenranta University of Technology, Lappeenranta, Finland, 2005.

[Pay 2002] B.S. Payne, S.M. Husband and A.D. Ball, “Development of condition monitoring techniques for a transverse flux motor”, in Proc. 2002 IEE Conf. Power Electronics, Mach. and Drives, pp. 139-144.

[Pol 1998] H. Polinder, “On the losses in a high-speed permanent-magnet generator with rectifier”, Ph.D. dissertation, Delft University of Technology, Delft, The Netherlands, 1998.

[Pol 2005] H. Polinder and J. Morren, “Developments in wind turbine generator systems”, Electrimacs 2005, Hammamet, Tunisia.

[Pol 2005] H. Polinder, S.W.H. de Haan, M.R. Dubois, J.G. Slootweg, ‘Basic operation principles and electrical conversion systems of wind turbines'. In EPE Journal, December 2005 (vol. 15, no. 4), pp. 43-50.

[Pol 2006] H. Polinder, F.F.A. van der Pijl, G.J. de Vilder, P. Tavner, “Comparison of direct-drive and geared generator concepts for wind turbines”, IEEE Trans. Energy Conversion, Vol. 21, pp. 725-733, September 2006.

[Pol 2007] H. Polinder, D. Bang, R.P.J.O.M. van Rooij, A.S. McDonald, and M.A. Mueller, “10 MW Wind turbine direct-drive generator design with pitch or active speed stall control”, IEEE Int. Electric Machines and Drives Conf., Antalya, Turkey, May 3-5 2007.

[Poo 2003] R. Poore and T. Lettenmaier, “Alternative design study report: WindPACT advanced wind turbine drive train designs study”, NREL, Golden, Colorado, report No. NREL/SR-500-33196, August 2003.

[Rah 2004] Z. Rahman, “Evaluating radial, axial and transverse flux topologies for ‘In-Wheel’ motors”, in Proc. 2004 IEEE Conf. Power Electronics in Transportation, pp. 75-81.

[Rah 2008] M.A. Rahman, “Advances of interior permanent magnet (IPM) wind generators”, in Proc. 2008 IEEE Conf. Electrical Machines and Systems (ICEMS), pp. 2228-2233.

[Ran 2002] Y. Rang, Chenglin Gu and H. Li, “Analytical design and modeling of a transverse flux permanent magnet machine”, in Proc. 2002 Conf. Power System Technology, Vol. 4, pp. 2164-2167.

[Rep 2006] Repower Systems AG, http://www.repower.de/index.php?id=12&L=1, last accessed September 2006.

[Rob 2000] Robert Harrison, Erich Hau and Herman Snel, “Large wind turbines Design and Economics,” John Wiley & Sons Ltd, 2000, ISBN 0471-494569.

[Run 2004] F. Runcos, R. Carlson, A. M. Oliveira, P. Kuo-Peng N. Sadowski, “Performance Analysis of a Brushless Double Fed Cage Induction Generator”, Nordic Wind Power Conference (NWPC04), Chalmers University of Technology, Göteborg, Sweden, March 1-2, 2004.

References

237

[Sah 2001] F. Sahin, “Design and development of a high-speed axial-flux permanent-magnet machine”, Ph.D. dissertation, Eindhoven University of Technology, Eindhoven, The Netherlands, 2001.

[Sch 2005] E. Schmidt, “3-D finite element analysis of the cogging torque of a transverse flux machine”, IEEE Transactions on Magnetics, Vol. 41, No. 5, May 2005.

[Shi 2003] J. Shi, Y. Li and W. Zheng, “Survey on the development of transverse flux machines”, in Proc. 2003 IEEE Conf. Elec. Mach. and Systems(ICEMS), Vol. 1, pp. 137-140.

[Sie 1998] S. Siegfriedsen and G. Böhmeke, “Multibrid technology-a significant step to multi-megawatt wind turbines”, Wind Energy, Vol. 1, Issue 2, pp. 89-100, December 1998.

[Sie 2006] Siemens AG, http://www.powergeneration.siemens.com/en/windpower/products/index.cfm, last accessed September 2006.

[Sin 2010] Sinovel Wind Group Co., Ltd., http://www.sinovel.com/Products.html, last accessed September 2010.

[Sit 2000] K. Sitapati and R. Krishnan, “Performance comparisons of radial and axial field, permanent magnet, brushless machines”, in Proc. 2000 IEEE Conf. Ind. Appl. Soc., pp. 228-234.

[Sle 1992] G.R. Slemon and X. Liu, “Modeling and design optimization of permanent magnet motors”, Elec. Mach. And Power Syst., Vol. 20, pp. 71-92, 1992

[Söd 1996] L. Söderlund, J-T. Eriksson, J. Salonen, H. Vihriälä and R. Perälä, “A permanent-magnet generator for wind power applications”, IEEE Trans. Magnetics, Vol. 32, No. 4, pp. 2389-2392, July 1996.

[Söd 1997] L. Söderlund, A. Koski, H. Vihriälä, J-T. Eriksson and R. Perälä, “Design of an axial flux permanent magnet wind power generator”, in Proc. 1997 IEE Conf. Elec. Mach. and Drives, pp. 224-228.

[Soe 2005] Joris Soens,“Impact of wind energy in a future”, Ph. D dissertation, ISBN 90-5682-652-2, Wettelijk depot,UDC 621.548, Dec., 2005.

[Spo 1992] E. Spooner and B. J. Chalmers, “TORUS : A slotless, toroidal-stator, permanent-magnet generator”, IEE Proceedings-B, Vol. 139, No. 6, pp. 497-506, November 1992.

[Spo 1996a] E. Spooner and A. C. Williamson, “Direct coupled, permanent magnet generators for wind turbine applications”, IEE Proc.-Electr. Power Appl., Vol. 143, No. 1, pp. 1-8, January 1996.

[Spo 1996b] E. Spooner and A. Williamson, “Modular, permanent-magnet wind-turbine generators”, in Proc. 1996 IEEE Conf. Ind. Appl., Vol. 1, pp. 497-502.

[Spo 1996c] E. Spooner, A.C. Williamson and G. Catto, “Modular design of permanent-magnet generators for wind turbines”, IEE Proc.-Electr. Power Appl., Vol. 143, No. 5, pp. 388-395, September 1996.

References

238

[Spo 1998] E. Spooner and A.C. Williamson, “Parasitic losses in modular permanent-magnet generators”, IEE Proc.-Electr. Power Appl., Vol. 145, No. 6, pp. 485-496, November 1998.

[Spo 2005] E. Spooner, P. Gordon, J.R. Bumby and C.D. French, “Lightweight, ironless-stator, PM generators for direct-drive wind turbines” , IEE Proc.-Electr. Power Appl., Vol. 152, No. 1, pp. 17-26, January 2005.

[STX 2010] STX Windpower B.V., http://www.stxwind.com/nl/products/7-products, last accessed June 2010.

[Suz 2009] Suzlon Energy, http://www.suzlon.com/products/l2.aspx?l1=2&l2=9, last accessed July 2009.

[Sve 2006a] D. Svechkarenko, J. Soulard and Chandur Sadarangani, “A Novel Transverse Flux Generator in Direct-Driven Wind Turbines”, in Proceedings Nordic Workshop on Power and Industrial Electronics, June 2006.

[Sve 2006b] D. Svechkarenko, J. Soulard and Chandur Sadarangani, “Analysis of a novel transverse flux generator in direct-driven wind turbines”, International Conference on Electrical Machines(ICEM), pp. 424.1-6, September 2006.

[Tav 2005] P.J. Tavner and J. Xiang, “Wind turbine reliability, how does it compare with other embedded generation sources”, in Proc. IEE Rel. Transmission Distrib. Netw. Conf., London, U.K., Feb. 2005, pp. 243-248.

[Tav 2006] P.J. Tavner and E. Spooner, “Light structures for large low-speed machines for direct-drive applications”, in Proc. of the International Conference on Electrical Machines(ICEM), pp. 421.1-6, September 2006.

[Tim 1968] S.P. Timoshenko and D.H. Young, “Elements of strength of materials”, 1968. [Tor 1993] D. A. Torrey, ‘Variable-reluctance generators in wind-energy systems,’ IEEE

PESC’93, 1993, pages 561–567. [Tor 2002] D. A. Torrey, “Switched Reluctance Generators and Their Control”, IEEE

Transactions on Industrial Electronics, Vol.49, No. 1, February 2002, pp. 3-14. [Tre 2009] www.treehugger.com/files/2008/06/queen-offshore-wind-turbine-power-clipper.php,

last accessed June 2009. [Ves 2006] Vestas Wind Systems,

http://www.vestas.com/vestas/global/en/Downloads/Downloads/Download_brochurer.htm, last accessed September 2006.

[Vih 2002] H. Vihriälä, “Control of variable speed wind turbines”, Ph.D. dissertation Tampere University of Technology, Hervanta, Finland, 2002.

[Viz 2005] D. Vizireanu, X. Kestelyn, S. Brisset, P. Brochet, Y. Milet and D. Laloy, “Polyphased modular direct-drive wind turbine generator”, in Proc. 2005 European Conference on Power Electronics and Applications, pp. 1-9.

References

239

[Viz 2006] D. Vizireanu, S. Brisset and P. Brochet, “Design and optimization of a 9-phase axial-flux PM synchronous generator with concentrated winding for direct-drive wind turbine”, in Proc. 2006 IEEE Conf. Ind. Appl. Soc., Vol. 4, pp. 1912-1918.

[Voy 2000] J.E. Voyce, S.M. Husband and D.J. Mattick, “The PM propulsion motor: from infancy to adolescence”, in Proc. 2000 All Elec. Ship Conf., pp. 134-139.

[Wan 2005a] F. Wang, J. Bai, Q. Hou and J. Pan, “Design features of low speed permanent magnet generator direct driven by wind turbine”, in Proc. 2005 IEEE Conf. Elec. Mach. and Systems(ICEMS), Vol. 2, pp. 1017-1020.

[Wan 2005b] F. Wang, Q. Hou, J. Bo and J. Pan, “Study on control system of low speed PM generator direct driven by wind turbine”, in Proc. 2005 IEEE Conf. Elec. Mach. and Systems(ICEMS), Vol. 2, pp. 1009-1012.

[Weh 1986] H. Weh and H. May, “Achievable force densities for permanent magnet excited machines in new configurations”, in Proc. 1986 Int. Conf. on Elec. Mach., pp. 1107-1111.

[Weh 1988a] H. Weh, “Permanentmagneterregte Synchronmaschinen hoher Kraftdichte nach dem Transversalflusskonzept”, etzArchiv, Vol. 10, No. 5, pp. 143-149, 1988.

[Weh 1988b] H. Weh, H. Hoffmann, J. Landrath, H. Mosebach and J. Poschadel, “Directly driven permanent magnet excited synchronous generator for variable speed operation”. European Community Wind Energy Conference, pp: 566-572, 1988.

[Weh 1988c] Weh H., Hoffman H., Landrath J. “New permanent magnet excited synchronous machine with high efficiency at low speeds”. Proceedings of the International Conference on Electrical Machines and Drives, pp: 35-40, 1988.

[Weh 1991] H. Weh, “Synchronous machines with New Topologies”, in Proc. 1991 Int. Conf. on the Evolution and Modern Aspects of Sync. Mach., pp. C1-C9.

[Weh 1994] H. Weh, “Multiple track transversal flux machine with ring coils”, German patent DE 4 300 440, German patent Office, 1994.

[Weh 1995] H. Weh, “Transverse-flux machines in drive and generator application”, in Proceedings of the IEEE Symposium on Electric Power Engineering (Stockholm Power Tech), Strockholm, Sweden, 1995, Vol. Invited speaker’ session, pp. 75-80.

[Wer 2005] U. Werner, J. Schüttler and B. Orlik, “Speed and torque control of a permanent magnet excited transverse flux motor for direct servo-drive applications”, in Proc. 2005 European Conference on Power Electronics and Applications, pp. 1-10.

[Wes 1996] A.J.G. Westlake, J.R. Bumby and E. Spooner, “Damping the power-angle oscillations of a permanent-magnet synchronous generator with particular reference to wind turbine applications”, IEE Proc.-Electr. Power Appl., Vol. 143, No. 3, pp. 269-280, May 1996.

[Wid 2006] S. Widyan, “Design, optimization, construction and test of rare-earth permanent-magnet electrical machines with new topology for wind energy applications”, Ph.D. dissertation Technische Universität Berlin, Berlin, Germany, 2006.

References

240

[Win 2006] Winwind Oy, http://www.winwind.fi/english/tuotteet.html, last accessed September 2006.

[Win 2009] Windflow Technology Ltd., http://www.windflow.co.nz/products/windflow-technology-ltd-products, last accessed July 2009.

[Wu 1995] W. Wu, E. Spooner and B. J. Chalmers, “Design of slotless TORUS generators with reduced voltage regulation”, IEE Proc.-Electr. Power Appl., Vol. 142, No. 5, pp. 337-343, September 1995.

[Wu 1995] W. Wu, E. Spooner and B. J. Chalmers, “Reducing voltage regulation in toroidal permanent-magnet generators by stator saturation”, in Proc. 1995 IEE Conf. Elec. Mach. And Drives, pp.385-389.

[Wu 2000] W. Wu, V.S. Ramsden, T. Crawford and G. Hill, “A low-speed, high-torque, direct-drive permanent magnet generator for wind turbine” , in Proc. 2000 IEEE Conf. Ind. Appl., Vol. 1, pp. 147-154.

[Zha 2005] Y. Zhao and J. Chai, “Power factor analysis of transverse flux permanent machines”, in Proc. 2005 IEEE Conf. Elec. Mach. and Systems(ICEMS), Vol. 1, pp. 450-459.

[Zha 2007] J. Zhang, M. Cheng and Z. Chen, “A Novel Stator Interior Permanent magnet Generator for Direct-Drive Wind Turbines”, in Proc. 2007 IEEE Conf. Elec. Mach. and Systems(ICEMS), pp. 723-728.

Appendix 1

241

Appendix 1 The magnetic reluctances in the equivalent circuit illustrated in Fig. 5-3-8 and Fig. 5-3-9 in Chapter 5 are formulated as the following equations.

1 15AR R R= + (A-1-1)

2BR R= (A-1-2)

3 4 5CR R R R= + + (A-1-3)

6DR R= (A-1-4)

7 8 9ER R R R= + + (A-1-5)

10FR R= (A-1-6)

11 12 13GR R R R= + + (A-1-7)

14HR R= (A-1-8)

20 21 22IR R R R= + + (A-1-9)

27 28 29JR R R R= + + (A-1-10)

17 18 19KR R R R= + + (A-1-11)

23 24 25LR R R R= + + (A-1-12) Considering the above reluctance equations and Kirchhoff’s voltage law, flux equations of the equivalent circuit in Fig. 5-3-9 in Chapter 5 are formulated as follows. ( ) 1 2 3 4 50 0 0 0C K CR R R+ ⋅Φ + ⋅Φ + ⋅−Φ + ⋅Φ + ⋅Φ = (A-1-45)

( )1 2 3 4 50 0 0 0G L GR R R⋅Φ + + ⋅Φ + ⋅Φ + ⋅−Φ + ⋅Φ = (A-1-46)

( ) ( )1 2 3 4 50 0C D C B A J A J AR R R R R R R R R⋅ −Φ + ⋅Φ + + + + + ⋅Φ + ⋅Φ + + ⋅−Φ = (A-1-47)

( ) ( )1 2 3 4 50 0G A I A H G F A IR R R R R R R R R⋅Φ + ⋅−Φ + ⋅Φ + + + + + ⋅Φ + + ⋅−Φ = (A-1-48)

( ) ( ) ( )1 2 3 4 50 0 A J A I E J A I mR R R R R R R R F⋅Φ + ⋅Φ + + ⋅−Φ + + ⋅−Φ + + + + ⋅Φ = − (A-1-49)

When the magnetic reluctances R and the magneto-motive force due to the magnets mF are known, the flux Φ is obtained by (A-1-50).

1mR F−Φ = ⋅ (A-1-50)

The flux equations of (A-1-45), (A-1-46), (A-1-47), (A-1-48) and (A-1-49) are transformed into (A-1-51) by (A-1-50).

Appendix 1

242

11 11 12 13 14 15

2 21 22 23 24 25

3 31 32 33 34 35

4 41 42 43 44 45

5 51 52 53 54 55

0000

m

R R R R RR R R R RR R R R RR R R R RR R R R R F

−Φ⎡ ⎤ ⎡ ⎤ ⎡ ⎤⎢ ⎥ ⎢ ⎥ ⎢ ⎥Φ⎢ ⎥ ⎢ ⎥ ⎢ ⎥⎢ ⎥ ⎢ ⎥ ⎢ ⎥Φ = ⋅⎢ ⎥ ⎢ ⎥ ⎢ ⎥Φ⎢ ⎥ ⎢ ⎥ ⎢ ⎥⎢ ⎥ ⎢ ⎥ ⎢ ⎥Φ −⎣ ⎦ ⎣ ⎦ ⎣ ⎦

(A-1-51)

The reluctances in (A-1-51) are determined by the following equations.

11 C KR R R= + (A-1-52)

12 0R = (A-1-53)

13 CR R= − (A-1-54)

14 0R = (A-1-55)

15 0R = (A-1-56)

21 0R = (A-1-57)

22 G LR R R= + (A-1-58)

23 0R = (A-1-59)

24 GR R= − (A-1-60)

25 0R = (A-1-61)

31 CR R= − (A-1-62)

32 0R = (A-1-63)

33 D C B A JR R R R R R= + + + + (A-1-64)

34 AR R= (A-1-65)

( )35 J AR R R= − + (A-1-66)

41 0R = (A-1-67)

42 GR R= − (A-1-68)

43 AR R= (A-1-69)

44 I A H G FR R R R R R= + + + + (A-1-70)

( )45 A IR R R= − + (A-1-71)

51 0R = (A-1-72)

Appendix 1

243

52 0R = (A-1-73)

( )53 A JR R R= − + (A-1-74)

( )54 A IR R R= − + (A-1-75)

55 E J A IR R R R R= + + + (A-1-76) Therefore, the fluxes Aφ , Bφ , Cφ , Dφ , Eφ , Fφ , Gφ , Hφ , Iφ , Jφ , Kφ , and Lφ in Fig. 5-3-9 are determined by

3 4 5Aφ = −Φ −Φ +Φ (A-1-77)

3Bφ = −Φ (A-1-78)

1 3Cφ = Φ −Φ (A-1-79)

3Dφ = −Φ (A-1-80)

5Eφ = −Φ (A-1-81)

4Fφ = −Φ (A-1-82)

2 4Gφ = Φ −Φ (A-1-83)

4Hφ = −Φ (A-1-84)

4 5Iφ = Φ −Φ (A-1-85)

3 5Jφ = Φ −Φ (A-1-86)

1Kφ = Φ (A-1-87)

2Lφ = Φ (A-1-88)

Appendix 1

244

Appendix 2

245

Appendix 2 The magnetic reluctances, the fluxes, the flux linkage and the no-load induced voltage in the equivalent circuit, Fig. 7-3-4 in Chapter 7 are determined by the following equations.

1 17AR R R= + (A-2-1)

2BR R= (A-2-2)

3 4 5 5 6C a bR R R R R R= + + + + (A-2-3)

7 8D aR R R= + (A-2-4)

8 9 10E b aR R R R= + + (A-2-5)

10 11F bR R R= + (A-2-6)

12 13 13 14 15G a bR R R R R R= + + + + (A-2-7)

16HR R= (A-2-8)

27 28 29IR R R R= + + (A-2-9)

34 35 36JR R R R= + + (A-2-10)

20 21KR R R= + (A-2-11)

30 31LR R R= + (A-2-12) From the above reluctance equations and Kirchhoff’s voltage law, the flux equations are formulated as follows. ( ) 1 2 3 4 5 .80 0K C D C D m aR R R R R F+ + ⋅Φ + ⋅Φ + ⋅−Φ + ⋅Φ + ⋅−Φ = ( A-2-13)

( )1 2 3 4 5 .100 0L F G G F m bR R R R R F⋅Φ + + + ⋅Φ + ⋅Φ + ⋅−Φ + ⋅−Φ = (A-2-14)

( ) ( )1 2 3 4 50 0C C B A J A J AR R R R R R R R⋅ −Φ + ⋅Φ + + + + ⋅Φ + ⋅Φ + + ⋅−Φ = (A-2-15 )

( ) ( )1 2 3 4 50 0G A I A H G A IR R R R R R R R⋅Φ + ⋅−Φ + ⋅Φ + + + + ⋅Φ + + ⋅−Φ = (A-2-16 )

( ) ( )( )

1 2 3 4

5 .8 .8 .10 .10

D F A J I A

E D J A I F m a m b m a m b

R R R R R R

R R R R R R F F F F

⋅−Φ + ⋅−Φ + + ⋅−Φ + + ⋅−Φ

+ + + + + + ⋅Φ = − − − − ( A-2-17)

When the magnetic reluctances R and the magneto-motive force due to the magnets mF are known, the flux Φ is obtained by (A-2-18).

Appendix 2

246

1mR F−Φ = ⋅ (A-2-18)

The flux equations of (A-2-13), (A-2-14), (A-2-15), (A-2-16), and (A-2-17) are transformed into (A-2-19) by (A-2-18). Using (A-2-19), the fluxes 1Φ , 2Φ , 3Φ , 4Φ and 5Φ are thus obtained.

11 11 12 13 14 15 .8

2 21 22 23 24 25 .10

3 31 32 33 34 35

4 41 42 43 44 45

5 51 52 53 54 55 .8 .8 .10 .10

00

m a

m b

m a m b m a m b

R R R R R FR R R R R FR R R R RR R R R RR R R R R F F F F

−Φ⎡ ⎤ ⎡ ⎤ ⎡ ⎤⎢ ⎥ ⎢ ⎥ ⎢ ⎥Φ⎢ ⎥ ⎢ ⎥ ⎢ ⎥⎢ ⎥ ⎢ ⎥ ⎢ ⎥Φ = ⋅⎢ ⎥ ⎢ ⎥ ⎢ ⎥Φ⎢ ⎥ ⎢ ⎥ ⎢ ⎥⎢ ⎥ ⎢ ⎥ ⎢ ⎥Φ − − − −⎣ ⎦ ⎣ ⎦ ⎣ ⎦

(A-2-19)

The reluctances in (A-2-19) are determined by the following equations.

11 K C DR R R R= + + (A-2-20)

12 0R = (A-2-21)

13 CR R= − (A-2-22)

14 0R = (A-2-23)

15 DR R= − (A-2-24)

21 0R = (A-2-25)

22 L F GR R R R= + + (A-2-26)

23 0R = (A-2-27)

24 GR R= − (A-2-28)

25 FR R= − (A-2-29)

31 CR R= − (A-2-30)

32 0R = (A-2-31)

33 C B A JR R R R R= + + + (A-2-32)

34 AR R= (A-2-33)

( )35 J AR R R= − + (A-2-34)

41 0R = (A-2-35)

42 GR R= − (A-2-36)

43 AR R= (A-2-37)

44 I A H GR R R R R= + + + (A-2-38)

Appendix 2

247

( )45 A IR R R= − + (A-2-39)

51 DR R= − (A-2-40)

52 FR R= − (A-2-41)

( )53 A JR R R= − + (A-2-42)

( )54 I AR R R= − + (A-2-43)

55 E D J A I FR R R R R R R= + + + + + (A-2-44) From the above calculation progress and the continuity of flux, the fluxes Aφ , Bφ , Cφ , Dφ , Eφ ,

Fφ , Gφ , Hφ , Iφ , Jφ , Kφ , and Lφ in Fig. 7-3-4 are obtained as the following equations.

3 4 5Aφ = −Φ −Φ +Φ (A-2-45)

3Bφ = −Φ (A-2-46)

1 3Cφ = Φ −Φ (A-2-47)

1 5Dφ = Φ −Φ (A-2-48)

5Eφ = −Φ (A-2-49)

2 5Fφ = Φ −Φ (A-2-50)

2 4Gφ = Φ −Φ (A-2-51)

4Hφ = −Φ (A-2-52)

4 5Iφ = Φ −Φ (A-2-53)

3 5Jφ = Φ −Φ (A-2-54)

1Kφ = Φ (A-2-55)

2Lφ = Φ (A-2-56) From the flux equations expressed above and Faraday’s law, the flux linkage λ and the no-load voltage pe are given as

cslotNλ φ= ⋅ (A-2-57)

pdedtλ

= (A-2-58)

Appendix 2

248

Appendix 3

249

Appendix 3 The procedure to determine the structural dimensions of the proposed generator in Chapter 8 is made as the following steps. (Step1) Determine the parameters by the electromagnetic design: actrotorm . , gB , )( dnF=σ , E , gl ,

rhl , ml , ryh , ssh , inactρ , soD , roD , riD , siD , side1gA . , rbn

(Step2) Define gl0250.max =λ , gl0250.max =δ

(Step3) Calculate rht , sht , rhm and shm

by ( )( )

3 4rb

3rhrirodn

rh 1nE812lddFt

max. δ+−

= , ( )3

3shsisodn

sh E812lddFt

max. δ−

= , ( )2

tDDm inactrh2ri

2ro

rhρπ −

= and

( )2

tDDm inactsh2si

2so

shρπ −

=

(Step4) Assume rhrb mm =

(Step5) Calculate rotorm by rbrhactrotorrotor mmmm ++= .

(Step6) Calculate airV by water

rotorair

m11Vρ.

=

(Step7) Calculate rbl by ( )2ri

2ro

airrb DD

V4l−

=π

(Step8) Calculate rbt by ( )( )

dn ro ri rbrb

rb max

F D D lt

4E n 2 λ−

=+

(Step9) Re-calculate rbm by

( ) ( )2 2

2 22 2 ro ri ro rirb inact rb ro ro rb ri rb ri rb rb

D D D Dm l D D 2t D 2t D n 2t4 2 2π ρ

⎧ ⎫⎡ ⎤+ +⎪ ⎪⎛ ⎞ ⎛ ⎞⎡ ⎤ ⎡ ⎤= − − + + − + − −⎢ ⎥⎨ ⎬⎜ ⎟ ⎜ ⎟⎣ ⎦ ⎣ ⎦ ⎝ ⎠ ⎝ ⎠⎢ ⎥⎪ ⎪⎣ ⎦⎩ ⎭

(Step10) Compare rbm calculated in (Step9) with rbm calculated previously. If,

( ) ( )rb Step9 rb previousm m accuracy− ≤ , then go to the next step. If not, go back to (Step5).

(Step11) Re-calculate rotorm , airV , rbl and rbt using the equations in (Step5), (Step6), (Step7) and

(Step8) (Step12) Calculate sbl by ( )2sb rb ss sy g m ryl l h h l l h= + + + + +

Appendix 3

250

(Step13) Calculate sbt by ( )max8

dn so si sbsb

F d d lt

Eλ−

=

(Step14) Calculate sbm by ( ) ( ) 2 22 2sb sb inact so so sb si sb sim l D D 2t D 2t D

4π ρ ⎡ ⎤ ⎡ ⎤= − − + + −⎣ ⎦ ⎣ ⎦

(Step15) Calculate inactm , fluidm and totalm

Table A-3-1 gives the structural dimensions of the rotating part and the stationary part obtained through both the calculation by the above procedure and (8-3) to (8-6) in Chapter 8 and the 3D FEA. The structure for the calculation and the 3D FEA is illustrated in Fig. A-3-1. The structural deformation and the structural mass obtained through both the calculation and the 3D FEA are given in Table A-3-2.

Table A-3-1: Structural dimensions of new generators by calculation and 3D FEA

5 MW 10 MW 20 MW

Cal. FEA Cal. FEA Cal. FEA

Axial length, lrb [m] 1.19 1.19 1.33 1.33 1.49 1.49

Front thickness, Rear thickness, trh [mm]

14.6 14.6 18.4 17 20.5 21

Outer diameter, Dro [m] 7.14 7.14 10.07 10.07 14.26 14.26

Center diameter, Dr [m] 6.3 6.3 8.8 8.8 12.6 12.6

Inner diameter, Dri [m] 5.46 5.46 7.69 7.69 10.94 10.94

Rotor

Outer thickness, Center thickness, Inner thickness, trb [mm]

1.02 1.5 1.16 1.6 1.27 1.8

Air gap length [mm] 6.3 6.3 8.9 8.9 12.6 12.6

Axial length, lsb [m] 1.55 1.55 1.72 1.72 1.9 1.9

Front thickness, Rear thickness, tsh [mm]

114.2 125 158.8 164 218.8 215

Outer diameter, Dso [m] 7.34 7.34 10.27 10.27 14.46 14.46

Center diameter, Ds [m] 6.3 6.3 8.8 8.8 12.6 12.6

Inner diameter, Dsi [m] 5.26 5.26 7.49 7.49 10.74 10.74

Stator

Outer thickness, Inner thickness, tsb [mm]

4.1 3.8 4.3 4 4.5 4.2

Appendix 3

251

Table A-3-2: Structural deformation and mass of new generators by calculation 3D FEA

5 MW 10 MW 20 MW

Cal. FEA Cal. FEA Cal. FEA

Air gap deformation

Rotor structure [mm] 0.315 0.311 0.444 0.427 0.63 0.6

Stator structure [mm] 0.315 0.314 0.444 0.44 0.63 0.627

Structural mass

Rotor structure, mrotor.inact [ton]

4.6 4.1 11.1 11.2 23.7 25.9

Stator structure, mstator.inact [ton]

38.1 42.3 97.9 102.9 253.7 254.4

Fig. A-3-1 Models of rotating part and stationary part for calculation and 3D FEA

Appendix 3

252

List of publications

253

List of Publications

Journal papers [1] D. Bang, H. Polinder, G. Shrestha, and J.A. Ferreira, “Promising direct-drive generator

system for large wind turbines”, EPE Journal, Vol. 13, pp. 7-13, September 2008. [2] G. Shrestha, H. Polinder, D. Bang, and J.A. Ferreira, “Structural Flexibility: A Solution for

Weight Reduction of Large Direct-Drive Wind-Turbine Generators”, IEEE Trans. Energy Conversion, Vol. , Issue: 99, pp. 1-9, 2010.

[3] J. Chang, J. Kim, D. Kang, and D. Bang, “Transverse flux linear machine with high thrust for direct drive applications”, Journal of Magnetics, Vol. 15, No. 2, pp. 64-69, June 2010.

Patents [1] D. Bang, H.G. Kim, and S.J. Cho, “Linear motor and control method”, Registration No.:10-

0728078, 13 pages, Korea patent office (KR) granted patent publication on 13-Jun.-2007. [2] D. Bang, Y.M. Yoon, I.H. Moon, and H.J. Kim, “Linear motor”, Registration No.:10-

0753460, 12 pages, Korea patent office (KR) granted patent publication on 31-Aug.-2007. [3] D. Bang, H. Polinder, J.A. Ferreira, and B.J. Kim, “Direct-drive electric machine”,

Registration No.:10-0969682, 26 pages, Korea patent office (KR) granted patent publication on 14-Jul.-2010.

Conference papers [1] H. Polinder, D. Bang, R.P.J.O.M. van Rooij, and A.S. McDonald, “10 MW Wind turbine

direct-drive generator design with pitch or active speed stall control”, in Proc. IEEE International Electric Machines and Drives Conferences (IEMDC), May 3-5 2007.

[2] D. Bang, H. Polinder, G. Shrestha, J.A. Ferreira, and R.P.J.O.M. van Rooij, “New active speed stall control compared to pitch control for direct-drive wind turbines”, in Proc. EWEC (European Wind Energy Conference & Exhibition), Milan, Italy, May 7-10 2007.

[3] G. Shrestha, H. Polinder, D. Bang, and J.A. Ferreira, “Direct drive wind generator with magnetic bearing”, in Proc. European Offshore Wind Conference & Exhibition, Berlin, Germany, December 4-6, 2007.

[4] D. Bang, H. Polinder, G. Shrestha, and J.A. Ferreira, “Promising direct-drive generator system for large wind turbines”, in Proc. Wind Power to the Grid – EPE Wind Energy Chapter 1st Seminar (EPE-WECS), Delft, the Netherlands, March 27-28, 2008.

List of publications

254

[5] D. Bang, H. Polinder, G. Shrestha, and J.A. Ferreira, “Review of generator systems for direct-drive wind turbines”, in Proc. EWEC (European Wind Energy Conference & Exhibition), Brussels, Belgium, March 31 - April 3, 2008.

[6] D. Bang, H. Polinder, G. Shrestha, and J.A. Ferreira, “Comparative design of radial and transverse flux PM generators for direct-drive wind turbines”, in Proc. IEEE International Conference on Electrical Machines (ICEM) , Vilamoura, Portugal, September 6-9 2008.

[7] G. Shrestha, H. Polinder, D. Bang, and J.A. Ferreira, “A new concept for weight reduction of large direct drive machines”, in Proc. IEEE International Conference on Electrical Machines (ICEM) , Vilamoura, Portugal, September 6-9 2008.

[8] D. Bang, H. Polinder, G. Shrestha, and J.A. Ferreira, “Design of a lightweight transverse flux permanent magnet machine for direct-drive wind turbines”, in Proc. 2008 IEEE Industry Applications Society Annual Meeting, Edmonton, Canada, October 5-9 2008.

[9] G. Shrestha, D. Bang, H. Polinder, and J.A. Ferreira, “Energy converters for future wind turbines”, in Proc. Global Wind Power China, Beijing, 2008.

[10] D. Bang, H. Polinder, G. Shrestha, and J.A. Ferreira, “Possible solutions to overcome drawbacks of direct-drive generator for large wind turbines”, in Proc. EWEC (European Wind Energy Conference & Exhibition), Marseille, France, March 16-19 2009.

[11] D. Bang, H. Polinder, G. Shrestha, and J.A. Ferreira, “Ring-shaped transverse flux PM generator for large direct-drive wind turbines”, in Proc. IEEE International conference on power electronics and drive systems”, Taipei, Taiwan, November 2-5 2009.

[12] J. Chang, J. Kim, D. Kang, and D. Bang, “A novel transverse flux machine for direct drive applications”, in Proc. Conference on the Computation of Electromagnetic Fields, Florianópolis, Brasil, Nov. 22-26 2009.

[13] D. Bang, H. Polinder, J.A. Ferreira and S.S. Hong, “Structural mass minimization of large direct-drive wind generators using a buoyant rotor structure”, in Proc. 2010 IEEE Energy Conversion Congress and Exposition, Atlanta, Georgia, September 12-16 2010.

[14] S. Ani, D. Bang, H. Polinder, J.Y. Lee, S.R. Moon and D.H. Koo, “Human powered axial flux permanent magnet machines: Review and comparison”, in Proc. 2010 IEEE Energy Conversion Congress and Exposition, Atlanta, Georgia, September 12-16 2010.

[15] D. Bang, S. Ani, H. Polinder, J.Y. Lee, S.R. Moon and D.H. Koo, “Design of portable axil flux permanent magnet machines for human power generation”, in Proc. 2010 IEEE International Conference on Electrical Machines and Systems, Incheon, Korea, October 10-13 2010.

Summary

255

Summary In order to maximize the energy harnessed, to minimize the cost, to improve the power quality and to ensure safety together with the growth of the size, various wind turbine concepts have been developed during last three decades. Different generator systems such as geared and direct-drive generator systems have been developed. Based on energy yield, reliability and maintenance problems, direct-drive generator systems have been regarded as a better concept than geared generator systems. Among different generator types, the permanent magnet synchronous generator (PMSG) is superior in overall efficiency. The PMSG has also been discussed as a concept with higher force density than the electrically excited (EE) generators. PMSG can thus be small and lightweight compared to induction generators and direct-drive electrically excited synchronous generators (EESG DD). Therefore, this thesis focuses on the research of the direct-drive permanent magnet synchronous generator system (PMSG DD) for large wind turbines. The main objective of the thesis is to find a design solution for large direct-drive wind generators with highest possible energy yield and lowest possible cost. In order to achieve the objective, the following issues are covered in the thesis.

Permanent magnet machines for direct-drive To identify the active mass-competitiveness of permanent magnet (PM) generators for large direct-drive applications, various configurations of PM machines such as the radial flux (RF), axial flux (AF) and transverse flux (TF) PM machines were surveyed. According to the survey of various PM machines, the following two PM machines were selected for the design of large direct-drive PM wind generators. • Slotted surface-mounted RFPM machine with inner rotor and rare earth magnets • Flux-concentrating TFPM machines with single windings

Mechanical structure of direct-drive wind generators To identify the structural mass-competitiveness and the total mass-competitiveness of direct-drive wind generators, an overview of different generators discussed in scientific literature was given based on the structural mass, total mass and toque rating.

Summary

256

In order to estimate the total mass of direct-drive and geared wind generators that are currently in use, the total mass/torque ratio of each generator concept was assumed to remain constant while scaling up the power rating of wind turbines. The doubly-fed induction generator with a three-stage gearbox (DFIG 3G) was taken as a reference generator. The direct-drive electrically excited synchronous generator (EESG DD) was addressed as the heaviest generator. The direct-drive permanent magnet synchronous generator (PMSG DD) with air gap bearings was addressed as the lightest generator among the other direct-drive generators. The total mass of the PMSG DD with air gap bearings was comparable with the mass of the DFIG 3G.

Electromagnetic modelling of PM machines A generalized electromagnetic design model was developed for both an RFPM machine and a TFPM machine. To verify the analysis model of TFPM machines, no-load induced voltage obtained from the analysis model was compared with the no-load voltages obtained through three-dimensional finite element analyses (3D FEA) and measurement. The no-load voltage per pole pair in the analysis model was 8 % higher than the voltage obtained through 3D FEA, and 2 % lower than the measured voltage. To validate the analysis model in the case with a load, the force obtained through the analysis model was compared with the force obtained in 3D FEA and static force measurement. At the nominal stator current, the force obtained by the analytical model is 10 % and 7 % larger than the force obtained by the 3D FEA and by the static force measurement, respectively.

Comparison of PM generators for large direct-drive wind turbines A comparative design of PM generators for large direct-drive wind turbines was represented to assess different configurations of the generators based on the active mass, losses and cost. The following five PM generators were selected for the comparative design. • RFPMG: a surface-mounted RFPM generator with full pitch windings • TFPMG-1: a single-sided, single winding flux-concentrating TFPM generator with a U-core • TFPMG-2: a double-sided, single winding flux-concentrating TFPM generator with a C-core • TFPMG-3: a single-sided, single winding flux-concentrating TFPM generator with a U-core

and passive rotor • TFPMG-4: a double-sided, single winding flux-concentrating TFPM generator with a C-core

and passive rotor Among the five different generators, the TFPMG-1 was addressed as the lightest generator for 5 MW wind turbines. The RFPMG was addressed as the lightest genertor for 10 MW wind turbines. The TFPMG-2 had the smallest loss and the lowest cost compared to the other generators for

Summary

257

both 5 MW and 10 MW turbines. The TFPMG-4 was addressed as the generator with the largest mass, the highest cost and the largest loss for both 5 MW and 10 MW turbines. The TFPMG-4 and TFPMG-3 were more expensive than the other generators, since both generators consist of large mass of permanent magnets which are the most expensive active material. The TFPMG-2 and TFPMG-4 were more complicated in constructing because these two genertors have double-sided air gaps. Therefore, the TFPMG-1 was selected as a suitable generator for large direct-drive wind turbines.

New configuration of TFPM generator for large direct-drive wind turbines A flux-concentrating transverse flux permanent magnet (TFPM) machine with multiple-modules segmented, multiple-slots per phase and racetrack-shaped copper windings was proposed to reduce active mass and to facilitate manufacture and maintenance. An analytical design model of the proposed TFPM generator was derived, and the model was verified by the experiment of a downscaled TFPM generator. Using the analytical design model, the proposed TFPM generators with various numbers of slots per phase were designed for 5 MW and 10 MW direct-drive wind turbines. The following ten different PM generator configurations were discussed in the analytical design. • RFPMG: a surface-mounted TFPM generator with full pitch windings • TFPMG-1: a single-sided, single winding flux-concentrating TFPM generator with a U-core • CP_1mod_ls: single-sided, single winding flux-concentrating TFPM generator with claw poles,








poles, eight slots per phase and limited pole area

Summary

258

Among different PM generators, CP_1mod_Ap was addressed as the lightest generator. CP_2mod_Ap was addressed as the second lightest generator. CP_8mod_ls was addressed as the heaviest generator for 5 MW and 10 MW wind turbines. From the design of the generators for 5 MW wind turbines, it was taken that the active mass of TFPMG-1, CP_2mod_ls, CP_1mod_Ap, CP_2mod_Ap and CP_4mod_Ap were smaller than that of RFPMG. The active mass of CP_1mod_Ap, CP_2mod_Ap and CP_4mod_Ap were smaller than that of RFPMG for 10 MW wind turbines.

Minimizing the structural mass of large direct-drive wind generators The following generator concepts were proposed to minimize the structural mass of large direct-drive wind generators. • a ring-shaped generator without a main shaft and torque arms • multi-sets of generator and converter • a generator with a buoyant rotor • a generator with a simplified bearingless drive and/or hydraulic bearings to guide the air gap The proposed generator integrated with the concepts described above was designed for large direct-drive wind turbines with 5 MW, 10 MW and 20 MW power ratings. From the design, it was concluded that the proposed direct-drive generator was lighter than traditional direct-drive PM generators at all power ratings. When scaled up to larger than 9 MW, the proposed generator becomes lighter than a geared generator (DFIG 3G).

Samenvatting

259

Samenvatting In de afgelopen 30 jaar is een verscheidenheid aan windturbineconcepten ontwikkeld, allen met het doel om het aandeel nuttig gebruikte energie te maximaliseren, om de kosten te minimaliseren, om de kwaliteit van de vermogensgolfvormen te verbeteren, en om de veiligheid te kunnen blijven garanderen bij toename [opschaling] van de turbinegrootte. Zo zijn verschillende machineconcepten met tandwielkast en zonder (direct-aangedreven) ontwikkeld. Direct-aangedreven machines worden [algemeen] gezien als een beter concept dan de versies met tandwielkast, als gemeten wordt in energieopbrengst, bedrijfszekerheid en onderhoudsproblematiek. De permanente magneet synchrone machine (PMSG) is [binnen deze klasse van machines] superieur wat betreft het totale machinerendement, en er wordt bovendien beargumenteerd [in de literatuur] dat dit concept een hogere krachtdichtheid heeft dan de elektrisch geëxciteerde (EE) machines. Het PMSG-concept kan dus [in theorie] leiden tot een meer compacte en lichtgewicht machine, in vergelijking met inductiemachines en direct-aangedreven elektrisch geëxciteerde machines (EESG DD). Om deze reden[en] is als focus van dit proefschrift gekozen onderzoek naar de direct-aangedreven permanente magneet synchrone machine (PMSGDD) voor toepassing in hoogvermogen windturbines. Het doel van dit proefschrift is om een ontwerpoplossing te vinden voor hoogvermogen direct-aangedreven windturbines, welke voldoet aan de eis van een zo hoog mogelijke energieopbrengst tegen zo laag mogelijke kosten. Om dit doel te bereiken zijn de volgende ontwerpaspecten behandeld in dit proefschrift.

Direct-aangedreven permanente magneet machines Om te kunnen bepalen of hoogvermogen permante magneet (PM) direct-aangedreven machines kunnen concurreren op actieve massa met andere machineconcepten, zijn verschillende PM-configuraties bestudeerd, zoals radiale flux (RF), axiale flux (AF) en transversale flux (TF) machines. Naar aanleiding van deze studie zijn de volgende twee PM-configuraties geselecteerd voor het [verdere, basis, detail] ontwerp van een hoogvermogen direct-aangedreven PM windturbine. • De gegroefde oppervlakte-gemonteerde RFPM machine met binnenrotor en zeldzame-aarde

magneten • De flux-concentrerende TFPM machine met enkele wikkelingen

Het mechanische ontwerp van direct-aangedreven windturbines Om de concurrentiepositie van het direct-aangedreven machineconcept op het gebied van constructieve massa, totale massa en nominaal krachtkoppel te bepalen, is een overzicht van de wetenschappelijke literatuur gemaakt voor verschillende machineconcepten. Voor dit overzicht is een schatting gemaakt van de totale massa van commercieel verkrijgbare direct-aangedreven

Samenvatting

260

machines en machines met tandwielkast, onder de aanname van een constante verhouding tussen massa en koppel bij opschaling van het nominale machinevermogen. De dubbelgevoede inductiemachine met 3-traps tandwielkast (DFIG 3G) diende hiervoor als referentiemachine. De direct-aangedreven elektrisch geëxciteerde synchrone machine (EESG DD) kwam bij deze analyse als zwaarste uit de bus, en de direct-aangedreven permanente magneet synchrone machine (PMSG DD) met lagering in de luchtspleet als minst zware [lichtste] machine. Het totale gewicht van deze laatste machine was vergelijkbaar met het gewicht van de DFIG 3G referentiemachine.

Elektromagnetische modellering van PM machines Voor zowel de RFPM als de TFPM machine is een gegeneraliseerd elektromagnetisch ontwerpmodel ontwikkeld. Om het analytische model van de TFPM machine te verifieren is dit model vergeleken met een 3-dimensionale eindige elementen computersimulatieanalyse (3D FEA) en met praktijkmetingen[, voor de situatie van een onbelaste machine]. De geinduceerde spanning voor een onbelaste machine was in het analysemodel 8 procent hoger dan de 3D FEA-simulatie liet zien, en 2 procent lager dan bij de praktijkmetingen. Om het model te valideren voor een belaste machine zijn de door het model voorspelde krachten vergeleken met de 3D FEA-simulatie en de praktijkmetingen. Bij nominale statorstromen voorspelde het model respectievelijk een 10 procent (3D FEA) en 7 procent (statische-kracht praktijkmeting) hogere kracht.

Een vergelijking van hoogvermogen direct-aangedreven PM machines onderling Een vergelijkend ontwerp is gedaan van de volgende 5 verschillende hoogvermogen direct-aangedreven PM machineconcepten, gebaseerd op actieve massa, verliezen en kosten. • RFPMG: een oppervlakte-gemonteerde RFPM machine met volledig-gegroefde wikkelingen • TFPMG-1: een flux-gecontreerde TFPM machine met U-vormige kern en een enkele

wikkeling aan een enkelvoudige kant • TFPMG-2: een flux-gecontreerde TFPM machine met C-vormige kern en een enkele

wikkeling aan beide kanten • TFPMG-3: een flux-gecontreerde TFPM machine met U-kern en een enkele wikkeling aan

een enkelvoudige kant, met passieve rotor • TFPMG-4: een flux-gecontreerde TFPM machine met C-vormige kern en een enkele

wikkeling aan beide kanten, met passieve rotor Als resultaat zijn respectievelijk de TFPMG-1 en de RFPMG concepten geïdentificeerd als de lichtste machines voor een nominaal ontwerpvermogen van respectievelijk 5 MW en 10 MW. De TFPMG-2 machine was zowel het meest efficiënt als het goedkoopst voor zowel 5 MW als 10 MW. De TFPMG-4 machine scoorde het slechtste van alle concepten wat betreft massa, kosten en verliezen, voor zowel 5 MW als 10 MW. De hoge kosten van TFPMG 3 en 4 worden verklaard door de grote massa van dure permanente magneten (actieve massa). De dubbelzijdige luchtspleten van TFPMG 2 en 4 leiden tot de relatief meest gecompliceerde constructie van deze machines. Om deze redenen is de TFPMG-1 machine gekozen als het meest geschikte basisconcept voor toepassing in hoogvermogen direct-aangedreven windturbines.

Samenvatting

261

Nieuw TFPM-concept voor hoogvermogen direct-aangedreven windturbines Het basis TFPM machineconcept is uitgewerkt tot een nieuw machineconcept met de volgende innovaties: • Een machineontwerp met meerdere segmenten (modules) • Meerdere groeven per fase • Racebaan-vormige koperen wikkelingen Deze innovaties moesten leiden tot een gereduceerde actieve massa, een betere fabriceerbaarheid en minder onderhoud. Een analytisch model van het nieuwe TFPM machineconcept is ontwikkeld, en dit model is experimenteel geverifieerd door een schaalmodel TFPM machine. Met het analytische model zijn de volgende 10 verschillende machineconfiguraties, met variabel aantal groeven per fase en variabel nominaal vermogen (5 en 10 MW), bestudeerd. • RFPMG: a surface-mounted TFPM generator with full pitch windings • TFPMG-1: a single-sided, single winding flux-concentrating TFPM generator with a U-core • CP_1mod_ls: single-sided, single winding flux-concentrating TFPM generator with claw poles,








poles, eight slots per phase and limited pole area De CP_1mod_Ap kwam hierbij als lichtste machinevariant uit de bus. CP_8mod_ls was voor zowel 5 MW als 10 MW het zwaarste. De actieve massa van TFPMG-1, CP_2mod_ls, CP_1mod_Ap, CP_2mod_Ap en CP_4mod_Ap varianten voor 5 MW was kleiner dan dat van de RFPMG-variant. CP_1mod_Ap, CP_2mod_Ap en CP_4mod_Ap waren ook nog bij 10 MW minder zwaar dan RFPMG-variant.

Samenvatting

262

Het minimaliseren van de constructieve massa van hoogvermogen direct-aangedreven machines De volgende machineconcepten zijn voorgesteld om de constructieve massa van hoogvermogen direct-aangedreven machines te minimaliseren: • Een ringvormige machine zonder middenas en krachtarmen • Een gesegmenteerde machine en vermogensomzetter • Een machine met een drijvende rotor • Een machine met een versimpelde magnetische lagering en/of hydraulische lagering ter

begeleiding van de luchtspleet. De voorgaande concepten zijn geintegreerd in een direct-aangedreven machineontwerp voor nominale vermogens van 5, 10 en 20 MW. Het resultaat van dit machineontwerp was voor elk van de voorgaande vermogens een lichtere machine, vergeleken met traditionele direct-aangedreven machines. Vanaf een nominaal vermogen van 9 MW [en hoger] is het voorgestelde PM machineconcept lichter dan een machine met tandwielkast (DFIG 3G).

Curriculum Vitae

263

Curriculum Vitae Deok-je Bang was born in Busan, Rep. of Korea. He received the B.Sc. and M.Sc. degrees in mechanical engineering from the Pukyong National University, Busan, Republic of Korea in 1996 and 1998, respectively. Since 2006, he has been working toward the Ph.D. degree at Electrical Power Processing Group, Delft University of Technology in The Netherlands. From 1998 to 2006, he was a researcher in the area of electrical machines and automatic transportations systems in Korea. His current research interests include direct-drive generators for wind turbines.

Design of Transverse Flux Permanent Magnet Machines for ...

Documents

Transcript of Design of Transverse Flux Permanent Magnet Machines for ...