University of WollongongResearch Online

University of Wollongong Thesis Collection University of Wollongong Thesis Collections

1997

(Bi,Pb)2Sr2Ca2Cu3O10+x/Ag : high Tcsuperconductors and their applications in anelectrical fault current limiter and an electronic highvoltage generatorJin Jian-xunUniversity of Wollongong

Research Online is the open access institutional repository for theUniversity of Wollongong. For further information contact the UOWLibrary: research-pubs@uow.edu.au

Recommended CitationJian-xun, Jin, (Bi,Pb)2Sr2Ca2Cu3O10+x/Ag : high Tc superconductors and their applications in an electrical fault current limiter andan electronic high voltage generator, Doctor of Philosophy thesis, , University of Wollongong, 1997. http://ro.uow.edu.au/theses/2074

(Bi,Pb)2Sr2Ca2Cu301o+x /Ag HIGH Tc SUPERCONDUCTORS AND THEIR APPLICATIONS

IN AN ELECTRICAL FAULT CURRENT LIMITER AND AN ELECTRONIC HIGH VOLTAGE GENERATOR

A thesis submitted in partial fulfillment of the requirements for the award of the degree of

DOCTOR OF PHILOSOPHY

from

UNIVERSITY OF W O L L O N G O N G

by

JIAN-XUN JIN

Bachelor of Eng., Master of Sci.

DEPARTMENT OF MATERIALS ENGINEERING, and CENTER FOR SUPERCONDUCTING A N D ELECTRONIC MATERIALS

July, 1997

CANDIDATE'S CERTIFICATE

This is to certify that the work presented in this thesis was carried out by the

candidate in the Department of Materials Engineering, the University of Wollongong, and

the School of Electrical Engineering, the University of New South Wales, and has not

been submitted for a degree to any other university or institution.

Jianxun Jin

i

ACKNOWLEDGMENTS

I would like to express m y gratitude to m y supervisor, Professor Shixue Dou, and

my co-supervisors Associate Professor Colin Grantham, and Associate Professor

Huken Liu for providing their support and advice to the project carried out.

Thanks to the staff of the Department of Materials Engineering, the University of

Wollongong, and the members of the Centre of Superconducting Materials and

Electronic Materials, the University of Wollongong. Special thanks to Prof. D. Dunne,

Dr. Y.C. Guo, Dr. J.N. Li, Dr. J. Horvat, Mr. R. Bhasale, Mr. N. Cui, Mr. M. Ionescu,

Dr. Q.Y. Hu, Mr. R. Kinnell, Mr. A. Bult, and Mr. B.D. Jong.

Some electrical measurements of the high temperature superconducting devices

mainly took place in the laboratory of the School of Electrical Engineering, the

University of NSW. My appreciation to the staff there, especially to Mr. B. Blears, Mr.

P. Partrigeon and Mr. Z.Y. Liu.

Thanks also to the following people for providing help, advice and discussion:

Prof. X.Y. Li, A/Prof. Z.J. Zeng, and Dr. X.P. Gu.

Finally, I would like to take this opportunity to express my appreciation for the

support from the APRA (Australian Postgraduate Research Award) scholarship and the

superconductivity scholarship from the University of Wollongong and its Centre of

Superconducting Materials and Electronic Materials.

ii

ABSTRACT

Ag-clad ceramic (Bi,Pb)2Sr2Ca2Cu30io+x high critical temperature

superconducting (HTS) wire has been developed at the University of Wollongong

with respect to its critical cunent density, long length and mechanical flexibility. In

the present work, the HTS wire was investigated for making a HTS coil, an electrical

fault cunent limiter and an electronic high voltage generator.

HTS coils were made using Ag-clad (Bi,Pb)2Sr2Ca2Cu30io+x HTS wire

prepared by the powder-in-tube technique. The HTS wire was characterised as a

conductor. The anisotropic HTS wire has a strongly magnetic-field-dependent critical

current, which causes critical cunent degradation when used in the form of a coil.

The magnetic field behaviour of the HTS coil was studied with respect to its critical

cunent and magnetic field properties. HTS solenoid and pancake coils were produced

using "react and wind" and "wind and react" procedures. Experiments which were

carried out included magnetic field measurements of the HTS coils in DC and AC

magnetic fields, and critical cunent ampere-turns when used as a magnetic core bias

winding. The magnetic field distribution of the HTS coils was also investigated using

a magnetic field analysis computer program. The magnetic field in the radial

direction (By)(//c-axis) at the HTS coil side's edge has the most influence on the HTS

coil performance at 77 K. The experimental results and analysis provide basic

information for the design and operation of a HTS coil made from the anisotropic

Ag-clad (Bi,Pb)2Sr2Ca2Cu30io+x wire.

The designed electrical fault cunent limiter was based on the principle of a

saturable magnetic core reactor, where a DC bias winding with a large number of

iii

ampere-turns is required to generate a high magnetic motive force (mmf) on a

magnetic core. The HTS coil was also studied for use in this device as the DC bias

winding. A small laboratory device rated at 3.5 kW was made and tested to

investigate the electrical behaviour of the saturable magnetic core fault cunent limiter

and the HTS coil in this application. The electrical behaviour of this fault cunent

limiter was also analysed using the modified device y-I characteristics, and also

considered for application in a 6 kV power distribution system. The results show that

this fault cunent limiter restricts fault cunents effectively both in the transient and

steady states.

A high voltage generator device was made and analysed with regard to using

a HTS wire inductor. A laboratory high voltage generator device was built, based on

a R-C-L resonant circuit, where large inductance and low resistance were required to

generate high voltages from a low voltage battery source. Analysis showed that by

using an HTS inductor to reduce the circuit resistance, the build-up voltage could be

significantly increased. 1.2 kV was generated using a small hybrid inductor from a 12

V battery source, and the potential build-up voltage when using a larger HTS

inductor was shown to be very high.

Both the electrical fault cunent limiter device and the electronic high voltage

generator were analysed from measurements and simulations. With improvements in

critical cunents and long lengths of Ag-clad (Bi,Pb)2Sr2Ca2Cu30io+x- wires, HTS

coils with high values of critical ampere-turns could be achieved at 77 K operating

temperature. These will enable the above HTS applications to be practicable in

electrical engineering.

iv

CONTENTS

CANDDDATE'S CERTIFICATE i

A C K N O W L E D G M E N T S ii

A B S T R A C T iii

C O N T E N T S v

1 INTRODUCTION 1

(LITERATURE SURVEY)

2 H I G H T c S U P E R C O N D U C T O R A N D ITS APPLICATIONS 6

2.1 SUPERCONDUCTIVITY 6

2.1.1 Background of Development 6

2.1.2 Basic Superconductivity in Physics 7

2.1.2.1 London and Ginzburg-Landau theories 7

2.1.2.2 Bardeen-Cooper-Schrieffer theory 8

2.1.2.3 Josephson effect 9

2.1.2.4 Type II superconductors and pinning 10

2.1.2.5 Non-metal superconductors 11

2.1.3 Electromagnetism in Superconductors 12

2.1.3.1 Critical-state model 12

2.1.3.2 Magnetic flux distribution and magnetisation 14

2.1.3.3 Hysteresis loss 14

v

2.1.3.4 Magnetic instability 16

2.1.3.5 Reversible fluxoid motion and flux creep 17

2.2 CONVENTIONAL LOW Tc SUPERCONDUCTORS AND APPLICATIONS 18

2.2.1 Development of Practical Superconducting Materials 18

2.2.2 Development of Practical Conductors 21

2.2.2.1 Conductor-techniques of low Tc superconductors 21

2.2.2.2 Factor for design of composite superconductors 25

3.2.3 Applications 28

2.2.3.1 Large scale applications 29

2.2.3.2 Superconducting electronic devices 33

2.3 DEVELOPMENT OF HIGH Tc SUPERCONDUCTORS 35

2.3.1 High Tc Superconductors 3 5

2.3.2 High Tc Superconducting Wires 36

2.3.2.1 Early stage of HTS wires 36

2.3.2.2 Development of Bi-2223/Ag wires 39

2.3.2.3 Achievement on the Bi-2223/Ag HTS wires 42

2.3.3 High Tc Superconducting Coil 45

2.3.4 Conductor-Design Consideration of High Tc Superconductor 46

2.4 POTENTIAL APPLICATIONS OF HIGH Tc SUPERCONDUCTORS IN ELECTRICAL ENGINEERING 47

2.5 DEVELOPMENT OF ELECTRICAL FAULT CURRENT LIMITER W I T H HIGH T c S U P E R C O N D U C T O R S 49

2.5.1 Fault Currents in Electrical Power Systems 49

2.5.2 Solutions to Fault Currents 52

2.5.3 Developing Fault Cunent Limiter in Electrical Power Networks 53

vi

2.5.4 High Tc Superconducting Fault Cunent Limiters 60

2.5.4.1 Developing HTS fault cunent limiters 60

2.5.4.2 High Tc superconducting shielding inductive reactor 63

2.5.4.3 Bi-2223/Ag HTS wire and its fault cunent limiter 64

2.5.4.4 Summary of the development of HTSFCLs 64

(PRACTICAL WORK)

3 Bi-2223/Ag HIGH Tc SUPERCONDUCTORS 67

3.1 INTRODUCTION 67

3.2 Bi-2223/Ag HIGH Tc SUPERCONDUCTING WIRES 68

3.2.1 Producing HTS Wires 68

3.2.2 Characteristics of the HTS Wires 70

3.2.2.1 Critical temperature Tc 70

3.2.2.2 Critical cunent L 73

3.2.2.3 Conductor magnetic field distribution 79

3.2.2.4 Critical cunent for AC operation 82

3.2.2.5 Mechanical properties 85

3.2.3 Evaluation of the Use of HTS Wire 89

3.2.3.1 Summary of the Bi-2223/Ag HTS wire 89

3.2.3.2 Application judgment 94

3.3 HIGH Tc SUPERCONDUCTING COILS 96

3.3.1 Introduction 96

3.3.2 Making Superconducting Coils with Bi-2223/Ag HTS Wire 98

3.3.2.1 Techniques 98

vii

3.3.2.2 Produced H T S coils 99

3.3.3 Experimental Results on the HTS Coil 103

3.3.3.1 The selected HTS coil 103

3.3.3.2 DC magnetic field 103

3.3.3.3 AC magnetic field 104

3.3.3.4 HTS coil used as a magnetic core bias winding and its ampere-turns 107

3.3.4 HTS Coil Magnetic Field Distribution Analysis 111

3.3.4.1 Analysis of the HTS coil magnetic field 111

3.3.4.2 Operated at 77 K with air core 111

3.3.4.3 Operated at 4.2 K with air core 120

3.3.4.4 Operated with magnetic core 120

3.3.5 Magnetic Field Properties of the HTS Coil 129

3.3.5.1 Critical cunent degeneration of the HTS coil 129

3.3.5.2 Generated magnetic field of the HTS coil 130

3.3.5.3 Basic HTS coil design consideration for magnet 131

3.3.5.4 Inductance of the HTS coil 133

3.3.5.5 Ampere-turns and magnetic core bias winding 134

3.6 SUMMARY AND CONCLUSIONS 13 5

4 HIGH Tc SUPERCONDUCTING SATURABLE MAGNETIC CORE FAULT CURRENT LIMITER 137

4.1 INTRODUCTION 137

4.2 EXPERIMENTAL DEVICE CONFIGURATION AND BEHAVIOR 138

4.2.1 Making a Testing Device 138

viii

4.2.1.1 Principle 138

4.2.1.2 Magnetic core 139

4.2.1.3 Device windings 141

4.2.2 Device Testing Results 145

4.2.2.1 Winding accuracy and transformer effect 145

4.2.2.2 Magnetic core B-H curves 146

4.2.2.3 Magnetic field and resistive loss of the normal DC winding 149

4.2.2.4 AC winding impedance 150

4.2.2.5 Tests on cunent limiting behaviours 152

4.2.3 High Tc Superconducting DC Bias 165

4.2.3.1 The HTS coil 165

4.2.3.2 The HTS coil critical cunent ampere-turns and self magnetic field 167

4.2.4 Device-U and its Results 168

4.2.4.1 Configuration 168

4.2.4.2 B-H curves 169

4.2.4.3 Tests on cunent limiting behaviour 171

4.2.4.4 Other results 174

4.2.5 Summary of Tests 176

4.3 ANALYSIS OF THE SATURABLE MAGNETIC CORE FCL 178

4.3.1 Operation Principle and \|/-I Characteristic of the FCL 178

4.3.1.1 Modified \|/-I curve 178

4.3.1.2 Section I of \|/-I curve 178

4.3.1.3 Section II of \|/-I curve 180

4.3.1.4 Section III of \|/-I curve 180

ix

4.3.2 Cunent Limiting Behavior 180

4.3.3 Magnetic Circuit Property 184

4.3.3.1 Winding relation and induced AC voltage 184

4.3.3.2 Magnetic core losses 185

4.3.3.3 Control-circuit impedance and transient response 186

4.3.4 DC Bias and HTS Winding 186

4.3.4.1 Design of the DC bias 186

4.3.4.2 HTS DC bias winding analysis 188

4.3.4.3 Application of the HTS wire for the FCL 191

4.4 CURRENT LIMITING BEHAVIORS OF THE HTS-FCL 192

4.4.1 FCL Samples and Equivalent Operation Circuit 192

4.4.2 Operation of the FCL 193

4.4.2.1 Normal operation 193

4.4.2.2 Short circuit 195

4.4.2.3 Periodic components of short circuit cunents in steady state 196

4.4.2.4 Maximum inrush cunent 197

4.5 POWER SYSTEM APPLICATION OF THE HTS-FCL 200

4.5.1 Nomenclature 200

4.5.2 Electrical Power System 202

4.5.3 System during Normal Operation 204

4.5.3.1 Equivalent circuit of the power system 204

4.5.3.2 Voltage drop and reactive power loss in normal operation 206

4.5.4 System under Fault Condition 208

4.5.4.1 Equivalent circuit for short circuit fault condition 208

4.5.4.2 Steady state short circuit cunent, 6.3 kV busbar residual voltage and inrush cunent 208

4.5.4.3 Power factor and DC time constant in short circuit transient state 213

4.5.4.4 Short circuit cunent 214

4.5.5 Appendix 216

4.6 DISCUSSION AND CONCLUSION 220

5 AN ELECTRONIC RESONANT CmCUIT WITH A HIGH Tc SUPERCONDUCTING INDUCTOR FOR HIGH VOLTAGE GENERATION 223

5.1 INTRODUCTION 223

5.2 PRINCIPLE OF OPERATION AND CONFIGURATION OF T H E HIGH V O L T A G E G E N E R A T O R 224

5.2.1 Principle of Operation 224

5.2.2 Configuration of the Generator 227

5.3 EXPERIMENT AND RESULTS 230

5.3.1 Testing Anangement 230

5.3.2 Inductor with the HTS Wire 230

5.3.3 Build-up Voltages 234

5.3.3.1 Testing of build-up voltage 234

5.3.3.2 Build-up voltages with different circuit resistance 236

5.3.3.3 Circuit cunent 236

5.4 CIRCUIT ANALYSIS 239

5.4.1 Resonant Circuit with a Superconducting Inductor 23 9

xi

5.4.1.1 Maximum build-up voltage

5.4.1.2 Ideal build-up voltage 240

5.4.1.3 Build-up voltage with a resistive circuit 242

5.4.2 Build-up Voltage and Circuit Cunent 245

5.5 DISCUSSION 251

5.5.1. Inductor with Bi-2223/Ag HTS Wire 251

5.5.2 High Voltage Generation 253

5.5.3 Applications 254

5.6 CONCLUSIONS 257

R E F E R E N C E S 258

PUBLICATIONS 278

xii

Chapter 1

INTRODUCTION

A number of superconducting copper oxide compounds have been found

since the discovery of high critical temperature (Tc) superconducting (HTS) ceramic

materials in 1986. At present, research on HTS materials is concentrating on

optimising materials for applications through processing techniques. The bismuth

strontium calcium copper oxide HTS material, (Bi,Pb)2Sr2Ca2Cu30io+x (Tc~110 K),

has been well developed in the form of a Ag-clad wire. Meanwhile yttrium barium

copper oxide, YiBa2Cu307-x (Tc~92 K), thallium barium calcium copper oxide,

TliBa2Ca2Cu309 (Tc~110 K), and also (Bi,Pb)2Sr2CaiCu208+x (Tc~85 K), have also

demonstrated suitable properties relevant to applications. These HTS materials, e.g.,

both YiBa2Cu307_x and (Bi,Pb)2Sr2Ca2Cu30io+x, have transition temperatures higher

than 77 K (boiling point of liquid nitrogen). For general electrical applications, a

HTS wire with high critical cunent density, long length, flexible winding property

and high Tc is required to produce a winding or a coil for use in electrical equipment.

One of the most promising ceramic high Tc superconductors, Ag-clad

(Bi,Pb)2Sr2Ca2Cu3Oio+x wire, has been produced at the University of Wollongong

using the powder-in-tube technique. In the present work, the wire has been tested

with regard to its electrical, magnetic and mechanical properties for making a HTS

coil. The HTS wire produced was then used to wind a number of solenoid and

pancake coils using "react and wind" and "wind and react" procedures. The HTS

coils made have been studied for magnet and magnetic core bias winding

1

applications, and also evaluated for applications to design an electrical fault cunent

limiting device and an electronic high voltage generator.

The H T S wire is characterised as a conductor. The anisotropic H T S wire has

strong magnetic-field-dependent critical cunent, which causes critical cunent

degradation when used in the form of a coil. The magnetic field behaviour of the

H T S coil has been studied with respect to its critical cunent and magnetic field

properties. Experiments carried out include magnetic field measurements of the H T S

coils, D C magnetic field, A C magnetic field, and critical cunent ampere-turns when

used as a magnetic core bias winding. The magnetic field distribution of the H T S

coils has also been investigated with a magnetic field analysis program. The

experimental results and analysis provide basic information for design and operation

of a H T S coil made from the anisotropic Ag-clad (Bi,Pb)2Sr2Ca2Cu30io+x H T S wire.

As electrical power systems grow in capacity, the potential increased fault

cunent must be managed. The newly developed high T c superconductors have

substantial advantages for the design of a fault cunent limiter for use in electrical

power transmission and distribution systems. Ag-clad (Bi,Pb)2Sr2Ca2Cu30io+x H T S

wire was selected to evaluate the possibility of building an electrical fault cunent

limiter. The designed electrical fault cunent limiting device is based on the principle

of a saturable magnetic core reactor, where a D C bias coil with a large number of

ampere-turns is required to generate high magnetic motive force .(mmf) on a

magnetic core. The H T S coil was studied for use in this device as its D C bias

winding. A small laboratory device rated at 3.5 k W was made and tested to

investigate electrical behaviours of the saturable magnetic core fault cunent limiter

and the H T S coil for this application. For the purpose of testing and comparison, the

device combines both normal copper wire and H T S wire D C bias windings. The

2

electrical behaviour of this fault cunent limiter was analysed by using the modified

device \|/-I characteristics, and it was also considered for application in a simplified 6

k V power distribution system. The simulation and calculated results show that this

fault cunent limiter restricts fault cunent effectively both in the transient and steady

states.

A novel method for generating high voltages from a low voltage battery

source by using an electronic R-C-L circuit was developed with the consideration of

using the Ag-clad (Bi,Pb)2Sr2Ca2Cu30io+x H T S wire. The high voltage generator was

analysed with regard to using an inductor made by the H T S wire, and a small

laboratory device was made and tested. The high voltage generator device built was

based on a R-C-L resonant circuit, where large inductance and low resistance are

required to generate high voltages. The analysis and testing results show that by using

the H T S inductor to reduce the circuit resistance, the build-up voltage can be much

higher than that which can be generated by a normal copper winding inductor. 1.2 k V

has been generated with a small inductor from a 12 V battery source. The potential

build-up voltage when using a larger H T S inductor is very high, and can be evaluated

by Vc(n) = (-l)n+1(VCo+2nVB), where V Co is the initial charge on the capacitor, V B is

the battery source voltage and n is the number of voltage build-up oscillations.

Both the electrical fault cunent limiting device and the electronic high

voltage generator were studied with measurements and simulations. It was

established that with practical rated cunents, the H T S coil will be required to have a

much higher value of critical cunent ampere-turns which will in turn generate a

higher self-magnetic field. The difficulty in producing long H T S wires having a high

critical cunent density under applied magnetic field at 77 K operation temperature is

identified as the greatest barrier for practical applications in electrical engineering.

3

The H T S wires are therefore required to be improved to be capable of transporting

high critical cunent under a high magnetic field at 77 K. A better mechanical

property must also be achieved, and longer lengths are required to form the final HTS

coils for practical applications. With the improvements in producing high critical

cunent and long lengths of Ag-clad (Bi,Pb)2Sr2Ca2Cu3Oio+x HTS wires, HTS coils

with high values of critical ampere-turns could be achieved at 77 K operating

temperature with liquid nitrogen in the near future. These will enable such electrical

applications to be practicable.

This thesis mainly discusses the subjects of Ag-clad (Bi,Pb)2Sr2Ca2Cu3Oio+x

HTS wires and coils, an electrical fault cunent limiter designed with the HTS wire,

and a method to generate high voltage with potential application of the HTS wire. It

is organised as follows. Chapter 2 is the literature survey part. Relative

superconducting phenomena, developing applicable superconductors, and the

application of the superconductors in building fault cunent limiters are summarized in

this chapter. Chapter 3 covers the Ag-clad (Bi,Pb)2Sr2Ca2Cu30io+x HTS wires and

coils studied with respect to their capability to make magnets and magnetic core bias

windings. Chapter 4 describes the saturable magnetic core fault cunent limiter, where

testing, simulation and calculated results are presented. Chapter 5 introduces a method

of generating high voltages with potential use of the HTS wire in the form of a coil

with large inductance.

4

LITERATURE SURVEY

5

Chapter 2

HIGH Tc SUPERCONDUCTOR AND ITS APPLICATIONS

2.1 SUPERCONDUCTIVITY

2.1.1 Background of Development

Since the discovery of superconductivity by Kamerlingh Onnes in 1911 [1],

with an experiment on the electrical resistance of mercury (Hg), the remarkable

characteristic of lossless cunent has attracted strong interest in the field of basic

physics as well as in industry. Following that discovery, studies have been made in

physics as evidenced by the appearance of Ginzburg-Landau (GL) and Bardeen-

Cooper-Schrieffer (BCS) theories on the mechanism of superconductivity as well as

the discovery of the Josephson effect [2-4]. Besides mercury, many superconducting

materials including metals, alloys and compounds have been found since then.

However, it took about 50 years to explicate the basic mechanism and to provide the

foundation for applications, until another period began in 1960, which was

highlighted by an emphasis on applications, and it continues up to the present.

Today, superconducting technology allows high field applications to be

considered, and some high field magnets and large electrical devices have been

constructed. In the field of electronics, extensive studies on Josephson computers and

the development of commercial superconducting quantum interference device

6

(SQUID) systems are under way. However, before arriving at this stage it was

necessary to accumulate fundamental knowledge of flux motion, flux pinning, AC

loss, instability in conductors and magnets, metallurgical and material problems, and

similar concerns.

Since 1986, the appearance of oxide superconductors with high critical

temperature Tc has produced great interest in the field of basic physics as well as in

practical applications [5, 6]. The third developmental period of superconductivity

started at this time, and considerable effort is cunently being expended in this high Tc

area.

2.1.2 Basic Superconductivity in Physics

2.1.2.1 London and Ginzburg-Landau theories

In 1933, the Meissner effect was identified [7], which describes the effect of

zero magnetic flux density in superconductors and has proved as important as zero

resistivity. The London equations were the first theoretical attempts to describe the

relation between magnetic field and superconducting cunent on the basis of these

fundamental features of superconductivity, by F. London and H. London, 1935 [8].

An important material parameter A,L, now known as London penetration depth, has

been derived by this theory: The magnetic field can penetrate a superconductor only

up to Xu which is of the order of IO"2 urn. However, this holds only for the type I

superconductors. For type II superconductors, to which superconducting wires for

7

magnets belong, the magnetic flux can penetrate superconductor mass in the form of

quantised flux (fluxoid) entities.

An understanding of type II superconductors had to wait until the appearance

of the Ginzburg-Landau (GL) theory in 1950 [2]. In this theory the order parameter is

used as a variable, the square of its absolute value being the density of

superconducting electrons. A material parameter £, the coherence length, is defined

in the theory as a width in which the density of superconducting electrons decays

gradually at the boundary between the superconducting and the normal conducting

states. Type I and II superconductors are distinguished by the difference in the

parameter K = XJ%, where X is a modified penetration depth which is not very

different from XL. The structure of a single fluxoid in type II superconductors is

related to X and % in such a way that it consists of a flux tube of radius A, with a

normal core of radius cj at the centre. The study of fluxoid motion provided the basis

for the development of theories of loss, critical cunent, and other characteristics,

which contributed greatly to progress in the development of practical

superconductors.

2.1.2.2 Bardeen-Cooper-Schrieffer theory

GL theory is phenomenological; the appearance of a microscopic theory, the

BCS theory, had to wait an additional six years [3]. The success of this theory is

based on the assumption of two electrons with opposite directions for both spins and

moments. These two electrons are attracted by the intermediation of phonons and

form a Cooper pair. This theory served to elucidate microscopic problems. The

superconducting electrons were proved to be electrons existing as Cooper pairs, the

coherence length in GL theory proved to be the size of the Cooper pair, and the

8

density of superconducting electrons or the order parameter in G L theory was found

to be proportional to the electron energy gap. This gap is proportional to the critical

temperature and to the critical frequency or upper frequency limit of the

electromagnetic wave, above which the superconductivity disappears. Three years

after the appearance of BCS theory, a more perfect microscopic theory was presented

by Gorkov [91. In contrast to BCS theory, it implies the case of a variation in energy

gap parameter with position in a superconductor.

2.1.2.3 Josephson effect

The next big step in basic theory came in 1962 with the discovery of the

Josephson effect [4]. This effect occurs in the junction where two superconducting

electrodes are separated by a very thin insulating layer or a very small weak link. The

effect implies that a superconducting cunent tunnelling through the junction

produces a phase difference between the superconducting electrons of the two

electrodes and that the time derivative of the phase difference generates a voltage

difference between the electrodes. From this effect two very interesting and useful

characteristics can be derived: the critical cunent changes periodically with the

magnetic flux in the junction and the V-I curve of the tunnelling cunent has voltage

steps whose width is proportional to the frequency of an electromagnetic wave

applied to the junction. Practical applications of the Josephson effect, -such as very

high speed switching elements in computers, very high sensitivity magnetometers,

and a very high precision voltage standard, are an outgrowth of these characteristics.

9

2.1.2.4 Type II superconductors and pinning

Some studies on superconductivity supplying direct bases for superconductor

applications began in the second developmental period and continued to support the

development of various applications. For practical materials, termed non-ideal type II

superconductors, the important phenomena are the critical cunent and the voltage

generated by a cunent exceeding the critical cunent. These were explained by Kim et

al., 1963 [10]. A fluxoid located at a point of cunent density J receives a driving force

J^o (<|)o-fiux quantum) which is counterbalanced by a force from pinning centres. This

force has a threshold value called a pinning force, Jcfy0. When J is larger than Jc, the

driving force exceeds the pinning force, which leads to fluxoid movement and a

voltage drop. Obviously, a high Jc value is required for practical applications of

superconductors. At the same time, however, high Jc-value material has a high loss

density when the applied field changes. In this case a movement of fluxoids is

generated by the introduction or removal of fluxoids from the surface of the sample

due to the change in applied field. A loss is generated by movement of the fluxoid

passing over a point of pinning. This loss is called a pinning loss and the point of

pinning is called the pinning centre. The loss is also known as a hysteresis loss

because the change of applied field in this case causes hysteresis in a magnetisation of

the superconductor. Where the field variation is very slow, the cunent density in

every point of the superconductor is thought to be kept just equal to Jc. This

assumption is known as the critical-state model. The concept for this model was

introduced by London [11], Bean [12], and Kim et al [13], in 1962. The simple

functions which they used for the pinning force are called the Bean-London model

and the Kim model. Other models have been presented that provide quantitative

10

descriptions of practical materials, in which the general form is presented by

Yamafuji and hie, 1967 [14].

The pinning force of a pinning centre experienced by one straight fluxoid is

called the elementary pinning force. The total pinning force due to pinning centres in

a unit volume, Pv, was originally thought to be a simple summation of elementary

pinning forces in this volume. However, because of the elastic nature of a fluxoid

lattice, the foregoing rule was found not to hold generally. It was pointed out by

Yamafuji and Irie [14] that the elastic property of the fluxoid lattice should be taken

into account in this problem, and as a result, Pv was found to be proportional to the

square of the elementary pinning force and inversely proportional to the elastic

constant of the lattice. The former is now known as the simple summation rule and

the latter as the dynamic definition of pinning force.

2.1.2.5 Non-metal superconductors

Although the second development period is characterised by increased growth

in applications starting around 1960, basic studies continued during this period,

especially on organic and oxide superconductors. Superconductivity in organic

materials was suggested theoretically by Little, 1964 [15], and experimental results

for organic materials were obtained in 1980 [16]. However, the Tc values obtained

were less than 10 K, which is much lower than that expected by theory, and the

mechanism was proved to be other than the mechanism suggested theoretically. The

first study of oxide superconductors appeared in 1975 by Sleight et al [17]. This was

preceded by a study of Ba-Pb-Bi-0 that reported a Tc value of 12 K, although this

11

work attracted little attention at that time. Another 11 years elapsed until the

discovery of the very high Tc oxides.

2.1.3 Electromagnetism in Superconductors

The basic electromagnetic phenomena in superconductors and some general

features are simply summarised with respect to the following important concepts [18-

22].

2.1.3.1 Critical-state model

Pinning interactions between quantized fluxoids and pinning centres in

superconductors are the origin of non-dissipative cunent, and their strength

determines the critical cunent density Jc. On the other hand, various ineversible

electromagnetic phenomena observed in superconductors, such as hysteretic

magnetisation, AC loss, and magnetic instability, are known also to be caused by flux-

pinning interactions. These phenomena can be described quantitatively by the critical-

state model in terms of macroscopic quantities such as Jc.

Electromagnetic phenomena in superconductors are described by Maxwell's

equations: rot H = J, and rot E = - SB/dt, with additional equations related to the

properties of superconductors. One of them is B = p0H. Another is the equation

relating J and E, and this is given by the critical-state model.

When a superconductor in a magnetic field carries a transport cunent, fluxoids

in the superconductor experience the Lorentz force FL = J x B, per unit volume. If

the fluxoids are driven by the Lorentz force, a damping force of the same magnitude

12

and opposite direction appears and a steady state is attained. The damping force

consists of two components: one is a pinning force Fp and the other is a viscous force

Fv. Thus the equation of force balance is: FL + Fp + Fv = 0, where Fp and Fv are the

pinning force density and viscous force density, respectively. The pinning force is

caused by the pinning potentials that fluxoids feel and the viscous force originates

from ohmic energy dissipation around normal cores of moving fluxoids. The

difference between the two components of the damping force is that Fp (= - 5 FP(BA))

is independent of the fluxoid velocity v, while Fv (= riB^o"1 v) is proportional to v.

These forces are therefore expressed as Uo"1 rot B x B - 8 FP(BA) - v uB^o"1 = 0,

where BA = IBI, 8 = v/lvl (unit vector directed to the fluxoid motion) and r| is a viscous

coefficient per unit length of a fluxoid. Fp and rj are functions of the temperature T.

One of the characteristic points of this model is that the pinning and viscous forces

always work opposite to the direction of fluxoid motion. This means that the

phenomenon described by this model is completely ineversible. There are two

unknown quantities, B and v. Hence another equation is necessary to obtain solutions

of these quantities.

The theoretical model based on the force balance expressed by the above

equation is the critical-state model; it can be rewritten as: IEI = pf(IJI - Jc) (Ul > Jc),

which represents the cunent-voltage characteristics in the flux flow state, and pf = §0

B/n is the flow resistivity. Superconductors are, in practice, used in the region of zero

or sufficiently low electric field. In such cases the viscous force is negligibly srnall in

comparison with the pinning force. The above equation means that cunent flows

locally with its critical cunent density, Jc. This is the critical state in a nanow sense.

13

3 0009 03204606

2.1.3.2 Magnetic flux distribution and magnetisation

In high-field superconductors, a variation of the flux distribution occurs

quasistatically except for cases such as an application of AC field with very high

frequency and the viscous force can be neglected. Consequently the force balance

equation contains B alone and can, in principle, be solved. As a simple case, a large

superconducting slab occupying 0 < x < 2d is put in a magnetic field Ho along the z-

axis. The flux distribution varies only along the x-axis and the critical state equation

is reduced to -(BVu*) dB7dx = 8 FP(BA), or J 8BA dBA / FP(B

A) = -po J dx, where 8 is a

sign factor (±) representing the direction of flux motion. When a functional form

FP(B ) is given, the flux distribution can be obtained.

In NbsSn, for example, we typically use UoHC2 ~ 20 T, Jc = 5 x IO9 A/m2, and

pf = (BA/UoHc2)pn = 8 x IO"

8 iim at BA = 5 T, where pn ~ 3 x IO"7 iim is the normal

resistivity. When the AC magnetic field is applied, the viscous force can be safely

neglected in most cases.

Here the theoretical expressions of magnetisation are derived for a

superconductor that does not carry a transport cunent. From the magnetic flux

distribution inside the superconductor, the averaged magnetic flux density Bav = (1/d)

Jo-d B dx is obtained and the magnetisation is given by M = Bav - PoH0.

2.1.3.3 Hysteresis loss

According to the critical-state model, the pinning force and the viscous force

act opposite to the direction of fluxoid motion. When the fluxoid motion occurs due

to a variation in the external field or the transport cunent, an electric field is induced

14

through a variation in the flux distribution as described by rot E = 3B/3t. The fluxoid

velocity v and electric field E are related to each other through E = B x v. The input

power density, JE, is JE = (J x B)-v, which can be regarded as the work done by the

Lorentz force in a second, and the critical state equation yields J-E = Ivl FP(B ) +

T|(B7<J>0) v2 > 0. The first and second terms in the equation above are the pinning loss

power density and the viscous loss power density, respectively. The characteristic

point of the pinning loss power density is that it is proportional to the velocity Ivl or

to a frequency f in an AC condition. On the other hand, the viscous loss power

density is proportional to v or f .

The origin of the energy dissipation is a motion of normal electrons around

normal cores of fluxoids. This is quite similar to the ohmic energy dissipation in

normal metals, which is similar to the viscous loss. The pinning interaction comes

from a pinning potential. The fluxoid motion becomes unstable and its velocity is

large when the fluxoid drops in and jumps over the pinning potential. Hence the

elastic energy associated with a deformation in the fluxoid lattice due to the pinning

interaction is almost dissipated at such instabilities. This means that dissipated

energy by one pinning centre is constant and is independent of the mean fluxoid

velocity v. Hence the pinning loss power depends only on the number of pinning

centres that the fluxoid meets in a second and is proportional to Ivl or f.

Hysteresis loss due to AC magnetic field is estimated in terms of the critical-

state model for the case where an AC magnetic field with an amplitude of Hm is

applied parallel to a superconducting slab of a thickness 2d. The pinning loss power

density therefore can be calculated, and an instantaneous value of the pinning loss

power density at an arbitrary position in the superconductor is given. This is contrary

15

to the calculation of the pinning loss from the area of the B - H or M - H hysteresis

curve. The latter method provides only the energy density averaged over the

superconductor during one cycle of the AC field. That is, the latter method is

available only for a periodic repetition of the external field where the hysteresis curve

closes. The pinning loss energy density is smaller for more strongly pinned

superconductors in the range Hm < Hp.

Dynamic resistance loss is considered when a small AC magnetic field is

applied to a cunent-carrying superconducting wire or slab in a DC magnetic field.

When the DC and DC magnetic fields are parallel to a superconducting slab of a

thickness 2d, the flux distribution can be given by the critical-state model. The loss

energy density per cycle supplied from the DC cunent source is called the dynamic

resistance loss. There is an additional loss energy that is connected to the fluxoid

motion at both surfaces of the superconducting slab. This is supplied from the source

of AC magnetic field. Therefore the total loss energy density is W =

2p0HpHm[l+(Jm/J)2] - (4/3)u0Hp

2[l-(Jm/J)2], where Jm is the mean transport cunent

density.

2.1.3.4 Magnetic instability

In the process of increasing magnetic field, the magnetisation changes

discontinuously, and this is a flux jump. The magnetic flux invades suddenly into the

superconductor when flux jumps occur. Such an instability comes from an essential

feature of flux pinning in superconductors. If a small increase in temperature is

assumed due to some cause, such as a local flux motion, the critical cunent density

decreases and this yields further flux motion because of weakened pinning force.

16

Thus the temperature increases again and the instability develops catastrophically by

repetition of such a positive feedback until the critical cunent density goes to zero. If

the flux jump occurs, superconductors cannot carry non-resistive cunent. It is

necessary to avoid flux jumps in superconductors. A reduction in the diameter of

superconducting wire was found to be effective for stabilisation. For this reason,

multifilamentary superconducting wires have been developed. Stabilisation against

the flux jumping has been considered, and a simple case of an isothermal condition is

used. This is approximately satisfied in a superconductor, since thermal diffusions

are sufficiently faster than magnetic diffusions. As a result the diameter of

superconducting wire should be smaller than a certain value.

2.1.3.5 Reversible fluxoid motion and flux creep

The pinning energy dissipation is an essence of ineversibility, and is brought

about when fluxoids jump over pinning potentials. These phenomena of the loss

energy in a thin superconducting filament originate from reversible motion of

fluxoids. It is expected that the phenomenon is almost reversible when the motion of

fluxoids is limited within pinning potentials. In the reversible regime, a variation of

the flux distribution penetrates from the surface up to a constant depth, X\ called

Campbell's AC penetration depth [23]. In the reversible region, the pinning force is

not always directed opposite to the fluxoid motion. This means that EJ contains a

component of variation in the stored energy. The loss energy can be obtained only

from an area of hysteresis loop. In very fine multifilamentary wires with

superconducting filaments smaller than X\ the reversible effect becomes prominent.

17

It has been found in various measurements that an ineversible magnetisation

or a persistent cunent due to flux pinning decays with time. This is noticeable in

high-temperature superconductors at high temperatures. The ineversible

magnetisation varies approximately linearly with log t, where t is time. Such

relaxation comes from a flux creep, i.e., a thermally activated motion of fluxoids

trapped in pinning potentials. That is because the trapped fluxoids jump over the

pinning barrier even for a cunent below the critical value. The resultant fluxoid

motion, mostly along the direction of the Lorentz force, yields the relaxation of

magnetisation or persistent cunent. It is concluded that the decay of persistent cunent

is remarkable even for the strongly pinned case, if the superconductor is used at

liquid nitrogen temperature.

2.2 CONVENTIONAL LOW Tc SUPERCONDUCTORS AND APPLICATIONS

2.2.1 Development of Practical Superconducting Materials

The 1960s was a time of substantial development in high field materials for

superconducting wires, and the foundation of practical materials was established. In

1961 for the first time, wires of NbZr were manufactured and wound into a small

magnet that generated a magnetic field of 4 T [24]. Such wires were later produced

commercially. However, they soon gave way to NbTi wires, which are more ductile

than NbZr and easier to handle. Today, NbTi is still the representative conventional

superconducting material for magnet use.

18

Another group of superconductors are the A15 compounds [25-33]. A m o n g

these, Nb3Sn was found by Kunzler et al, 1961 [34], to have the highest critical field

of materials known at that time. They made wires by drawing and heat treatment of a

Nb tube filled with Nb and Sn powders. A magnet wound by this wire, which

produced 7 T, was the first truly high field magnet to be developed [34]. However,

because of the brittle nature of NbsSn, the magnet was very unstable. Several years

later Nb3Sn tapes with better mechanical characteristics were developed. These

consist of NbsSn deposited onto stainless steel tapes in the gas phase. Magnets

wound by this tape produced fields greater than 10 T [35]. They were supplied

commercially and also used for experiments in physics.

The idea of multifilamentary composite wires was presented in the late 1960s,

and it was realised using NbTi [36-38]. However, for application of this idea to

Nb3Sn, some metallurgical problems had to be solved. The first multifilamentary

composite of NbsSn was realised by Tachikawa in 1970, using the bronze method

[39]. Although these wires are more sensitive than NbTi wires to stress, they are now

used for high-field magnets above 10 T [35, 40]. By the use of Nb3(SnTi) the field

has been raised to 20 T [41]. Another A15 compound, VsGa, has a critical cunent

higher than that of Nb3Sn in high fields [42]. The highest field obtained by V3Ga

wire is close to 20 T [43]. Recently, Nt^Al has attracted attention because of two

excellent characteristics: higher cunent density and less sensitivity to stress [44].

Reports on Chevrel-phase compounds at first attracted considerable attention because

the estimated upper field was 60 T by Chevrel et al., 1971 [44]. However, the critical

cunent density was still too low (~104 A/cm2), and they are, at present, too brittle for

practical use.

19

Extensive metallurgical studies have been made on pinning centres in alloys

and compounds. For example, with regard to the enhancement of critical cunent

density for the NbTi, it is found that both precipitates of ct-Ti and dislocations in NbTi

grain boundaries in compounds act as pinning centres [45-54]. Through an optimised

heat treatment, Jc has recently been raised to 105~106 A/cm2 in alloys [55, 56].

Table 2-1 is a summary of the critical parameters of some conventional

superconducting materials [18-22].

Table 2-1. Critical parameters of superconducting materials.

Material T (K) poH=(0) (T) UoHc^O) (T) UoHc2(0) (T)

Hg

Pb

Nb

Nb37Ti63

Nb3Sn

Nb3Al

Nb3Ge

V 3Ga

PbMo6Sg

4.15

7.20

9.25

9.08

18.3

18.6

23.2

16.5

15.3

0.041

0.080

0.199

0.53

0.63

~

~

0.174

~

—

0.404

15

29

33

38

27

60

20

2.2.2 Development of Practical Conductors

2.2.2.1 Conductor-techniques of low Tc superconductors

(a) Instabilities

Today superconductors can be used for a variety of practical magnets.

However, before arriving at this stage, much effort had to be spent to remove causes

of instabilities. The first cause encountered was an effect of flux jumps, a sudden

invasion of flux into a superconductor during its magnetisation. The effect develops

in a vicious cycle: A temperature rise leads to a flux invasion by the resulting

weakened pinning force, which in turn leads to a temperature rise to be added to the

original one. By this effect superconducting magnets in the early stage of ramping up

the cunent often became unstable, even at a cunent much lower than the critical

cunent. The mechanism of flux jumps was first studied by Wipf and Lubell, 1965

[57], whose work was followed by many other studies [58-65]. Of practical

importance was the finding that flux jumps should never take place if the diameter of

the superconducting filaments in wires is below a certain value. The development of

practical multifilament wires is based on this idea. The next cause was the occunence

of instabilities in cunent-carrying conductors. Again, this is caused by a vicious cycle

of Joule heat generation in a trace of the normal region and subsequent spreading of

the region due to this heat. Stekly and Zar, 1965 [66], proposed a practical solution to

this problem: enclosing the superconductor in copper. Stable superconducting

magnets were not realised until the appearance of wires of this type.

(b) Multifilamentary composite wires

21

For realising stable conductors with high-cunent-carrying-capacity,

multifilamentary composites were proposed in 1970 [36-39]. These wires have many

thin superconducting filaments embedded in a copper or aluminium matrix.

However, wire of this type has a severe problem: a non-even distribution of cunents

among filaments when a perpendicular magnetic field is suddenly applied to the wire,

which is the normal condition in magnets. This non-even distribution occurs due to

the shielding cunent generated, which flows in a loop through filament and then

through the copper matrix in between filaments near the ends of the wire, and is often

called the coupling cunent. Again, this effect leads to wire instability, because this

shielding cunent is similar to that of a thick superconductor with the same diameter

as that of the composite wire. The same phenomenon is related to large AC loss in

composite wires placed in an AC magnetic field. However, the shielding cunent does

not operate permanently but decays by a time constant related to the length of the

wire. In theory, this time constant is proportional to the square of the conductor

length and to the conductivity of the matrix. For large magnets this time constant may

easily exceed a year. A new idea was soon born: twisting the conductor. In this case

the twist pitch conesponds to the full conductor length, i.e., the length of each

shielding cunent loop is restricted to one pitch. By this method the time constant

becomes very much smaller, i.e., of the order of milliseconds for thin wires with a

short twist pitch. Very stable magnets and even pulsed magnets were realised using

such wires. This was really the beginning of the age of stable practical magnets.

For a cable application, to further increase the cunent-carrying capacity of a

wire, the number of filaments in the conductor should be increased; as a result, the

wire diameter increases. However, in such a case, the twist pitch necessarily becomes

22

larger because the allowed pitch should be several times larger than the wire diameter.

Again instability arises. One of the possible countermeasures is the use of a metal with

high resistivity, such as CuNi, to weaken the coupling cunent in the matrix of

conductors. The other is the cabling of a number of strands of small diameter. The

former is used for slow-pulse magnets and the latter for fast-pulse magnets. Pulse

magnets with subsecond operation have been realised by using cable conductors.

(c) Stress effect

Magnet windings are subjected to magnetic pressure proportional to the

square of the magnetic field in the winding. The tension applied to magnet conductors

is proportional to the radius of their winding. Therefore, conductors in large or high-

field magnets are subject to high levels of stress. Stress in conductors also arises

during the winding process of a magnet through the bending of wire. This stress is

larger for smaller winding radiuses. The critical cunent of superconductors is

degraded by such stress, called the stress effect or the strain effect.

There are two types of stress effect, one with a reversible stress-strain

characteristic and one with an ineversible characteristic. The former takes place in a

low-stress region and was fully studied by Ekin, 1980 [67], who presented a general

formula applicable to any material having B and Bc2 as variables. The latter depends

not only on material but also on the filament diameter and structure. In magnets,

design stress or strain in conductors should be limited so that the critical, cunents do

not degrade notably in the reversible region.

Superconductors are subjected to significant stress in many applications, e.g.,

in magnet applications, stress can arise from three sources, (a) Fabrication stress: In

magnet construction, the superconducting wire is subjected both to bending stress as

23

it is wound into the magnet coil and to uniaxial stress from pretensioning of the wire

during winding, (b) Thermal-contraction stress: As the wire is cooled from heat-

treatment to liquid-helium temperature, different materials within the composite wire

and magnet structure contract at different rates. Because of this mismatch in thermal

contraction rates, significant stress on the superconducting material can be generated

during cool-down, (c) Magnetic stress: When a superconducting winding is

energised, the Lorentz force acting on the superconductor can be quite large,

particularly in large magnets. For example, the axial magnetic hoop stress along each

wire in a solenoid magnet is given by O// = JBR, where J is the transport cunent

density, B the magnetic field strength, and R the radius of the winding.

Stress and strain introduced into multifilamentary composites by these

sources have three important consequences, (a) Training: Senated yielding of the

superconducting material, fracture of magnet potting material, or wire movement can

generate enough local heat to initiate thermal runaway. This results in a process

known as magnet training, (b) Matrix degradation: The ability of the matrix material

to stabilise the superconductor can be more than halved by cyclic strains as low as

0.4%; the purer the matrix material, the lower its tolerance to fatigue, (c) Critical

cunent degradation: The critical cunent of compound superconductors, such as

Nb3Sn and V3Ga, can be lowered (reversibly) more than 30% by a static strain as low

as 0.4% at 12 T; the effect becomes more severe at high fields [21].

(d) Fabrication techniques and copper ratio

The development of conductors outlined above was realised only by

sophisticated metalworking techniques such as cladding of a superconducting rod by

matrix metal (Cu or Al), repeated bundling of the clad rods, twisted cabling, and

24

sometimes heat treatments. To obtain an Nb 3Sn multifilamentary composite a

combination of filaments and a bronze matrix is often used, first worked to make a

wire, and then heat-treated to generate the Nb 3Sn compound near the surface of the

filaments. Conductors recently developed for power frequency applications require

filaments of submicrometer diameter embedded in a CuNi matrix. This necessitates

the use of highly developed techniques, because breaking only a few filaments in a

conductor seriously increases A C loss.

The stability of a cunent-carrying conductor is determined by the quantity of

copper, the copper ratio, as well as by heat transfer from the conductor to the coolant.

Stabilisation by the Stekly criterion [66], or condition of full stabilisation, requires a

large quantity of copper and perforce provides the entire conductor with a low

average cunent density.

2.2.2.2 Factor for design of composite superconductors

(a) Flux jump stability

If a high field superconducting material is exposed to a magnetic field the

field inside the superconductor will stay zero except in a layer adjacent to the surface

in which shielding supercunents are induced. This shielding layer will grow in

thickness as the field increases. W h e n the field exceeds some critical value, the

shielding supercunents become unstable and magnetic flux suddenly rushes into the

sample and the associated temperature rise is large enough to drive the sample to the

normal state. This breakdown process is called flux jump (see p. 16). Experimentally,

flux jumps have been observed in all technically important high field

superconductors. The flux jumping is now well understood and a stable conductor

can be designed with conventional superconductors under all practical conditions of

25

operation [68-71]. The basic conditions are adiabatic flux jump, stability of tape

conductors, multifilamentary composite, and self-field instability.

(b) Dynamic stability

The adiabatic theory of flux jumping has neglected the movement of heat

developed by flux motion [72-74]. The dynamic stability theories take account of

time dependence in the movement of the heat, cunent, and magnetic flux.

(c) Cryogenic stability

The multifilamentary composite conductor based on the criteria adiabatic

stability and dynamic stability is stable against flux jumping. The actual magnets

wound from these wires are not free from disturbances. The examples are conductor

motion due to the electromagnetic forces, internal friction in the wire, and cracking

of the potting materials. If the dissipated energy is not removed efficiently, the

normal zone caused by the disturbance will grow in time and result in a quench of the

magnet. The techniques developed to ensure stable operation of magnets are called

cryogenic stabilisation [75-89]. The criterion of cryogenic stability is given as the

Stekly parameter [90]. The cryogenic stabilisation is provided by the forced

circulation of helium coolant.

Cryogenic stabilisation is usually effected by cooling the conductor with

boiling helium at 4.2 K contained in channels through the magnet winding. If the

disturbance is small and the temperature rise is less about 1 K, the full.advantage of

high heat transfer in the nucleate boiling can be utilised. Large disturbances,

however, are quite likely in large magnets.

(d) A C loss

The changing magnetic field generates an electromotive force within the

superconductor. If the associated electric field is E, the power generation per unit

26

volume becomes EJC. The electric field E drives the cunent density beyond its

critical value Jc into the flux flow regime, but this resistance line is very steep.

Therefore, it may be approximately assumed that the cunent remains at critical

density. The loss mechanism is thus of a resistive nature. The resistance is highly

non-linear, however, and the loss per cyclic field change is independent of the cycle

time. For this reason, the A C losses in a superconductor are often called hysteresis

loss Qh [91-95]. W h e n the sample geometry is complicated, it is more convenient to

calculate loss in terms of the magnetisation M , defined as the total magnetic moment

per unit volume of sample. Provided that the external field H is uniform over the

sample, the loss per cycle per unit volume becomes: Q h = J H d M = J M dH. In a

multifilamentary composite, the superconducting filaments are coupled

electromagnetically by changing magnetic fields. This coupling can greatly increase

the A C losses. Since the extra losses are of an eddy cunent nature, the loss per cycle

depends on the cycle time, unlike hysteresis losses [96-107].

(e) Practical conductors

A multifilamentary composite conductor is used as a basic strand to construct

large-scale, heavy cunent conductors. The design of such large-scale conductors

depends strongly on the cooling mode of the winding and also on the design

philosophy of the winding or magnet builder. NbTi multifilamentary composite has

been used widely to construct magnets that can generate fields less than 8 T [67].

Various types of NbTi multifilamentary composites are basically made with standard

copper matrix and CuNi matrix or CuNi alloy jacket around each filament. The CuNi

increases effective transverse resistivity. Nb3Sn multifilamentary composites are now

becoming an engineering material. The Nb 3Sn filaments are embedded in a bronze

(CuSn) matrix. Therefore the transverse resistivity is high and the loss time constant

27

is small. To ensure low longitudinal resistance, both conductors have pure copper

separated by the diffusion banier of Ta or Nb. At present, the number of filaments

within a strand is at most about 10,000 for a single stack due to the limitation of the

conductor fabrication process. The practical conductor can be designed for different

applications, such as a general-purpose D C magnet, a D C magnet for persistent-mode

operation, a pulsed magnet and various A C applications [108-118].

2.2.3 Applications

With regard to achieving better energy efficiency, high quality and efficiency

products or systems, the applications of superconductivity rapidly increase [119,120].

The applications can be divided into two categories: power applications, such as the

superconducting generator, and electronic applications, such as SQUIDs. The major

power applications of superconductivity are related to magnet technology. The most

fully developed application, for instance, is magnetic resonance imaging (MRI), used

primarily as a medical instrument, in which the large volume and homogeneous

magnetic field required can be supplied by means of a superconducting magnet.

Other exciting devices and facilities applying superconductivity are now at key stages

of development, such as the maglev car (Japan), superconducting electromagnetic

thrust ship (Japan), superconducting generator (USA, Japan, Germany, Russia and

France), superconducting magnetic energy storage (USA and Japan),

superconducting supercollider (USA), and fusion reactor (USA, Europe, Russia and

Japan). The use of superconductivity for these devices produces tremendous benefits.

The high-field superconducting magnet will be an indispensable component for the

28

realisation of a commercial fusion reactor for the generation of electricity.

Superconducting materials, in order to be utilised on an industrial scale, must have

both superconducting and mechanical properties. W e now have a long history of

developing reliable and industrial-scaled superconducting products. The strand

design that satisfies the requirements of stability, durability, and high performance is

the composite superconductor. The basic elements of this design are fine and

continuous superconducting filaments embedded in a metallic material with high

electrical and thermal conductivity.

2.2.3.1 Large scale applications

Most large scale applications of superconductors are related to

superconducting magnets. Since the initial superconducting materials had a very low

critical field, the idea of building strong electromagnets using the superconducting

wires soon after the discovery of superconductivity, had to be given up. It was not

until the early 1960s that high-field magnets really appeared. These were NbZr and

Nb 3Sn magnets, which at that time were unstable. Soon after NbTi became

representative material for superconducting magnets. Magnetic technology took its

first big step early in the 1970s, when bubble-chamber magnets for studies in high-

energy physics played a leading role.

By the end of the 1970s, the leading role for magnet development shifted to

the experimental study of controlled thermonuclear fusion systems [69]. M a n y types

of superconducting magnets were studied for use in this application. The

development of the toroidal magnet used in tokamak reactors, which confines plasma

in a magnetic field, began in the early 1980s [70]. In 1985 a toroidal magnet system

was built and tested [71]. This was a large coil (LCT) project promoted by the

29

International Energy Agency (IEA), aimed at assembling and testing six coils. Those

coils were manufactured independently by the United States, Euratom, Switzerland,

and Japan, and installed for testing in the United States, each having a different

design for coil shape and cooling. Materials used for coils were mostly NbTi, with

NbsSn for one coil. A total stored energy of 800 M J was achieved by the test [72].

(a) Magnets for high-energy physics

Elementary particle accelerators in high-energy physics is another field in

superconducting magnet application that has contributed to the development of

magnet technology of a different type. The advantages of an accelerator ring of

superconducting magnets are that the magnetic field can be greatly increased, with a

reduction in the ring diameter, and greatly reduced power consumption.

Consequently, superconducting coils are indispensable for very high energy

accelerators. The first example of this application was established at the Argonne

National Laboratory in 1968 [121]. The first accelerator with superconducting

magnets was installed at the Fermi National Laboratory (FNAL) in the United States

in 1983 [122-124]. The requirements for these magnets are high field uniformity and

low cost. The superconducting super collider (SSC) project in the United States is the

largest accelerator ever built. The circumference of its ring is 83 k m and it has

approximately 7700 and 1800 superconducting dipole and quadrupole magnets,

respectively [21].

(b) Generators and motors

Studies of superconductors for industrial use began with an investigation of

their applicability to electric motors or generators [125-128]. The essential features of

superconducting machines are high efficiency, high power rating, and comparatively

small size. The first superconducting machine, which was homopolar, was

30

manufactured in the United Kingdom and was operated successfully in 1970 [74].

Since then many large machines, mostly alternators, have been manufactured for tests

in many countries. The largest machine ever planned was rated up to 400 MVA [75].

However, most of these machines have superconducting coils only for field windings,

not for the armature windings. Based on the development of AC superconductors,

studies are now being made on the realisation of complete superconducting

machines.

(c) Superconducting magnet energy storage

Magnetic energy storage in electric power systems can be realised only by

superconducting magnets. There are two types of superconducting magnetic energy

storage (SMES): one is for load levelling of power systems and the other is for power

system stabilisation. The former was proposed in the latter half of the 1960s, and

design studies were begun in the 1970s in the United States and Japan [129-131]. If

pumped hydropower stations were to be replaced by this system, much energy would

be saved because of its high storage efficiency. A magnet for this purpose should

have an energy of several gigawatthours (GWh) and a diameter of several hundred

meters, and for its construction very high levels of magnet technology, far beyond the

present level, will be required. Some systems above 20-MWh have been constructed

since 1989 [131, 132]. Although this energy is less than a hundredth of the SMES

required for diurnal load levelling, it is still very large with a diameter of 150 m. The

SMES for power line stabilisation was studied early in the 1980s in the United States,

and a 30-MJ (10 kWh) pulse magnet was manufactured and used together with a

controlled converter system in an experiment to stabilise the power oscillation in a

long power line [133].

31

(d) Magnetically levitated transportation

The magnetically levitated (maglev) transportation system suggested by

Powell and D a m b y in 1967 attracted the attention of researchers interested in a very

high speed means of transportation [134]. It requires light high-field magnets that can

only be realised using superconducting coils. Large-scale experiments have taken

place in Germany and Japan, and a speed record of 500 km/h was achieved in 1979

[135, 136] and tests were carried out on a 7-km test track in 1983. The next

development began in 1990 for commercial tests.

(e) Magnetic resonance imaging

For some applications, superconducting magnets are now being produced

commercially [137-139]. The most remarkable example is the use of such magnets

for magnetic resonance imaging (MRI) systems used for medical diagnosis.

Commercial superconducting M R I devices were first developed by Oxford

Instrument early in 1980 [83], before which time the magnets used were of

permanent-magnet type or normal coil type. In the meantime, however,

superconducting magnets have become dominant, because superconducting M R I has

a high image resolution due to its high field, high stability, and high uniformity of

field.

(f) Magnetic separator

Magnetic separation is a powerful technique which can be used to separate

widely dispersed contaminants from a host material. This technology can separate

magnetic solids from other solids, liquids or gases. It uses large magnetic fields or

large magnetic field gradients to separate fenomagnetic and paramagnetic particles.

A high-field magnet is basically required for this application. Several

superconducting magnetic separation systems are in operation [140-154].

32

(g) Features of superconducting magnets

There are several limits for designing magnets with the superconductors [21,

22]. The Jc limit comes from superconducting material characteristics, while the full

stabilisation limit depends on the copper ratio used for stabilisation. The energy

damping limit results from the voltage limit between energising terminals and also

from the heating of the coils by residual energy in the case of magnet quenching. The

stress limit depends on the mechanical property of the conductor. However, the

energy damping limit is valid only when the magnet has only two terminals for

releasing energy, and the stress limit is valid only when the coil is supported by a

cold structure and is not applicable when supported by a warm one. The most serious

barrier to the development of industrial applications is the high cost of

superconducting magnet systems. However this problem might be solved, through a

broader market for the product and cost reduction through mass production.

2.2.3.2 Superconducting electronic devices

(a) History

Although present superconducting electronic devices are, for the most part,

based on the Josephson effect, those studied around 1960 were based on the simple

characteristic of the superconducting normal transition. In the first suggestion for a

superconducting computer, presented in 1956 by Buck [155], such a-transition in

bulk superconductors is utilised as a switch, called a cryotron. The first cryotron was

wire-wound and had a low switching speed. Extensive studies were canied out, and

thin-film cryotron systems that included memory devices were developed. They were,

however, soon replaced by the Josephson junction, which has very much higher

switching speed. A brief instruction to Josephson devices is given below.

33

(b) Superconducting quantum interference device

Superconducting quantum interference devices (SQUIDs), a notable

application of the D C Josephson effect, were devised by Jaklevic et al., 1964 [156].

A S Q U I D has a very high sensitivity as a flux meter. This characteristic originates

from the cyclic change of critical cunent through the S Q U I D with the trapped flux in

its ring, where the period of the cycle conesponds to a quantized flux. Magnetic field

measurement with an extremely high resolution of field (IO"9 G/Hz1/2) is possible

using this device. It is IO3 to IO4 times higher than the common existing measuring

devices. Therefore, the magnetic cardiograph and even magnetic encephalography

can be realised. Commercial multichannel S Q U I D systems with thirty-seven or more

SQUIDs are under development for clinical diagnosis of the brain, and simple units

for scientific use have already become popular [86].

(c) Voltage standard and detectors

As a manifestation of the A C Josephson effect, the cunent-voltage

characteristic of a junction has many voltage steps when a high-frequency

electromagnetic field is applied. A voltage standard is obtained using the

proportionality between the voltage difference in adjacent steps and the applied

frequency. Such studies started in 1966 by Langenberg et al. [157], and since 1975

the Josephson standard has been adopted in many countries. There are also many

studies of detectors, mixers, amplifiers, and oscillators made with-a Josephson

junction [89]. Some of them are already in use in radio-astronomical observations in

the submillimeter range.

(d) Computers

The Josephson junction as a switching device in computers has been under

study since the 1960s [158]. At that time the switch was controlled by a magnetic

34

field. Many types of switching circuits have been developed since then. The record

switching speed obtained some years ago was of the order of 10 ps, which is very

much faster than that of semiconductor switches. In addition to their high speed,

Josephson switching devices have a very low switching loss, which allows a high

packing density of elements and results in conespondingly small propagation delays.

Therefore, the Josephson computer was expected to be a promising practical device.

23 DEVELOPMENT OF HIGH Tc SUPERCONDUCTORS

2.3.1 High Tc Superconductors

Shortly after the discovery of a new oxide superconductor (La,Ba)2Cu04 by

Bednorz and Mueller (1986), an unexpected increase in critical temperature Tc was

made in the development of high Tc superconductors. Until now the highest Tc value

has been approximately 125 K obtained by Tl2Ba2Ca2Cu30x [159], and 133.5 K

obtained by HgBa2Ca2Cu308 + x [160]. Other oxides, including YBa2Cu307-x (-92 K)

[6, 161], Bi2Sr2CaiCu208+x (~85 K ) [162], and Bi2Sr2Ca2Cu30io+x (-110 K ) [162,

163] are also being studied extensively as candidates for practical materials with high

Jc. This situation has aroused strong interest in industry as well as in pure science,

and much effort is being put into basic as well as developmental studies, forming the

beginning of the third period in the development of superconductivity.

The crystal structures of oxide high Tc superconductors have a common

feature which is an oxygen-deficient multiple perovskite with some Cu-0 planes,

35

which leads to a strong anisotropy of superconductivity. The basic superconducting

phenomena of metal superconductors, such as zero resistivity, the Meissner effect,

and quantization of the magnetic flux, are confirmed also to be valid for oxide

superconductors. Concerning microscopic points, electron pairing seems to be

backed by the experiment of flux quantization. However, what type of interaction is

really working for electron pair attraction is not yet clear, although new models have

been presented and are being discussed together with the older electron-phonon

interaction model. The existence of an energy gap, which should be ascertained by its

observation as a Josephson junction, is also not yet well confirmed. Looking at this

situation, the new high T c superconductivity is phenomenologically very similar to

metal superconductivity, although the basic mechanism is not yet clear.

23.2 High Tc Superconducting Wires

2.3.2.1 Early stage of HTS wires

Since the discovery of new oxide ceramic high critical temperature (Tc)

superconductors in 1986, considerable effort around the world has been expended

towards fabricating commercial products from these materials. In terms of electrical

applications, practical superconducting wires are required with high critical transport

cunents & ) under high magnetic fields. The techniques to produce the highest critical

cunent density (Jc) Ag-clad (Bi,Pb)2Sr2Ca2Cu3Oio+x superconducting wires, now are

being commercialized. The c o m m o n final product is normally a Ag-clad tape with half

a millimeter in thickness, about half a centimeter in width, and available in lengths now

exceeding one kilometer with about 12,000 A/cm2 critical cunent density Jc [164].

36

In the early high-Tc days, wires made of these ceramic materials seemed unlikely

to be useful because of their brittleness. However, remarkable progress was made soon

after that. Bi-based (n=3) wires, with Pb replacing the Bi, are being made in kilometers

lengths with fairly high critical cunent density. Bi-based superconductors are very

micaceous, and the rolling and sintering causes grain alignment of the microstructure so

that cunent in the wire flows mostly through ab planes, with apparently minimal effects

from weak links between grains. Grain alignment in Bi-based materials occurs much

more readily than in YBa2Cu307. Thus, at present, the former appears more promising.

Meanwhile the Tl-based superconductor shows a good magnetic behavior, but is very

toxic.

High T c superconducting oxides, such as YBa2Cu307, are ceramic materials

which are mechanically brittle and are not easily formed into desirable wire geometry.

However, attempts have been made to overcome the difficulty, using several

fabrication methods to produce high T c superconducting wires in different

configurations. There have been extensive efforts to form wires using YBa2Cu307.

(a) Bare ceramic H T S wire. Bare wires have been prepared by a number of

different processing methods, such as extrusion of powder-in binder slurry [165],

molten-oxide spinning [166], or metallic precursor melt spinning followed by oxidation

[167].

(b) Metal-core-composite wire. This is a metal-core composite in which the

superconductor material envelops the normal metal wire in the core. This configuration

may not offer surface protection of the superconductor, but for Y B a 2 C u 3 0 7 wire, it

makes the control of oxygen stoichiometry much easier, with the diffusion of oxygeh

[168]. A metal (e.g. Ag) wire is coated with a superconductor slurry, then is wound into

a coil shape and then heat treated.

37

(c) Metal-clad composite wire. Commercially useful bulk superconductors in

general need to be drawn into fine wires and wound into a solenoid configuration. The

superconducting wires typically require stabilization using a normal metal cladding

which serves several important functions in such a composite structure: (i) a parallel

electrical conduction path to carry cunent in case of local loss of superconductivity; (ii)

thermal stabilization (heat sinking) to minimize local heating and prevent catastrophic

loss of superconductivity; (iii) mechanical protection of the generally brittle

superconducting core against the Lorentz force during operation, as well as many other

unavoidable stresses occurring during fabrication, handling, or use of the

superconductor wire; and (iv) in addition, the normal metal shell provides

environmental protection from the air, which is especially important for the YBa2Cu307

superconductor as it is known to be sensitive to humidity and carbon dioxide gas. The

choice of proper cladding metal has been found to be critical in fabrication of the

composite superconductor wire. A g and A u have been found to be compatible with the

compound at high temperature. A g is relatively inexpensive, compared to gold, and

could be an economically acceptable material as a thin-walled tube or thin layer coating

of normal metal in the composite wire fabrication.

One of the major problems facing the wires is the "weak-link" behavior, i.e., the

relatively low Jc values at 77 K of these wires and their severe field dependence, similar

to the sintered materials. For the high T c superconductor wires to be useful for major

applications, they have to exhibit Jc values of at least 104~105 A/cm2 at a field of

several teslas. Obviously, the incorporation of microstructural control, such as

densification and grain alignment into the wire fabrication technique, is required to

achieve useful oxide superconductor wires. The pros and cons of the various wire

fabrication methods will have to be balanced, based on the superconducting properties,

38

ease of fabrication, long term reliability, and cost. Substantial progress is needed before

the eventual practical wire fabrication method is determined.

2.3.2.2 Development of Bi-2223/Ag wires

Metal-clad composite wires of high Tc superconductors can be made by the

powder-in-tube method, which was firstly applied by the Vacuumschmelze and

Sumitomo Electric Company in Japan [169-172]. This is a major technique which is

mainly used to produce long Bi-based wires for the high Tc superconducting coils. At

present, the Ag-clad Bi-based high Tc superconducting wire has the best magnetic

behavior among the cunent high Tc superconducting families. Poor coupling or "weak

link" is more severe in YBCO than in the other high Tc families. Toxicity of both

thalhum and mercury has been a major block to the development of TBCCO and

HBCCO. The powder-in-tube process of the Ag-clad (Bi,Pb)2Sr2Ca2Cu3Oi0+x (Bi-

2223/Ag) HTS wire, is started by filling a silver tube with the high Tc superconducting

material. The tubes are then drawn, rolled and sintered, step by step, resulting in a cross

section, e.g. 0.2 mm x 4 mm, with the ratio of Ag/superconductor between 2 to 10.

There are different variants of this general procedure, such as different heat-

treatment time, temperatures, precursor powders, and different mechanical or chemical

procedures, to obtain ideal microstructures with a higher Jc or Ic. In the general powder-

in-tube routine of producing Bi-2223/Ag superconducting wires, precursdr powders are

required, and can be prepared, e.g., by the solid-state reaction of high purity oxides and

carbonates of Bi, Pb, Sr, Ca, and Cu with a certain cation stoichiometry [173]. The

polyphase powders are then mixed and calcined in a range of 800-850 °C for about 50

h to obtain a mixture of Bi-2212, calcium cuprate, and secondary phases. The precursor

39

powder is packed into a high purity silver tube, swaged, drawn through a series of dies,

and rolled to a certain thickness. The tape is then subjected to a series of pressing and

heat-treatment cycles in the temperature range of 800-850 °C until the Bi-2223 phase is

completely formed with a desired good aligned microstructure, and the highest Jc.

Like conventional low T c superconductors, multifilament wires will be future

candidates for large scale application of the high T c superconductors in electrical

engineering such as A C transmission cables and A C generator windings. At present,

multifilament high T c superconducting wires and even a twisted form can be produced

[174], and they have a better performance than the single core H T S wire with regard to

mechanical properties.

Ag-clad Bi-based multifilament high Tc superconducting wires have also been

developed in Australia since 1990 [175-180]. As an example, the multifilament high Tc

superconducting wire produced in the superconductivity group of the University of

Wollongong, has good performance, both electrically and especially mechanically. It is

produced by putting a number of drawn Ag-clad high Tc superconducting thinner wires

in a larger diameter A g tube, then drawing this composite tube down to a final thin

wire. For example 7 cores/per group x 7 groups = 49 filaments with the cross-section =

3.5 mm x 0.5 mm and ratio of Ag to superconducting core equals about 12 [181].

Generally, in the Bi-based system, there is a weak coupling problem in the c-axis

direction of its microstructure. There are also a number of factors which'influence the

Jc-H behavior in Ag-clad (Bi,Pb)2Sr2Ca2Cu3Oio+x wires. These include grain alignment,

temperature, duration and atmosphere of heat treatment, doping and substitutions, plus

phase assemblage [182-184]. It is clear that densification and grain alignment, derived

from repeated rolling, pressing, and sintering, are critical for raising the Jc. Other

factors, such as Ag additions [185], Ca2Cu03 excess [186], uniform distribution of

40

impurities, elimination of carbon [187], and partial melting, are also beneficial to the Jc-

H characteristics. However, some factors are still not well understood. For example, the

low Tc phase (Bi,Pb)2Sr2CaCu20g+x and various defects have been observed in Ag-clad

tapes, but the effect of these phases on the Jc is not very clear.

For the Bi-based system, the compositions of these two phases are

approximately Bi2Sr2CaiCu20&« (Bi-2212) and Bi2Sr2Ca2Cu3Oio+x (Bi-2223), with Tc

values of about 85 K and 110 K respectively. Bi-2212 and Bi-2223 phases exist over a

wide range of stoichiometrics, but neither has been prepared in phase-pure form. The

Bi-2212 phase can be formed more easily than the Bi-2223 phase. The kinetics of the

reaction and stability of the Bi-2223 complicate its development. It has been shown that

the formation of the Bi-2223 can be enhanced by the partial substitution of bismuth

with lead. Synthesizing techniques such as solid-state reaction, two-step process,

coprecipitation, freeze drying, etc., have been developed to produce powders with

desired properties [188-190].

2.3.23 Achievement on the Bi-2223/Ag HTS wires

There are various techniques developed with regard to making a HTS wire; the

powder-in-tube (PIT) process with the bismuth-strontium-calcium-copper-oxide

BiSrCaCuO system seems to be the most promising and industrially viable method for

fabricating long lengths of high Tc superconducting wires. Significant progress has

been made in the development of high Tc monofilament and multifilament Bi-based

superconductors by the PIT technique. In the PIT process, texturing and grain

alignment, which dominate the superconducting properties of the Bi-based tapes, can

be easily obtained through a series of thermomechanical treatments.

41

Very high Jc values for short length samples, subjected to a series of uniaxial

pressing and heat-treatment cycles, have been achieved in both single core and

multifilament superconductors. The highest Jc of short Ag-clad Bi-2223 tape, 69,000

A/cm2 was achieved at 77 K and with self-field [191].

The critical cunent density Jc is dependent on the applied magnetic fields and

operation temperature as shown in Figure 2-1, which indicates the situation in 1993.

The relationship between Jc and applied magnetic fields is anisotropic. Long lengths of

superconductors, up to several hundred meters or even kilometers have been produced,

and high Jc has been achieved in the long-length Bi-2223 superconducting wires. As

reported by K. Sato (Sumitomo Electric Industries Ltd, Japan) 1993 [192, 193], a 100

m single-core wire and a 114 m 61-filament wire were made with full length Jc (10"13

Qm criterion) 6570 A/cm2 and 9700 A/cm2 respectively, at 77 K and with self-field. An

850 m 37-filament wire has achieved a Jc of 1.2 x IO4 A/cm2 (L = 16 A), at 77 K with

zero applied field, reported by U. Balachandran (Argonne National Laboratory, USA)

1994 [194-197]. A Ag-clad Bi-2223 61-filament wire has achieved 53,700 A/cm2 at 77

K with self-magnetic field, and can maintain 42,300 A/cm2 at 0.1 T and 12,000 A/cm2

at 1 T [193]. It exhibits a Jc of 1.03 x 105 A/cm2 at 4.2 K and 23 T, and 5.5 x IO4 A/cm2

at 20 K and 19.75 T. These Jc - B properties at 4.2 K - 20 K are much better than the

values of conventional metallic superconductors at 4.2 K.

Table 2-2 summaries the achievement on the HTS wires by various groups in

Japan [21].

42

f Jc (A/mm2!

Ag/BiPbSrCaCuO Wire

5.4x10 A/mm

\0\ Nb3Sn

r4.2lT^'v^x N \

•> 20K \ \

50 x\l00 IV

/x0H(T)

Figure 2-1. Critical surface (Jc-B-T) of the Bi-2223/Ag H T S wire measured using transport current technique [211].

43

Table 2-2. Critical current density of oxide superconducting tapes [21].

HTS Materials

YiBa2Cu307.x

Bi-Sr-Ca-Cu-0

Bi-Pb-Sr-Ca-Cu-0

Tl-Ba(Sr)-Ca-Cu-0

Process

Ag-sheathed

tape, heavy cold

work

Ag-sheathed

wire, and tape

with heavy cold

work

Doctor blade

casting, Ag tape

Ag-sheathed

tape, heavy cold

work

Ag-sheathed

tape, heavy cold

work

JCR (A/cm2)

3.3 x 10J at B = 0,

T = 77.3K

1.0xl03atB = 15T

T = 4.2K

2 x 104 at B = 0,

T = 77.3K

4 x IO4 at B = 26 T,

T = 4.2K

2.1xl05atB = 30T,

T = 4.2K

1.6xl05atB = 17T,

T = 4.2K

4.7 x 104 at B = 0,

T = 77.3 K

1.3xl05atB = 23T,

T = 4.2K

1.6xl04atB = O,

T = 77.3K

Main Ref.

Okadaetal. (1987)

(Hitachi) [198]

Osamura et al. (1989)

(Kyota Univ.) [199]

Krauthetal. (1990)

(Vacuumschmelze)

[200]

Enomoto (1990)

(Furukawa) [201]

Toganoetal. (1990)

(NRIM) [202]

Sato etal. (1990)

(Sumitomo) [203]

Aiharaetal. (1990)

(Hitachi) [204]

44

2.33 High T c Superconducting Coil

The H T S wires, having a material-processing defined configuration which aims

to have the highest Jc, have been used to prepare primary H T S coils. Apart from the Jc

of the H T S wire, the mechanical property of the wires is another main concern relative

to its practical applications. Silver is widely used as the sheath material in the PIT

process, but its mechanical properties seem insufficient to withstand the stresses

introduced during fabrication and service. Consequently, various techniques are

considered to improve the strain tolerance of this wire [205, 206]. Multifilamentary

techniques have been used in conventional metallic superconductors to avoid

electromagnetic instability. The multifilament configuration has more reliable geometry

for Bi-2223 high T c superconducting wire because of its better mechanical properties

and improved cunent distribution. Multifilament Bi-2223 superconductors have shown

a much better mechanical flexibility [207, 208]. Testing shows that a 0.7% strain (30

m m diameter mandrel) does not reduce the Jc of a 1296-filament wire, and it is

significantly better than the 0.2% measured for monocore tapes [209, 210]. The

bending strain tolerance of the Bi-2223 multifilament wire enables a high Tc

superconducting coil to be made by a react and wind procedure.

Magnetic fields generated by a primary H T S coil made by the Ag-clad B-2223

superconducting wire have achieved 0.21 T at 77 K, 20.35 T at 20 K (with background

field Be x = 20 T), and 23.37 T at 4.2 K (with Be x = 23 T), in Japan [211]. In the U S A ,

another research group has reported that a test magnet, fabricated by stacking ten

pancake coils, generated a field of 2.6 T at 4.2 K with zero applied field. At 27 K, 64 K,

and 77 K, the magnet generated fields of 1.8 T, 0.53 T, and 0.36 T respectively, and

also another magnet generated a field of 1 T at 4.2 K in a 20 T background field [212].

45

The results indicate that this high T c superconducting magnet is capable of use as an

insert in a low Tc superconducting magnet to form a hybrid magnet for generating very

high fields.

23.4 Conductor-Design Consideration of High Tc Superconductor

The advantage of a HTS is its ability to be used at temperatures beyond 4.2 K

(the liquid-He range), for example at 20 K (liquid hydrogen) and 77 K (liquid nitrogen),

with very high upper critical field and high theoretical Jc values. Therefore it has been

identified as a candidate material for a conductor to carry large cunents and to produce

high magnetic fields, with less energy consumption compared with resistive

conductors.

The HTS has some special characteristic features, such as anisotropy of Jc, weak

link at grain boundaries, and flux creep. It has been observed that for the transport

cunent there is a ratio J^ I Jcc - 102-105 for the HTS thin film and single crystal [213].

Due to weak links at the grain boundaries, the transport Jc is a few orders of magnitude

smaller than the Jc estimated from magnetization hysteresis. There is a flux creep effect

in the high temperature region near 77 K which causes a broadening of resistive

transition [214], relaxation of magnetization [215], and a degradation of Jc [216].

Apart from critical cunent density Jc, properties to be addressed by the conductor

designer are mainly the HTS wire stability and AC loss. The HTS parameters have

been used for conductor design with the HTS by considering flux-jump stability,

cryogenic stability, and AC loss [217-225]. Using material processing technology, the

multifilament and even twisted multifilament wires [174] have been produced with

configurations which are required to make practical conductors.

46

2.4 POTENTIAL APPLICATIONS OF HIGH Tc SUPERCONDUCTORS

IN ELECTRICAL ENGINEERING

The technology of conventional metallic superconductors in production and

application has been significantly developed, as some of them perform very well with

good electrical, magnetic and mechanical properties. So far, there are a number of

extra-high field magnets, transmission cables and generators which have been designed

and built, based on these superconductors. The present commercial superconductors for

large-scale application are typically NbTi or NbsSn superconducting strands, embedded

in a matrix like Cu. The cunent densities are up to IO5 A/cm2, in magnetic fields of up

to several teslas (T). However, the barrier to operation with this kind of material is

extremely great, given its working environment and the huge cost of cooling it down to

about 4.2 K (-269 °C i.e. liquid helium temperature). The new oxide ceramic materials

have critical temperatures which can be higher than 77.3 K (-196 °C) that of liquid

nitrogen temperature with applicable critical cunent density Jc and very high upper

critical fields H^. The coolant of liquid helium costs generally about $25 a litre

cunently, compared with only 60 cents a litre for liquid nitrogen. For cooling in a

practical system the refrigerator will need more than 1 k W to remove 1 W thermal

energy at the liquid helium temperature, and less than 20 W at the liquid nitrogen

temperature. Not only is the cost of cooling reduced but the engineering design with

these superconductors is simplified.

It is expected that the use of high T c superconductors at 77 K liquid nitrogen

temperature will greatly affect such fields as power transmission, power generation,

energy storage, motors, transportation, medical technology, and also electronics. These

47

high T c superconductors have the potential to be operated with liquid nitrogen (N2),

whose boiling temperature at 1 atm is 77.3 K. The advantages of the use of liquid

nitrogen as coolant are not only is it inexpensive and easy to use, but also simplifies the

cryostat and allows a large heat flux.

The high Tc superconductors now are considered for a number of electrical and

electronic applications. For small-scale application, i.e., where the cunents are much

smaller, the more specialised properties of superconductors tend to be exploited and

examples include detection systems like superconducting quantum interference devices

(SQUIDs) as well as analogue and digital processing. Comparing with small-scale

applications, a large-scale application generally means larger cunents and lengths of

superconductors are required in working environments where the magnetic field may

be several teslas. Examples include magnets, power transmission lines, transformers,

energy storage, generators, fault cunent limiters and other devices in electrical

engineering. For large scale applications, the basic requirement of the high Tc

superconductor is that they must be able to carry a large cunent with zero resistance

under a high magnetic field and that they must be long enough, with high flexibility, in

order to achieve high values of cunent ampere-turns in coils. In these applications, the

high Tc superconductor critical cunent density Jc is normally required larger than IO4

(A/cm2), with a magnetic field from about 0.2 T for a transmission cable to about 4 T

for a generator.

The Jc - B - T characteristics of the Bi-2223 superconductors have allowed the

wires to be designed at 77 K for low field applications, and at 4.2 K to about 40 K for

high field applications. At 40 K and above, Jc decreases drastically when the magnetic

fields are applied perpendicular to the tape (H // c).

48

While the processing techniques and critical cunent of the high Tc

superconductors are being improved, there are some laboratory research prototype

devices which have been built or designed with high T c superconductors. With the Bi-

2223/Ag high T c superconducting wires, the following primary research prototypes

have been built and considered, besides the high T c magnets mentioned before, cunent

lead [226-230], busbar or power cable [231-233], transformer [234, 235], fault cunent

limiter [236,237], and generator [238,239].

2J DEVELOPMENT OF ELECTRICAL FAULT CURRENT LIMITER

WITH HIGH Tc SUPERCONDUCTORS

2.5.1 Fault Currents in Electrical Power Systems

In an electrical power transmission or distribution network, a fault cunent occurs

basically due to any kind of short circuit. Enormous cunents, much higher than the

normal value, will flow through the system instantaneously. This short circuit, as an

example, m ay be a lightning strike temporarily ionising the sunounding air.

Consequently, in an electrical power transmission or distribution network, problems

arise when any kind of short circuit occurs. The electrical equipment used in the

distribution network must be able to withstand the fault level associated with these

faults and equipment must be rated accordingly. As new generating capacity is added to

the system, the fault level increases and the existing equipment, such as a switchgear,

may no longer be able to withstand the increased fault level. In many cases, it is much

49

too expensive to replace the existing electrical equipment, so in order to keep the same

switchgear and circuit breakers, and avoid failures of the equipment to withstand and

clear this extra fault level, a way of limiting this increased fault level is needed.

If a fault current occurs it could cause destruction, such as burning out the feeder

line and creating explosions at the fault site or at the switchgear. Consequently all the

in-line components, such as switches and breakers, must be suitably rated in the design

so that they are not destroyed by the fault. High potential fault levels mean an increase

in the capacity of these components, i.e. more cost and more complicated design.

Broadly speaking, the faults in electrical systems can be classified as (i) Shunt

faults (short circuits); (ii) Series faults (open circuits).

Shunt faults involve conductor-to-ground conditions or a short circuit between

conductors. When circuits are controlled by fuses or any device which does not open all

three phases, one or two phases of the circuit may be opened while the other phase or

phases are closed. These are called series types of faults. Shunt faults are characterised

by an increase in cunent and a fall in voltage and frequency whereas series faults are

characterised by an increase in voltage and frequency and a fall in cunent in the fault

phases.

Shunt faults are classified as (a) Line-to-ground fault; (b) Line-to-line fault; (c)

Double line-to-ground fault; and (d) 3-phase fault. Of these, the first three are

asymmetrical faults as the symmetry is disturbed in one or two phases.- The 3-phase

fault is a balanced fault which could also be analysed using symmetrical components.

The series faults are classified as: (a) One open conductor, and (b) Two open

conductors. These faults also disturb the symmetry in one or two phases and are,

therefore, unbalanced faults.

50

As an example, in a 12 k V distribution line, a normal operating cunent is 300

A/per phase; subtransmission lines (66 - 115 kV) and transmission lines (above 230

kV) with a typical 700 A normal operation cunent, fault cunents in such systems can

exceed 20,000 A. If a fault cunent occurs it could cause destruction, such as burning

out the feeder lines, creating explosions at the fault site, and destroying in-line

equipment, e.g. switchgear. Consequently all the in-line components must be suitably

rated in the design so that they are not destroyed by a possible fault. The growth in

power demand and the interconnection of the power transmission networks lead to

very high short-circuit cunents. The fault cunent levels are becoming higher in the

existing power systems. The increased fault cunent level means a higher potential

destructive cost, and an increase in the capacity of the in-line components, i.e. more

cost and more complicated design.

Circuit breakers are normally used to protect transmission or distribution

systems. Their principle is to cut off the circuit at zero cunent because the energy of

the electrical arc is reduced at this time, therefore all the in-line components have to

withstand the destructive effects of the short circuit cunents for a period of at least 10

ms in the case of a 50 Hz symmetrical fault, and for about 20 ms in the case of a 50

Hz asymmetrical fault. Even for the most rapid of circuit breakers [240], which

typically operate within 1 cycle, all the in-line components have to be designed to

cope with the higher potential fault level.

51

2.5.2 Solutions to Fault Currents

As electrical power systems grow in capacity, so does their potential fault

cunent. The existing electrical equipment in the system has to be able to cope with the

increased fault level. There are two ways of achieving this. One, the existing

equipment can be replaced by equipment with a higher rating. Alternatively, the fault

level can be reduced by the use of a suitable fault cunent limiter. The cost of replacing

the existing equipment can be enormous, and the second option is usually employed.

For low voltage equipment, the cunent can be limited almost instantaneously by

using electronic power components like the GTO (Gate turn off thyristor) or by using

cunent hrniting circuit-breakers equipped with electrodynamic repulsion contacts and

magnetic blast coils, or by using fuses etc. For medium voltage equipment, up to 36 kV,

fuses are frequently used as cunent limiters. This method means that the fuses must be

changed after each fault, which constitutes a major inconvenience. Furthermore the use of

fuses is not applicable in HV networks. For high rated cunents, above 2000 A for

example, the fuse or cunent limiting breaker is often shunted by a contact opened very

quickly during the fault [241, 242]. Cunently circuit breakers are normally used for

protection from the fault cunents.

Some methods [243] considered as possible solutions to the problem of increased

fault levels are

(i) Splitting the grid and providing interconnection through higher voltage AC or

DClinks.

(ii) Increasing the short circuit withstand capability of the substation by replacing the

substations.

52

(iii) Limiting the short circuit cunent by operational methods like sectionalizing

and sequential network tripping using a sophisticated control system etc.

(iv) Developing and introducing cunent limiting devices.

Because the cost to upgrade the existing systems is enormous, interest in the

development of fault cunent limiting devices has increased over last decade.

2.5.3 Developing Fault Current Limiter in Electrical Power Networks

Fault cunent levels can be reduced by adding fixed reactance to the network,

such as a linear inductor, which is one of the present solutions, but this inductive load

may cause severe stability problems or excessive use of transformer tap changers.

Ideally a fault cunent limiting device is required with a low reactance under healthy

network conditions and a high reactance under fault conditions. Therefore a fault

cunent limiter (FCL) is a device which has a low inductance during normal operation, a

necessary requirement for better power system stability, and a high inductance, and

thereby impedance during fault conditions. The fault cunent limiter can protect the

breaker and the transmission network and reduce the fault level to which the breaker,

busbar, transformer and switch are subjected. It can also stabilise the power system.

Normally a fault cunent limiter is a series device; it presents a low impedance to

cunent flow under normal conditions. W h e n a fault occurs, this impedance rapidly

increases to limit the cunent flowing into the fault. It may be viewed as a normally

closed switch in parallel with a resistor. For D C cunent, the switch transfers cunent

abruptly into the resistor at a fixed time after the fault occurs, regardless of the phase

angle of the A C voltage. A n alternative is an A C switch or fuse that increases the

53

impedance more gradually than a D C switch and therefore shares cunent with the

resistor until the first zero cunent, when full cunent transfer into the resistor is made.

In power system networks, breakers are the devices which interrupt fault cunents.

In a normal circuit, fault cunent usually will be stopped after a delay due to mechanical

inertia of a few cycles by the opening of circuit breakers. However, these breakers need

to be sized to take the enormous potential fault cunent levels and still be capable of

opening. Breakers could be downsized by the fault cunent limiter, which means the

fault cunent limiter may reduce the capacity to which the circuit breaker can be

subjected.

Breaker interrupting times have been improved from 8 to 5, then down to 3 and

finally 2 cycles. Now even a one-cycle circuit-breaker with a minimum interrupting

time of 12.5 msec, associated with a 1/4 cycle relay, has been built successfully [240].

It is well-known that all AC high voltage (HV) circuit-breakers require zero cunent

passage in order to interrupt the fault. To interrupt or limit a fault cunent at HV,

however, well before its first peak, is a much more difficult problem to solve as it

requires a very short total functioning time (detection time + communication time +

function time + limiting time + recovery time) of less than, or equal to, 2 milliseconds

in order to limit the fault cunent's amplitude to a fairly low pre-determined value.

Superconductors have been employed to design electrical fault cunent limiters.

The design of fault cunent limiters with superconductors was considered initially with

conventional superconductors. There are few operation principles to design an

electrical fault cunent limiter with superconductors. Under the superconducting

quenching model, the fault cunent limiters can be separated into two kinds i.e. the

resistive [244-246] and the inductive [247, 248] types. For the non-quenching models

54

of the fault cunent limiter, there are solid-state methods [249], the magnetic core

saturable reactor method [243], and so on.

(a) Superconductor-quenching resistive switch

This is an essential idea to use a superconductor in the superconducting fault

cunent limiting device [245]. A superconductor-resistive fault cunent limiter is based

on a large resistance transition from the superconducting state to the normal state. The

principle of operation of the resistive superconducting fault cunent limiter is to

commutate the cunent after the fault, from a superconducting element into a shunt

resistor, which limits the fault cunent to an acceptable level. The commutation is

accomplished by switching the superconducting element from zero-resistance state into

its resistive state, by increasing I above L, or T above Tc.

In the simplest design the superconducting element is placed directly in series

with the line and the limitation is purely resistive due to the resistance of the

superconductor being quenched. T o avoid excessive dissipation, and possible

destruction of the superconductor, other designs use the superconducting element to

"trigger" the hmiter and the fault cunent is redirected into a limiting resistor or

inductor. This also allows the superconductor to recover in time for another operation.

The primary conclusion is that a superconducting fault cunent limiter (SCFCL) is

technically feasible for a wide range of parameters for the required electrical power

source and material properties of the superconductor available, assuming that the

superconductor can be switched without fusing or producing excessive overvoltages.

The following are the specific advantages

(i) Solid-state switching. There are no moving parts or materials in the

commutating element, only the electrons change their state. In other words, nothing

wears out in a SCFCL.

55

(ii) L o w losses. During normal operation the losses are entirely due to heat leaks

into the cryostat and constitute only about 0.01% of the transmitted power.

(iii) Fast recovery. A computer simulation indicates that the superconducting

element is ready for load cunent within 15 m s of the fault, i.e., even before the circuit

breaker has opened. This could be very useful to quickly restore service to

interconnected systems.

(iv) High voltage. There are no particular problems with the superconductor at

high voltages, the element is just designed to be longer. High voltage insulation

problems will be a major concern, especially at the edges of the superconducting film.

There is a considerable cost advantage if the S C F C L is used in conjunction with

other superconducting or cryogenic equipment, since that eliminates the cost of the

cunent feedthroughs and the major expense of cooling them. In an all-superconducting

system (generator, transformer, and transmission line) there would be a considerable

saving if a low-temperature fault limiter is available. Along with these are the following

disadvantages

(i) L o w temperature. Dissipation of power at low temperature costs about 450

times as much as at room temperature. A closed-cycle refrigerator, with possible

maintenance is required.

(ii) Sensing. The device is not self-sensing.

(iii) Load cunent. W h e n not used with other cryogenic equipment, the costs

become much greater as the continuous load cunent is increased due to the refrigeration

for the high-current feedthroughs.

The advent of superconducting cunent limiting devices seems certain to open the

way to new functioning and protection designs for the large electric networks in the

near future. Compared with previous cunent limiting devices, the superconducting ones

56

allow the three functions of detection, commutation and limitation, to be regrouped in

the superconductor itself. This entirely static functioning should greatly increase the

reliability of the apparatus.

For future networks equipped with superconducting A C generators and

transformers, the presence of superconducting cunent limiting devices will be more

necessary in order to avoid harmful transitions on these large machines following faults

on the line.

Indeed, the instantaneous limitation of the short-circuit cunent to a relatively low

value, the self-protection and very short fault time which these limiting devices allow,

should greatly improve the network's function (increased transmitted power, improved

stability, reduced stresses, extended material life, etc.).

Ob) Superconductor-quenching inductive limiters

For the resistive limiter, the superconducting winding triggers and effects the

limitation, or switches to a parallel shunt. For the inductive limiter, the superconducting

winding acts only as the trigger while the fault cunent is limited by another

superconducting or normal winding. In this case it is envisaged that there is limited

heating of the superconducting coil and then a very short recovery time. It will enable

quick restoration of the service. O n the other hand, the cunent in the trigger coil may be

selected lower than the rated cunent so that the choice of the superconducting wire may

be made easier. The use of a judicious magnetic circuit may improve the performance

as it decreases the voltage drop in rated operation and the heating of the

superconducting coil during a fault.

The objective is the design and conception of a superconducting cunent limiter

with an intrinsic short reclosing time, typically 0.3 s. During the reclosing time three

phenomena will occur successively [247]

57

(i) Ultra fast limitation of the cunent triggered by a quench;

(ii) Break of the limited fault cunent by a classical two or three cycle circuit

breaker;

(iii) Recovery of the superconducting state.

As an example, an inductive limiter essentially consists of two coils connected in

parallel, and in series in a line [247]. The first coil, superconducting or normal, is called

the hmiting coil, and the other is the trigger coil, which is superconducting. The two

coils are closely coupled in normal operation to offer a low voltage drop, and are well

uncoupled under fault conditions.

Compared to the resistive limiter this inductive one has a better failure

characteristic; under accidental quench conditions the operation of the lines could not

be reduced. O n the contrary, the line has to be switched off after a quench of a resistive

limiter without another device in parallel.

(c) Saturable magnetic core reactor

The saturable magnetic core reactor can be employed to design a fault cunent

limiter [243]. It consists of a laminated iron core, rather like a transformer, with an A C

winding in series with the power network system, and a D C winding which is energised

and completely saturates the core with magnetic flux. Under normal network

conditions, the system load cunent in the A C winding is unable to drive the magnetic

core out of saturation, and therefore this winding is effectively air cored, which results

in a low voltage drop. W h e n a fault occurs, the cunent in the A C winding starts to rise,

and beyond the peak of the m a x i m u m normal load cunent, begins to drive the magnetic

core out of saturation; when this happens the voltage drop across the A C winding

increases instantaneously, and therefore the fault cunent is limited. It is necessary to

have two cores per phase because it is not known in which half cycle the fault will

58

occur, so one core will limit the fault cunent which occurs in the positive cycle, while

the other one will function in the negative cycle. This is not a new idea but if the DC

bias winding is made of a normal conductor the losses are prohibitive for high cunent

rated practical application, and a high level DC power supply has to be provided to

drive the magnetic core into saturation.

(d) Other types of superconducting fault current limiters

There are also different designs of superconducting fault cunent limiter, e.g. (i)

3-phase single fine to ground superconducting fault cunent limiter [250]; (ii) short-

circuit superconducting secondary winding reactor [241]; (iii) solid-state

superconducting fault cunent limiter [249]; (iv) DC superconducting fault cunent

limiter [251]; (v) stable system with superconducting energy storage [252-257].

The fault cunent limiting device built with a superconductor may provide the

following possibilities for the network which need to be considered

(i) Fault cunent amplitude reduced. Influence on circuit-breakers,

disconnecting switch, busbars, cunent transformers, lines, generators

and transformers;

(ii) Reduction of fault cunent duration;

(iii) Auto-detection and auto-limitation of the fault cunent;

(iv) Limitation of over-voltage at interruption.

Therefore the superconducting fault cunent limiters offer several operational

advantages as follows

(i) Reduced rating requirements for downstream circuit breakers;

(ii) Fewer system changes;

(iii) Limit voltage sags and minimize brownouts;

59

(iv) Enhance the system (more stable and not easy to collapse);

(v) A single device can handle multiple substations.

2.5.4 High T c Superconducting Fault Current Limiters

2.5.4.1 Developing high Tc superconducting fault current limiters

Developing FCLs is one of the important applications of HTS in electrical

engineering [258-261]. A fault cunent limiter could be constructed using low Tc

superconductors but high Tc superconductors offer potentially several essential

advantages. The most important is that a cryocooler can be easily built to attain the

envisaged operating temperature which could be much higher than the liquid helium

temperature, and other consequent advantages are: (a) Significant refrigeration

savings; (b) Convenient cryocooler operation; (c) Higher stability (with a larger AT =

t c" t operation)-

The family tree of advanced fault cunent limiters has been depicted by E.M.W.

Leung et al. from Lockheed Martin Corporation, American Superconductor

Corporation et al., and it is summarised in Table 2-3 and Figure 2-2 [261}.

60

Table 2-3. Fault current limiter comparisons [261].

Project

company/country

ABB,

SWITZERLAND

LMC, USA

TOSHIBA, JAPAN

GEC, ALSTHOM,

FRANCE

Device concept

Inductive control

with shielded core

Inductive control

with electronic circuit

Inductive control

with trigger coil

Resistive control

Superconductor

High Tc superconducting cylinder,

YBa 2Cu 30 7 bulk

Low or high Tc superconducting coil,

(Bi,Pb)2Sr2Ca2Cu3CWx wire

Low Tc superconducting coil, NbTi

Low Tc superconducting coil, NbTi

61

s/c

v© si

i I load

7777777777 No fault - Low inductance Fault - High inductance

(a)

RN=13,500Q [loadl

Pulse field

' Q_<3 CD

V

No fault -1 thru s/c coil Fault -1 ihru shunt If I reaches Ic line will fuse $295K, just coil - Argonne

(b)

Limiting Coil, s/c

Trigger Coil, S C -^JTmTYTJL—

v @

No fault - Low inductance Fault - Trigger coil quenches high inductance thru limiting coil

(c)

circuit breaker

t i

ii

Limiter 1 I loadl

////////

metal winding

&

H T S C shield iron core

ii i I loadl

7777777777 i

SC winding secondary

metal winding primary

No fault - Low inductance, no resistance Fault - Lots of resistance

(d)

No fault -HTS shields magnetic flux Fault - Flux penetrates HTS shield

(e)

No fault - S C saturates iron Fault - Alternate cycles cancel SC flux

(0

(a) Conceptual, (b) Triggerable at A N L , (c) Inductive at Toshiba, (d) Resistive at G E C -

Alsthom, (e) Screened core at ABB and (f) Saturated core at Wollongong.

Figure 2-2. T h e various high T c superconducting fault current limiters [261].

62

2.5.4.2 High T c superconducting shielding inductive reactor

This FCL is based on using a HTS cylinder which can be made by textured

YBa2Cu307 or Bi-based superconductors [258-260]. The basic idea is to use a HTS

cylinder between the AC circuit winding and the magnetic core used. The AC winding is

series-connected to an AC circuit. In its normal condition, i.e., AC cunent is below the

reactor operation threshold, the magnetic flux generated by the AC cunent is shielded by

this HTS cylinder and does not penetrate into the magnetic core. In this case, the reactor's

inductance is low and therefore the impedance is low during normal operation, and it

depends on the leakage flux in the winding. Under fault conditions the cunent in the circuit

increases and leads to a higher magnetic field on the surface of the HTS. When the field

penetration depth reaches the shield thickness, the magnetic flux begins to penetrate the

magnetic core. This brings a non-linear dramatic rise in the reactor's inductance and

conespondingly to its impedance, and the fault cunent is thus limited by the reactance.

An analysis of this design has been discussed [258] for the voltage and cunent up to

25 kV rms (3-phase) and 2000 A rms, and shows the recuperation times for the

superconductors operating at liquid nitrogen temperature are acceptable. As high

temperature superconductivity progresses it should become possible to fabricate the large

diameter cylinders, tubes, or stacks of rings, with the required critical cunent density

greater than IO4 A/cm2 at 77 K. With respect to the development of superconducting

materials, a specific need of this design is a high Jc in low fields, below critical screen field,

and low Jc at the high fields during fault conditions. This will reduce the overall

temperature rise of the superconductor and speed up the recovery time for subsequent

operations.

63

2.5.4.3 Bi-2223/Ag high T c superconducting wire and its fault current limiter

Using the Bi-2223/Ag HTS wire, HTS coils or windings can be produced.

Consequently some fault cunent limiters can be designed with the HTS wire. A prototype

fault cunent limiting device has been designed by using the HTS wire and an electronic

switch [261]. This is the only prototype device of fault cunent limiter which uses the Bi-

2223/Ag HTS wire, with the exception of the concept reported in this thesis. When the

HTS wire is further improved, it can be used to replace the LTS coils in the fault cunent

limiters where the LTS wire is used, e.g. some inductive and resistive devices [244-248],

Although the cunently available properties of the high Tc superconducting wires still

fail to justify commercial applications, some devices can still be based on these materials.

With their performance improved, possibly in the next few years, due to increased critical

cunent densities up to 104~105 A/cm2 in fields up to 2 T at 77 K, the HTS coils can be

made in reasonable scale with the HTS wire for such an application.

2.5.4.4 Summary of the development of HTSFCLs

Conventional fault cunent Umiting devices are designed with copper windings

and produce continuous losses which reduce the efficiency of the transmission and

distribution systems in which they are used. The conventional copper conductor has

not been suitable to design a high efficiency electrical fault cunent limiting device.

Metallic low Tc superconductors have been considered as the basis of a FCL and

designs based on operation at helium temperature have been proposed. Most of the

designs are based on the transition from superconducting state to a progressively more

resistive state as the cunent increases beyond the critical value. A superconducting fault

cunent limiter with conventional superconducting film was proposed by Gray in 1978

64

[245]. In this case superconducting thin film is stable in the superconducting state and

will quench to a high resistance normal state due to the fault cunent. The basic

requirements of superconductors in this case are high Jc, low AC loss, high normal

resistance, fast response and ability to recover.

The recently developed high Tc superconductors have substantial advantages and

potential capability to replace those conventional low Tc superconductors.

Consequently the operation cost can be significantly reduced by using the high Tc

superconductors. As summarised in Figure 2-2, there are different methods to design or

build HTSFCLs with different formats of high Tc superconductors. The HTSFCLs can

be categorised mainly as three types which all the HTSFCLs belong to: (i) Air core

inductive reactor; (ii) Magnetic core inductive reactor; and (iii) Resistive reactor. In

terms of the HTS format, there are (i) Bulk YBCO or BSCCO shielding cylinders or

rings; (ii) YBCO thin films or BSCCO thick films; and (iii) Bi-2223/Ag wires.

To limit fault cunent with superconductors is not a new idea, but it becomes

more attractive with the development of the high Tc superconductors. At present, the

situation is to establish new concepts of fault cunent limiters, and to link HTS materials

with electrical engineering. The theory and technology of producing practicable high Tc

superconductors have been under exploration since the liquid nitrogen high Tc

superconductors were discovered in 1987. So far, the electromagnetic behaviour of the

high Tc superconductors is still not reaching practical requirements, especially at 77 K

operation in liquid nitrogen. The hope is that the critical cunent value of the high Tc

superconducting long wire and the high field shielding capability of bulk high Tc

superconductors are becoming close to some practical requirements. The FCL will

become commercially available based on improved high Tc superconductors.

65

PRACTICAL WORK

66

Chapter 3

Bi-2223/Ag HIGH Tc SUPERCONDUCTORS

3.1 INTRODUCTION

High Tc superconductors (HTS) demonstrate potential in electrical

applications with respect to energy saving, low operation cost and high efficiency

when compared with conventional low temperature superconductors. Cunently the

Ag-clad (Bi,Pb)2Sr2Ca2Cu30io+x HTS wire is one of the most promising candidates

for high power magnets and other electrical applications [262-270]. This HTS wire

has the capability to be used to produce a HTS coil or a winding, because of its high

critical cunent density and long length with fairly good mechanical flexibility [271-

281]. The critical cunent density of this HTS wire at 77 K liquid nitrogen

temperature has achieved about 69,000 A/cm2 in short length samples [191], and

about 12-18 kA/cm2 for lengths over 1.2 km at 77 K [282, 283].

The Ag-clad (Bi,Pb)2Sr2Ca2Cu3Oio+x HTS wires are selected to develop HTS

coils, properties of the Ag-clad (Bi,Pb)2Sr2Ca2Cu30io+x HTS wires are identified with

respect to their performance for practical applications. The winding capability of the

Ag-clad (Bi,Pb)2Sr2Ca2Cu3Oio+x HTS wire is related to strongly magnetic field-

dependent critical cunents, and the magnetic field-dependent critical cunents are

essential in the use of the HTS wire as a conductor. Other issues which

influence the design of a suitable practical conductor include mechanical properties

67

and A C loss when used for A C operation. Experiments carried out on the Ag-clad

(Bi,Pb)2Sr2Ca2Cu3O10+x wire in this work include its critical cunent related with

applied magnetic fields, magnetic field distributions of the HTS wire, mechanical

bending properties, operation temperatures and transport AC cunents. The Ag-clad

(Bi,Pb)2Sr2Ca2Cu3Oio+x HTS wires are judged, based on the experimental results and

analysis obtained, for developing HTS coils.

Using the Ag-clad (Bi,Pb)2Sr2Ca2Cu3O10+x HTS wires, HTS coils are made by

using different methods and techniques. The HTS coils fabricated include the

pancake, double pancake and solenoid coils, which are wound by using both single

core and multifilament Ag-clad (Bi,Pb)2Sr2Ca2Cu30io+x HTS wires, and by using the

react and wind, and the wind and react procedures. Experiments are carried out to

study the HTS coil with regard to its critical cunent ampere-turns and magnetic field

properties. The HTS coil magnetic field distributions are also analysed using a

computer calculation program. The testing results and analysis are with a view to

achieve a practical design of a HTS coil with the anisotropic Ag-clad

(Bi,Pb)2Sr2Ca2Cu3O10+x HTS wire.

3.2 Bi-2223/Ag HIGH Tc SUPERCONDUCTING WHUES

3.2.1 Producing HTS Wires

The Ag-clad (Bi,Pb)2Sr2Ca2Cu3Oio+x HTS wires used are produced by the

normal powder-in-tube techniques, and these Ag-clad (Bi,Pb)2Sr2Ca2Cu3Oio+x wires

68

include a single core wire and multi-core (multifilament) wires. The powder-in-tube

techniques are started by packing the processed HTS powders into a Ag tube. The

tube was then drawn to a thin wire and rolled to a tape. Subsequent roll-sinter or

press-sinter cycles on the tape were repeated several times.

The HTS wires used for developing coils or windings, and also for use in the

project of electrical fault cunent limiter covered in this thesis, are mainly produced at

the University of NSW and the University of Wollongong by a superconductivity

research team. As an example, a single core HTS wire used is briefly introduced here.

The Ag-clad (Bi,Pb)2Sr2Ca2Cu30io+x HTS single core ribbon has been produced

through the normal powder-in-tube techniques, and three rolling-sintering procedures

with total 240 hours at 830 °C. The HTS wire is about 20 m long with final cross-

section CSwire = 0.18 x 4 mm2, and superconductor-core CSHTS / CSwire ~ 1:6. The

short samples have total critical transport cunents Ic in a range of 5.25 A to 7.85 A,

determined by the 1 pV criterion, at 77 K liquid nitrogen temperature and with

applied magnetic field Ha = 0.

Another example is a 27-core multifilament HTS wire produced by

Wollongong University. The powder used for the 27-filament wire has cation ratio

Bi:Pb:Sr:Ca:Cu = 1.85:0.35:1.90:2.03:3.05. During the rolling-sintering procedure,

the wire was sintered on a A1203 tube having diameter 4> = 75 mm at 830 °C to 834

°C in four cycles for a total of about 300 hours. The final critical cunent of the 46.3

m long wire is about 10 A at 77 K. A maximum critical cunent was achieved at a

previous step i.e. with total 220 hours sintering, for which sectional critical cunents

vary between 13.7 A to 16 A at 77 K over the entire 46.3 m length of wire.

69

3.2.2 Characteristics of the H T S Wires

Practical performances of the Ag-clad (Bi,Pb)2Sr2Ca2Cu3Oio+x HTS wire are

investigated with a view to make a HTS coil. The following characteristics of the

HTS wires are essential to design a practical HTS coil, and also related to the HTS

conductor design: (a) Magnetic field and operation temperature dependent critical

cunent density Jc; (b) Mechanical bending properties; and (c) AC transport cunent

characteristic for AC operation.

3.2.2.1 Critical temperature Tc

The Ag-clad (Bi,Pb)2Sr2Ca2Cu30io+x HTS wire critical temperature Tc is

identified by using the four-point resistive measurement and the AC susceptibility

measurement. Short length samples are selected, and results are shown in Figure 3-1

(a) for the resistive measurement, and Figure 3-1 (b) is the AC susceptibility measured

result. In Figure 3-1 (b), the single core sample AC susceptibility signal x' + ix" ls

measured by using the lock-in technology, and its real part x' is recorded as shown in

the figure.

Typical operation temperatures of the HTS with liquid coolants are 77.3 K - liquid

nitrogen, 63.1 K - triple point of nitrogen, 90.2 K - liquid oxygen, 27.1 K - liquid neon,

20.4 K - liquid hydrogen, and 4.2 K - liquid helium.

The temperature margin between the Tc of the HTS sample and operation

temperature Top which is 77 K for operation at liquid nitrogen temperature, is AT = Tc -

T„p = 110 - 77.3 = 32.7 K which is much greater than 13.85 K (18.05 K - 4.2 K) for

NkSn and 5.3 K (9.5 K - 4.2 K) for NbTi operated in liquid helium. The latent heat of

vaporization for liquid nitrogen and helium is 198.64 U/kg and 20.41 kJ/kg respectively.

70

1

(0

0.08

0.06-

0.04

f 0.02 4

a:

0.00

Liquid nitrogen 77.4 K

J Bi-2223/Ag short tape

60 80 100 120 140 160

Temperature (K)

(a) Tc from resistive measurement.

180 200

0.00016-

> ^ 0.00014 ro

c '</) CD ro 0.00012

0.00010

-\ r

„<*

x signal „S" 0JOJ.|'JJV,'.W).U..\JJJ.J.UJO J.V.JO,.UU JJOJ*A.-JUOQU JJJJ.

Bi-2223/Ag single core tape

90 95 — I r-

115 120 100 105 110

Temperature (K)

(b) Tc from A C susceptibility measurement.

Figure 3-1. Critical temperature Tc of the Ag-clad (Bi,Pb)2Sr2Ca2Cu30i<HX wire.

71

Considering the flux-jumping in a superconductor, which is an intrinsic materials

property, the flux-jump threshold H*j, initial flux-jump field, is important to select the

operation temperature, and its value can be evaluated by [284-287]

Hfi2 = V[7t3Cv(T)(Top-Tc)] (3-1)

and quench field HQ is [284-287]

HQ2 = 87tj(Tc.T0p)Cv(T)dT (3-2)

where the Cv(T) is the volume specific heat. According to the superconductor stabilization

against flux jumping, the maximum stable filament diameter <|) under the adiabatic criterion

is [287]

4 = VtlO^/^C^TXTo-T^yj^T) (3-3)

The specific heat Cv of (Bi,Pb)2Sr2Ca2Cu3Oio+x is about 0.8 Jcm^K"1 at 77 K

and 0.835 Jcm^K"1 at 120 K. For HTS YBa2Cu307, its Cv is about 1.005 Jcm^K'1 at 80

K. The typical conventional low Tc superconductor (LTS) NiTi has Cv value about

5.76 x IO-3 Jcm'3K_1 at 4.2 K. The specific heat ratio of the HTS

(Bi,Pb)2Sr2Ca2Cu3Oi0+x to LTS NiTi is CvBi / CvNm ~ 139. According to Equations (3-

1) to (3-3), the larger Cv and Tc values lead to larger H* HQ and (J) values for HTS

conductors.

72

3.2.2.2 Critical current Ic

A. Critical current with applied magnetic field

A short piece tape of 20 mm long was cut from the Ag-clad

(Bi,Pb)2Sr2Ca2Cu3Oio+x single core long HTS wire mentioned in Section 3.2.1. The

transport cunents are then determined by using the four probe technique with 10 mm

voltage signal interval in magnetic fields applied in 3 directions. The results in Figure

3-2 (a) show the critical cunent reduces dramatically with applied magnetic fields at

77 K. The a or b-direction in the figure denotes that the applied field is parallel with

the tape broad face, and the b-direction is also the direction of self field generated by

a general coil wound from this wire. This anisotropic HTS wire has a better critical

cunent value when the applied magnetic field is parallel to the ab plane of the HTS

wire compared to the field parallel to the c-direction.

Figure 3-2 (b) shows the result for another short sample from the HTS 27-

filament Ag-clad (Bi,Pb)2Sr2Ca2Cu3Oio+x wires. From Figure 3-2, there is a dramatic

change in the Ic - H curve gradient at low fields, i.e. about 100 mT for the samples used.

It is suggested that only the strong linkage of the HTS material maintains the transport

cunent above this magnetic field value [288]. The Ag-clad (Bi,Pb)2Sr2Ca2Cu3Oio+x

wires exhibit strongly magnetic field-dependent and anisotropic critical cunents. At 77

K operation, the 27-filament sample has Ic(lT) / Ic(OT) ~ 20% (B//ab); consequently the

Ag-clad (Bi,Pb)2Sr2Ca2Cu3Oio+x wire has to be improved to maintain a higher

percentage critical cunent under practical magnetic fields for general applications.

73

0.00

Applied magnetic field (Tesla)

0.20

(a) Ic of a short piece single core sample with applied magnetic field.

25

20-

<T c 15-a> i— i—

2 104 ro

o

5 5 OH

-, . 1 • 1 < r

% h ft i <

A\

Bi-2223/Ag tape at 77 K

va

a B//ab o BAc=45° A B//c

> t>-.

S. --fl-xi...

a. E- ,

^ 5 . . -A - = & a = -&°" *.-. --6

T "

0

1 • I ' T

200 400 600 800 1000

Applied magnetic field (mT)

(b) Ic of a short piece 27-filament sample with different applied magnetic fields.

Figure 3-2. Critical current Ic of HTS samples.

74

The critical current density Jc of the single core sample at 77 K is Jc = Ic / C S H T S

= 7.6 x IO3 A/cm2 or conductor cunent density J = Ic / CS^ = 1.1 x IO3 A/cm2.

According to the highest critical cunent density Jc (6.9 x IO4 A/cm2, at 77 K) achieved

with this type of wire, the above mentioned dimension of the HTS wire as a conductor

has potential to achieve Ic = 0.71 x 102 A or conductor cunent density Ic / CS^ = 9.86

x 103 A/cm2 for 77 K operation, which is higher than that of the normally used Cu

conductor having current density JCu about JCu « 3 x IO2 A/cm2 as the limit for most

normal applications. The 10"13 Qm criterion is normally used for the HTS, which is

Q

much less than 1.68 x 10" Q m for a normal copper conductor at room temperature.

B. The Ic-H-T relation

When a superconducting wire is used as a coil for a magnet, it is exposed to a

magnetic field when the coil is energised. For a practical superconductor in the mixed

state, the magnetic flux penetrating the superconductor is quantized. Usually, the

cunent I and the resultant magnetic field B are perpendicular to each other and a

driving force of a similar form to the Lorentz force acts on the fluxoids. When the

fluxoids are driven by this force, a macroscopic electric field E is induced as

E = B x v (3-4)

where v is the velocity of the fluxoids. Inhomogeneities, such as dislocations, grain

boundaries, voids and precipitates as flux pinning centres act against the Lorentz

force to stop the fluxoid motion. Hence the critical cunent density Jc is given by [18-

22]

75

JC = F P / B (3-5)

where the macroscopic force density of the pinning centres, Fp, is called the pinning

force density. Based on the two dimensional nature of the Bi-based materials, the

pinning potential, U0, can be expressed as follows [289]

U0 = kTc[ln(fo/f)(l-t)3/2]a'3/2B'1/2 (3-6)

where f0 and f are the flux creep velocities in the presence and absence of pinning

respectively, t is T / Tc, a is a numerical constant, and k is the Boltzmann constant.

Consequently the HTS critical cunent is related to its operation temperature and

applied magnetic field.

Samples of the Ag-clad (Bi,Pb)2Sr2Ca2Cu3Oio+x HTS wires were tested with

different temperatures and applied magnetic fields, and the results are shown in Figure

3-3. Figure 3-3 (a) and (b) represent the relations of critical cunent Ic - temperature T,

and critical cunent Ic - magnetic field B (PoH) - temperature T of the single core short

sample respectively. Figure 3-3 (c) shows the 27-filament sample critical cunents

changing with different operation temperatures, where the background residual

magnetic field is B = 5.5 mT. The 4.2 K critical cunent in Figure 3-3 (c) is a calculated

value which is estimated from the experimental values from a HTS coil made by this

wire, and the value is given by Impe>77K (W.2K I WTHC) = 168 A. Figure 3-3 (d) shows

the relation of Ic - B -T for the 27-filament sample. The Ic - B - T phase diagram sets the

limits for operation of the Ag-clad (Bi,Pb)2Sr2Ca2Cu3Oio+x HTS wire. The

76

experimental result shows when the operating temperature is reduced down to near

the liquid nitrogen triple point, 63.1 K, by reducing its vapour pressure, the Ag-clad

(Bi,Pb)2Sr2Ca2Cu3Oio+x wire critical cunent increases to Jc(64 K) ~ 2 x Jc(77 K).

c CD

t 13 O ro

o O

.

11-

10-

9-

8-

7-

6-

5-

4-

, ,_ ._,_ ._, ,_

O.

H(apply) = 0

i | i |

i — •''"

o.

I •

1

°-.

i

i • i • i — "

Bi-2223/Ag single core tape

-

'_

.

-

'•o.

"o-... '"•o -

-

1 1 1 1 > ' 1 1

62 64 66 74 76 78 68 70 72

Temperature (K)

(a) Critical current Ic of a short single core sample at different temperatures.

10

c a> 3

(0

o o

6-

4-

2-

0-

,o..

cK -a.

B Nab

— i —

0.0 1 0.1

Bi-2223/Ag single core tape

' — o B

-T~* 0.2

n 77.4 K

o 72 K

© 65 K

o

o

a

- i —

0.3 - 1 — 0.4

— I —

0.5 0.6

Applied magnetic field B=w,H (T)

(b) Ic magnetic field and temperature dependencies of a short single core tape.

77

200

< 150 -

c « 100

U

"a .y so

o

^ Bi-2223 27-filament Ag clad short tape

Calculated minimum value at 4.2 K

©

ooooo B II c M O - M B 11 ab A^A-A-A B Z c - 45°

"*>

1—I—I—I—r~ 0 20

i i i | r 40

~ 1 — I — i — i — i — | — i — r 60 80 100

Temperature (K)

(c) Ic of a short 27-filament sample at different applied fields and temperatures.

Bi-2223 Ag clad tape

ooooo 27-filament tape at 77.4 K O D O O D 27-filament tape at 65.8 K A-AA-JS A Single core tape at 77.4 K

100 200 300 400

Applied magnetic field p,0Y\ (mT)

500

(d) Ic magnetic field and temperature dependencies of a short 27-filament tape.

Figure 3-3. Ic with applied magnetic field and different operation temperatures.

78

3.2.2.3 Conductor magnetic field distribution

For high cunent applications, the self-field needs to be considered for the Ag-

clad (Bi,Pb)2Sr2Ca2Cu3Oio+x HTS wire which has anisotropy of the superconducting

properties and a high degree of texture. The self-field generation was studied with a

Ag-clad (Bi,Pb)2Sr2Ca2Cu3Oio+x HTS single core tape which had a final cross-section

- 0.18 x 4 mm , superconductor-core/wire cross-section ~ 1:6, and 6.85 A critical

transport cunent at 77 K.

Figure 3-4 shows the self-field distributions for a hypothetical tape with a

homogeneous Jc distribution at 77 K. X in Figure 3-4 represents the horizontal position

of the HTS tape, and its centre position is at X = 590500 um. The lines a, b and c in

Figure 3-4 (a) represent the co-ordinates to analyze the field distributions. The

maximum self-field generated along the Y direction, By (//c), is about 1 mT near the

tape edge with this configuration. The results show the self-field reaches its maximum

value at the tape sides. The result obtained has a good agreement with the experimental

result obtained by using a scanning Hall probe [290-292]. It is suggested that the self-

field generated by the HTS tape can be reduced by using two conductor strips which are

placed near and parallel to the HTS tape to increase its critical cunent [293].

(a) Lines a, b and c. (Figure 3-4)

79

Flux Density

B( 10"3 T)

(b) B map.

Flux Density

Bu ( 10"4 T)

9.02

7.21

5.40

3.59

1.78

-0.03

-1.84

-3.65

-5.46

-7.27

-9.08

(c) B y m ap.

Flux Density

By ( 10'3 T)

1.720

1.372

1.024

0.676

0.328

-0.020

-0.368

-0.716

-1.064

-1.412

-1.760

(d) Bxmap.

80

(e)

c CU

OJ

c CJ) CO

1.5x10-3-

1.0x10-3-

5.0x1 a* -

0.0-

5.0X10-4-

1.0X10-3-

1.5X10-3-

-2.0x10"3-

i - i 1 1 1

/

^\y **

\

t 1 1 1 1

1 1 1 • 1

? c c c « c ^ .

X y^\

1 -r

0 B

* Bx o By

b^

-

-

J? ******* -

^t^^tf^

Bi-2223/Ag single core tape at 77 K' - i ' 1 1 1 1 1

587000 588000 589000 590000 591000 592000 593000

X (nm)

(0

c TJ X 3

CD C O) CD

i.oxicr3-

8.0x10"* -

6.0X10"4-

4.0x10"4 -

2.0x10-4-

0.0-

-2.0x10"4-

-4.0x10"4-

-6.0X10-4-,

-8.0x10"4-

-1.0X10"3-

• i i | - i — , , ,

A / •*"*„. * " ' • / : * '

o o

Bi-2223/Ag sin

1 i ' i ' i ' i

• | • -i — p —

O B

• Bx

- By

gle core tap

1 i

-

e at 77 K

I I

587000 588000 589000 590000 591000 592000 593000

X (nm)

(g)

2.0x10"3

1.5x10"3

fc 1.0x10'3-

>. « 5.0x10"" c tu TJ X 3 o -5.0x10"1

S> -i.oxio3

-1.5x10'3-

-2.0x10

0.0

T ' r

B—-B—B-?-n-

/ 'D-n-i B-n-n-o-a

A-A-A-A-A-A-A-A-A-5~A—A-A-A—A-A-A-A-A

— D — B

—*—Bx —A—By

\ !_•-*-

Bi-2223/Ag single core tape at 77 K

1 1 1 1 1 1 , 1 1 586000 58S500 587000 587500 588000

Y (urn)

(a) Line a, b and c; (b) B map; (c) By map; (d) Bx map;

(e) X = Line a; (f) X = Line b; (g) X = Line c.

Figure 3-4. Self-field distributions of a HTS tape.

81

3.2.2.4 Critical current for A C operation

AC loss of the HTS has been studied related to the design of a practicable

conductor [294-303]. The AC power loss of a superconductor placed in a transverse

magnetic field consists of a hysteretic loss Ph and an eddy cunent loss Pe. At low

frequencies of the applied magnetic field, Ph and Pe of the HTS composite strand, in

c.g.s. practical units, are given by [285]

Ph = (8xl0"8/3rc) H4Hmf W/cm

3 (3-7)

Pe = (10-7/2)(Lp

2/piiicm)Hm2f2 W/cm3 (3-8)

where Hm is the amplitude of a sinusoidal field of frequency f, d is the filament

diameter, and X is the filling factor, Lp is the twist pitch and pmcm is the transverse

resistivity of the strand in nQcm. The formula for AC loss under AC self-field

conduction can be derived from Maxwell's equations [304], and the loss is caused by

the periodic penetration of the magnetic field into the conductor in a

thermodynamically ineversible manner. The resulting loss is given by (in MKS units)

Q = 2AfIcJc [i(2-i) + 2(l-i)ln(l-i)]xl0"7 (W/m) (3-9)

where A is the conductor cross sectional area, f is the frequency and i is the ratio of

operating cunent density Jm to critical cunent density Jc.

Practically the HTS wire and coil are normally required to work in an AC

circuit for electrical applications, and the AC transport cunent characteristic of this

82

Ag-clad (Bi,Pb)2Sr2Ca2Cu3Oio+x wire is important for such applications. A short

sample was selected from the 27-filament HTS wire. It had about 59 mm V-V

terminal interval, and was tested at 77 K with a lock-in amplifier and AC cunent

supplied by a 50 Hz AC power source. The resistive signal real part x' is separated

from the lock-in signal %' + i%" and the result is shown in Figure 3-5 (a), in which the

original DC critical cunent of this short sample is 17.4 A with 1 uV criterion. The I-

V curve of the HTS short sample at room temperature is also plotted. Compared with

the room temperature Ag resistance, the resistance of the HTS sample is dramatically

reduced in its superconducting state at 77 K. To calibrate the lock-in measuring

system used, a pure Ag sheath, having roughly the same size as the HTS sample, is

also measured, and the result is shown in Figure 3-5 (b).

At 77 K, the resistive AC loss of the HTS tape is AV(%') / AIAC(rms) « 0.52

pV/A at 5 A (50 Hz, rms) and about 12.72 pV/A at IAc(rms) ^ 20 A which is the

value for the normal state at 77 K. From the measurement with low frequency AC

transport cunent, the AC critical cunent rms value is close to its DC critical cunent.

Consequently for a HTS conductor carrying both DC and AC transport cunents, its

rated cunent can be evaluated by iAc(rms) + IDC ^ Ic, which has a good agreement

with the previous superimposed AC measurement [305]. The HTS tape used as a

conductor for 5.0 A (rms) and 25 A (rms) 50 Hz operation in liquid nitrogen has AC

losses of 4.4 x 10"8 J/cycle-cm and 9.3 x IO"6 J/cycle-cm respectively.

83

2000

c ri

o >

1500-

1000-

500-

0-

A

A

A

A

• H I i • • « 1

B-2223 27-filament Ag clad tape

AC current character (Real part •/')

10

77 K

Room temp.

20 I 30

(Critical DC current Ic=17,4 A) i

40 I 50

A C current (rms, 50 Hz) (A)

(a) A C I-V curve of a short sample with 50 H z transport current.

4000

3000-

^ 2000

0 Oi (D ^ 1000 >

•

*

ir-r""^.

i • i

Ag tape at 77 K

.*•-

^ ^,~-'

i • i

• i

/ " " /

» 1

!

/ - -

1 '

/ ' "

X'

X"

1

y

•

-

^

0 10 20 30 40

Current (50 Hz) (rms) (A)

50

(b) Calibration of the A C measuring system with a pure A g sheath

and 50 Hz transport current.

Figure 3-5. AC transport current measurement.

84

3.2.2.5 Mechanical properties

To form a HTS coil, the HTS wire is subjected to bending stress as it is

wound into the coil and to uniaxial stress from pretensioning of the wire during

winding. Besides this fabrication stress, other sources of stress include thermal-

contraction stress and magnetic stress. When the HTS wire is cooled from room

temperature to liquid-nitrogen temperature, for example, different materials within the

composite HTS wire and its structure contract at different rates. Because of this

mismatch in thermal contraction rates, significant stress on the superconducting

material can be generated during cooldown. When a superconducting winding is

energised, the Lorentz force acting on the superconductor can be quite large,

particularly in large magnets. One important consequence of stress is the HTS critical

cunent degradation.

The mechanical property of this HTS wire is very important both to form a

coil with bending under the react and wind procedure, and to be a conductor carrying

high cunent under high magnetic field with electromagnetic force stress. The

mechanical bending property has been tested with the single core sample and also with

the 27-filament sample [207]. The mechanical property of the HTS sample was tested

with respect to its bend strain eb = (1+2M)'1 and tensile strain et = (L-LD)/L0, where R

is the radius of the bending curvature, t is the thickness of sample, L0 is the original

length of sample in the tensile test, and L is the length of sample after being extended

by the applied uniaxial tensile force. The mechanical strength a therefore is

a = eb E = (l+2R/t)"1 E * (t/2R)E (3-10)

85

where E is the elastic-modulus.

For the magnetic stress, the axial magnetic hoop stress Gf/ along each wire in a

solenoidal coil is given by [18]

C// = JBR (3-11)

where J is the transport cunent density, B the magnetic field strength, and R the

radius of the winding. Meanwhile the transverse stress a± with the conductor

thickness t is

G± = JBt (3-12)

which is in the direction perpendicular to both the cunent and magnetic field, or the

radial direction for a solenoid. The normalized Jc versus tensile strain curve for a

sample of Ag-clad (Bi,Pb)2Sr2Ca2Cu30io+x wire with a total thickness of 0.2 mm has.

shown that the Jc remains unchanged up to a strain of 0.15% [207]. The Ampere force

dF of a conductor dl carrying a cunent I under magnetic field B, is given by

dF = IdlxB (3-13)

The F is in the radial direction and its value is estimated byAF = AlxIxB(lN=l

TAm).

From Equation 3-11, if J = 10s A/m2, B = 10 T and R = 0.1 m, the o„ is about

100 MPa, which is less than 150 MPa as estimated [278]. If B = 10 T and I = 100 A,

86

the AF on a part of the wire, Al = 0.01 m, is about 10 N. At room temperature, the

yield strength of a similar Bi-2223/Ag HTS conductor is about 20 N [277].

Table 3-1 is a brief summary of the testing results. The bend strain e used is

defined as s = t/2R, where t is the HTS tape thickness, and 2R is the diameter of the

former on which the HTS tapes were bent. A flexible bending curvature (1/R) up to

0.24 cm"1 can be used in the react and wind procedure to form a HTS coil with the

single core HTS wire. Within this curvature, the critical transport cunent maintains

90% of the original value. The multifilament HTS wire has a better mechanical flexibility

as shown in Table 3-1.

Table 3-1. Mechanical properties of the Ag-clad (Bi,Pb)2Sr2Ca2Cu3Oi(RX tapes.

Winding tolerance

Bend strain e = t/2R

Radius curvature 1/R

H T S Bi-2223/Ag Tape

Single-core sample

~ Jc90% at e = 0.2%

> Jc90% within 0.24 cm"1

in curvature

27-core sample

~ Jc90% at e = 1 %

> Jc90% within 0.9 cm"1

in curvature

87

2000 —

0

2 4 6

DC current (A

Figure 3-6. I-V characteristics in mechanical crack-conditions.

88

During the react and winding process to form a coil, cracks in the H T S core of

the HTS wires are difficult to avoid. However, since liquid nitrogen is cheap, the HTS

wire still works in a cracked condition as a low loss conductor and therefore can be

considered in a practical application. A 50 mm short tape of the single core HTS wire

with a 40 mm voltage-terminal interval cut from a coil was tested with a series of

severe bendings applied. Figure 3-6 shows different I-V curves at each stage of the

bending test. Each step in the figure represents a 90° bend on a sharp 90° metal block

and subsequently straightening at one point of the tape. The next step of bending and

straightening occurs at a different point. Finally the cracked ceramic superconducting

core was taken out from the Ag sheath. This result shows that even the single core

tape with a number of cracked sections can still produce a low loss coil since a large

area of the interface of the Ag-sheath / HTS-core is still intact after bending. The

critical cunent density for transmission through the HTS grain boundary Jgb is lower than

the intragranular Jc and decreases much more rapidly with field than Jc [306], consequently

the cracks have to be avoided.

3.2.3 Evaluation of the Use of HTS Wire

3.23.1 Summary of the Bi-2223/Ag HTS wire

The technology of using the Ag-clad (Bi,Pb)2Sr2Ca2Cu30io+x wire, with regard

to its critical cunent, long length, and possibility to form a coil, is summarised with

the results of Figure 3-7, which have been achieved by the University of Wollongong.

Figure 3-7 (a) shows a typical Ag-clad (B^Pb^S^CaaCusOio+x wire produced at the

89

University of Wollongong (by a research team, Y.C. Guo and J.N. Li et al), which

demonstrates that high critical cunent, long length and uniformity can be achieved

with this HTS wire produced through the normal powder-in-tube techniques. Figure

3-7 (b) shows the capability of the 27-core HTS wire to form a HTS coil. The critical

cunent degradation during forming a coil is mainly caused by the increased self field

and the mechanical bending procedure.

The Ag-clad (Bi,Pb)2Sr2Ca2Cu30io+x wire can be joined to increase the wire

length, and the joining method is illustrated in Figure 3-7 (c). For a single core to

single core joining, 80% Jc of the original value of the tapes can be retained [305,

307]. The potential critical cunents of a Ag-clad (Bi,Pb)2Sr2Ca2Cu30io+x sample is

shown in Figure 3-7 (d) with regard to its magnetisation critical cunents for different

operation temperatures and magnetic fields [275]. Figure 3-7 (e) shows the

ineversible line which was obtained from the measurements on the sample AC

susceptibility. The AC magnetic susceptibility (% = %' + ft") was measured in a

constant DC field H as a function of temperature. The temperature at which the

component of x" reaches its peak, is identified as TP(H) and gives a point on the Tirr -

HOT line as shown in Figure 3-7 (e).

The characteristics and testing results both of the single core HTS wire and

the 27-filament HTS wire are listed on Table 3-2. With small scale laboratory

facilities, the length of the HTS wire can easily achieve a few tens of meters, and the

properties are reproducible.

As introduced in Chapter 2 the Ag-clad (Bi,Pb)2Sr2Ca2Cu3Oio+x wire has

reached high critical cunent density and long length. The greatest difficulty in using

this HTS wire is that the material flux pinning strength at 77 K is not strong enough;

90

consequently it can not be used at liquid nitrogen temperature under high magnetic

fields. Practical application requires improving the material's intrinsic property, and at

present one solution for the application under high magnetic fields is to reduce the

operation temperature.

Table 3-2. High Tc superconductors.

H T S 27-Filament Tape

HTS: (Bi,Pb)2Sr2Ca2Cu30io+x Ag-

clad tape

CStape* = 4.15 m m x 0.22 m m

CSHre**/CStape = 0.21

Normal short tape:

Ic = 21 A at 77 K in 5.5 m T applied field

Ic = 33 A at 66 K in 5.5 m T applied field

At 77 K and B//ab:

Ic (0.15 T) / Ic (0 T) « 4 3 %

Ic (0.5 T) / Ic (0 T) « 2 7 %

Winding tolerance:

> Jc90% within 0.9 cm"1 curvature,

or at 1 % bend strain***

H T S Single Core Tape

HTS: (Bi,Pb)2Sr2Ca2Cu30io+x Ag-clad

tape

CSupe= 3.8 mm x 0.4 mm

CSnTs/CStape = 0.17

Normal short tape:

Ic = 5.25 A at 77 K in 5.5 m T applied field

Ic = 9.4 A at 66.2 K in 5.5 m T applied field

At 77 K and B//ab:

Ic (0.01 T) / Ic (0 T) * 8 0 %

Ic (0.15 T) / Ic (0 T) * 28 /o

Winding tolerance:

> Jc90% within 0.24 cm"1 in curvature

*CStape is the cross-section of the H T S tape.

**CSHTS is the cross-section of the H T S core.

***The critical cunent density is determined by an average value of five H T S tapes at

77 K. The bend strain is defined by t/2R, where t is the H T S tape thickness and 2R is

the diameter of the bend former.

91

30

~ 20 < o

O

o

5 °

-10

.

290

5- 200

I '50 n

- § 100

SO

0

I 1

n

I-V curve of 10 m wire at 77 K

o

,o

) 5 10 15 20 25

Current (A) 1 , 1

I 1 1

lc Uniformity of 9603w Long Tape

(after 2nd sintering)

D a u

0

/ 'J

1 0 1 4

.

°*. * ' • . HlU *

J °' • O

0.01

30 35 »""

1

°0 o o o

Mil. „ 0

0

3 200 400 600 eoo K

H(mT) 1 , 1

o.t

0.01 -

0 001 oo

4 6 Length of Tape (m)

10

(a) Performance of a laboratory produced Ag-clad (Bi,Pb)2Sr2Ca2Cu30i<n.x wire.

60

5 0 -

4 0 -

c

k_

ZJ u o u

30

20

u 10 —

i—i—r

Bi-2223 27-core Ag clad HTS wire

O Q Q Q O Coil at 77 K • oDoo Coil at 4.2 K ** a a A Wire (220 h) at 77 K o-oooo Wire (300 h) at 77 K

~i—i i — r 10 20

i — i — | — i — i r 30

HTS wire length (m)

1—I r 40 50

(b) Solenoid coil formed from a 27-core Ag-clad (Bi,Pb)2Sr2Ca2Cu3Oi,HX wire

under react and wind procedure.

92

Ag sheath

HTS core i >

Bi-2223 wire

V Press

(c) A joining scheme to increase the H T S wire length.

10'

1 0° I • • • • L. • i i i i i _

2 3

H (T)

(d) Magnetisation critical current at different temperature [275].

10

b 4 -

2 -

0 -27-filaments

cs. = 0.2mm x 3 m m

20 40

-> r

Bi-2223/Ag multifilament tape

60

TirT W

80 100 120

(e) Irreversible line of a Bi-2223/Ag tape.

ure 3-7. Properties of the Ag-clad (Bi,Pb)2Sr2Ca2Cu30i<»x HTS wires.

93

Some important values of the Ag-clad (Bi,Pb)2Sr2Ca2Cu3Oio+x H T S wire,

achieved recently by U S A research groups [308], are summarised in Table 3-3. These

are significant achievements which show the capability of the H T S wire.

Table 3-3. S u m m a r y of transport current properties of short and long

monofilament and multifilament Ag-clad (Bi,Pb)2Sr2Ca2Cu30io+x conductors

[308],

Conductor

Monofilament Short Pressed Short Rolled Long Length Long length

Multifilament Long length Long length Long length Long length

Length (m)

0.003 0.03 70 114

20 90 850 1,260

Ic (A)

51 51 23 20

42 35 16 18

Core Jc (A/cm2)

45,000 29,000 15,000 12,000

21,000 17,500 10,500 12,000

Overall Jc (A/cm2)

9,000 7,800 3,500 3,200

6,800 5,600 2,500 3,500

Fill factor %

20 27 24 27

32 32 24 30

3.2.3.2 Application judgment

The application of the Ag-clad (Bi,Pb)2Sr2Ca2Cu30io+x H T S wire requires both

technical and commercial judgments. Very expensive pure silver is cunently required

to produce the Ag-clad (Bi,Pb)2Sr2Ca2Cu3Oio+x wire as a metal sheath. The silver

tube used is cunently about A $ 500-600 / kg. Comparing with other chemical metallic

oxide compounds used to produce the H T S materials, the silver material represents the

most percentage of the expense for the raw materials used to produce the Ag-clad

(Bi,Pb)2Sr2Ca2Cu30,o+x wire. Fortunately the silver used could be recyclable. The

94

amount of silver required can be estimated as follows: a A g tube of 1 m length with <fon

= 4.2 mm, (jw = 6.48 mm, after a few cycles of drawing and pressing, finally becomes

about 28 m in length with cross-section 2.85 x 0.35 mm2.

Commercially useful superconductors in general need to be drawn into fine wires

and wound into a solenoid configuration. The superconducting wires typically require

stabilization using a normal metal cladding which serves several important functions in

such a composite structure: (a) a parallel electrical conduction path to carry cunent in

case of local loss of superconductivity, (b) thermal stabilization (heat sink) to minimize

local heating and prevent catastrophic loss of superconductivity, arid (c) mechanical

protection of the generally brittle superconducting core against the Lorentz force during

operation, as well as many other unavoidable stresses occurring during fabrication,

handling, or use of the superconductor wire, (d) in addition, the normal metal shell

provides environmental protection from the air, which is especially important for the

materials like the YBa2Cu3(>7 superconductor, as it is known to be sensitive to humidity

and carbon dioxide gas. The choice of proper cladding metal has been found to be

critical in fabrication of the composite superconductor wire. Ag and Au have been

found to be compatible to the compound at high temperature. Ag is relatively

inexpensive, compared to gold, and could be an economically acceptable material, as a

thin-walled tube or thin layer coating of normal metal in the composite wire fabrication,

unless some other metal or alloy becomes available.

Related to the HTS material properties, 77 K operation temperature does not

allow a high critical cunent to be maintained under an applied magnetic field over 1 T,

but at 4.2 K (i.e. liquid helium temperature) it has superior electrical transport

properties than the conventional low temperature superconductors. Since the

95

cryogenic cost will dramatically reduce if the operation temperature can be shifted

from 4.2 K up to 77 K, improvement of the HTS material with respect to its critical

cunent at 77 K is essential for large scale electrical applications.

The mechanical property of the HTS wire is another problem related to its

application. Cracks are introduced by the mechanical bending procedure. The studies of

critical cunent density Jc in high Tc oxide superconductors revealed that Jc for a

polycrystalline specimen is much lower than the intragranular Jc [306]. The critical

cunent density for transmission through the boundary Jgb is lower and decreases much

more rapidly with field than the intragranular Jc. The wind and react procedure can

avoid cracks produced by the final winding of the HTS wire to form a coil.

To achieve practical applications of the Ag-clad (Bi,Pb)2Sr2Ca2Cu3Oio+x HTS

wire in the near future, the present material-processing-defined HTS configuration

which aims to achieve a high Jc, requires to be considered to convert to application-

defined HTS conductors with practical conductor criteria.

3.3 HIGH Tc SUPERCONDUCTING COILS

3.3.1 Introduction

High Tc superconducting (HTS) Ag-clad (Bi,Pb)2Sr2Ca2Cu3O10+x wire has

been produced with high critical transport cunent density and flexible mechanical

property. It has advantages compared with conventional low temperature

superconductors, with respect to energy saving, low operation cost and high

96

efficiency. A length of this wire, over 1.2 km, has been achieved, with 1.2-1.8 x IO4

A/cm2 critical cunent density and 0.18 x IO2 A critical cunent at 77 K [282, 283].

Cunently the Ag-clad (Bi,Pb)2Sr2Ca2Cu3O10+x HTS wire is one of the most

promising candidates for magnets and electrical applications. In most of these

applications, a winding form of the HTS wire is required.

Ag-clad (Bi,Pb)2Sr2Ca2Cu3O10+x HTS wires have been investigated in very

high field applications with regard to their critical cunent capability in the presence

of a magnetic field. At 4.2 K, this HTS wire can maintain about 105 A/cm2 critical

cunent density under a 23 T magnetic field [193]. As a hybrid magnet coil, a HTS

coil of this superconducting wire has been made to generate 1 T in a 20 T

background field and 0.37 T in a 23 T background field at 4.2 K [193, 194].

However, due to the flux pinning property of this HTS material, this HTS wire

cunently is not suitable for applications where it is necessary to generate a high

magnetic field or to operate in the presence of a high magnetic field at 77 K liquid

nitrogen temperature. For most electrical applications, the HTS wire is not usually in

a very high field situation like a magnet, but used to replace the normal copper

winding in electrical devices. The generated magnetic field of the HTS winding is

related to its value of critical ampere-turns and is normally much less than that of a

magnet.

By using the Ag-clad (Bi,Pb)2Sr2Ca2Cu30io+x wires, HTS coils are developed

by using different methods and techniques. The HTS coils fabricated include pancake

coil, solenoid coil, double pancake coil, which are made by using both single core

Ag-clad (Bi,Pb)2Sr2Ca2Cu30io+x wire and multifilament Ag-clad (Bi,Pb)2Sr2Ca2Cu3

Oio+x wire, and by using the react and wind, and the wind and react procedures.

Experiments are canied out to study the HTS coil properties, and include magnetic

97

field measurements of the H T S coil, D C magnetic field, A C magnetic field, and the

value of critical cunent ampere-turns when used as a magnetic core bias winding.

The HTS coil magnetic field distribution is also evaluated, with regard to using the

HTS wire, by using a computer analysis program. The experimental results and

analysis provide direct evidence for the design and operation of a practical HTS coil

and winding made from the anisotropic Ag-clad (Bi,Pb)2Sr2Ca2Cu3O10+x HTS wire,

with regard to its magnetic field, both magnitude and direction, dependent critical

cunents.

3.3.2 Making Superconducting Coils with Bi-2223/Ag HTS Wire

3.3.2.1 Techniques

HTS coils can be produced both by the react and wind, and the wind and react

procedures. The HTS coils made include solenoid coils and pancake coils. Increased

self-field and mechanical strain or cracks of the HTS core lead to the critical cunent

reductions when forming a coil under the react and wind procedure.

The capability of the Ag-clad (Bi,Pb)2Sr2Ca2Cu3Oio+x HTS single core wire

to form a coil is identified. The critical cunent is reduced from its short sample to the

final solenoid coil. Figure 3-8 (a) shows critical cunent changes during the making of

a coil through the react and wind procedure. The sintered long wire has been

transfened from an initial A1203 long tube with diameter d> = 75 mm into a solenoid

coil with final diameter <fcd = 57 mm. The critical cunent reduction is related to the

winding procedure used and its increased self magnetic field in a coil format.

98

Figure 3-8 (b) shows the critical cunent of a double layer pancake coil made

by a (Bi,Pb)2Sr2Ca2Cu3Oio+x 30-filament HTS wire. Varnish, Teflon tape and a

Urethane seal coat is used for insulation of the coils.

Another HTS coil (HC-22CWRS in Table 3-4) has been made through the

wind and react procedure, and after final winding and sintering, it maintains nearly

100% of the previous Jc, and the test on the coil also indicates that its field is

symmetrical along the length of the coil. KAOWOOL paper, which has composition

of A1203 (47%) + Si02 (53%), and melting point of 1760 °C, was also used for

making this coil. The final coil is varnished for insulation purposes.

3.3.2.2 Produced HTS coils

Table 3-4 summarises the HTS coils developed in this research work.

Photographs of the HTS coils produced are shown in Figure 3-9.

99

> E

D) CO — O >

7 5 -

^. 50

cu cn D o 25

BBBBQ Bi AAAAA Bi oeee©Bi

-2223 long wire -2223 coil -2223 short tape

0 f M A 1^1 M I | lD[Tl | Pi I 0.0 0.3 0.5 0.8

DC—current/lc

1~TT

(a) A solenoid coil made by a monofilament Ag-clad

(Bi,Pb)2Sr2Ca2Cu30i(Hx wire under react and wind procedure.

40

30-

20

10-

0-

O

/ o

Double-layer pancake coil / o /

/

o

•

•

•

o ,0' ,o

o

anirirrici2QQQ8DDDDnDQnanDnD azP1

D u Bi-2223/Ag 30-filament tape

-1 ->—r

0

n — ' — i ' i • i

2 3 4 5

DC current (A)

T" 6 8

(b) A double pancake coil made by a 30-muItifilament Ag-clad

(Bi,Pb)2Sr2Ca2Cu30io+x wire under react and wind procedure.

Figure 3-8. Critical currents of the Ag-clad (Bi,Pb)2Sr2Ca2Cu30i(>+x H T S wires

during forming a coil.

100

C/3 Qi

s-%

H E +

d n 3

u «

u <s u c«

a* — •

ao < JS

cu T3 C5

E o u

H as 0)

JS

H • I

OS

H

w r-c

<

H?

^

s v_/

c o •-* C <u

B • »-H

Q

• P H

O CJ (4-1

o CJ CL, >, E-i

bO c • —

g ,o <4-l

1—4

"3 U

s

6, .a bO

c JU

CJ

CO

L PC

o o CO

PC

> v

s

,2 'C 9J 'C u

E > 3.

/-—s.

g. CA

E 3 TJ C

CJ

TJ CJ O

o I—,

Cu,

|

ZZ X

B CO

U

CO

vd

8 vo "c 2

CO „

C N r~

'L M • #

TJ • -H O c CJ o CA

TJ a • H

£ TJ C ca •4—1 CJ OS

a

CN

3 CN CN CO CN

a o CJ t -CN

PH CO

U CN 1

U PC

cs cs

Si" o V S2 4 A cn _.-II VO

% II -e-

TJ

'3 c CJ o CA

4-*

o 03 H TJ C S3 TJ C

I

X 00

o o cs

o CJ 1

cs cs

CO oi

u cs cs 1

u PC

cs

£}" <^ in ^t II II •- y

-e- ~ ^t CJ

oo >,

'L i2 3 1

-J? CN

•e- ^

<4 •a o c cd Cu,

CJ «—4

XJ 3 O TJ

TJ C

"1 TJ C OS •4-4

o a

co X cs CO

ON

a o o 1

o cn

cu

1 U o cn 1

U PC

^t in

^ ^ VO f~

£ II

-e1 2 oo" o" vo 00

'L II = l-J •e-

TJ •*H

o c J1J

"3 CA

TJ C

• P H

£ TJ C 03 •4-4

CJ

CJ 41

Tfr

00

o ^ —H

a o CJ 1

__CJ "bO

.s CO

CO

oi U CO

U PC

cs in

© cn II -e1

in vn II. a o -e-

TJ • •4-H

O C CJ

"3 CA

TJ C •*H

£ T3 a 03 •4—1

o 03 CJ

CS

X cs cs p VO TJ-

a o CJ 1

f-cs

s CO

1 u cs

i U PC

VO o cn II

z in VO II

^^

'••>.-. ass

M

kj •k. 1 ;' ^fc.

\» •' Y-

' ' S*. v \

\^v. »

1 «Bfe«fco*

1. V

11 L '

X V % • ' i x

, x •

SPK&EII

III I '"'^^j B 1

«Siw ^c^

• <4>. .••*.<

*

'"ii

*

saip

-^g?^

mm

(a) HC-22CWRS HTS coil.

(b) HC-30CRWP HTS coil.

Figure 3-9. The produced HTS coils.

102

3.3.3 Experimental Results on the H T S Coil

3.3.3.1 The selected HTS coil

The magnetic field properties of the HTS coils are studied in this section. The

Ag-clad (Bi,Pb)2Sr2Ca2Cu301o+x 27-core multifilament HTS wire and its associated

coil HC-27CRWSM, as shown in Table 3-3, have achieved promising results and

therefore are selected as samples to carry out magnetic field measurements and

analysis.

Using the wire described in the previous section, a HTS coil was wound

under the react and wind procedure, and insulated with a Teflon tape. The initial HTS

wire had a fairly uniform critical current of about 10 A along the length with a

variation of -10% determined by multiprobe measurements. The properties of the

HTS coil are summarised in Table 3-4 where the 1 pV/cm criterion is used.

The critical current of the HTS solenoid coil made from the 27-filament HTS

wire was measured to be ^ = 5.2 A at 77K, and Ic = 45.6 A at 4.2 K without external

magnetic field applied (measured by CSIRO [317]). Consequently, the value of

critical ampere-turns of the HTS coil are 1591 A-t at 77 K and 13954 A-t at 4.2 K.

3.3.3.2 DC magnetic field

With regard to the HTS coil magnetic field properties for operation at 77 K,

experiments were carried out on the HTS coil DC magnetic field. The critical current

and generated DC magnetic field of the HTS coil operated in liquid nitrogen were

measured, and are shown in Figure 3-10.

103

The magnetic field generated at 77 K by the H T S coil with 1 pV/cm criterion

is about 0.025 T at the HTS coil centre. A magnetic flux meter was used to detect the

generated magnetic field. From Figure 3-10 the voltage drop across the HTS coil

becomes linear when its transport current is larger than about 17 A. From the I-V

curve in Figure 3-10, the Ic between 5.2 A (1 pV/cm, criterion) and 17 A (linear

increasing point) is a low loss range for operation of the HTS coil. When I > 17 A,

the Ag sheath becomes the main current path and results in a linearly increasing

voltage drop.

3.3.3.3 AC magnetic field

The AC magnetic field of the HTS coil was tested at 77 K and with 50 Hz

transport current, and the results are shown in Figure 3-11 where the magnetic field

in rms value is indicated. The 50 Hz AC transport current property of a 59 mm short

piece sample is also shown in Figure 3-11; this sample has DC critical current L, =

17.4 A with 1 pV criterion. The voltage in the figure is the resistive voltage which is

the real part x'of the voltage signal x' + i%" measured with a lock-in amplifier. The

generated magnetic field is about 0.023 T with 5.2 A (rms) AC transport current.

104

0.20

DC current (A)

Figure 3-10. Generated D C magnetic field and voltage drop for the H T S coil.

105

0.10

0.08-

0.06-

0.04-

0.02-

0.00

o Generated A C field A A C voltage (x') of short sample 0

• * Q

240

200

160

120

80

- 40

—r-

10

A A 1 1 1 r

Bi-2223/Ag HTS coil

5 10 15 20

AC current (A)(rms)(50Hz)

25

Figure 3-11. Generated A C magnetic field for the H T S coil

and short sample resistive loss at 77 K.

106

3.3.3.4 H T S coil used as a magnetic core bias winding and its ampere-turns

Operation with a magnetic core is a common requirement for electrical

applications of coils. The HTS coil is studied as a magnetic core bias winding with

regard to its value of critical current ampere-turns and magnetic flux distributions.

When the HTS coil is added to a magnetic core, the magnetic flux in the HTS coil

will be re-distributed. To investigate the magnetic core influence on the HTS coil, the

HTS coil is tested, with respect to its I-V characteristics, for different magnetic core

gaps and without the magnetic core. The magnetic flux of the HTS coil as re­

distributed by the magnetic core used, will change the I-V curves of the HTS coil.

A HTS coil which consists of five pancake units made by the Ag-clad

(Bi,Pb)2Sr2Ca2Cu3Oio+x single core HTS wire, was used as a bias coil of a magnetic

core, and its critical current was measured at 77 K with varying magnetic core gaps.

The I-V curves of the HTS coil in different conditions, without the magnetic core,

and with the magnetic core but without air gaps and with a 1.8 mm air gap, are

shown in Figure 3-12 (a). In Figure 3-12 (a) a magnetic ferrite core is used which has

cross-section (CS) = 25 x 25 mm2, magnetic path length (MPL) = 135 x 2 + 75 x 2

mm, and a saturable flux density of about 1.3 T. Figure 3-12 (a) shows that the HTS

coil critical current is affected by the magnetic core used and the air gap of the

magnetic core.

The magnetic field distribution is also related to the saturation of the

magnetic core used. For a closed magnetic core, the current F required to saturate the

magnetic core is related to the magnetic circuit by

107

I' = [MPL x (B/p0p)] / Ncoii < Ic (3-14)

where p is the magnetic core permeability.

To further investigate a HTS coil used as a magnetic core bias winding, the

HTS coil, HC-27CRWSM, was selected. Figure 3-12 (b) shows the generated

magnetic flux density, in the pair of ferrite magnetic C-cores used, with the solenoid

HTS coil as a bias winding and for different values of ampere-turns. The magnetic

circuit used has cross section CS = 13 x 11 mm2, and magnetic path length lm = 140

x 2 + 70 x 2 mm without air gaps. A magnetic flux search coil is used to measure the

magnetic core flux density. The HTS coil DC resistive voltage drops with the

magnetic core having different core gaps and with an air core are measured. Figure 3-

12 (b) shows that the value of critical ampere-tums of this HTS coil is capable of

saturating the magnetic core within its superconducting range. From this figure, the

value of critical current ampere-tums of this HTS coil is capable of saturating the

magnetic C-core with up to 1.3 mm double magnetic core gaps.

Figure 3-12 (c) shows I-V curves for the HTS coil with an air core, two

magnetic C-cores with different core gaps, and an iron bar core to present the

magnetic core effect on the HTS coil made by the 27-filament wire. The two C-cores

are the same size, and the iron bar is 16 mm in diameter and 150 mm in length. From

Figure 3-12 (c), it is indicated that there is a slight critical current change of the HTS

coil with a closed magnetic core and an air core. Figure 3-12 (d) shows the testing

arrangement of the HTS coil used with the selected magnetic core.

108

300-

?00-

100-

0-

i • i • i - i 1 1

A single-core HTS coil on a magnetic core at 77 K ° /o

// o — •-• Magnetic core with 1.8 m m gap // y' — O — Magnetic core without gap >'/' y — O—Without magnetic core . Q y' - - -A -Single unit on magnetic core without gap ,',' //'

ss' y

_ t f . e ^ ; : ; . . A " "

I i . | i | i i -. - , -

0

Current (A)

(a) I-V curves of the 5-unit pancake H T S coil with different magnetic cores.

2.0

1.5 -

-4->

'co c CD

•o X =3

1.0 -

o '-M

C

o

0.5 -

0.0

Magnetic core flux with HTS coil

0

ooooo No gap a^aAA 0.25 m m gap a-ooao 1,3 m m gap /

i — i — i — i — | — r r ^ i i i r 1000 2000 3000

160

— 120

•200

> E

o u

— 80 o

— 40

CD

cn a O >

DC ampere-turns (A-t)

(b) Magnetic core flux density generated by the HTS coil

made by the 27-filament wire with different magnetic core gaps.

109

>

100

75

O

75 >

50-

25

0

HTS coil I-V with differen

0

ooooo Air core • •ono C-core &*&&& C-core 00000 C-core ***** Iron bar

DC current

(c) I-V curves of the H T S coil made by the 27-filament wire

at 77 K and with different media.

(d) The H T S coil ( H C - 2 7 C R W S M ) with a magnetic core.

Figure 3-12. The HTS coil used as a magnetic core bias winding.

110

3.3.4 HTS COIL MAGNETIC FIELD DISTRIBUTION ANALYSIS

3.3.4.1 Analysis of the HTS coil magnetic field

The well-known Biot-Savart law describes the magnetic field distribution of a

conductor carrying current, .and the generated magnetic field in space by

B(w)=^nji^ (3.15)

where j is the vector of current density of the coil, r is the vector of distance variable,

V is the volume of the coil, and X is the filling factor. Consequently the magnetic field

distribution of the HTS coil can be calculated.

The magnetic field distributions of the HTS coil, HC-27CRWSM, are analysed

with an air core and a magnetic core for 77 K .and 4.2 K operation. The well-known

Quickfield program is employed for the analysis. The scheme used for the analysis of

the HTS coil magnetic field is shown in Figure 3-13, where the lines A, B, C and D

represent the position used as co-ordinates of the magnetic fields.

3.3.4.2 Operated at 77 K with air core

The magnetic field distribution around the HTS coil was then analysed for 77

K operation with maximum critical transport current of 5.2 A and along four typical

lines as shown in Figure 3-13.

Figure 3-14 (a) shows the magnetic field B, Bx and By values of the HTS coil

along the coil axis marked as line A in Figure 3-13. In Figure 3-14 (a), Bx and By are

111

the X (HTS coil axial direction) and Y (HTS coil radial direction) components of the

generated magnetic field by the HTS coil.

Figure 3-14 (b) shows the magnetic field values B, Bx and By along the radius

of the HTS coil centre marked as line B. Figure 3-14 (c) shows the magnetic field

values B, Bx and By along the HTS coil inner wall and parallel with the axis and

marked as line C. Figure 3-14 (d) shows the magnetic field values B, Bx and By along

the HTS coil side and parallel with the coil radius and marked as line D.

In Figure 3-14 (a) and (c), the broad lines on the L axis represent the HTS coil

position and length. In Figure 3-14 (b) and (d), the broad lines on the L axis represent

the HTS coil position and diameter.

The magnetic field distribution maps are also presented by Figure 3-15 (a)-(f),

which show (a) Magnetic flux density B (T); (b) Magnetic flux density Bx (T); (c)

Magnetic flux density By (T); (d) Magnetic field H (A/m); (e) Magnetic field potential

A (VVb/m); and (f) Magnetic energy density W (J/m3).

112

Figure 3-13. The schematic used for magnetic field distribution analysis.

113

0.020-

£ 0.015-m >»

'1 0.010-

2 0.005-u_

0.000-

.... , , , , . , ,

<

—o—B

-+-B(x) -x-B(y) : / \:

•0 «$.

/ \ • j"' ' 2 ' • \ : / : : \

X-XX-X-X-X-X-XXXTCX-X-^^-XX-XX^X-XX-XXX -X-X-X-;:

1 . 1 ' I ' 1 0.0 6.5 13.0

L (cm)

19.5

(a) B, Bx and By values along the coil axis marked as line A.

0.020-

0.015

0.010-

c 0.005 H 0)

B 0.000

-0.005-

-0.010

—o-—+-

—x-

-B -B(x)

-B(y)

ooO°N xxxxxxxxxx

+ +

+ / +

/ / /

?ooo

0

o 0 0

<>o ooo 1 j$xxxxx><xx^xxxxxxxxx4:

. + + + "

+ + +

0.0 5.5 11.0

L (cm)

16.5

(b) B, Bx and By values along the coil radius marked as line B.

114

CO

c 0 TJ X

0.020

0.015-

0.010-

0.005-

0.000-

-0.005-

19.5 6.5 13.0

L (cm)

(c) B, Bx and By values along the coil inner wall marked as line C

0.012

0.008-co

f 0.004 c CD

"5 0.000-I -0.004-

-0.008-

—O—B

-+-B(x) -X-B(y)

o /+' / - • '

o v\

_ovi + &

ioo<> *X.

o^< / >+

&

/

xx X

+ :

X /

-H

XvX

X /

X

X /

!+ +++ +++++

0.0 16.5 5.5 11.0

L (cm)

(d) B, Bz and By values along the coil side edge marked as line D

Figure 3-14. Magnetic field distributions of the HTS coil

at 77 K along typical defined lines.

115

Flux Density

B(T)

~H 0.01940

0.01746

0.01553

0.01359

0.01166

0.00972

0.00779

0.00585

0.00392

0.00198

0.00004

Magnetic flux density B (T).

•'."•• •".-'

Flux Density

Bx(7)

-

— I -H

mm

0.01941

0.01678

0.01415

0.01151

0.00888

0.00625

0.00362

0.00099

-0.00165

-0.00428

-0.00691

Magnetic flux density B* (T).

(a)

(b)

116

Flux Density B ( 10"3

\ ^ • \ ,-„V»»

Wfy ml

frTV : • : \-t±^y /$SKm

. 'taj". •"•' ':.

— - ... .,

Magnetic flux density B y (T).

Strength H( 104 A/m)

Magnetic field H (IO4 A/m).

(c)

09

117

Potential

A( 10"* Wb/m)

3.720

2.982

2.244

1.506

0.768

0.030

-0.708

-1.446

-2.184

-2.922

-3.660

Magnetic field potential A (10"4 Wb/m).

Energy Density

V\( J/m

p™

PH

J>

150 135 120 105 90 75

60

45

30

15

Magnetic energy density W (J/m3).

(e)

(0

118

Page 116:

(a) Magnetic flux density B (T);

(b) Magnetic flux density B x (T).

Page 117:

(c) Magnetic flux density B y (T);

(d) Magnetic field H (A/m).

Page 118:

(e) Magnetic field potential A (Wb/m);

(f) Magnetic energy density W (J/m3).

Figure 3-15. Magnetic field distribution maps of the H T S coil at 77 K.

119

3.3.4.3 Operated at 4.2 K with air core

Operated at 4.2 K liquid nitrogen temperature with 45 A critical transport

current, the magnetic field distributions of the HTS coil are summarised in Figure 3-

16 (a) to (d). At 4.2 K liquid helium temperature, magnetic field distribution maps are

shown in Figure 3-17 (a) to (f). The analysis is the same as for the 77 K operation,

except the critical current changes to 45 A from 5.2 A at 77 K.

3.3.4.4 Operated with magnetic core

The HTS coil was analysed as a DC bias winding of a magnetic core and

arranged as shown in Figure 3-18. The magnetic field distributions of the HTS coil

with the magnetic core were analysed along the typical line 1 to line 4 as shown in

Figure 3-18, and the results are shown in Figure 3-19 for 77 K operation and in

Figure 3-20 for 4.2 K operation. The HTS coil centre is co-ordinated at (X=500 mm,

Y=500 mm). The selected magnetic core is the one mentioned in Section 3.3.3.4 used

with the HC-27CRWSM HTS coil.

120

0.20

0.15-

H

CQ

c D x

0.10-

0.05 ---

-O-B -+-B(x) -X-B(y)

o^^o

&

O

HTS Coil at 4.2 K

o.oo -x x-xx-x-x x-x-x-xkx-x-x x-x-xx-xxx-x-x-x XXX-X X^ (

— I — 0.0 6.5 13.0 19.5

L (cm)

(a) B, B x and B y values along the coil axis marked as line A.

0.20-

0.15-

0.10-

-O—B -+—B(x) -X-B(y)

,ooo<> 0<A

<E> /...T..T...V

\

<£> 3> <>°<>0 oo oooi

0.05

ooo 4 -xx-xxxxxxxx^xxxxxxxx.§x><x.xxxxxx4

16:5

(b) B, Bx and B y values along the coil radius marked as line B.

121

0.16-

0.12-

0.08-

0.04

0.00

-0.04 -(

-0.08

O-B -+-B(x)

-X-B(y) #? «>****4>

fxxxxx x

,$

./X7 t • ! f t : + xix+c x*<i>

X ^

X x-

x.xAfifc> X

\ •X x

/..

HTiS Coil at 4.2 K

0.0 6.5 13.0 19.5

L (cm)

(c) B, B x and B y values along the coil inner wall marked as line C.

0.12

0.08-

CO .-t? 0.04-j w c <u Q x 0.00-1 _2 u.

-0.04 -

-0.08

-o-—+-—x-

-B -B(x)

-B(y) o. <y

6ooo 0 <

• / :

X

+

X

><£><£,

dr. &:..... 'v\ O :

+ X + :

:+

X /

X

X / -H

•K-x X x

X

X X X

HTS

+ +++

Coil at 4.2 K

+++++^

0.0 5.5 11.0

L (cm)

16.5

(d) B, B x and B y values along the coil side edge marked as line D.

Figure 3-16. Magnetic field distributions of the HTS coils at 4.2 K.

122

Flux Density

B(T)

(a)B.

I

0.1700

0.1530

0.1361

0.1191

0.1022

0.0852

0.0682

0.0513

0.0343

0.0173

0.0004

Flux Density

By(T)

0.0710

0.0575

0.0439

0.0304

0.0169

0.0034

-0.0102

-0.0237

-0.0372

-0.0508

-0.0643

(b)By.

(a) Magnetic flux density B (T); (b) Magnetic flux density By (T).

Figure 3-17. Magnetic field distribution maps of the H T S coil

for 4.2 K operation.

123

Figure 3-18. The arrangement of the H T S coil and the magnetic core.

124

m >> c CD T3 X

1.50-

1.25-

0.1

P*b

-0.5-

-1.0-

-1.5

• ; w •l-l l>l4 |!-|l-lrM4,.U.ull.l.l.l.U.LU.|.|.t;l-L;i.t-mn:c ,-. j^-TFH

—n - - +-

-B - Bx •By

475 500 525 550 575

L (mm)

(a) B, B, and By values along the coil radius marked as line 1.

unm

JWOOO0 . nrm "ooooo

+++++

475 500 -> 1—

525 55C 575

L (mm)

(b) B, Bx and By values along the coil side edge marked as line 2.

125

CD

1.5-

1.0-

0.5

§ 0.0-•• x

^ -0.5-

-1.0

1 1 1 1 .

0

6 6 i 6

K Ul. i w '

•

t • _ +

A 1 1

i

0

1

S • -to •

> 9 c

+

-D-B -- o-- Bx • •-•*.. .-By

.

•

-

-

-

-

400 450 500

L (mm) 550 600

(c) B, B x and By values along the coil axis marked as line 3.

c CD TJ X

1.50.

1.25-

1.00

0.0

-0.5

-1.0-

-1.5-

400

ao

450

rTtTTl

aBaffl:^ <' >T7--^n-^'a a i a w" w' a^^

-n-B --o-- Bx ....+.... By

500

L (mm)

550 600

(d) B, Bx and By values along the coil inner wall marked as line 4.

Figure 3-19. Magnetic core flux density generated by the HTS coil

as a bias winding at 77 K.

126

1.75-

1.50;,

C. 0.0l?9W?J

-0.5-

CQ >> t7) c tu •o

X ^ -1.0'

-1.5-

CCD

'o o ok Jvk cipipiii

450 500 550 600

L (mm)

(a) B, B x and By values along the coil radius marked as line 1.

-0.1-

°°°TOra

450 500 550 500

L (mm)

(b) B, Bx and By values along the coil side edge marked as line 2.

127

2.0

1.5-

1.0-

0.5-

0.01

-0.5-

-1.0-

5E0oS!EI

e CIS

rfW

•intTJTWtfl 1111111111111111111111111 I + I-I-I-I i i-n fe'-rmpwi

+ +

+

+

+

— • -Q. .

"- +-

-B Bx

-By

400 450 550 500

L (mm)

(c) B, Bx and By values along the coil axis marked as line 3.

600

1.75

1.50- rxn / \

-1.0-

-1.5-

^HnidddcldcldcldHig

o.o jpaaaa^o^gsee00^'" ., J.+++++++++4

-0.5-

ocidj^cid^a

—•-o..

- - +-

-B Bx -By

400 450 500

L (mm)

550 600

(d) B, Bx and By values along the coil inner wall marked as line 4.

Figure 3-20. Magnetic core flux density generated by the HTS coil

as a bias winding at 4.2 K.

128

3.3.5 M A G N E T I C FIELD PROPERTIES O F T H E H T S COIL

3.3.5.1 Critical current degeneration of the HTS coil

From Figure 3-2, the critical current degeneration of the HTS coil AIc(coii),

which is the difference between the HTS wire critical current L,(wire) and the HTS coil

critical current Ic(coii), is related to the generated magnetic field and winding

procedure in forming the HTS coil. The HTS coil critical current degeneration can be

generally expressed by

Alc(coil) = Alc(f) + Alcon) + AIc(c) (3-16)

or

Alc(coil) / Ic(wire) = [Alc(f) / Ic(wire)] + [Alc(m) / Ic(wire)] + [AIc(c) / Ic(wire)] (3-17)

where Mc^ is the degeneration caused by increased self magnetic field to form a coil

from the HTS wire, <4Jc(m) is the degeneration caused by mechanical processing to a

coil made from the HTS wire, and ^(C> is the degeneration caused by changing

cryogenic parameters due to forming a coil from the HTS wire. The AIc(m) can be

evaluated by

Alc(m) = l/2[IC(wire) " I c(wire)] (3-18)

where I c(wire) is the long HTS wire critical current after being rewound from the HTS

coil to the initial long HTS wire configuration. The result from a short sample of the

129

H T S wire used shows ^Mc(m) « 0 for a winding former having diameter larger than

about 10 cm. For the multifilament H T S wire with a large A g sheath percentage cross

section .area and operated by immersion in liquid nitrogen, the AIc(m) and AIc(C) both

can be neglected, therefore Equation 3-8 can be simplified as

Alc^oil) / Ic(wire) ~ [Alc(f) / Ic(wire)] (3-19)

which means for the Ag-clad (Bi,Pb)2Sr2Ca2Cu3Oio+x 27-core multifilament wire and

operated by immersing in liquid nitrogen, the magnetic field degradation of the

critical current is a major factor for the total critical current degeneration of the H T S

coil. For the 27-filament H T S wire used, the H T S coil critical current degeneration is

Alc(coii) / Ic(wire) - 65.3% [(15-5.2) A / 15 A ] , which is controlled by the H T S material

flux pinning property, i.e. the magnetic field-dependent critical current.

3.3.5.2 Generated magnetic field of the HTS coil

The self magnetic fields generated by the H T S coil were measured with

critical transport D C current and also with 50 H z A C transport current. The magnetic

fields in both magnitude and direction were also analysed with respect to the

application of this anisotropic H T S wire. The magnetic field analysis has a good

agreement with the measurement results. The error is caused mainly by (a) the

analysis being based on an ideal uniform configuration which is slightly different

from the practical nonuniform situation; (b) the search coil configuration and

position.

W h e n the magnetic field B x along the X axis is at the level 0.2 T and in a

favourable direction to the H T S wire, i.e. B//ab, according to the Figure 3-2 (b) the

130

critical current of the H T S short sample maintains 37 % of the zero field critical

current, which is the limitation of this HTS wire at this magnetic field. According to

Figure 3-2, at 77 K the magnetic field generated by the HTS coil is about 0.025 T at

the HTS coil centre, and at this value, B = Bx//ab at the HTS coil centre allows 81%

(17.5 A / 21.5 A) of the HTS short s.ample critical current to be maintained. At the

side of the HTS coil, i.e. the line D as shown in Figure 3-13, the By component is

about 0.007 T, which is equivalent to B = 0.025 T at the HTS coil centre, and in an

unfavourable direction to the HTS wire, which allows the HTS short sample to

maintain 65% (14 A/ 21.5A) of the zero field critical current. Consequently the By at

the HTS coil side will determine the HTS coil critical current.

The AC magnetic field rms value also has good experimental agreement with

the DC magnetic field result. At 77 K to generate the rms magnetic field with an

equivalent DC magnetic field of about 0.025 T, the AC resistive voltage drop of the

HTS coil is evaluated by a short sample, and is within the low loss range, i.e. at

transport AC current Iac(ims) = 5.2 A (rms), AV(x') / AIac(nnS) « 0.52 pV/A « 12.72

pV/A which is the value for the normal state at 77 K, i.e. Iac(rms) ^ 20 A.

3.3.5.3 Basic HTS coil design consideration for magnet

Consideration for designing a HTS coil requires the relation between I, N and

H for the HTS wire. As a basic consideration, the magnetic field at the HTS coil

centre B(0,0,0) is given by

B(0,0,0) = p0NCOHl-[2(R2-R1)]-1-ln{[R2+(R2

2+L2)1/2]/[R,+(Ri2+L2)1/2]} (3-20)

131

where Rt and R 2 are the H T S coil inner and outer radius respectively, 2L is the H T S

coil length, and N is the number of HTS coil turns. Since the critical current of the

HTS wire is degenerated with increasing magnetic fields, the HTS coil ampere-turns

values NI are limited by the HTS coil magnetic field. The generated magnetic field H

is proportional to the HTS coil NI; meanwhile the maximum HTS coil working

current Ic is also a function of the H. For a (Bi,Pb)2Sr2Ca2Cu3Oi0+x HTS coil with

fixed dimension, the value of By(x,y,z) at the HTS coil side edge, as well as the

B(0,0,0), should be used to design the HTS coil. For simple evaluation, B(0,0,0) [=

kB(x,y,z)] = 0.1 T can be selected to calculate the NI for 77 K, since I^O.IT) / W) ~

50% at liquid nitrogen temperature. The present (Bi,Pb)2Sr2Ca2Cu3O10+x HTS coil

operated at 77 K is limited by its low field critical current, which field is lower than 1

T, for designing a HTS coil as shown by the sample investigated. For 4.2 K

operation, B (0,0,0) can be selected much higher than 20 T.

To calculate the maximum value of ampere-turns (NI)max of the HTS coil, Ic =

f(B,T) is required. The experimental data on this (Bi,Pb)2Sr2Ca2Cu3O10+x HTS wire

have been fitted and expressed as an exponential function [309]

JC(B,T) = cc(T)exp(-B/P(T)) (3-21)

where parameters oc(T) and P(T) are temperature dependent and 50 K is suggested to

be a crisis point [309]. Consequently a HTS coil can be designed by considering its

magnetic field and temperature dependent critical current, and the coil configuration

dependent magnetic field generated.

132

At 77 K low field application, considering the weak link path of the

(Bi,Pb)2Sr2Ca2Cu3O10+x HTS material, -0.2 T magnetic field is the point for the

sample used; above this value only strong grain-to-grain linkage maintains the critical

current. If the working current of the HTS coil is set according to the critical current

given by the strong link application, the relation between the critical current and

applied magnetic field is given by [310]

Ic^Ico^xpKH/Hor1] (3-22)

where L/5 is the strong linkage critical current, and Ic0s, H0, and a are defined in

reference [310].

3.3.5.4 Inductance of the HTS coil

The HTS winding inductance is related to its particular application. As a coil

the inductance can be evaluated by [311]

L = Uo (1/47C) [(Ri+R2)/2] N2 P F (3-23)

where P, F are winding dimension dependence coefficients. For the HTS coil made

by the 27-multifilament wire, the constants P and F are selected as 26.8 and 0.46

respectively according to the HTS coil configuration. Therefore the 306 turns HTS

coil, HC-27CRWSM, has about 5 mH inductance. Equation 3-23 describes only the

configuration specified inductance. For the inductor made by the Ag-clad

(Bi,Pb)2Sr2Ca2Cu3O10+x wire, it has specifications of (a) operation at low frequency

133

(50 H z test result in Figure 3-5 shows it is an acceptable value experimentally), and

(b) working current I < 5.2 A for 77 K application.

3.3.5.5 Ampere-turns and magnetic core bias winding

At low field range and 77 K operation, the HTS critical ampere-turns values

NIc are affected by the magnetic field distribution of the applied magnetic core. For a

closed magnetic core, the magnetic field generated by the HTS coil is

Hs = NIc/lm (3-24)

where Hs is the critical magnetic motive force (mmf) generated by the HTS coil, and

lm is the magnetic path length. The HTS coil used generates 3.8 x 103 A/m mmf on

the magnetic core used. Its NIc is large enough to saturate the magnetic core at about

1.4 T, which is a common magnetic core flux density value for a normal ferrite

magnetic core used in electrical engineering. For a magnetic core with an air gap, the

mmf required to saturate the magnetic core becomes much larger and its relation to

the magnetic flux generated by the HTS coil can be evaluated by [312]

O = Ncoillc [lm/UoPrA + UPcA]"1 (3-25)

where Ncou is the number of the HTS coil turns, lm is the magnetic core length, la is

the air gap length, O is the magnetic flux, A is the cross sectional area of the

magnetic core used, u^ is the permeability of free space, and ur is the relative

permeability of the magnetic core.

134

3.6 SUMMARY AND CONCLUSIONS

(Bi,Pb)2Sr2Ca2Cu3O10+x H T S compound has high intrinsic values of critical

current density and can be produced in long lengths in the form of a Ag-sheathed

tape. The material crystal structure of (Bi,Pb)2Sr2Ca2Cu3O10+x is highly anisotropic;

however the plate-like shape and growth of the HTS grain is such that rolling and

annealing the tape naturally aligns them with their c-axis perpendicular to the plane

of the tape. Unfortunately the critical current density of the Ag-clad

(Bi,Pb)2Sr2Ca2Cu3O]0+x wire is significantly below that of thin films of the same

HTS materials. In particular at high temperature (e.g. 77 K) and in high magnetic

fields the critical current density is presently still low over long lengths for useful

application at liquid nitrogen temperature, though recent improvements have led to

the development of small prototypes operated at lower temperature (e.g. 4.2 K).

HTS coils have been developed by using the Ag-clad (Bi,Pb)2Sr2Ca2Cu3Oio+x

as both single core and multifilament core HTS wires. The HTS coils can be made in

the form of a solenoid or pancake, through the react and wind and the wind and react

procedures, and maintain a fairly high percentage of their short tape values when

operated in a low field at 77 K. This enables the Ag-clad (Bi,Pb)2Sr2Ca2Cu3Oio+x

wire to be considered for some low field applications at liquid nitrogen temperature,

and for high field magnet applications at 4.2 K, where a much higher critical current

value is maintained by strengthening the flux pinning of the HTS.

A HTS coil made by a Ag-clad (Bi,Pb)2Sr2Ca2Cu3OI0+x 27-core multifilament

HTS wire, through a react and wind procedure, has been studied with regard to its

135

magnetic field properties. The magnetic field degeneration due to forming a coil,

with this mechanical flexible HTS wire having a large Ag sheath cross sectional area,

is mainly caused by increased self magnetic field and its distribution in the HTS wire.

Consequently the HTS coil magnetic field properties have been studied to help with

the design of the HTS coil. The magnetic fields generated by the HTS coil have been

measured experimentally, both for DC and AC field. The magnetic field distributions

have also been analysed. The results show that the radial field at the HTS coil side

edge determines the HTS coil performance at 77 K. The magnetic field information

obtained is essential for future design of a HTS coil with this anisotropic HTS wire

having strongly magnetic field-dependant critical currents at 77 K. The present

(Bi,Pb)2Sr2Ca2Cu3O10+x HTS coil operated at 77 K is limited by its low field critical

current. Before dramatically improving the (Bi,Pb)2Sr2Ca2Cu3O10+x HTS material

pinning property at 77 K to suit practical higher level magnetic fields, it is necessary

to reduce the operation temperature well below 77 K to increase the critical current.

This represents the greatest problem in the use of HTS coils for some electrical

applications.

In addition the use of the HTS coils to produce a higher magnetic field can be

achieved by reducing the temperature, such as to 64 K, 27 K and 4.2 K. However, to

produce a HTS coil with a large number of critical current ampere-turns for high field

applications at 77 K, further improvements in the critical current are necessary. The

production of long length HTS wires with uniform properties is also required.

136

Chapter 4

HIGH Tc SUPERCONDUCTING SATURABLE MAGNETIC

CORE FAULT CURRENT LIMITER

4.1 INTRODUCTION

High Tc superconductors (HTS) have been used to design electrical fault

current limiters (FCL) with different operation principles [261, 313-315]. The

saturable magnetic core reactor, is one of the techniques used to design a FCL, and

has been studied with respect to using the well developed Ag-clad

(Bi,Pb)2Sr2Ca2Cu30io+x H T S wire which has been discussed in Chapter 3.

The principle of a saturable magnetic core reactor is employed to develop a

fault current limiter device, in which a D C bias winding with a large value of

ampere-tums NI is required. The Ag-clad (Bi,Pb)2Sr2Ca2Cu3Oio+x H T S wire is used

to make a H T S coil and is considered for this application. A small laboratory device

is made and tested at a low current level for the practical investigation.

The H T S coil is made by using a Ag-clad (Bi,Pb)2Sr2Ca2Cu3Oi0+x H T S wire,

and studied for use as the D C bias winding of the HTS-FCL device. The testing on

the H T S coil made for this application includes its critical current, critical ampere-

turns value, and generated magnetic field. The required values of D C winding

ampere-turns and generated magnetic field, related with its rated A C current and

137

required A C impedance, were studied to design a H T S coil for the D C bias winding.

The results from the H T S coil and the device electrical behaviour provide

information to evaluate the application of the H T S and this type of F C L in electrical

power systems.

The electrical current limiting behavior of this type of HTS-FCL is analysed by

using modified \|/-I characteristics and different device samples. The application of

such a H T S - F C L is then considered and calculated for use in 6 k V electrical power

systems. The well known "Electro-Magnetic Transients Program" (EMTP) is used to

investigate its electrical fault current limiting characteristics, and a comparison with a

conventional air-core linear inductor normally used in present networks as a fault

current limiting reactor is also carried out. The No. 98 pseudo non-linear inductor

element in the program is selected for the E M T P analysis of this FCL. The results

prove that this new HTS-FCL can reduce the short circuit currents effectively in a

practical electrical transmission and distribution system.

4.2 EXPERIMENTAL DEVICE CONFIGURATION AND BEHAVIOR

4.2.1 Making a Testing Device

4.2.1.1 Principle

The designed F C L is based on the principle of a saturable magnetic core

reactor, which consists of a pair of magnetic cores (for single phase power system)

with a normal A C limiting winding and a D C bias winding on each of the cores. A

138

H T S coil is prepared to be the D C bias winding. The polarity of the two reactor A C

windings is set to limit fault currents in both positive and negative cycles. The

normal DC biases are adjusted to let both cores saturate while the rated AC current is

passing through the AC limiting windings. Whenever a fault occurs, the fault current

drives one of the magnetic cores out of saturation, thus increasing the inductance and

thereby reactance in the AC circuit. Therefore, this fault current limiting device

behaves like an air-core low inductance in normal operation, and as a high impedance

during fault condition. Since the DC bias is required to have a large number of

ampere-turns with this type of FCL, a normal copper winding is prohibitive for this

application when the device is rated at practical values.

The HTS DC bias coil prepared by using the Ag-clad (Bi,Pb)2Sr2Ca2Cu30io+x

HTS wire, is operated in a liquid nitrogen cylinder Dewar which surrounds one or

even six (in the case of a three-phase power system) branches of the magnetic cores.

High DC ampere-turns values are required on the HTS coil to provide the necessary

magnetic motive force (mmf) on the magnetic core.

A small laboratory current limiting device was designed and built. Its

operating principle and configuration are shown in Figure 4-1.

4.2.1.2 Magnetic core

The magnetic cores selected for the current limiting device are common ferrite

magnetic "C" cores, the specification of which are summarized in Table 4-1.

139

F C L

1—Transmission or distribution line;

2,3—AC limiting windings;

4,5—Magnetic cores;

6,7~HTS DC bias coils in liquid nitrogen bathes.

Figure 4-1. Principle of the HTS fault current limiter.

140

Table 4-1. Magnetic core parameters.

Rating

Window length

Window width

Strip width

Total strip thickness

Core cross-section

Mean length

Stacking factor

Flux density

3500 V A

254 m m

101.6 m m

38.1 m m

38.1 m m

1452 mm2

830 m m

96%

1.7 T (50 Hz) (1.5 T for

quiet applications)

4.2.1.3 Device windings

The device windings consist of AC windings in one branch of each magnetic

core, and on the other branch there is a DC winding. There are also magnetic field

search coils on each branch of the magnetic cores. The configuration of the device

normal copper wire windings is shown in Figure 4-2, where <|) is the wire diameter.

141

0,

100,

1000

2824

10

100

DC

1 1

-: : • • • • • .

.>•••••'j

AC

yy -' • >

" • - ' •

* ' • ' < •

[ [

0

,10

,20

,50

.100

,227

10

100

(|)Ac (0-227 t) = 3 x 2.7 m m

fyac (0-2824 t) = 1.6 m m

tacsearch-coii (10 t, 100 t) = 0.3 m m

focsearch-coil (10 t, 100 t) = 0.3 Him

Figure 4-2. Normal windings of the F C L device.

142

(a) A C windings

To calculate the AC winding turns nAc, a simple model has been used [316], and it

can be estimated from

V„ns=2V2 7cfnAcOmax (4-1)

where O.^ is the maximum magnetic flux change through one of the AC windings, V^ is

the source voltage in rms value and f is the frequency. The Gw for this practical device is

O^ = BmaxxS = 2xl.7Tx 14.5 x IO"4 m2 = 4.93 x IO"3 Wb, where S is the area of the

winding cross-section. The nAC for the system having V,™ = 240 V and f = 50 Hz, is about

110 turns.

Practically the n^ has termmals at 5, 10, 20, 50, 100 and 227 turns. The AC

power source used can be variable from 0 to 240 V at 50 Hz. The rated AC current is

selected at about 100 A. Three 2.7 mm wires are used in parallel for the AC windings.

(b) Search coils

There are also magnetic flux search coils linked with the magnetic cores,

which are 10 turns and 100 turns coils wound by an enamel wire having a diameter (j)

= 0.3 mm on each set of the magnetic cores. As the search coil is passed through a

magnetic field, a voltage is generated in the coil that is directly proportional to the

strength of the field, and inversely proportional to the length of time that it takes to

pass the field through the coil. This voltage signal, e oc dO/dt, from the search coil is

fed into a magnetic fluxmeter to monitor the magnetic core flux change.

(c) DC windings

143

To take the A C current effect into account, the D C winding needed to keep the

magnetic core in saturation is calculated by

nix: = IiimitiUc / IDC (4-2)

where noc is the DC winding turns, nAC is the AC coil turns, and Iiimit is the AC current

at which the AC current is limited initially. The normal DC copper winding has iDcriDc

= 2 A x 28241 ~ 5.6 kA-t. The Iumi, is then equal to 5.6 kA-t / nAC. The maximum DC

current IDC is limited to 10 A for the normal operation.

The mmf, H, generated by a winding on a magnetic circuit is

H = Ixn / MPL (A-t)/m (4-3)

where Ixn is the winding current .ampere-turns value, and MPL is the magnetic core

mean path length. From the B-H curve of the selected magnetic core, the Ixn required

to saturate the magnetic core depends on the magnetic core saturable flux density Bs

which corresponds to a value of 1.5 T.

Based on the available HTS sample, only a low current ampere-turn coil can be

designed at present with the Ag-clad (Bi,Pb)2Sr2Ca2Cu3Oi0+x HTS wire, limited by its

length and critical current. To evaluate the application of the HTS wire in the device, a

HTS coil, HC-27CRWSF, is prepared for one set of the magnetic cores as part of the

DC bias winding. The final HTS coil used has critical ampere-tums of about 0.86 kA-t.

A normal Cu coil is used in series with the HTS coil to form a hybrid DC bias winding

providing the balance of the DC bias required.

144

4.2.2 Device Testing Results

4.2.2.1 Winding accuracy and transformer effect

The windings of the device were tested after the device was made. To confirm

the precise rum-ratio of the normal copper windings, an AC voltage of 50 V (rms) was

added to the 100 turns DC winding; then the voltages (rms) were measured at different

windings, and results are as follows.

(a) DC windings: V(0-1000 t) = 500 V. [V(0-2824 t) = 564 V when 20 V was

added to the 1001 DC winding].

(b) AC windings: V(0-10 t) = 4.87 V; V(0-20 t) = 9.74 V; V(0-50 t) = 24.7 V;

V(0-1001) = 49.6 V; V(0-227 t) = 110.8 V.

(c) Search coils on DC winding side: V(0-101) = 5 V; V(0-1001) = 50 V.

(d) Search coils on AC winding side: V(101) = 4.88 V; V(0-1001) = 49.2 V.

The device windings were therefore confirmed to be as designed.

For a saturable magnetic core electrical fault current limiter, the AC coil and

the DC coil are equivalently coupled with an air-core since the magnetic core is

deeply saturated during its normal operation. The coupling effect between AC and

DC windings due to the flux linkage between these two coils is measured without the

magnetic core. While the magnetic cores between the device windings were absent,

AC currents (IAc, 50 Hz, rms) were supplied to the 100 t AC winding, and AC

voltages (rms) are measured across the 100 t DC winding (VDC). The results obtained

are: IAC = 2 A ~ VDC = 0.44 mV; IAC = 20 A - VDC = 5.2 mV; IAC = 40 A - VDC =

10.5 mV.

145

4.2.2.2 Magnetic core B-H curves

(a) DC B-H curve

The B-H curve of the magnetic circuit used, as shown in Figure 4-3, was

tested with a DC power supply. A magnetic flux meter was used in the test to

monitor the magnetic core flux change AO, and AO against (nDC x IDC) was recorded.

The flux density B can be calculated by B = AO / nS, where AO is the total magnetic

flux change in a n-turn flux search coil, and S is the magnetic core cross-section. In

Figure 4-3, the units for B and H are given by 0.68 T and 120 (A-t/m) respectively.

The ideal magnetic core for this application should have high saturable flux

density Bs and low Hs; therefore the condition of high AC impedance and low DC bias

could be satisfied.

(b) AC B-H curve

The B-H curve was also tested with a 50 Hz AC current using a R-C

integration circuit, and the result is shown in Figure 4-4. There are magnetic core

losses, i.e. hysteretic loss and eddy current loss, and the B-H curve of the magnetic

core therefore has a greater area with the AC exciting current compared with the DC

operation. From the R-C integration circuit connected with the secondary winding of

the magnetic core, based on the selected R, C components, the magnetic core flux

density is given by

B = Vy x (RSCS) / nsS (4-4)

= 1.1 x (81810 ft x 2.33 pF x IO'6) / [1001 x (0.038 m)2]

= 1.45 T

146

3.00

.00

CD

1.00

-3.00 4.00

TT

£

TT TT

•2.00 0.00

H 2.00 4.00

H: 120 (A-t/m)/d; B: 0.68 T/d.

Figure 4-3. D C B-H curve of the magnetic core.

147

The magnetic field generated by the exciting winding is

H = V x x n p / ( R pMPL) (4-5)

= 20 V x 100 t / (10 Q x 0.83 m) = 241 (A/m)

where the subscripts s and p used represent the secondary and primary circuit, and the

subscripts X and Y present the voltage signals in X and Y direction respectively.

^<^^m^^^^r^!;T*:mT^m^m^^m

(Vx: 20 V/d; Vy: 0.5 V/d)

Figure 4-4. AC (50 Hz) B-H curve of the magnetic core.

148

4.2.2.3 Magnetic field .and resistive loss of the normal D C winding

The normal DC winding magnetic field and its power loss are tested with

different DC NI. The DC coil was made from a enamel Cu wire with total 2824 turns

on a 46 mm x 47 mm x 252 mm former. Figure 4-5 shows this DC coil magnetic

field and its power loss. The loss curve shows that the coil has a high energy loss and

is heated by increased transport currents. This is the reason why a normal Cu wire is

not appropriate for use in a practical fault current limiting device. The measured

magnetic field at the coil centre is shown in the figure to be about 0.15 T at 20 kA-t.

rO

O

o —

CO 00

o

o

FCL dc Cu coil magnetic field and loss

o — A A A A A Magnetic field ooooo Loss

° fT i T ~ T M I I I | I I I I II I I I | I I I I M I I I | M I I

0 5 10 15

Dc coil current ampere-turns (kilo-AT)

rO

d

33 cu

<N 'C d .2

-> cu c cn o d "S

CU

c CD

o

20

Figure 4-5. Losses of the normal C u D C winding and magnetic field generated.

149

4.2.2.4 A C winding impedance

(a) AC reactance

The AC winding of the FCL device is equivalent to a variable inductor L in

series with a resistor R. Therefore its winding impedance is Z = [R2 + (X)2]1/2, .and

reactance X = caL. As a FCL, the L of the device is not a constant; rather it is a

function of the circuit current. The normal inductance L is related with the rate of

magnetic core flux changing dO/dt. Since Z = Vpeak / Ipeak (= V™ / Irms), the AC

winding reactance can be calculated by its voltage drop and circuit current. The

voltage drop across the AC winding is then measured with variable circuit current IDC

and IAC.

A 20 turns AC winding is selected for testing its AC reactance with

increasing DC bias ampere-turns, and the magnetic core is temporarily energised by a

1000 turns Cu winding, and the AC current is 20 A rms. The reactance measurement

of a 20 turns AC winding with different DC bias was measured from AC voltage

drops with AC currents for different DC bias currents. Figure 4-6 shows one set of

the magnetic core AC winding reactance changing with different DC biases. As the

DC bias is increased, the AC winding reactance is decreased. The rated AC current

should be set within this low reactance range in the normal operating condition. The

reactance calculated for the situation with an open DC winding is approximately 0.6

Q (12.1 V / 20 A), which is much higher than that with a close DC winding circuit

having zero DC current.

(b) R-C-L bridge measurement

The inductance of the A C winding is also measured with a R-C-L bridge. The

AC winding nAC selected has nAC = 100 t, and the DC winding has nDc = 1000 t

connected with a 12 V DC battery. The inductance of the 100 t AC winding is shown

150

in Figure 4-7, where 100 H z testing frequency and the R-L series model were used.

On the other hand, the situations with open and short DC terminals have inductance

of 57.35 mH and 4.47 mH respectively.

0.12-

,-N 0.11 -

S. i 0.10-

Reactance

p

p

0.07-

0.06-

•

• \

•,

' -B^

1

Reactance of one magnetic core ac winding

i 1 1 1 1 1 1 1 1 1

1000 2000 3000 4000 5000

Dc bias coil ampere-tums (AT) 6000

Figure 4-6. A C winding reactance of the device.

D C current (A)

Figure 4-7. A C winding inductance measured by a R-C-L bridge.

151

(c) Circuit current

Figure 4-8 is a steady state low current testing result of the device. One set of

the two magnetic cores with 20 turns AC winding is connected to a 50 Hz - 5.1 V

(rms) AC power supply without any additional resistor in series, and the 0 - 1000

turns of the normal DC coil on the magnetic core is supplied from a DC power

supply. The rms AC circuit current is then measured with increasing DC current. The

result shows that the AC circuit reactance is reduced by increasing the DC bias value.

The fairly flat part of the curve is expected where the magnetic core is saturated.

50-j ,

. Testing one magnetic core ac winding 45 - in an ac circuit •

• • •

< 40- "

I 35" 3 30-

O •

< » - •

20-1 , , , , j 1 1 1 1 1 1 , 1 0 1000 2000 3000 4000 5000 6000

Dc ampere-turns

Figure 4-8. AC circuit current with different DC bias ampere-turns.

4.2.2.5 Tests on current limiting behaviours

(a) Testing circuit arrangement

Figure 4-9 shows a laboratory circuit .arranged for reactance, short circuit tests

and operation of the device. Tests on the device performance .are carried out, such as

B-H curves, I-V curves winding reactance, and short circuit behaviour.

152

Figure 4-9. Testing circuit.

153

(b) Steady state tests

With an oscilloscope, the voltage across the device AC winding VAC and

circuit current IAc were measured, and the results are shown in Figure 4-10, in which

100 t AC winding, nAC = 100, and DC windings, nDC = 100 and 1000 were used.

Curve numbers used in the following figure captions are arranged from top to bottom

of each figure. The circuit current is measured by using a current transformer (CT)

having a scale of 140 A/V, therefore the circuit current IAC = (140 A/V) x (CT

voltage signal).

(a)

Open DC circuit

Curve-1 is the voltage across the 1001 A C winding,

10 V/d.

Curve-2 is the A C circuit current, 2 mV/d (I-

0.002x140 A/d).

Short 1001 DC winding.

Curve-3 is the voltage across the 100 t A C winding,

10 V/d.

Curve-4 is the A C circuit current, 10 mV/d.

Short 1001 DC winding by a 22 Q resistor.

Curve-5 is the voltage across the 100 t A C winding

5 V/d.

Curve-6 is the A C circuit current, 10 mV/d.

AC circuit: VAC = 36 V (rms), and a 3.5 Ci resistor

in series.

(b)

With DC bias circuit 0.2 A x 100t

Curve-1 is the voltage across the 1001 A C winding,

20 V/d.

Curve-2 is the A C circuit current, 0.1 V/d.

With DC bias circuit 13 A x 1001

BBBBBBBSBaB • •• • • • • • • • • • • • •

355S5BBB1

WSSMWKBR ffiS SB SS SS SS ;|

HIIII " IIIIIII S IS 51S5 IS SB SS 5= 51 154

Curve-3 is the voltage across the 100 t A C winding,

20 V/d.

Curve-4 is the A C circuit current. 0.1 V/d.

With DC bias circuit 9.85 A x 100 t.

Curve-5 is the voltage across the 100 t A C winding,

20 V/d.

Curve-6 is the A C circuit current, 0.1 V/d.

With DC bias circuit 1.97 A x 100 t.

Curve-7 is the voltage across the 100 t A C winding,

20 V/d.

Curve-8 is the A C circuit current, 0.1 V/d.

With DC bias circuit 9.9 A x 1000 t.

Curve-9 is the voltage across the 100 t A C winding,

20 V/d.

Curve-10 is the A C circuit current, 0.1 V/d.

With DC bias circuit 6 A x 28241

Curve-11 is the voltage across the 100 t A C

winding, 20 V/d.

Curve-12 is the A C circuit current, 0.1 V/d.

AC circuit: VAC = 36 V (rms), and a 3.5 Cl resistor

in series.

mm

3.3&Si3SiS£ii£i

(c)

With D C bias circuit 0.06 A x 1000 t.

Curve-1 is the voltage across the 100 t A C winding,

20 V/d.

Curve-2 is the A C circuit current, 0.1 V/d.

With DC bias circuit 7.54 A x 1000 t.

Curve-3 is the voltage across the 100 t A C winding,

20 V/d.

Curve-4 is the A C circuit current, 0.1 V/d.

AC circuit: VAC = 36 V (rms), and without a

resistor in series.

| ~-

1

tv E R I

I* H--

/

\

M M

"ft / \

V !- *V ': —f\

\f\f] i J /

'J' M M

/

JI

11

1 M 1 M M

_ f\ :: A

um

r "t 1

\J y J 1111

t

1 1

\^f

<=v+- M V v A 1111 M 111

A i i \

x. LLJ, 4-

UM Figure 4-10. Steady state I-V characteristics of the device.

155

Figure 4-11 shows other measured results, where data were acquired at a

sampling frequency of 1000 Hz and sampling interval of 0.4 ms. The AC winding

voltages across one set of the magnetic cores of the device having source voltage VAc

= 25 V (peak) and circuit current IAc were measured, and the results are shown in

Figure 4-11, in which nAc = 100 t AC winding, and nDC = 1000 t DC winding were

used. Since the positive cycle is not limited, the current peak in this cycle is higher

than the negative cycle which is limited by the device.

The current peak values are recorded and shown in Figure 4-12. The positive

cycle is not limited, and the AC current in this period drives the magnetic core

towards a higher degree of saturation. In the case of having a constant circuit current,

the relation between applied AC voltages and DC bias values is shown in Figure 4-

13, which has IAc = 20 A (positive peak).

30

20 >

CD

o > 0-\ T3 C (0

C? -10 c a? -20-O

-30 Open DC Winding

0.00 0.05 0.10

t (s)

0.15

Voltage

0.20

(a) Open D C winding.

156

JU-

20-

10-

0-

-10-

-20-

-30-

I

Voltage ;

Current, ; / '. i

l(DC) = 0.14A I

r

•:/ A KJ ••

• i ' i ' i

: A : A :' A : A : A i .•

,\/\\/l\t\\/ :\ i/ i\ 71

1 • I » i

1 1

:' ': :'

'•

ww ; : ; :

i j ;; ;. . -

1 1 0.00 0.05 0.10

t (s)

(b) IDC = 0.14 A.

0.15 0.20

30'

CD

cn CD

o > •o

c CD

20-

10

0

-10-1

£ -20-3

o -30

Voltage -Jl

l(DC) = 0.54A 1

0.00 0.05 0.10

t (s)

(c) IDC = 0.54 A.

0.15 0.20

30

ID •) CD O

> •o

c ca

c

E o

20

10

-10

-20

-30 l(DC) = 4.43 A

0.00 0.05 0.10

t (s) (d) IDC = 4.43 A.

0.15 0.20

Figure 4-11. I-V characteristics of the device.

157

c cu

U O <

_

25-

20-

15-

10-

5-

0-

-5-

-10-

-15-

i

o

/

o / o

fi -*— Ope

A

l '

I ' 1

-"° ^O"""

n D C winding

~A A

1 ' 1

1

. o—

,

i | i

. o

— o — + peak values

— A — - peak values

A

1 1 '

_

-

.

;

-

-

-

-

-

0 1 2 3 4

D C Bias Ampere-tums (kA-t)

Figure 4-12. Current peak values for single magnetic core.

70

60-

| 50 0_

> 40-(D D) CO

% 30 >

-i r n ' r

A C Applied Voltage and DC Bias

- o- -o- -20 - |(AC) = 20 A [Peak(+)]

-o

1 2 3

D C Ampere-turns (kA-t)

Figure 4-13. D C bias values and applied A C voltages.

158

(c) Short circuit and transient measurements

The device has been tested at low voltage and low current levels to investigate

its fault current limiting behaviour. The AC source has voltage VAc (rms) = 34 V, and

the circuit resistance R = 3 Q. R is shorted at different voltage points on the

waveform. Figure 4-14 is the short-circuit test results for one (negative cycle) of the

AC windings with 100 turns and 300 A-t DC bias. Figure 4-15 has the same condition

as Figure 4-14 but with 2000 A-t DC bias.

(a)

'I I '"•- •'•• jft A- • "•• L'V V •'••

i'. .Mill (I (1 I1. /Dl is I I i ! • ) ! II M M M M M (

;•;.(( M n !•( is n-Mft•'• ! 1 ! I ! ', ! M l ! I I M 'I M ' II i'i !J li:',f 1) !;l Si !l •; -j v *.' v" v" y ' v "v v •

(b) X./i..A.:A.A..A A A A A. A A A... «./ V ^ V "vV • ^ -< •_• - •-• V -M

(c)

H i\ ;iiii /i n (i n m fi ii n n\ :,-, « .> n ,-,:A „ a i! ,'liH f! H II /I'll !'i .'1 I! M i ->.H,! , i-,-.4..;• ,., |..; | ?.(,.,,..( f ! r ( v.+ .|v.;;.l:4.V-.(,-.t; + t-f-i:-Hi-V'tHi

,' •/ v.v y y y y-iU n u n iiii,' i n n i M M ; W t Ml !| !,' i! in II || !( II (Ii 1/ i| t

- - 7 — j;v '•> V v in-» V v v v-v tf 1

. J W L A.A...A,.A,.J.,.A..A A A A . A - .: v *-> v ^ w • *-< •-» y w > --J • •

:A ii ii n n :n li I n !'• II n n ' A * !/, A ... .« ,. 5)1 II (I J! ft i/l M I, I •/ J

^./4../v/..v/••vy..•^/•^•/v-/••^•f^••:H•|•it^f4-^Vf1lltlft^i

V V 'Vi V V U V V! tl M l I 1 I I,1 II If ) W I ! ! i »• ,1 II 'II If II I) 11 '.II If I II II I •? ; T ; v - v-rtf •"-•*,* y—.^-.-^-TV V- v, V t/-i

(a) - (c) are the situations of the shorted circuit at different voltage points on the waveform. The higher curve is

the AC current (1.2 x 140 A/d), the lower one is the voltage across the device (40 V/d), and 50 ms/d for the

time.

Figure 4-14. Short circuit currents with the FCL device having 300 At DC bias.

159

(a)

!-.NjAV\..».>V".^A.,fcr!V\..^i.**r^

•H ft Hi I! 1 u Ii i s IS ft f,i j:..i..4i ii.,ji„a,..ijk.,j\ jt..j.uujJU.uii

J v >v V \J' V W V' '•-'' V 'V

I. :' i : i :....-* .f. jn. n .i..!....n. n JA. r. n..:...^ r. -j

• l\ v I! ;! n (! i1 (' (l 'i1; f! '!| f i A .: ! A .4. A .i A I • * ... i ,.. A I fi M in / I ] \ I t I i\ ii 'l l' I J 11 I 'l / I ji ;u u' it u v ;u u v4 u y IU y ^ 1IUIW II \!( )i \\\\\ Ii .il ii II! i ! ! ! ! t I; ilMI ,1 lil II iliil II 'il II Hi

r ! i i t t it \ r r u vt y y ift-y « v y y-i

(b)

i i i i i •!• n ft i A A /I I, ft ! Jl II ft A ft : : : : T (I Ji :|| A II II ,'l : M || IS II ||

: 1 i' W I :.' I 1 \ I • 1 / '. >' \ -I \ .' i I A ) I /

:.?. n ir.i L. A :A a AL....S, i. :J J, A.i

ri n y n )i t\ i\ ii n n n /, i\ , :n /i n n II n n n n /i n n /H

/ IMf 1/ U if 1/ II if Ni i! Ii ' u n y u III w u M ii in if II i •1! \\ ii/ tl Vi"t t) f V .It"t tl i

A ^ : . A

? V 1/ I U V \)f U V ! U 1/ ¥ U U f | / IJ if ) { \ P i f I M ! '"' ' '= u '' !

(a) and (b) are the situations of the shorted circuit at different voltage points on the waveform. The higher curve is the A C

current (1.2 x 140 A/d) and the lower one is the voltage across the device (40 V/d), and 50 ms/d for the time.

Figure 4-15. Short circuit currents with the FCL device having 2 .kA-t DC bias.

Various situations have been investigated to understand the device behaviour.

The circuit currents with an open DC winding, a short-circuit DC winding, and

without the FCL device are also measured and shown in Figure 4-16 to Figure 4-18.

There is a 3 Cl resistor in series with the testing circuit for the above measurements.

160

Short circuit current with the F C L device having an open D C winding is

shown in Figure 4-16, which is the situation of a normal iron core inductor without

the DC bias. It has very high reactance in the normal operating situation, which can

not be accepted by practical applications as a fault current limiter. Since the device

reactance is much higher than the circuit resistance, there is no significant current

change when the 3 Q resistor was shorted, as shown in Figure 4-16.

When the DC winding terminals are shorted, the AC winding reactance is

reduced. Figure 4-17 shows the short circuit current with the device having a shorted

10001 DC winding.

When the testing circuit used was shorted without the device in series, the

short circuit current became very high, and the circuit breaker opened about 10 ms

after the circuit was shorted. This is shown by Figure 4-18.

Figure 4-19 shows the short circuit currents and also the induced AC currents

on the DC winding through the coupling by transformer effect. Different DC biases

were applied for the tests. When the DC bias increases, the coupling reduces to a very

low level, which is the situation for the device at normal operation.

i A. .A. i\. A. A\ L\. A. .-I",. A. A\ A. A. A...A. Aj,A^l.,,A:..Ai;..AiiA:..A;.A;..A,j^ u~\7 \7 PdTf\7^"'^rM'''^''^'"sj'''\T;iV"^"^J"^7"M'I\r\THJT V \JiXi \ M

1 U hi i ii <•) ii it./1 ii ti n ihn ii -. M fhri i) \\ i, MM, h i ,i.-.U j,i;l,i i l(:t.X.( Si. j i.,JX,.!4 ii.,I C'-f-i J-U4J.4J i f l-i i u !4 l.-i'14=11.i, ill M !•!/ i J N l i M i M M / ! Hi f I \ Hi i M K ( I M i n \ i) ii ( i M O i ! i '/: a/ ,1 \i VI \'i\)i \] in (/ \)W'i u iii W U r n \"t U if i/Hi If U if if; rv v 1? v y 4 •?•-••& y tt tH-y y y y \fftt--y y y #71? *i tt V" tfi

The higher curve is the AC current (10 x 0.14 A/d) and the lower one is the voltage across the device (40 V/d).

It is 50 ms/d for the time, and VAC = 34 V (rms).

Figure 4-16. Short circuit current with the FCL having an open DC winding.

161

! ! I ! -y I, 11 n I', /'i IA l\ Ai A ft U\ l\ ,M A . A . A . A . A . A . A . A . A . A A . A . J .1 J4 ! ] j i ; I i I .,' i J \ .M i'l ii 1 j 1 ] 1

i/ (,J v; v V V V Vi V v/ \7 \/ ui- w u it M 7 i f u i J U '('/ \ 1 \ M /1 i ! I I T '"' '"' ii' V ,/j )i \i iii i./ ',/i V V I

i .i. ii A n fi m n M I\ n M ; n. ft j j .A A 1 A rt yi A rt i ,', A ,1 A .4-, I i i f i ii i I l / i i i i f i Jl] i i j i i '} i iif!

a J iv i\ J v I U A iv J i, i V J v Uvt v.J V JA i ,. u i i n U i , t, j i M / I J A i, if U u if liAi if u A if u 1/ i i' ii ii u if if , i nil,1 u 11 i / ,r i' - - " * i - - 1 • • i" XJ yi \jM,i ii [H \\ ijiu \} ii] V! IJi : : T I j ^ '<' V-l-lj •/ •!* tf V'f-y 'V \* V ''.A

The higher curve is the A C current (300 x 0.14 A/d) and the lower one is the voltage across the device (40 V/d).

It is 50 ms/d for the time, and VAC = 32 V (rms).

Figure 4-17. Short circuit current with the F C L device

having a shorted 1000 t DC winding.

I \ 1 ii f'TT ii I M

M ? , n i \ i I' /: \ A \ /j \ J: \ !\

\ / i. \ /1 \ /1 I /1 \ I! iii i f i I f : 1 M !ii I /! 1: in rn trt tri xr\

The higher curve is the A C current (1.5 x 140 A/d) and the lower one is the voltage across the 10 A breaker used (40 V/d).

It is 50 ms/d for the time, and VAC = 35.8 V (rms).

Figure 4-18. Short circuit current without the FCL device.

162

(a) With D C bias circuit 0.2 A x 10001.

(ai)

Curve-1 is the AC circuit current IAc, 2 x 140 V/d.

Curve-2 is the voltage across the 100 t A C winding

V F C L , 200 V/d.

Curve-3 is the voltage across the breaker V B , 10 V/d.

Curve-4 is the voltage induced in the D C winding,

400 V/d.

(a2)

Curve-1 is the AC circuit current IAc, 2 x 140 V/d.

Curve-2 is the voltage across the 100 t A C winding

VFCL, 200 V/d.

Curve-3 is the voltage across the breaker V B , 10 V/d.

Curve-4 is the voltage induced in the D C winding,

400 V/d.

AC circuit: VAc = 36 V (rms), and with a resistor in

series.

163

(b) With D C bias circuit 1 A x 1000 t.

Oh)

Curve-1 is the A C circuit current IAC, 2.5 x 140 V/d.

Curve-2 is the voltage across the 100 t A C winding

VFCL, 200 V/d.

Curve-3 is the voltage across the breaker V B , 10 V/d.

Curve-4 is the voltage induced in the D C winding,

400 V/d.

Oh)

Curve-1 is the A C circuit current IAc, 2.5 x 140 V/d.

Curve-2 is the voltage across the 100 t A C winding

VFCL, 200 V/d.

Curve-3 is the voltage across the breaker V B , 10 V/d.

Curve-4 is the voltage induced in the D C winding,

400 V/d.

AC circuit: VAC = 36 V (rms), and with a resistor in

series.

F\

J-

"Oi ' ' "Oi V 2

---------^i^H^^r^^H •V^jH.

(c) With D C bias circuit 3 A x 1000 t

Curve-1 is the AC circuit current IAC, 2.5 x 140 V/d. i

Curve-2 is the voltage across the 100 t A C winding j

VFCL, 200 V/d.

Curve-3 is the voltage across the breaker VB, 10 V/d.

Curve-4 is the voltage induced in the D C winding, T

400 V/d.

J.l.^:l^^My^iy^Jy^ J 1

A:j:-ya'c'A:y A^M-AM—I

j'\ /'\..A. /iA. ./~"\. ./A 3

—": i/ F: r/ ' IM rr;

A C circuit: V Ac = 36 V (rms), and with a resistor in

series.

Figure 4-19. Short circuit currents with the F C L device

having different D C biases.

164

4.2.3 High Tc Superconducting D C Bias

In the saturable magnetic core FCL, a DC bias winding is required to generate

a high mmf on the magnetic core used. The well developed Ag-clad

(Bi,Pb)2Sr2Ca2Cu30io+x HTS wire provides a new opportunity for this type of FCL,

and substantially reducing operation cost by using liquid nitrogen rather than using

previous metallic superconductors with liquid helium. The HTS coil used is simply

introduced in this section; meanwhile more detailed discussions will be carried out in

Section 4.3.4.

4.2.3.1 The HTS coil

A HTS coil, used as the superconducting part of a hybrid DC bias winding,

has been made with a Ag-clad (Bi,Pb)2Sr2Ca2Cu30io+x 27-filament HTS wire,

produced through the normal "powder-in-tube" technique. The flexible wire allows

the HTS coil to be made by the "react and wind" procedures. The final HTS wire has

cross-section of 0.22 x 4.2 mm2, and CSHTS / CS(HTs+Ag) ~ 21%. Figure 4-20 shows the

HTS coil, which has dimensions §m = 93 mm, (J>out = 123 mm, 37 mm in thickness, and

total turns n = 1361.

Multiprobe measurements show the HTS wire used has fairly uniform Ic along

the complete length of 42 m, having 15 A average value over the entire length within

a variation of 10%. The short sample of this wire has a critical current Ic = 21 A (1

\xV, 77 K) at Ha= 0, and 10 A at Ha = 0.15 T, where Ha is the applied magnetic field.

Compared to the wire, the Ic is reduced in the final coil format by increased self

165

magnetic field. The maintained Ic by the H T S coil is about 4 2 % (6.3 A / 15 A) of the

original wire Ic value, with 15 mT self magnetic field and 0.86 kA-t at 77 K.

Figure 4-20. The H T S coil.

166

4.2.3.2 The H T S coil critical current ampere-turns and self magnetic field

The critical current Ic of the HTS coil is 6.3 A according to the 1 uV/cm

criterion at 77 K. Its critical current ampere-turn is about 0.86 kA-t at 77 K and self

magnetic field, as shown in Figure 4-21. The measured self magnetic field along the

coil inner wall has a value of 15 mT at centre point.

300-

250-

> 200-E

a 150 H CD ro

0.0

—i • ! ' r

o- Voltage drop

°-- Magnetic field

i a r

o-

0.2 0.4 0.6 0.8 1.0

Coil ampere-turns (KA-t)

o

1.2

20

CO

15 % 3

0) CQ

10 | o ^ i

Q.

5 3* H

0 1.4

Figure 4-21. V-H-IN characteristics of the H T S coil.

167

4.2.4 Device-II and its results

Another small fault current limiting device previously made was used to

conduct basic tests and to evaluate the ceramic HTS wire for building an electrical fault

current limiter [316]. It mainly consists of a pair of magnetic cores, normal copper

auxiliary windings, coolant containers and a HTS coil. Some additional tests with the

device are summarized in this section.

4.2.4.1 Configuration

A pair of laminated transformer iron cores ("C" core), which are commonly used

in electrical engineering laboratories, were selected. The magnetic core used has cross-

section 2.5 cm x 2.5 cm, and window 5 cm x 11 cm.

For test purposes a normal Cu enamel wire was used to wind coils for the device.

Cu wires were used to wind the AC coils, magnetic flux search coils and DC auxiliary

coils in this prototype. The arrangement of the Cu wire windings is (l)--Flux search

coil, <|> = 0.3 mm and 101; (2)--Auxiliary DC winding, <j) = 0.72 mm and 75 t; and (3)~

AC windings, <)) = 0.97 mm and 51 + 101 + 301, where <|) represents the wire diameter.

The DC bias consists of a HTS coil in a Dewar bottle. The HTS coils were

made from single core Ag-clad Bii.8Pbo.4Sr2Ca2Cu30io+x wire, which has cross-section

of 0.2 mm x 4 mm. Five pancake coils are used in series, and each has i.d. = 57 mm,

o.d. = 85 mm, and about 28 turns. The critical current Ic of the final coil is about 2 A at

77 K and self field [316]. A thermally insulated container was prepared to operate the

HTS DC coil with liquid nitrogen. The liquid nitrogen container is made with a

polythene cylinder with foam sheet coated inside for insulation against heat loss. The

168

size of the container is about 10.2 cm high and 10.8 cm for outside diameter. For the

comparison with normal copper wire windings, only one set of the magnetic core was

designed to use the H T S coil, for the other set of magnetic cores only normal copper wires

were used.

4.2.4.2 B-H curves

(a) D C B-H curve

The B-H curves of the magnetic core used for the device were measured and

shown in Figure 4-22. The testing circuit has a primary winding with 75 turns having

2.24 A (rms), and secondary winding with 10 turns. The R-C components used in the

secondary are Rs = 81.81 K Q and Cs = 2.33 uF.

The left curve is for one magnetic core. (Y: 0.05 V/d; X: 88 A-t/d).

The curve in the centre of the screen is for two magnetic cores in series.

Figure 4-22. B-H curves of the used magnetic cores.

169

(b) B-H curve with D C bias

The magnetic core B-H curve is changed in shape when a DC bias is added.

Figure 4-23 shows how the B-H curve of one of the magnetic cores used changes

with different DC biases applied. The testing circuit was the same as the arrangement

used in section (a), except for the applied DC biases.

1 ! !

4—-+-H—+-

Curve-1: IDC = 0.5 A .and nD C = 45 t.

Curve-2: IDC = 1 A and nD C = 45 t.

Curve-3: IDC = 5.28 A and nDC = 45 t.

Curve-4: IDC = 2.24 A .and nD C = 45 t.

5mV :&Y=C|.32DJ

«^,*r«'>'*^*.>>W;i«:,>^^.k^-.tf*EJ^

(Y: 0.05 V/d; X: 88 A-t/d)

Figure 4-23. B-H curves of the magnetic core used with different DC bias.

170

4.2.4.3 Tests on current limiting behaviour

The device with both positive and negative magnetic cores in series was

tested for current limiting behaviour. The windings of nAC = 45 t and noc = 75 t on

each of the magnetic cores were used, and the AC source voltage is expressed by

VAc- The DC current was supplied by a 12 V battery having a variable resistor in

series. A 4 Q resistor was in series with the AC circuit for normal operation and was

shorted for a short circuit situation. The AC circuit current was measured by using a

current transformer having a scale of 140 A/V.

To demonstrate the effect of the current limiting device, the AC circuit was

initially shorted without the device in the circuit, and the resultant current is shown in

Figure 4-24.

Next the AC circuit with the FCL device in place was shorted with a source

voltage VAc = 15 V (rms) with different DC bias currents. The AC circuit current and

the voltage across the device AC winding are shown in Figure 4-25. The AC circuit

current IAc and the voltage drop VFCL across the device is also shown in Figure 4-26

for VAC = 20 V (rms) and IDc = 3.5 A.

These tests are the same in principle as those mentioned in Section 4.2.2.5,

but in this case the complete connection was in place, both for limiting positive and

negative cycles. From Figure 4-25 and Figure 4-26, it is clear that the short circuit

currents can be effectively limited for both cycles, and the AC impedance is

controlled by the DC bias applied. Figure 4-26 shows that the AC voltage across the

device increases significantly when the AC circuit is shorted. This increased voltage

is related to the high impedance introduced to the circuit automatically by the short

circuit current.

171

1 "Ni^i".

T

>«-t'"T7r:'kT-

7

,0

* • • i

........

/i

i i

? 1 1 J.

(VAC = 1.5 V rms, shorting 4 Q RAC, 4 A breaker, and 40 x 0.14 A/d.)

Figure 4-24. Short circuit current without a FCL.

(a) IDC = 0.1 A. I U , j , i ! ft rt hi ft ft! (•. ft 111 II I'I A i\ \f\ I'I ft rt A Ai ft it

I /• ,'1 iA A A A AT A /I A ,'l j"i. \}\ i\ l\ A l\ \!\ A /'< 1,4ca™^Accu™,(1o»o.,4M,> ffw«-^

11/ v y; v V j v v *;. v ./ : v y v. -I,I y ; y -y IJJ y y ;

4 A A I A A A A A t A A A A AIA A A A A! A A i\ (D)IDC = O.5A. iv v v v VJV-V f v vjy v f v v|v v f v y;

The higher curve is the AC current (20 x 0.14 A/d) j

and the lower one is the voltage across the device

(40V/d)- \ F\ A h A n f\ A :§A A A A A iA A A A A \f\ A A T r T ; r v r r r r T r

:V v v; v V ;'.' v v.{. V V IV '1/ U! V V ;U V V| V V j

sA /\ i. .\ A! A A S n t\\h n \\ h h\h h A ft fil , XT HVjvvy v vi v vj/VVSUUI/V-JiDV! (c)iDC = i A . | | t j :j : r

The higher curve is the AC cunent (40 x 0.14 A/d) i==:^==::=i===^

and the lower one is the vohage across the device j \ 1 i \

(40V/d). IA A A A A IA A ft A AlA A A A AlA A f\ A f\\ \ •' v ; v v v.; v v .;.U V V ,J V ! V U V V V ! V V \i

172

(d) IDC = 2 A.

The higher curve is the A C current (40 x 0.14 A d )

and the lower one is the voltage across the device

(40 V/d).

(e) IDC = 5 A.

The higher curve is the A C cuiTent (100 x 0.14

A/d) and the lower one is the voltage across the

device (40 V/d).

'1 A A A Ai •'! A A •"

(f) Open D C winding.

The higher curve is the AC current (20 x 0.14 Ad)

and the lower one is the voltage across the device

(40 V/d).

M/V \A A.AAAAAAAAAAA/W v v:;:V V V V V:V V v' v u,V V

AW\/WVWWVW¥\^

Figure 4-25. Short circuit currents with the F C L device

having different DC bias currents.

173

(a) Short circuit.

The higher curve is the A C cunent (100 x 0.14

A/d) and the lower one is the voltage across the

device (20 V/d).

<' \ r \ .' \ i .' \ • /" \

: / ' • V J ' ' V' / ' ' V i .' ' ' V j > ' V

: i J

1 V

P \

1 i j

ii j t\ 1 i'\

/ i ' '( '

/ u ; / V 1/ • 1/ ':

j \ ;

v • ' • 1

/

.=. <v_ i-i r\ r\ 4 .-\ ' l •, . / \ . ' . / i t ; i

1 \ : 1 '. . 1 ( It • 1 I

; ; 1 / \ ; ; \ ; / \ \ 1 \ \ I \

; •• i i I ; | i :i 1 =1 1 :i i

v. SI > V \ i-

(b) Short circuit with a 2 Q

resistor added into circuit.

The higher curve is the AC cunent (100 x 0.14

A/d) and the lower one is the voltage across the

device (20 V,d).

•: "^^J' ' • "•> _.-•'" • "*• " >'' 1 '\__/' • I " A " ; "1' ' .' V /' •4. /

1 rh r:

1 / \ \ /

'• . • h ': /< : A .; A

fl : /! : ,'1 : H = f'i ! -t .' !. t '.; ( \: } i.: ) i..\

• >'. / :\ / :\ ! -\ ( :\ ,' •

_ _ -.-™ : — J l J :l J -1 .' 1| :i -'

'" •! .V ' \ -"' : \ > = \ --1 = 'l -•' : r ; 1 .' ; H : -i 1 11 •" II : \i 11 11

V A C = 20 V (rms) and IDC = 3.5 A

Figure 4-26. Short circuit current with the F C L device.

4.2.4.4 Other results

(a) Impedance characteristic

Figure 4-27 shows the voltage drops of the AC limiting winding, which were

measured against the AC current for different DC bias currents. The low impedance

range increases with the increased DC bias current. The rated AC current should be

set within this range for the normal condition. In Figure 4-27,1(s) is the magnetic core

saturation DC current. As a FCL, this impedance changes instantaneously with the

174

fault current as the A C fault current drives one of the magnetic cores out of

saturation.

I-V curves of the FCL prototype

Figure 4-27. Impedance of the F C L device.

(b) H T S coil loss

A HTS coil made with the Ag-clad (Bi,Pb)2Sr2Ca2Cu30io+x single core HTS

wire was used as a bias coil of a magnetic core [316, 317], and its I-V curves were

measured at 77 K with varying magnetic core gaps. The calculated resistive losses (I *

V) of the HTS coil and generated magnetic flux densities with the different core gaps

are shown in Figure 4-28. In Figure 4-28 a magnetic ferrite core with cross-section

(CS) = 25 x 25 mm2, and magnetic path length (MPL) =135x2 +75 x2 mm, and a

175

saturable flux density about 1.3 T was used. A search coil on the magnetic core was

linked with a flux meter for the test. Figure 4-28 shows that for different core gaps

there is a H T S coil low loss region to generate the magnetic fields prior to the losses

increasing dramatically. For the closed magnetic core used, it can be saturated by the

H T S coil under its superconducting state, and there is a dramatic loss increase in the

H T S coil above 1.35 T.

8000-

g 6000-

0) £ 4000-O

HTS c

oil

O O

O

O

Magnetic core with singl e core HTS coil

at77K

if T in; inn o

i

0.0 0.2

D

D

D

A°

i |

0.4

•

•

0

A

O 0

A

1 1 '

0.6

° A

-With 1.8 mm core gap

-With 0.4 mm core gap -Without core gap

o

A

A

A A A AAA7V>2F

1 • , • , • , • 1 0.8 1.0 1.2 1.4 1

Magnetic core flux density (T)

Figure 4-28. Losses of the 5-unit pancake H T S coil at 77 K

with different magnetic core gaps.

4.2.5 Summary of Tests

A Ag-clad (Bi,Pb)2Sr2Ca2Cu30io+x H T S wire has been used to make a H T S

coil and the sample coil has been tested with regard to its critical current, value of

critical current ampere-turns, and generated magnetic field. The H T S wire can be used

176

to prepare a H T S coil under the react and wind techniques, and operated in liquid

nitrogen. The practical application requires a consideration of the Ic-B-T relation of

the HTS wire and its coil. The sample coil has 0.86 kA-t at 77 K having 15 mT

generated self magnetic field at the coil inner wall centre point and having a total

conductor critical current density of about 690 A/cm2. At 77 K the HTS coil

maintains 42% (6.3 A / 15 A) of the initial HTS wire critical current.

To investigate the device current limiting properties, tests arranged mainly

include measuring I-V characteristics of the device AC winding with different DC

biases. Effects of changing DC bias currents, open DC winding, short DC winding,

different short circuit angles, transformer effect influence on the DC winding, and

circuit breaker are investigated.

Conclusions which can be drawn from the tests relative to the device current

limiting characteristics are (a) The AC winding reactance decreases as the DC bias

value increases until the magnetic core is deeply saturated; (b) The required DC bias NI

is related to the rated AC winding current and given by Equation 4-2; (c) The AC

winding impedance is automatically increased and limits the fault currents during the

first half cycle of the fault current; (d) The AC winding has much higher impedance

when the DC winding is open circuited, and this proves that the device has much higher

impedance when the AC current drives the magnetic core out of saturation at fault

conditions; (e) The device has low impedance when its DC winding is shorted; (f) The

device limits fault currents effectively, and this can be substantiated by comparing the

results from one magnetic core with the second magnetic core in series; (g) Some AC

components will be induced in the DC bias circuit through the transformer effect, and

high additional DC circuit reactance will limit this influence.

177

4.3 ANALYSIS OF THE SATURABLE MAGNETIC CORE FCL

4.3.1 Operation Principle and \|f-I Characteristic of the FCL

4.3.1.1 Modified y-I curve

The saturable magnetic core FCL is analysed with respect to its magnetic

flux-linkage characteristics. Curve 1 in Figure 4-29 shows flux-linkage x\r versus

winding current I for a single magnetic core without DC bias (IDC = 0). The flux

linkage is represented approximately by two straight lines with different gradients.

When \\\f\ < \|/i, the magnetic core is normal, and this produces a higher AC reactance

Li = tgcti; when |\j/| > \|/i, the magnetic core becomes saturated, the AC inductance

equals the smaller gradient which is equivalent to an air core inductor, i.e. L2 = tga2,

where (Xi and cc2 are as defined in Figure 4-29.

The magnetisation curves of the fault current limiter are also shown in Figure

4-29. Curve 2 and 3 represent the two magnetic cores to which the DC biases have

been applied. Curve 4 represents total magnetisation of the fault current limiter and is

the combination of the curves 2 and 3. Curve 4 represents the \}/-I relationship of a

non-linear inductor which can be simplified to three linear sections with different

gradients.

4.3.1.2 Section I of \|/-I curve

In Section I (a - a') of the \|/-I curve 4, there are |\|/| < \|/a or |I| < Ia = (IDc - Ii)

which are required for the normal rated current value. The two magnetic cores are

both saturated by the superconducting DC bias windings; therefore the inductance La

178

of this fault current limiter is equal to the sum of two air core A C winding

inductances

La = 2L2 = tga2 (4-6)

Figure 4-29. \|/-I curve of the F C L .

179

4.3.1.3 Section II of \|/-I curve

When the fault current occurs and increases to reach \|/a < |\j/| < \|/b, or Ia < |I| <

lb = (IDC + Ii), the two magnetic cores come out saturation alternately in each half

cycle. Therefore the FCL behaves like a high transient reactance Lb to the fault

current, and Lb is

Lb = Li +L2 = tgcci + tgcc2 (4-7)

4.3.1.4 Section III of \|/-I curve

If the fault current is not totally limited during the region marked II (a - b, or

a' - b'), it will increase to |\|/| > \j/b, or |I| > lb, and drive the magnetisation curve to be

in region HI. The reduced reactance in this section can not effectively limit the fault

current further. The inductance Lc of the FCL during the fault current in Section III is

Lc = La = 2tga2 (4-8)

4.3.2 Current Limiting Behavior

The FCL is also analysed from derivation of current waveforms through the

\|/-I curves of this type of HTS-FCL [318,319].

180

Figure 4-30 is used to analyse the short circuit current waveforms and the

fault current limiting effect. When the AC circuit is shorted, the voltage across the

FCL is equal to the source voltage, therefore

dy / dt = VmCos(©t + a) (4-9)

where a is the switching angle of the voltage at the instant of short circuit. Assuming

t = 0,1 = 0, the above equation can be integrated and

i|/ = \j/m[Sin((Dt + a) - Sina] (4-10)

where v|/m = VJ(o. If the short circuit occurs at a = 0, therefore

v|/ = v|/mSino)t (4-11)

The dashed line in Figure 4-30 represents the periodic waveform derived from

Equation 4-11 and \|/-I curves for different samples (as also shown in Figure 4-35). In

Figure 4-30, there are samples S2 to S6 selected for the FCL. Since their v|/-I curves

satisfy V|/m / vj/b < 1, and the changes of their vj/ are limited in Section I and II as shown

in Figure 4-29, so the optimum limiting result can be obtained for the periodic

component of the short circuit current, and the result is similar to the simulation result

obtained (from Figure 4-39). The flux linkages from Equations 4-10 and 4-11 are the

integrated values of the voltages across the FCL, and they are only the apparent and

equivalent flux linkages of the FCL.

181

From Equation 4-10, when the short circuit occurs at a = -90°, the F C L flux

linkage increment is maximum

vj/ = V|/mSin(cot-7t/2)+v|/m (4-12)

From Equation 4-12 the maximum flux linkage vj/m appears after half a cycle (i.e. 0.01

s), in which case

M/max = 2ij/m (4-13)

The solid lines in the Figure 4-30 represent the figure analysis results of the

short circuit inrush current waveforms for S2 to S6, and these are similar to the

results shown in Figure 4-38. From Figure 4-30, when v|/m / \j/b > 0.5 (i.e. \j/b < 2v|/m =

2Vm/o) the top part of the vj/ waveforms will be into Section III where the V|/-I curve

gradient is smaller.

182

(Wb-t)

(b)/ \

(a)/ / \ \

/ / \

0° 180°

(a) Fault at a = Q °

(b) Fault at a=-90°

30-

20 -

10-

CD o

ao -CD O

SQ^^SS^" |B6 ^A ^

(B5 - ^ .,

M W ^^"^

(B3 ^ -"""

lA'

B2

0 10

1* 1 ' 1 1

/ /

J

S2-S6~ -

36

S*^^ 83^- S2

i (kA)

20 30

"*—»„S1

^ /

S2

Figure 4-30. Derivation of current waveform of six samples.

183

4.3.3 Magnetic Circuit Property

4.3.3.1 Winding relation and induced AC voltage

A saturable magnetic core reactor is a control device which consists of a

ferrite-magnetic core having two windings, one carrying alternating and the other

direct current. The AC winding is connected in series with an AC supply and a load,

the current through which is to be controlled. By varying the current in the DC

winding, the AC impedance of the circuit can be changed and the load current

controlled. The polarity of the two reactor AC windings is set to limit fault currents

in both positive and negative cycles.

The FCL AC impedance is controlled by a DC bias winding which is made

with the HTS wire. The DC control winding current is limited by the HTS DC

winding which has critical current Ic, and its relation with the device windings by

IDC = IACIIAC / noc < Ic (4-14)

where IAC^C stands for the value of AC ampere-turns per core. IDcnDC represents the

control ampere turns per core. The magnetic field generated with the HTS DC coil is

proportional to IDCIIDC ampere-tums and is related to its configuration.

The voltage induced by the fault current in the AC load winding is estimated

using Faraday's law V = ndO/dt, which gives

V(rmS) = 4.44 f nAC S ABmax (4-15)

184

where f is the system frequency, nAc is the number of turns of the A C limiting

winding, S is the magnetic core cross-section, and ABmax is the maximum change of

the magnetic core flux density.

4.3.3.2 Magnetic core losses

For the AC winding related to the magnetic circuit, the core power loss

dissipated in the core, mainly consists of hysteresis and eddy current losses.

Hysteresis loss Ph (watts) is the energy used to align and rotate the atomic magnetic

moments of the core material and can be written as [320]

Ph = V f (0.4 Tt)-1 10-8 J H dB (4-16)

where V (cm3) is the volume of the core. H is in oersteds and B in gauss. Eddy

current loss is caused by circulating currents in a conductive magnetic core and its

average power loss Pe (watts) value is [321]

Pe = V f2 TC2 r B2max (6 p)"

1 (4-17)

where T is the thickness (m) of the individual slab or lamination, p is volume

resistivity (Q/m3), f is frequency (Hz), and Bmax is sinusoidal flux density peak value

(Wb/m2).

Since the magnetic core used is well saturated for normal operation, the Bmax

for rated AC current can be simply estimated by

Bmax = 2u0(IRnAC-lDcnDc)/MPL (4-18)

185

4.3.3.3 Control-circuit impedance and transient response

A general analysis can divide the problem into two cases; one in which the

control-circuit impedance is assumed to be zero, and the second in which it is

assumed to be infinite. The latter case is subdivided into one in which the load has

zero impedance and a further case in which it is finite.

For an applied voltage VmSin((Ot + a), the total change in flux density for the

case one during the period 0 to o/co is [322]

AB = (Vm / 2conDCA)( 1 - Cosa) (4-19)

and the transient response which depends on the change in the mean flux density, is

evaluated from the control circuit time constant x [322]

x = (l/4f)(nDc/nAc)2(RAc/RDc) (4-20)

To simplify the problem an assumption is made for the simulation study. The

control voltage and impedance are assumed to be very large so that the control

current remains constant despite the effects of the transformer action through the

saturable reactor.

4.3.4 DC Bias and HTS Winding

4.3.4.1 Design of the DC bias

186

In the saturable magnetic core FCL, a D C bias winding is required to generate

a high mmf on the magnetic core used. When a simple operation is considered where

there is no shielding screen for the DC coil, the required DC bias can be estimated

based on the magnetic core B-H curve by [323]

iDcnoc = IumitiiAC + HSMPL (4-21)

where the IDC is DC bias current, noc is the number of turns in the DC winding, Iijmit

(peak) is the AC fault current value at which the device starts to limit the increased

AC current, nAc is the AC winding turns, Hs is the saturable field of the magnetic

core, and MPL is the magnetic path length.

The limiting behaviour or the required AC impedance of the FCL device is

related to the voltage drop across the AC limiting winding, which is proportional to

nAc, and this has been evaluated from Equation 4-1.

Based on the magnetic core B-H curve, the DC bias current IDC can be set

according to the rated AC transport current Uc (rms) [323]

(nDClDc-V2 nACIAc)/MPL=Hs+AH (4-22)

where noc is the number of turns in the HTS DC winding and MPL is the magnetic

path length. Hs is the saturable field of the magnetic core. AH as a coefficient is used

to design the starting limiting current level Iumit (rms) of the prototype, and it is

related to Iumit by [323]

187

AH=V2(Ilimit-IAc)nAC/MPL (4-23)

4.3.4.2 HTS DC bias winding analysis

(a) Magnetic field distribution

The scheme used for the analysis of the HTS coil magnetic field is shown in

Figure 4-31, where the line 1 represents the position used to co-ordinate the magnetic

field. The HTS coil centre is co-ordinated at the point of (X=500 mm, Y=500 mm).

The magnetic field distribution around the HTS coil is then calculated for 77 K

operation with a maximum critical transport current of 6.3 A, and along the typical

line 1 as shown in Figure 4-31.

Figure 4-32 shows the magnetic field values of the HTS coil without the

magnetic core, and along line 1 in Figure 4-31. In Figure 4-32 Bx and By are the X

(HTS coil axial direction) and Y (HTS coil radial direction) components of the

magnetic field generated by the HTS coil.

Figure 4-31. The scheme used for analysis.

188

(b) Magnetic field distribution with added magnetic core

Figure 4-33 shows the magnetic field values B, Bx and By along the same line

as shown in Figure 4-31, but with the magnetic core in position. The magnetic core

used has dimensions as specified in Section 4.2.1.

350 400

L (mm)

Figure 4-32. The magnetic field of the HTS coil along line 1

without the magnetic core.

650

1.8

1.6:

0.006'

P ^" 0.004 >. w 0.002 c ~° 0.000 x JD U- -0.002 -0.004

-0.006

1

DDDDDD

• OA QH.D D

.o. 'A.° •. •

•

,°aD°Dci po.rj

A q •Ag •°°°a

A

A

...-•..

— o ••

....A-.

•B •Bx

•By

b o o

A'

A 'A A A A

8

400 450 550 500

L (mm)

Figure 4-33. The magnetic field of the HTS coil along line 1

with the magnetic core.

600

189

Flux Density

T)

10

8

6

4

2

0

-2

-4

•e -8 10

(a) Magnetic field density Bx.

Flux Density

B, { 10*

1 .

====

T)

10 8 6 4 2 0 -2 -4 -6 -8 -10

(b) Magnetic field denaty By,

Figure 4-34. The magnetic field map around the corner of the H T S coil

with the magnetic core applied.

190

4.3.4.3 Application of the H T S wire for the F C L

The HTS wire has been considered to form a DC bias winding for the FCL. A

coil can be made by a Ag-clad (Bi,Pb)2Sr2Ca2Cu3O10+x multifilament HTS wire,

through a react and wind procedure. The critical current degeneration due to forming a

coil, with this flexible HTS wire, is mainly caused by increased self magnetic field

and its distribution in the HTS wire. The DC bias winding magnetic field distribution

is the most important factor to consider the application of the anisotropic Ag-clad

(Bi,Pb)2Sr2Ca2Cu30io+x HTS wire having strongly magnetic field-dependant critical

currents at 77 K. The HTS coil magnetic field distributions, with and without an

added magnetic core, have been analysed. The magnetic field generated by the HTS

coil is re-distributed by the added magnetic core. Since the main flux is bounded

within the magnetic core, the influence on the HTS coil critical current is not

significant. The radial field By at the HTS coil side edge, which corresponds B//ab for

the HTS wire, determines the HTS coil performance at 77 K. For the application of

building an electrical fault current limiter, the AC loss of the HTS wire has to be

considered since there is a transformer effect and induced AC currents superimposed

on the DC current. According to the previous tests, the 50 Hz AC transport current has

a low loss when its rms value is less than the HTS wire DC critical current Ic. From

the HTS wire tests of k-B-T and irreversibility line, the present Ag-clad

(Bi,Pb)2Sr2Ca2Cu30io+x HTS wire can only be used to design a coil whose magnetic

field is less than 1 T at 77 K. Reduction in the operation temperature of the HTS DC

bias winding has to be considered as a solution based on present technology.

191

4.4 CURRENT LIMITING BEHAVIORS OF THE HTS-FCL

4.4.1 FCL Samples and Equivalent Operation Circuit

Six samples are selected for the analysis, as shown in Figure 4-35. Sample 1

(SI) is the conventional air core linear inductor. Samples (S2 - S6) have different i|/-I

curves and different \\f\, or Ib values which are defined by Figure 4-29. They have the

same inductance in three sections of the v|/-I curves as shown in Figure 4-29, and in

the Section I and HI they have the same inductance as the conventional linear air core

inductor (SI).

c s _

CD .Q CD

CO Oi ro

c x

3 4 5

Current (kA)

Figure 4-35. \|/-I curves of the selected six samples.

192

Figure 4-36 represents a per-phase equivalent circuit of the fault current

limiter in a three phase system. This 6.3 kV infinite system has a voltage peak value

Vm = (^2 x 6300) / V3 = 5143.9 V. The switch S is to simulate a short circuit on one

of the phases.

FCL

rvpt^r\-

© V Vmcos(cot+a)

(R = 2.7434 Q; L = 6.5495 m H )

Figure 4-36. Equivalent circuit for simulation.

4.4.2 Operation of the F C L

4.4.2.1 Normal operation

When the switch S in Figure 4-36 is open, a load is applied, and the circuit

has the equivalent values as shown in the figure. The maximum rated current i(t) and

193

the corresponding voltage drop across the F C L Au(t) were calculated and s h o w n in

Figure 4-37. T h e results s h o w that: (a) For the samples SI to S 6 from Figure 4-37,

the curves i(t) and Au(t) are the same for each sample. Because the F C L is operated

in Section I of Figure 4-29 during normal operation, the six samples have the same

transient inductance L a = \j/a / Ia (= 0.5459 m H ) ; (b) T h e peak value of the voltage

drop across the F C L A u p = 2 5 0 V for the six samples. T h e voltage drop due to the

F C L in the transmission line is A u p % = 250/(^2x6300) = 2.8%, which agrees with

the calculated result for a conventional linear inductor reactor.

o.oo 0.01 0.02 0.03

t (s)

0.04 0.05 0.06

Figure 4-37. N o r m a l operation of the six samples.

194

4.4.2.2 Short circuit

For a short circuit as the applied voltage passes through zero, which is

expected to be the worst current limiting situation, the fault currents from the

maximum inrush current to the steady state values are shown in Figure 4-38, in

which samples SI to S6 are used. A small circuit resistance (~ 20 mQ) was used to

damp the DC component of the waveform.

From Figure 4-38, it is clear that the FCL has the advantage over the

conventional air core linear reactor by reducing fault current levels significantly. The

FCL is a non-linear inductor device, the limited short circuit current becomes

seriously distorted; therefore its short circuit inrush coefficient K = Is / Ip is greater

than 2, where Is is the maximum short circuit inrush current, and Ip is the peak value

of the short circuit periodic component. A normal linear reactor has 1 < K < 2.

195

6.0x10

4.0x104 -

2.0x10

-2.0x10* -

-4.0x10 0.05 0.10 0.15 0.20 0.25 0.30 0.35

t (s)

Figure 4-38. Short circuit currents for different samples.

196

4.4.2.3 Periodic components of short circuit currents in the steady state

When the short circuit occurs at the source voltage at peak value (i.e. a = 0°),

the free (DC) component of the short circuit current is zero; therefore the periodic

component can be obtained directly from the short circuit current curve.

Figure 4-39 shows the short circuit current periodic component waves for

different FCL samples. According to Figure 4-39, the HTS-FCL limits the fault

current periodic component efficiently. The periodic components of the fault current

curves with different samples S2 to S6 are about 1/10 of the value with the linear

reactor. The results shown in Figure 4-39 are similar to the periodic component from

the figure analysis shown in Figure 4-30.

4.0x10

2.0x10

-2.0x104 -

-4.0x10 0.280 0.290

Figure 4-39. The periodic component of the limited short circuit currents.

197

4.4.2.4 M a x i m u m inrush current

For a linear reactor (SI) without (negligible) resistance, the maximum short

circuit inrush current Isi is obtained when the short circuit occurs when the source

voltage passes zero, and Isi = Ki x Ipl = 2 x 28.443 = 56.886 kA.

The inrush current waveforms of S2 to S6 are shown in Figure 4-40. The

relation of the inrush currents, inrush coefficients K and \j/m / \|/b were obtained and

shown in Figure 4-40. The K is the ratio of the Is of a sample to the short circuit

periodic current peak value Ip = 2.756 kA. The per unit maximum inrush current Is* =

Is / Isi is based on the inrush current of a linear reactor Isi.

. . i . - i " « •

5 0.020 0.025 0.030 0.035

t (s)

Figure 4-40. Inrush current waveforms.

198

30

25-

e. 20

c CD

§ 15 o I 10-1

S6

0.5

S5

0.6

S4

0.7

S3

K(X)—I<x)/2.8

— i —

0.8 0.9

T m / T b

S2

1.0

Figure 4-41. The relationship between inrush current and \|/m / \|/b.

199

4.5 POWER SYSTEM APPLICATION OF THE HTS-FCL

4.5.1 Nomenclature

Definitions of the symbols used in this section are summarised as follows.

(a) Currents

IN -Normal load current through HTS fault current limiter (FCL)

IR -Rated current of air-core inductor

IP -The periodic component of short-circuit currents

Is -Short-circuit current in steady state

Ippu -The p.u. value of the periodic component of short-circuit currents

IRS -Inrush short-circuit currents

IK -Short-circuit current for the duration of short circuit

(b) Voltages

VN -Normal operating voltage (Line rated voltage)

VR -Rated voltage of equipment

V^ -Rated voltage of air-core inductor

V^ -Rated voltage of HTS-FCL

VRG -Rated voltage of generator

VRTI -Rated voltage of transformer (T-1)

VB -Base value of voltage

AVa -Voltage loss of air-core inductor

200

AVf -Voltage loss of HTS-FCL

Vp -Busbar (6.3 kV) residual voltage during short circuit

Vy\R -Average rated voltage of the network

VT(%) -Nominal short-circuit voltage (%)

(c) Powers

PT -Rated short-circuit power loss of transformer

PR -Rated capacity of generator

SR -Rated capacity of transformer

S -Load capacity of 6.3 kV line (S2 = P2 + Q2 )

SRG -Rated capacity of generator

SB -Base values of power capacity (assuming SB = 100 MVA)

AQa -Reactance loss of air-core inductor

AQf -Reactance of HTS-FCL

(e) Resistance, Reactance and Impedance

RL -Unit resistance

XL -Unit reactance of transformer line

XRJ -Reactance air-core inductor

XR% -Reactance percentage of air-core inductor

XRf -Reactance of HTS-FCL

Xj" -Subtransient reactance of a generator

RK -Magnitude of the system short-circuit resistance

XK -Magnitude of the system short-circuit reactance

201

Z K -Magnitude of the system short-circuit impedance (ZK = R K + jXK)

(f) Various

KT -Nominal transformer ratio on main tap

Tg -DC time constant (= XK / G>RK = LK / RK)

VJ/ -Switching angle of voltage at the instant of short-circuit

(PK -Phase angle of the system short-circuit impedance

Coscp -Power factor

p.u. -Per unit value (pu)

4.5.2 Electrical Power System

The saturable magnetic core type high Tc superconducting (HTS) electrical

fault current limiter (FCL) is considered for use in an electrical power transmission

and distribution network as illustrated in Figure 4-42. The HTS-FCL is analysed in the

system, both for normal operation and short-circuit conditions, and it is also

compared with the normally used air-core linear inductor.

As shown in Figure 4-42, the fault current limiting device is connected in a 6

kV distribution line. The main parameters of the power system used, as shown in

Figure 4-42, are given in Table 4-2.

202

£> 10kV

Generator

110 kV HOkV

Transmission line I] T-2 •] R=0.17Q/km

10.5 kV/125 M W 121/10.5 kV X=0.4 C2/km cos<r0.8, Xd"=0.125 (p.u.) 120 M V A

6kV

•l T-l 110/6.3 kV 15 MVA

H FCL

X%=5 (Air core inductor)

Load

Figure 4-42. An electrical power network.

Table 4-2. Power system parameters.

Generator

Transformer-1

Transformer-2

Transmission Line

Air Limiter

H T S Limiter

PR(MW)

or SR (MVA)

125 M W

15 MVA

120 MVA

VR

(kV)

10.5

110/6.3

121/10.5

110

6

6

IR

(kA)

0.3

Coscp

0.8

Xd"

(p.u.)

0.125

PT

(kW)

120

450

VT(%)

10.5

10.5

203

4.5.3 System during Normal Operation

4.5.3.1 Equivalent circuit of the power system

The equivalent circuit of the power system during normal operation is shown

in Figure 4-43, which is calculated based on the 6.3 kV voltage level as shown in

Figure 4-42. The calculation results for the equipment in the network are summarised

in Table 4-3.

Table 4-3. Resistance and reactance of the network devices.

Transformer T-l

Transformer T-2

Transmission line

Air core inductor

HTS-FCL

Resistance (Q/phase)

0.021

0.0015

0.0223

0

0

Reactance (H/phase)

0.2778

0.042

0.0524

0.577

0.02

204

0.0015+J0.042 0.0223+J0.052 0.O21O+JO.278 jO.606 (Air inductor) jO.021 (HTS-FCL)

G Load

Generator Busbar Fault current limiter

Figure 4-43. Equivalent circuit for the power system.

For the air core inductor fault current limiting device in a 6.3 kV system, the

reactance percentage XR% = 5 is normally selected, and there is [324]

X R % = V 3 I R X R a 1 0 0 / V r (4-24)

therefore

XR* = V R X R % / (100 IRV3) (4-25)

= 6000 x 5 / (100 x 300 x V3) = 0.577 Q/phase

From the previous study on a small FCL device, if the reactance of the HTS-

FCL is selected as 29.43 times of the air-core inductor [319], therefore for the HTS-

FCL there is XRT = XRa x 29.43 = 16.98 H/phase.

205

4.5.3.2 Voltage drop and reactive power loss in normal operation

For the power system with different loads on the 6.3 kV busbar, the voltage

loss and reactive power loss are calculated as shown in Table 4-4.

Table 4-4. Reactive power loss and voltage drop across the fault current

limiting devices on 6.3 kV busbar in normal operation.

Load

S ( M V A )

0.2+0.12j

0.4+0.247J

0.6+0.372J

0.8+0.495J

1.0+0.62J

1.2+0.74J

Normal current

IN (A)

22.4

45.24

67.93

90.524

113.22

135.66

Voltage drop A V

(V)

Air-core inductor

AV a

11.54

23.75

35.77

47.60

59.62

71.16

HTS-FCL

AVf

0.40

0.82

1.24

1.65

2.07

2.47

Reactive power loss A Q

(KVAr)

Air-core inductor

AQa

0.87

3.54

7.98

14.19

22.19

31.86

HTS-FCL

AQf

0.03

0.12

0.28

0.49

0.77

1.11

206

30-

20-

10-

-

0-

i •

cf

— i — i —

i ' i

, • •

a

...A'"'

. • : : : : : $ : : : : : : : : : : :

1 ' 1

i g

fl'

....Q.-

...-0--

....A--

•—v--

.A'

— ,

i ' i > i ' i

FCL in normal operation . A

-airV • FcIV

• air Q • FclQ

A'

.•'

A' y

•

-

o v

T '

::::::::v v

1 ' 1 ' 1

20 40 60 80 100

Current (A)

120

40

30

- 20

- 10

- 0

140

ure 4-44. The change of AVa, AVr, AQa and AQf with different load currents.

207

4.5.4 System Under Fault Condition

4.5.4.1 Equivalent circuit for short circuit fault condition

It is assumed that a three phase short circuit occurs at the site of the fault

current limiting devices, and the system has 100 MVA base value of capacity, and

base values of voltages [VB = (1 + 5%)VN)] which are the average rated voltage of

the network at different levels, i.e., 6.3 kV, 10.5 kV, and 115 kV.

The p.u. (per unit) values are used which are based on a base value of capacity

(MVA), SB = 100 MVA, and base values of voltage (kV), VB. The system is

represented by p.u. values as shown in Figure 4-45.

Load

X1=0.08 X2=0.0875 X3=0.121 X4=0.7 Xa=1.4547 Xf=42.78

Figure 4-45. Equivalent circuit for short circuit condition.

4.5.4.2 Steady state short circuit current, 6.3 k V busbar residual voltage and

inrush current

(a) With air core inductor

As shown in Figure 4-45, the total reactance, Xz (p.u. value), from source to

short circuit point, under the short circuit condition of the system with the air-core

inductor is

208

XI = X, + X 2 + X 3 + X 4 + X 5 (4-26)

= 0.08 + 0.0875 + 0.121 + 0.7 + 1.4547 = 2.4432

As an example the short circuit occurs at t = 0 s, the p.u. value of the short

circuit current periodic component is Ipu = 0.418 given by the XSa - Ipu curve.

Therefore the short circuit current is

Is = IPu x SRG / (VR V3) (4-27)

= 0.418 x (125/0.8) / (6.3 V3) = 0.418 x 14.319 = 5.985 kA

The residual voltage, p.u. value, at 6.3 kV busbar VPpu = IPpu x X5 = 0.418 x

1.4547 = 0.608; therefore at t = 0, the 6.3 kV busbar residual voltage is

VP = VPpu x VB = 0.608 x 6.3 = 3.83 kV (4-28)

Since the short circuit occurs at the distribution line of the system, it is far

away from the generator of the system, and the evaluated inrush current is IRS = 2.55

x IP, i.e. IRS = 2.55 x 5.985 = 15.26 kA.

Table 4-5 shows the 6.3 kV level short circuit current periodic components Ip

(kA), and remaining busbar voltages during the fault condition Vp (kV), which are

calculated versus the duration of short circuit time.

209

Table 4-5. The periodic component of short-circuit currents and residual

voltages under fault conditions with air core inductor.

Duration

of short circuit

(s)

0

0.01

0.06

0.1

0.2

0.4

0.5

0.6

1

2

4

Periodic component

of short circuit current

Ip(rms) (kA)

5.985

5.957

5.785

5.856

5.762

5.898

5.985

6.01

6.01

6.01

6.01

6.3 k V busbar

residual voltage

V P (kV)

3.830

3.812

3.740

3.748

3.688

3.775

3.830

3.846

3.846

3.846

3.846

(b) With HTS-FCL

For the system with HTS-FCL, Xf = 29.43 x 0.577, and the p.u. value is

XRfpu = XRfSB/VB2 (4-29)

= 16.98 x 100/6.32 = 42.78

Therefore the total p.u. reactance of the shorted circuit is X^ = 0.08 + 0.0875 + 0.121

+0.7 + 42.78 = 43.77.

Since X^ > 3.45, the value of the short circuit periodic component Ip can be

assumed not to change with time. Consequently the p.u. value of the short circuit

periodic component is

Ippu = 1/ X1 (4-30)

= 1/43.77 = 0.0228

and Ip is

IP = IPPUXSRG/(VBV3) (4-31)

= 0.0228 x (125/0.8)/(6.3 V3) = 0.326 kA

The p.u. voltage of the periodic component Vppu is

VppU = IppUxXfp„ (4-32)

211

= 0.0228 x 42.78 = 0.975

Therefore the 6.3 k V busbar residual voltage is

VP = VPpu x VB (4-33)

= 0.975x6.3 = 6.14 kV

The evaluated inrush current IRS is IRS = 2.55 x IP = 2.55 x 0.326 = 0.831 kA.

Table 4-6 shows the comparison of the steady state values by using the air

core inductor and the HTS-FCL.

Table 4-6. Comparison of using the air core inductor and the HTS-FCL.

Short circuit current

in steady state Ip (kA)

Busbar residual voltage

V P (kV)

Inrush current

IR(kA)

Air core inductor

5.985

3.83

15.26

HTS-FCL

0.326

6.14

0.831

212

6.0

5.5

<

c

§ | ° £ 3 | o E ? 8 3.9 o o ^ ? 3.8

i_

a> a.

3.7

3.6 0.00

• Periodical component of short circuit current * Busbar residual voltage during short circuit

With an air-core inductor

0.25 0.50

t (s)

0.75 1.00

6.0

5.5

3.9

3.8

3.7

3.6

CD c c/> CT CD -1

—( CD

w CL

c cu_ < o cu CQ CD

<

Figure 4-46. Periodic component of short circuit current and 6.3 k V busbar

remaining voltage under the short circuit condition.

4.5.4.3 P o w e r factor and D C time constant in short circuit transient state

It is assumed that the source voltage V = (l/V3)V2VBSin(cot + i|/) kV/phase,

the three phase short circuit occurred at \|/ = 0, and without load before the short

circuit. From Figure 4-43, the system parameters are calculated and results are shown

in Table 4-7.

213

Table 4-7. System parameters of short circuit in transient state.

Total resistance R K (.Q/phase)

Total reactance X K (Q/phase)

Impedance angle

(pK = arctgXK/RK

D C time constant (s)

T g = LK/RK = XK/(314RK)

Air Core Inductor

0.0448

0.9492

87.3°

0.0675

HT S - F C L

0.0448

17.35

89.85°

1.233

4.5.4.4 Short circuit current

The relation of instantaneous values of the short circuit current IK and the

period of short circuit t is given by [324]

/^^^-Si^^+V^-^J-^^-Sin^-^)-^ (4"34)

where VAR is the average value of rated system effective voltage (rms), \|/ is the

switching angle of the voltage V at the instant of the short-circuit, <pK is equal to

arctg(XK/Rfc), and Tg is DC time constant.

Therefore the short circuit current with the air core inductor is

214

IK = 5.413Sin(1007ct 1807TT, - 87.3°) - 5.413Sin(- 87.3°) e Os -t/0.0675 (4

The short circuit current with the HTS-FCL is therefore

IK = 0.296Sin(1007it 180°/7C - 89.85°) - 0.296Sin(- 89.85°) e ON -t/1.233

(4

12-r

10

8

6-

4-

2-

0-

-2-

-4-

-6

Air-core inductor

HTS-FCL

-, . 1 1-

t (ms)

~0~~10 20 30~ 40 50 60 70 80 90

Figure 4-47. Short circuit currents with the HTS-FCL

and the air core inductor.

215

4.5.5 Appendix

A. System equivalent circuit for normal operation

(a) Transformer T-l

Rn = PTI VRTI2 / [1000 SRi2] = 120 x 6.32 / [1000 x 152] = 0.021 Q/phase

Xn = UTi VRTI2 / [100 SRI] = 10.5 x 6.32 / [100 x 15] = 0.2778 Q/phase

(b) Transformer T-2

RT2 = Pr2 VRT22 (I/KT2)2 / [1000 SR22]

= 450 x 1212 (6.3/1 IO)2 / [1000 x 1202] = 0.0015 Q/phase

Xn = VTI VRTI2 (l/K^)2 / 100 SRI

= 10.5 x 1212 (6.3/1 IO)2/ [100 x 120] = 0.042 Q/phase

(c) 110 kV transmission line

RL' = RL x (I/KT2)2 = 0.17 x 40 x (6.3/1 IO)2 = 0.0223 Q/phase

XL' = XL x (l/K-n)2 = 0.4 x 40 x (6.3/1 IO)2 = 0.0524 Q/phase

(d) FCL device

XR% = A/3 IR XRa 100 / VR

XRa = VR XR% / 100IR V3 = 6000 x 5 / [100 x 300 x V3] = 0.577 Q/phase

216

B. Voltage drop and reactive power loss in normal operation

As an example, the calculation for the load S = 0.2 + 0.12j MVA is given as

follows.

(a) Normal rated current

IN = S / (VN V3) = V [0.22 + 0.122] / (6V3) = 22.4 A

(b) Voltage loss

(i) Air-core inductor

AVa = (PRRa+ QXRa) / V = (0.2 x 0 + 0.12 x 0.606) / 6 = 12.12 V

(ii) HTS FCL

AVf = (PRRf + QXRf) / V = (0.2 x 0 + 0.12 x 0.02) / 6 = 0.4 V

(c) Reactive power loss

AQa = (P2 + Q2) XRa/ VN

2 = (0.22 + 0.122) x 0.577/62 = 0.87 kVAr

AQf = (P2 + Q2) XRf / VN

2 = (0.22 + 0.122) x 0.02 / 62 = 0.03 kVAr

C. Equivalent circuit (p.u.) for fault current

(a) The reactance p.u. values for the generator

Xt = Xd" (VRG2/ SRG) (SB/ VB

2) = 0.125 x [10.52/( 125/0.8)] x 100/10.52 = 0.08

217

(b) The reactance p.u. values for transformer

(T-l): X4 = VT(%) (SB/ SR) = 0.105 x 100/15 = 0.7

(T-2): X2 = VT(%) (SB/ SR) = 0.105 x (100/120) = 0.0875

(c) The transmission line

X3 = XL SB/ VB2 = 0.4 x 40 x 100/1152 = 0.121

(d) The air-core inductor

X5 = XR(%) [VR/ (IR V3)] SB/ VR2 = 0.05 x [6/(0.3 V3)] x 100/6.32 = 1.4547

(e) The HTS FCL

XRT = 29.43 x XRa = 29.43 x 0.577 =16.98 Q/phase

XRfpu = XRTSB/ VB2= 16.98 x 100/6.32 = 42.78

D. Steady state short circuit current, 6.3 kV busbar residual voltage and inrush

current

(a) Air-core inductor

Xz = Xi + X2 + X3 + X4 + X5 = 0.08 + 0.0875 + 0.121 + 0.7 + 1.4547 = 2.4432

Is = IPu x SRG/ (VR V3) = 0.418 x (125/0.8)/(6.3 V3) = 5.985 kA

Vpu = Ipu x X5 = 0.418 x 1.4547 = 0.608

VPe = Vpu x VB = 0.608 x 6.3 = 3.83 kV

218

(b) H T S FCL

Xs = 0.08 + 0.0875 + 0.121 + 0.7 + 42.78 = 43.77 Q

IPP„= 1/ X2= 1/43.77 = 0.023

Ip = Ippu x SRG/ (VB A/3) = 0.023 x (125/0.8) / (6.3 x A/3) = 0.33 kA

Vpp„ = Ipp„ XRfpU = 0.023 x 42.78 = 0.98

VPe = Vpp„ x VB = 0.98 x 6.3 = 6.17 kV

IRS = 2.55 xIP = 0.84 kA

E. Short circuit in transient state

(a) Using air-core inductor

RK = 0.0015 + 0.0223 + 0.021 = 0.0448 Q/phase

XK = 0.042 + 0.0524 + 0.2778 + 0.577 = 0.9492 Q/phase

(PK = arctgXK/RK = arctg 0.9492/0.0448 = 87.3°

Tg = L/R = XK/(314RK) = 0.9492/(314 x 0.0448) = 0.0675

(b) Using HTS FCL

RK= 0.0015 + 0.0223 + 0.021 = 0.0448 Q/phase

XK= 0.042 + 0.0524 + 0.2778 + 16.98 = 17.35 Q/phase

(pK = arctgXK/RK = arctgl7.35/0.0448 = 89.85°

Tg = L/R = XK/(314RK)= 17.35/(314x0.0448)= 1.233 Q

219

4.6 DISCUSSION AND CONCLUSION

The saturable magnetic core electrical fault current limiter designed with the

Ag-clad (Bi,Pb)2Sr2Ca2Cu30io+x HTS wire has been studied with respect to the HTS

wire behaviour, the device current limiting performance, and the required DC bias

winding for this application.

The Ag-clad (Bi,Pb)2Sr2Ca2Cu30io+x HTS wire has good applicable

properties, such as high critical current, long length, and fairly good mechanical

flexibility. The HTS wire therefore can be used to produce a HTS coil such as by "react

and wind" techniques. For the HTS-FCL device designed, a high DC mmf. is required

to be generated by the HTS coil as the DC bias winding. The DC bias ampere-turns

iDcnrx: is related to the maximum rated system AC current (peak value) Imax and AC

winding reactance (<* nAc)- The HTS wire is required to produce a HTS coil which

provides a high IocriDc value to allow the FCL to operate at practical high rated

current levels with the designed \|/b as shown in Figure 4-29. The high percentage Ic

of the HTS wire must be maintained in a high ampere-turn coil format, and this can

be achieved and satisfied only at a temperature below 77 K, e.g. 4.2 K or 50 K, for

the present Ag-clad (Bi,Pb)2Sr2Ca2Cu3Oio+x HTS wire. To increase the HTS coil

critical current ampere-tums at liquid nitrogen temperature, the intrinsic property of the

HTS magnetic field dependent Ic needs to be improved for a practical FCL device with

a HTS coil operated at 77 K.

An ideal and modified flux linkage pattern of the FCL device has been

introduced for the analysis of the current limiting behaviour by using the well-known

EMTP program. The selected FCL parameter XRf and samples S1 - S6 have also been

220

introduced into a power transmission and distribution network and compared with the

normally used linear reactor. As a general conclusion, this FCL can limit fault currents

effectively both for inrush currents in the transient state and periodic components in

the steady state, and with much less voltage drop during normal operation.

When the FCL v|/-I curve satisfies v|/m / \j/b < 1 (i.e. vj/b > v|/m), the periodic

component of the limited fault current receives an effective limiting result. When the

FCL has 0.5 < v|/m / \|/b < 1, and a short circuit occurs as the source voltage passes

through zero, the flux linkage v|/ is partially entering Section III of Figure 4-29 having

lower transient inductance, then the inrush current Is and its coefficient K increase

dramatically with the increasing of v|/m / v)/b. When v|/m / v|/b < 0.5, the FCL has an

unnecessary high DC bias, which causes an uneconomic operation of the FCL device

and makes it more difficult to prepare the HTS DC bias winding. Further investigation

is required to optimise \\fm I \j/b values for a practical FCL device.

Compared with the normally used linear reactor, the saturable magnetic core

FCL has lower voltage loss at rated current, and higher inductance under fault

conditions. The ratio 29.43 of the FCL reactance to the air core inductor reactance

has been used for analysis of the short circuit currents with the FCL and with a normal

air core inductor. When a three phase short circuit occurs on a 6 kV distribution line,

the FCL reduces the short circuit current by 94.6%, both with respect to the periodic

component in the steady state and the inrush current in the transient state, compared

with the air core inductor. Meanwhile the residual busbar voltage under short circuit

condition remains about 97% of the rated value. By selecting the FCL inductance

value properly, the FCL can reduce the fault current effectively.

221

The general advantages of this saturable magnetic core F C L compared to other

FCL designed with the HTS therefore are: (a) This FCL is designed with the

(Bi,Pb)2Sr2Ca2Cu3Oio+x HTS wire, which currently is the most promising and

practicable HTS; (b) The rated AC system current does not pass through the HTS,

therefore the HTS will not need to be rated up to the power system rated current

levels. In addition there will be ideally no AC loss generated in the HTS; (c) The

required impedance to limit the fault current can be easily achieved by controlling the

number of AC winding turns; (d) The cryogenic system for operation of the HTS is

much simplified since the high voltage and high current will not pass through it.

222

Chapter 5

AN ELECTRONIC RESONANT CIRCUIT WITH A HIGH Tc SUPERCONDUCTING INDUCTOR

FOR HIGH VOLTAGE GENERATION

5.1 INTRODUCTION

In the past some methods have been studied and developed with regard to

generating high voltages, for applications in electrical engineering, e.g., insulation

tests of power cables [325-329]. A novel method for generating high voltages from a

low voltage source will be introduced in this chapter. By using a resistor-capacitor-

inductor (R-C-L) resonant circuit, a method to generate high voltage has been

developed with the consideration of using a resistanceless inductor made by a high Tc

superconducting (HTS) coil [330-332].

A R-C-L series resonant circuit is investigated with regard to high voltage

generation. Resistance in the circuit limits the voltages that can be achieved. By

replacing an inductor which has normal copper windings by another inductor with

superconducting windings, the magnitude of the circuit quality factor Q can be

dramatically increased in the R-C-L series resonant circuit. Consequently the

magnitude of achievable voltages is dramatically increased.

The Ag-clad (Bi,Pb)2Sr2Ca2Cu3Oio+x HTS wire has potential capability to

make a coil or winding. This HTS wire has high critical current density, high

magnetic field tolerance, mechanical flexibility, and long length, and can be

223

considered for design of a H T S inductor. A Ag-clad (Bi,Pb)2Sr2Ca2Cu3Oio+x H T S

wire is selected to make a superconducting inductor with very low resistance, and is

employed in an electronic resonant circuit used to generate high voltages.

To generate high voltages, a low voltage DC power source can be used in the

circuit, and its polarity is required to be reversed at a certain frequency, e.g. at each

half of the resonant cycle, and this control is implemented by an electronic switch. A

voltage generating device built for testing purpose, mainly consists of a capacitor, an

inductor, a battery DC source, and an electronic switch. The inductor required is a

small HTS inductor operated in a liquid nitrogen cooling system, and a normal Cu

winding inductor is also used for tests and circuit analysis. The resonant circuit is

tested and analysed with regard to using a HTS inductor for high voltage generation.

This high voltage generator is capable of generating a very high voltage for

any high impedance-load application, such as insulation partial discharge tests in

electrical engineering.

5.2 PRINCIPLE OF OPERATION AND CONFIGURATION OF THE HIGH VOLTAGE GENERATOR

5.2.1 Principle of operation

In a series resonant circuit with resistor (R), capacitor (C) and inductor (L), as

shown in Figure 5-1, the resonant frequency is given by ox, = [(LC)" -(R/2L)"] .

Reducing the circuit resistance to zero leads to an infinite circuit quality factor Q of

the circuit at resonance. The principle of this high voltage generator is based on

224

such a R-C-L resonant circuit. Consequently the idea is considered with a view to

making a HTS inductor to reduce the circuit resistance. When a HTS inductor is used,

the circuit of Figure 5-1 can be simplified by assuming the circuit resistance R to be

zero with circuit resonant frequency © = 1/A/(LC) rad s"1.

Instead of using an AC power source, a DC power supply can be used and the

circuit configuration with a DC battery source is shown in Figure 5-2. This

configuration consists of a DC battery as power supply and an electronic switch with

variable switching frequency capability. To generate high voltages from the low

voltage DC power source, the electronic switch is used to reverse the polarity of the

power source at a certain frequency, e.g. at each half of the resonant cycle.

With zero initial charge on the capacitor, when the switch S in Figure 5-2 is

closed, the voltage across the capacitor C will be built up to an equilibrium value of

2VB after t = 7cV(LC), as illustrated in Figure 5-3, where Vc is the voltage across the

capacitor C. The circuit current is also shown in the Figure 5-3. A diode linked with a

DC source in Figure 5-2 is used to block reverse current flow and illustrate the circuit

current and voltage in the first half resonant cycle.

For the method using a HTS inductor with a DC power supply having voltage

VB, if at the point t = 71A/(LC), the battery source a-b in Figure 5-2 is disconnected,

and then reconnected in the opposite polarity for the next half cycle, then the

capacitor voltage is increased by 2VB. If the battery polarity is continued to be

reversed each half resonant cycle, the voltage magnitude across the capacitor will

increase by a further 2VB after each reversal. Consequently the voltage across the

capacitor C after n cycles will be

225

vn+1 Vc(n) = (-ir

i(Vco + 2nVB) (5-1)

where n is the switch reversing number, and V C o is the initial capacitor voltage.

If the circuit resistance of Figure 5-1 and Figure 5-2 is not zero, it causes

energy dissipation by the resistance, and for each reversal of the battery polarity the

corresponding increase in voltage is less than 2 V B , and the magnitude of the voltage

increase gets smaller with each iteration. N o theoretical maximum voltage is reached

by this method; however, a practical maximum voltage will be achieved when the time

taken to increase the voltage by a quantifiable amount is no longer feasible.

irYTr\_Rr—>

VpO c = Vr

Figure 5-1. Basic R-C-L resonant circuit for the high voltage generation.

Figure 5-2. Principle of generating high voltage with a D C power supply.

226

1

^/

J ^ VB(IX)-«

X

2VB

t=7t(LC)1/2

1 -i*-1

0 t

Figure 5-3. Voltage and current curves after switch S in Figure 5-2 is closed.

5.2.2 Configuration of the Generator

The electronic high voltage generator designed with the HTS mainly consists

of a HTS inductor (L), a capacitor (C), a DC power source (VB) and an electronic

switch (S), and its principle schematic diagram is shown in Figure 5-4 (a).

The HTS inductor is prepared using a Ag-clad (Bi,Pb)2Sr2Ca2Cu3Oio+x wire

which has inductance but no resistance; therefore it has an advantage for this

application. The HTS inductor is operated in a liquid nitrogen Dewar vessel.

The high voltage Vc is built up across the capacitor C, so a high nominal

voltage rating is required for the capacitor. The value of the capacitor C is chosen so

that the current flowing in the high voltage generator is limited to a value less than

227

the critical current of the H T S inductor at the natural frequency of the resonant

circuit, for the range of voltages to be generated.

The DC power source is required to be a low impedance voltage source. A

common 12 V lead-acid battery or a set of these batteries can be one of the choices.

To achieve the battery switching, a thyristor bridge is employed as shown in

Figure 5-4 (a) and (b). In Figure 5-4 (b), the electronic controlled switch is to reverse

the battery polarity, and this can be achieved by using an electronically controlled

thyristor bridge or alternatively an electronic relay. The frequency of the signal

generator, for supplying the thyristor firing pulses, is set at the natural frequency of

the high voltage generator or lower. By firing the thyristor pairs at the natural

frequency, the voltage generated is maximised. If the thyristors are fired at a slower

frequency, then some of the capacitor charge will leak away between the voltage peak

and the firing of the next thyristor pair. The generated triggering pulses are used to

control the thyristor firing. A voltage sensing circuit can be connected to the thyristor

controller; once the voltage across the capacitor reaches a pre-set value, the circuit

disables the thyristor triggering. Initially all thyristors are off. Thyristors T, and T2 are

fired simultaneously. The circuit conducts for half a resonant cycle, at which time the

current is about to pass through zero, resulting in the thyristors turning off naturally.

Thyristors T3 and T4 are then fired simultaneously and the process repeats itself.

An AC source can also be selected as the low voltage supply, and in this case

a configuration with a power amplifier added to the output of a signal generator can

be designed.

228

HTS inductor

Vc

(a) The generator schematic.

Thyristor-Trigger Signal Controller

T1 T2 T3 T4

-6- 6-DC

Power Supply

HVG AUTO-SWITCH

(b) The electronic switch.

-O

HVG Switch

AA

ure 5-4. Principle schematic of the H T S high voltage generator.

229

5.3 EXPERIMENT A N D RESULTS

5.3.1 Testing Arrangement

The R-C-L resonant circuit is tested with the specially built electronic switch

and a small HTS inductor with regard to its voltage generation capability. The

electronic switch is protected by a voltage dependent resistor for thyristor

overvoltage, and a semi-conductor fuse in series in the circuit provides overcorrect

protection. A capacitor bank with high rated voltage was selected, and 1200 V rated

thyristors were used. The thyristor bridge is controlled by an electronic circuit for

switching battery polarity. The maximum desirable voltage across the capacitor bank

is 2400 V, due to each pair of thyristors working in series. The DC power source was

a 12 V lead-acid battery. An oscilloscope was used to record the voltage built up

across the capacitor.

5.3.2 Inductor with the HTS Wire

Based on the available HTS material received, a HTS inductor, L-l, is made

by a 22 filament Ag-clad (Bi,Pb)2Sr2Ca2Cu3Oio+x wire via a wind and react

procedure with final wire cross-section 4 x 0.18 mm2. A wire sintered for 300 h at

830 °C to 838 °C was selected and wound to a coil, and an additional 60 h / 830 °C

treatment was applied. It has a total 180 turns wound with 20 layers and 9 turns per

layer. The dimensions of the coil are L = 60 mm, Oin = 29 mm, and Oout = 53 mm.

230

The H T S inductor is operated at 77 K in a liquid nitrogen container. The inductance

measured by a R-C-L bridge is 0.55 mH (at 100 Hz), and its calculated value is about

0.5 mH.

Figure 5-5. The electronic switch used.

231

Table 5-1. The experimental inductors.

Conductor

L

R

Inductor L-l

Bi-2223/Ag HTS wire

0.55 m H

0.72 Q at room temp.

0 Q (ideal) at 77 K

Inductor L-2

Cu wire

188 m H

35.6 Q at room temp.

4.74 Q at 77 K

The resistance of the H T S inductor L-l is related to its transport currents, and

its DC resistance at 77 K as shown in Figure 5-6 (a). The inductance of the inductor

L-l was also tested at 77 K and room temperature, the testing result at room

temperature is shown in Figure 5-6 (b), where Z = [R2 + (©L)2]1/2. The typical results

from a "Impedance Analyser" are 556.8 uH - 854.5 mQ - 63.1 Hz and 542.5 uH -

858.8 mQ - 100 Hz, where the resistance was caused by normal connections. The

magnetic flux density B measured at the centre of the inductor L-l is 3.4 mT at a

current of 1 A.

Because of the limitation of HTS materials at the time when this experiment

was carried out, a larger copper winding induetor L-2 had to be selected to increase

the inductance for the experimental investigation. The inductance of L-2 is 0.188 H

with resistance 35.6 Q at room temperature.

232

800-

700-

600-

500-

400-

300-

200-

100 J

0-

' i l ! 1 1 1

A A A A A

_

H T S inductor resistance

-

-

O At77K

A At room temp. -

r, O O ° ° 3ggggg%)0 0 9 9 9 ? ? — • 1 ' 1 i 1 •

1 2 3 4 5 6 7 8

Current (A)

(a) DC resistance.

a: j3

1 I 1 I I N I I I 1 I I I I I I I I I I- I I I

LL I I I I I I I I [ 1 [ I I I I I I I I I I I I 1 I

1 3 3

Frequency (Hz)

(b) Impedance.

Figure 5-6. Testing results of the HTS inductor L-l.

233

5.3.3 Build-up Voltages

5.3.3.1 Testing of build-up voltage

The build-up voltages with inductor L-l and L-2 were recorded using a

storage oscilloscope. The results are summarised and shown in Figure 5-7.

As shown in Figure 5-7 (a), the build-up voltage (peak-peak value) with the

small HTS inductor L-l after reversing the DC power supply polarity 20 times is

about 240 V. Since the practical circuit has some resistance, the build-up voltage

with the small HTS inductor L-l is limited by the circuit quality factor Q.

Inductor L-2 has a larger inductance. The build-up voltage with the inductor

L-2 is tested for 77 K operation of the inductor, and shown in Figure 5-7 (b), where

Rj„ represents the inductor resistance.

The circuit resistance is caused by the inductor as well as the DC source,

thyristors, wires and connections, whose resistance affects the circuit quality factor

significantly for the situation when the inductor used has small inductance. In

practice the final voltage across the capacitor would reach a finite maximum value,

and this value could be very low when the circuit has high resistance. This would

occur as a result of the small voltage increase per cycle being balanced by the same

magnitude voltage loss per cycle due to leakage effects.

234

5 V A=f 24.6v X lb-S • i i

• ; ' I - . 1 4 ! 1 .. 'I

»••_!___..; i L.-^.-f'-i

j ! j - - 1 ' \ ';;' r :- - i i

i i i

j 1 : 1 1

L___T_. 1 ! . !. ••

;~. 1. i «-! * *|

: . . . . _ . i •_• . i -

I ! ' T 1 1 ' ! I

** -J

i r FT ". ,

1 _V=j3.0mS i

i ; 1 -r-

*

-

'-*-

f --i * r

• i

•

• i * J

- *- j - *

l//!lT=3_t3H2 ||

(Y - 50 V/d, X - 10 ms/d and R^ = 0 Q)

(a) The voltage build-up with the small HTS coil at 77 K.

(Y - 200 V/d, X - 100 ms/d and Ri„ = 4.74 Q)

(b) The voltage build-up with L-2 at 77 K.

Figure 5-7. The generated voltage with the inductors.

235

5.3.3.2 Build-up voltages with different circuit resistance

The build-up voltages with the inductor L-2 and different circuit resistance are

measured and shown in Figure 5-8 (a), where curve (1) has 4.74 Q inductor-

resistance, and curve (2) has 36.32 Q inductor-resistance respectively. In Figure 5-8

(b) the maximum build-up voltage (peak-peak) is recorded with change in the

resistance of the inductor L-2.

As evident in Figure 5-7 and 5-8, the voltage can be practically built up from a

low voltage DC source, and its value increases with decreasing circuit resistance. The

circuit resistance has a significant influence on the final voltage achieved.

5.3.3.3 Circuit current

Figure 5-9 shows the circuit current build-up with the inductor L-2 at room

temperature, and the R-C-L circuit has values of R - 36.6 Q, C - 2.2 uF and L - 0.19

H. The circuit current was measured with a 1 Q resistor which is in series in the

circuit. The circuit current of the generator builds up as well as the build-up voltage.

The amplitude of the maximum build-up current is also required to design the circuit

and in particular the HTS inductor; in addition the circuit di/dt is also limited by the

thyristors used in the switch.

236

(1) 77 K (200 V/d, Rj. - 4.74 Q ) ; (2) Room temperature (100 V/d, Rj- = 35.6 Q )

2000

(a) Maximum build-up voltages with L-2.

Inductor-resistance (a)

(b) Build-up voltages with L-2 for various inductor-resistances.

Figure 5-8. Build-up voltages with different circuit resistance.

237

(1)

(2)

jxmssmsswsB^^^^sss^mm

i i i i

ti • i f h l ! t t i

i M

(1) Build-up voltage 50 V/div; (2) Build-up current 0.2 A/div.

Figure 5-9. The circuit current with the build-up voltage.

238

5.4 C I R C U I T ANALYSIS

5.4.1 Resonant Circuit with a Superconducting Inductor

5.4.1.1 Maximum build-up voltage

The build-up voltage in a resistive resonant circuit is related to the circuit

quality factor Q. By considering the quality factor Q = COoL/R of a resonant circuit,

sometimes called the magnification factor, the potential difference across the

capacitor at resonance is Q times as great as the applied emf Vp (rms). For a

sinusoidal power supply, the voltage across the capacitor VcmaX at resonance is given

by

Vcma. = Q Vp (5-2)

For a DC power supply with an electronic switch, the maximum voltage

Vcmax can be expressed as

Vcma. = Q VB' (5-3)

where VB' is the voltage of the low voltage power source. The switching controller

used has rectangular switching wave form with low voltage source polarity switching

frequency f. Therefore the build-up voltage wave form F(t) can be expressed by

Fourier series as

F(t) = (4VB/7i)[sincot + (l/3)sin3cot + (l/5)sin5cot + ...] (5-4)

239

The generator only resonates on the first harmonic with amplitude of 4VB/7C.

Therefore

VB' = 4 VB/TC (5-5)

and the maximum build-up voltage is related with the R-C-L resonant circuit Q value

by

VCmax = Q(4VB/7t) (5-6)

Reducing the R-C-L resonant circuit resistance is achieved by introducing the

superconducting inductor, the circuit Q value will be very high, therefore leading to a

very high voltage across the capacitor.

5.4.1.2 Ideal build-up voltage

In a R-C-L resonant circuit as shown in Figure 5-10, when switch S is closed

in this circuit, the instantaneous current and voltage solutions respectively are [333,

334]

i(t) = < -Rt 2L

(Vco + VB)sin<»

Q)L (5-7)

vc(t) = ( v C 0 + v B ) --^( R . N

1-e u cosco t + ——sin*y t V 2Lco J

-V co (5-8)

240

where R < 2(L/C)1/2, V c o is the initial capacitor voltage, and co = [(1/LC)-(R2/4L2)]1/2

is the resonant frequency in rad s1. Both equations describe decaying sinusoids, with

Vc(t) approaching a steady state value of V B , and i(t) approaching a steady state value

of zero.

Now in a circuit using a HTS inductor and no separate resistor, then R will

become very small. If R = 0 Q, then Equations (5-7) and (5-8) can be simplified to

.(t)=(vco+VB)sina,. G)L v '

Vc(t) = -(vco+VB)cos<ut + VB (5-10)

where co = (LC)" - rad s" . These two equations describe constant magnitude

sinusoids, with the average values of i(t) and Vc(t) being zero and V B respectively.

W h e n switch S in Figure 5-10 is closed, Vc(t) = Vc(0) = -VCo- One half a

resonant cycle later, this voltage will have increased to

Vc(t) = Vc(7c/co) = -(Vco + VB)(-1) + VB = Vco + 2VB (5-11)

If at this point of time the battery is disconnected, and then reconnected in the

opposite polarity for the next half cycle, then the initial capacitor voltage Vco is

changed to Vco-new, and is given by

VcO-new = -Vc(t) = -(Vco + 2VB) (5-12)

241

Haifa cycle later, Vc(t) = Vc(27c/a>) becomes

Vc(t) = -[-(Vco + 2VB) + (-VB)](-1) + (-VB) = -(Vco + 4VB) (5-13)

If the battery is reversed every half cycle thereafter then

Vc(t) = (Vco + 6VB); -(VCO + 8VB); (Vco + 10VB); etc (5-14)

Therefore the positive and negative peak voltages can be described by

Equation (5-1). This is the build-up voltage for an ideal circuit without resistive loss.

5.4.1-3 Build-up voltage with a resistive circuit

From Equation (5-7), when t = rcAo, i = 0, if the thyristor bridge in Figure 5-4

changes polarity, the capacitor voltage is given by

VCi = (Vco+VB)(l+e-Rn/2L<a)-Vco (5-15)

After the polarity is changed n times, the capacitor voltage becomes

242

A/W R

Vc VB

1 i(t) — « —

Vc(t)

Figure 5-10. Series R-C-L resonant circuit.

ure 5-11. High voltage generator with a thyristor bridge switch.

243

V c„ = (VCn-i+VB)(l+e-R7I/2L£0)-Vc„.i (5-16)

If V c o = 0, then

Vc,=VB(l+e-R,t/2Ll0)

Vc2=(Vci+VB)(l+e-R7t/2L(0)-Vcl=VB(l+2e-

R'c/2Lco+e-2R7I/2L£0)

VCn=(Vc„-i+VB)(l+e-R,l/2Lco)-Vcn-i

=V (T+2e"Rjt/2L(0+2e"2R,t/2Lco 4.2e"(n"1)R7t/2Lto+e"nR,t/2Lco>l

=(1+e-nR)V2L_)+2VB ^e-i*xl2U» (5.,7)

Assuming that the thyristor bridge changes the polarity at t = njc/co, after the polarity

is changed n times, the build-up voltage VCn is given by

-R/T/2L<D _ -(n-l)R;r/2L<u

VCn = VB(l + e-R^) + 2VB TZ^J^ (5-18)

W h e n n -> °sV C n -> V C m a x , therefore

V C M , = l i m V 0 l = V , ( l + to/2Ut ) (5-19) n-**» c — I

From Equation (5-19), when R -> 0, V c m a x -» *=. W h e n R -> °s V C m a x -» VB.

The quality factor Q of a R-C-L resonant circuit is Q = coL/R; therefore

Equation (5-19) can be expressed as

244

2VB VCmax = VB + -^nQ—T (4-20)

When VCn is known, the circuit current in can also be calculated. The circuit

current is required for the design of a HTS inductor, and the circuit di/dt is also

required for design of the thyristor switch.

5.4.2 Build-up Voltage and Circuit Current

The high voltage generator is analysed to predict the voltage build-up with the

HTS inductors. The values of the circuit components have been chosen to be of a

similar size to what will be possibly used in a small prototype. It is assumed that the

bridge thyristors commutate precisely at the natural half cycle, i.e. the DC source

polarity is switched at the circuit resonant frequency [27tv'(LC)]"1, and without any

additional voltage drop and capacitor leakage. It is also assumed that the circuit

current is well below the critical current of the HTS inductor, and AC loss in the HTS

inductor is negligible. Computer programs such as "SPICE" or "MATLAB" can be

employed for this study.

To model a circuit containing a HTS inductor, the values chosen were L = 20

mH, C = 20 uF, R = 0.05 Q, VB = 12 V and Vco = 0. Using this model the voltage

across the capacitor has increased steadily from zero to 700 V in 60 ms, as can be

seen in Figure 5-12, and then to 1149 V in 100 ms (50 iterations). After 6 sec. (3020

iterations) the voltage achieved is 9.66 kV. There is only a 15 V increase of Vc

between t = 5 s and t = 6 s.

245

To model a circuit containing an inductor with copper windings, the values

chosen were L = 20 mH, C = 20 uF, R = 3 Q, VB = 12 V and Vco = 0. Figure 5-13

shows the voltage across the capacitor for this model. The voltage achieved after 30

iterations (-60 ms) is 159 V, and after 50 iterations (-100 ms) is 161 V.

By comparing the build-up voltages in the circuit with same L (20 mH) and C

(20 uF) values but different R, it can be seen that the build-up voltage with the HTS

inductor at 100 ms is more than 7 times higher than the copper inductor at the same

instant. This is achieved by reducing the circuit resistance from 3 Q to 0.05 Q. After 2

s, the build-up voltage with the HTS inductor is about 45 times higher than the value

with a Cu inductor.

Since this model has a small value of resistance, then the voltage increase

reduces with time. Figure 5-14 shows the envelope of the capacitor voltage for the

same model but over a longer period of time. After 6 seconds (3020 iterations) the

voltage achieved is 9.66 kV. There is only a 15 V increase that occurs between t = 5.0

seconds and t = 6.0 seconds.

The circuit current is built up with the build-up voltage, and the result for R =

0.05 Q, C= 20 uF and L = 20 mH is shown in Figure 5-15. From the result, a current

generator can also be considered in the future, and the maximum current I„ux = Vp / R

can be used for the evaluation.

Based on the available HTS material produced, a 5 mH inductor can be

produced which has critical current Ic = 5.2 A at 77 K and Ic = 45.6 A at 4.2 K [312,

317]. Figure 5-16 (a) shows the build-up voltages by using a 5 mH inductor with a

2.2 uF capacitor, and with variable circuit resistance. By considering the HTS critical

current limitation, the maximum build-up voltages are given by Figure 5-16 (b).

246

1500 T

20 40 60

Time (ms)

80 100

HTS Inductor: R = 0.05 Q - C = 20uF-L = 20mH

Figure 5-12. Voltage Vc(t) build-up with a HTS inductor.

20 40 60

Time (ms)

80 100

Normal Inductor: R = 3 Q - C = 20uF-L = 20mH

Figure 5-13. Voltage Vc(t) build-up with a normal Cu inductor.

247

10.00 -

8.00 -

6.00-

4.00 •

> 2.00-

X o.oo -g -2.00 •

-4.00 -

-6.00 -

-8.00 •

in ev\ -lU.UU •+

0.

I ' i i i r

- " _ T - - " i

{ i 1 1 i 1

0 1.0 2.0 3.0 4.0 5.0 6.0

Time (s)

Figure 5-14. Envelope of Vc(t) for R = 0.05 Q, C = 20 uF and L = 20 mH.

248

40 60

Time (ms)

(a) With R = 3 Q, C= 20 uF and L = 20 mH.

0.0 1.0 ZO 3.0 4.0

Time(s)

5.0 6.0

(b) With R = 0.05 Q, C= 20 uF and L = 20 m H (envelope).

Figure 5-15. Circuit current i(t).

249

>

1.5x103

1.0x103

5.0x102 -

OJ cn ro

o > -5.0x10" -

-1.0x10J -

-I.SxIO'

High Voltage Generator

16 4 8 12

t (ms)

(a) Build-up voltages with different circuit resistance.

20

3x10'

2x10'

1x1 Cr

0

•1x10v

-2x10'

-3x10'

i r

26.6ms-2503V-44.6A<lc=45.6A(4.2K)

5.6ms-283V-4.9A<lc=5.3A(77K)

Current

Voltage

0 5 10 15 20 25

t (ms)

(b) Critical current limitation for the build-up voltage.

30

Figure 5-16. Build-up voltages with a small H T S inductor.

250

5.5 D I S C U S S I O N

5.5.1. Inductor with Bi-2223/Ag H T S Wire

The Ag-clad (Bi,Pb)2Sr2Ca2Cu30io+x HTS wire is prepared to make an

inductor with much less resistance than a normal resistive inductor. For a multilayer

close-wound cylindrical HTS coil, the desired inductance can be designed according

to

L = 0.019739(2a/b)aN2K' (uH) (5-21)

where a (cm) is the coil mean radius, b (cm) is the coil length, and K' is the coil

dimension dependent coefficient [335].

Since the flux linkage \j/ of the HTS coil is \|/ = NO = LI, therefore

B = LI/S (5-22)

where S represents the HTS winding area. The large inductance and large rated

working current will generate a high magnetic field within the HTS inductor itself.

To design a suitable inductor for this application with the HTS wire, the magnetic

field-dependent critical current needs to be very high. Consequently high working

current can be obtained when a large number of ampere-turns are formed for a high

transport current and large inductance.

The design of a HTS inductor with the Ag-clad (Bi,Pb)2Sr2Ca2Cu3Oio+x wire

has been discussed in Chapter 3. The Bi-based Ag-clad wire shows that it can be used

251

to make a compact H T S winding since it has achieved considerable high critical

current density and flexible mechanical property. The presently produced Ag-clad

(Bi,Pb)2Sr2Ca2Cu30io+x wire is a good candidate for this application, but

improvement is required for 77 K use. The critical current, the magnetic field

generated by the winding itself, and the A C loss of the H T S wire are the limitations

in the design of a H T S inductor. For a circuit with high transport current, the H T S

inductor is required to operate below 77 K to increase the H T S material flux pinning

property, and this is the present solution for use of this H T S wire before it is further

improved for its magnetic field dependent critical current at liquid nitrogen

temperature. With the improvement in the flux pinning properties of the H T S

materials, there is potential to make H T S coils with a large inductance and operated

at 77 K.

Choosing circuit component values is also related to the H T S inductor used.

To maintain a low circuit current magnitude the inductance should be large. To

reduce A C loss of the H T S , it is advantageous to keep the natural or switching

frequency low, by having a large capacitance and large inductance. To increase the

inductance of the H T S inductor, several such small inductors can be used in series.

The value of the capacitor is chosen so that the current flowing in the circuit is

limited to a value less than the critical current of the H T S inductor at the natural

frequency of the circuit, for the range of generated voltages.

About 0.2 T magnetic field (B//c) can be selected to design the H T S coil

ampere-turns with the Ag-clad (Bi,Pb)2Sr2Ca2Cu3Oio+x H T S wire, at which - 5 0 %

1^=0 can be maintained at 77 K [312, 317]. Working at 4.2 K, the B can be selected

to be much greater than 20 T for designing the inductor with this H T S wire. Based on

the resistivity criterion of 10"13 Q m for the H T S , the Q ratio of the H T S inductor to a

252

normal C u (p = 1.7 x IO"8 Q m ) winding inductor, QHTS/QCU, is approximately of the

order of IO5.

5.5.2 High Voltage Generation

The maximum theoretical generated voltage of the circuit as shown in Figure

5-11 with a perfect circuit which has no circuit resistance, no capacitor leakage, and

switching frequency equals to half circuit resonant frequency, is infinite and its value

with iteration is 2nVB, as shown by Equation (5-1). In practice the maximum build-up

voltage reaches a finite maximum value which limitation is mainly determined by

circuil resist.ance caused by the inductor resistance as well as the DC source,

thyristors, wires and connections. For such a resistive circuit, the maximum build-up

voltage is equal to Q(4VB/7I). This limitation occurs as a result of the small voltage

increase per cycle being balanced by the same magnitude voltage loss per cycle due to

resistive and leakage effects.

With a superconducting inductor, the generated voltage can reach a very high

value. The high voltage generator can generate voltages of almost any magnitude with

the appropriate choice of circuit components. By limiting the number of iterations, or

by using some form of voltage sensing, it is possible to generate a pre-determined

voltage.

Circuit components of the generator decide the resonant frequency of the

circuit, and therefore the switching frequency, the voltage magnitude after a certain

253

number of iterations, the maximum voltage and the circuit currents. To maintain a low

current magnitude the inductance should be large according to Equation (5-9). Also

since the superconductor critical current reduces with increasing AC frequency, it is

advantageous to keep the natural frequency low, by having a large capacitance and

large inductance.

5.5.3 Applications

The high voltage generator can generate voltages of almost any magnitude

with the appropriate choice of circuit components. By limiting the number of

iterations, or by employing some form of voltage sensing, it is possible to generate a

pre-determined set voltage.

The voltage generator can only generate high voltages at low load current.

When the resistance load across the capacitor C is low, the maximum build up voltage

is considerably reduced. Consequently the device is not capable of supplying a low

resistance or low impedance load.

Since very high voltage can be generated with a HTS inductor, the possible

applications in electrical engineering may include partial discharge testing and

pressure testing of electrical insulation systems, which do not require low impedance

loads to be driven.

This device is also possible to be a useful tool for studying HTS. Since the

maximum build-up voltage occurs only at resonant frequency, the inductance L of the

HTS coil can be obtained by co0 = 1/V(LC). The circuit current in the generator

254

increases with the build-up voltage, and the build-up circuit current in both frequency

and amplitude can be controlled. Consequently it can be used as a method to test a

superconductor at various frequencies with regard to its critical current as evident

from Figure 5-17. When the superconducting coil becomes resistive, the build-up

voltage will drop and its build-up voltage curve will shift away from the ideal zero

resistance build-up voltage curve. Correspondingly the circuit current at that shifting

point J is related to the HTS critical current. As shown in Figure 5-18, the maximum

build-up voltages by a small HTS inductor with different circuit resistance at t = 26.93

ms where the circuit current peak value reaches the HTS inductor critical current, are

reduced with increasing circuit resistance.

The AC loss of a HTS can also be tested with the generator at different

frequencies. The high build-up voltage across the HTS inductor can be used to test

the HTS winding insulation capability. Consequently the voltage generator described

can be further developed for studying the HTS coils and wires with respect to their

transport currents and AC losses at different frequencies, by comparing the voltage

measured and the theoretical zero-resistance build-up voltage.

255

Vc

Zero-resistance voltage peak value

Figure 5-17. Principle of testing a superconductor with the generator.

3_

-i 1 — i — i i i i 11 1 1 — i — i i i i 11 1 1 — i — i i i i 11 -i 1 — i — r

2.5x10

2.0x10J-

03 CD T

CL 1.5x103J

>

<D

CO

1.0x10*-

> 5.0x102

0.0-

Voltages with different circuit resistance

at t=26.93ms

l(27ms)=lc(45.6A-4.2K) JJ.

— i — i — i — i 1 1 1 1 1 —

0.01 0.1 • r " I ~ T i i i i | -T 1 — I I I I I |

10

R (Q)

(L = 5 mH, C = 2.2 uF and different R.)

Figure 5-18. The maximum build-up voltages by a small HTS inductor at t

26.93 ms with different circuit resistance.

256

5.6 C O N C L U S I O N S

Based on an available Ag-clad (Bi,Pb)2Sr2Ca2Cu30io+x H T S wire, a small

inductor has been made. By considering an electronic resonant circuit with a

superconducting inductor, a method of generating high voltage from a low voltage

source has been developed. This method has been analysed both from experiments

and theoretical calculations.

By reversing the battery polarity in a R-C-L resonant circuit, it is possible to

generate high voltages of many orders of magnitude greater than the low voltage DC

source. By using a superconducting inductor to reduce the circuit resistance, the

maximum generated voltage is much higher than in a resistive circuit with a normal

copper winding inductor. Built with a HTS inductor, the high voltage generator could

be portable and easy to operate, and has a much lower cost due to operation with

liquid nitrogen rather than liquid helium as with conventional low Tc superconductors.

With the prospect of manufacturing large HTS inductors, this method of voltage

generation has the potential to be utilised in industry. Applications such as partial

discharge testing and insulation pressure testing, that do not require low impedance

loads to be driven, could use this method of voltage generation.

257

REFERENCES

[1] H. K. Onnes, Cornmun. Phys. Lab. Univ. Leiden, No. 119b, 120b, 122b (1911).

[2] V. L. Ginzburg, and L. D. Landau, Zh. Eksp. Teor. Fiz., 20:1064 (1950).

[3] J. Bardeen, L. N. Cooper, and J. R. Schrieffer, Phys. Rev., 108:1175 (1957).

[4] B. D. Josephson, Phys. Lett., 1:251 (1962).

[5] J. G. Bednorz, and K. A. Mueller, Z. Phys. B Condensed Matters, 64:189 (1986).

[6] C. W . Chu, P. H. Hor, R. L. Meng, L. Gao and Z. J. Huang, Y. Q. Wang, Phys. Rev. Lett., 68 (1987) 405 908.

[7] W . Meissner, and R. Ochsenfeld, Naturwiss, 21 (1933) 787.

[8] F. London, and H. London, Proc. R. Soc. London Ser., A 149:71 (1935) Physica, 2: 963 (1935).

[9] n. H. ayaapeB, y<I>H, 71 (1960) 71.

[10] Y. B. Kim, C. F. Hempstead, and A. R. Strand, Phys. Rev., (1963) 129:528-(1965) 131:2486.

[11] H. London, Phys. Lett, 6:162 (1962).

[12] C. P. Bean, Phys. Rev. Lett., 8:250 (1962).

[13] Y. B. Kim, C. F. Hempstead, and A. R. Strand, Phys. Rev. Lett., 9:306 (1962).

[14] K. Yamafuji, and F. Irie, Phys. Lett, A25:387 (1967).

[15] W. A. Little, Phys. Rev., 134:A1416 (1964).

[16] D. Jerome, A. Mazaud, M. Ribault, and K. Bechgaard, J. Phys. Lett, 41, L-95 (1980). (For first polymeric superconductor synthesised: R. L. Greene, G. B. Street, and L. J. Suter, Phys. Rev. Lett, 34:577, 1975.)

[17] A. W . Sleight, J. L. Gillson, and P. E. Bierstedt, Solid State Commu., 17:27 (1975).

[18] H. Brechna, Superconducting Magnet Systems, Springer-Verlag, Berlin, 1973.

[19] M. Brian Maple, "The New High Temperature Superconductors-The Materials and Their Properties", in High Temperature Superconducting Compounds: Processing & Related Properties, edited by S. H. Whang, and A. DasGupta, the Minerals, Metals & Materials Society, 1989.

[20] M. Cyrot, and D. Pavuna, Introduction to Superconductivity and High-Tc Materials, World Scientific Publishing Co. Pty. Ltd., 1992.

[21] Composite Superconductors, edited by K. Osamura, T. Matsushita, P. J. Lee, and S. Ochiai, Marcel Dekker, Inc., 1994.

[22] M. Tinkham, Introduction to Superconductivity, McGraw-Hill, New York, 1975.

258

[23] A. M. Campbell, and J. E. Evetts, Adv. Phys., 21:199 (1872).

[24] A. R. Kantrowitz, and Z. J. J. Stekly, "A New Principle for the Construction of Stabilized Superconducting Coils", Appl. Phys. Letters., Vol. 6, No. 3, 56-7,

Feb., 1965. (I. N. Goncharov, "Niobium and 75 Percent Zirconium Superconducting Magnet for 86 kG", Cryogenics, 8:158-61, June 1968.)(E. D. Hoag, "Experimental Investigation of Advanced Superconducting Magnets", Sci. Tech. Aerospace Rept, 4(7), 1088, 1965.)

[25] G. F. Hardy, and J. K. Hulm, Phys. Rev., 87:884 (1953); 93:1004 (1954).

[26] B. T. Matthias, T. H. Geballe, S. Geller, and E. Corenzwit, Phys. Rev. 95:1435 (1954)

[27] J. J. Hanak, K. Strater, and R. W. Cullen, RCA Rev., 25:342 (1964).

[28] B. T. Matthias et al., Science, 156:645 (1967).

[29] P. R. Sahm, and T. V. Pruss, Phys. Lett, 28A:707 (1969).

[30] G. W. Webb, L. T. Vieland, R. E. Miller, and A. Wicklund, Solid State Commu., 9:1769(1971).

[31] J. R. Gavaler, Appl. Phys. Lett, 23:480 (1973).

[32] J. R. Gavaler, M. A. Janocko, and C. K. Jones, J. Appl. Phys., 45:3009 (1974).

[33] L. R. Testardi, J. H. Wemick, and W. A. Royer, Solid State Commnu., 15:1 (1974).

[34] J. E. Kunzler, E. Bachler, F. S. L. Hsu, and J. E. Wemick, Phys. Rev. Lett, 6:89 (1961).

[35] W. A. Fietz, E. F. Mains, P. S. Swartz, E. G. Knopf, W. D. Markiewicz, and C. Y. Rosner, IEEE Trans. Magn., MAG-11, 559 (1975).

[36] F. P. Levi, J. Appl. Phys., 31:1469 (1960).

[37] R. M. Rose, H. E. Cline, B. P. Strauss, and J. Wulf, J. Appl. Phys., 37:5 (1966);

Trans. Am. Soc. Met, 1:132 (1966).

[38] M. N. Wilson, C. R. Walters, J. D. Lewin, and P. F. Smith, J. Phys., D3:1517

(1970).

[39] K. Tachikawa, Proc. ICEC 3 (Third Int. Cryog. Conf.), p. 339 (1970).

[40] W. A. Fietz, and C. H. Rosner, IEEE Trans. Magn., MAG-10, 239 (1974).

[41] Composite Superconductors, edited by K. Osamura, T. Matsushita, P. J. Lee, and

S. Ochiai, Marcel Dekker, Inc., 1994.

[42] L. H. Wemick, F. J. Morin, F. S. L. Hsu, J. P. Malta, D. Dorsi, and J. E. Kunzler, "High Magnetic Fields" (H. H, Kolm, B. Lax, F. Bitter, and R. G. Mills, eds.), p. 609, MIT Press, Cambridge, Massachusetts and Wiley, New

York, 1962.

[43] D. Dew-Hughes, Cryogenics, 15:435 (1975).

[44] R. Chevrel, M. Sergent, and J. Prigent, J. Solid State Chem., 3:515 (1971).

[45] J. B. Vetrano, and R. W. Boom, J. Appl. Phys., 36:1179 (1965).

259

[46] A. D. Mclnturff, and G. G. Chase, J. Appl. Phys., 44:2378 (1973).

[47] R. G. Hampshire, and M. T. Taylor, J. Phys., F2:89 (1972).

[48] C. S. Panda, and M. Suenaga, Appl. Phys. Lett, 29:443 (1976).

[49] R. G. Hampshire, J. Phys., D7:1847 (1974).

[50] A. V. Narlikar, and D, Dew Hughes, Phys. Status Solidi., 6:383 (1964).

[51] D. Dew Hughes, Mater. Sci. Eng, 1:2 (1966).

[52] A. V. Ngrlikar, and D. Dew Hughes, J. Mater. Sci., 1:317 (1966).

[53] A. D. Mclnturff, and A. Paskin, J. Appl. Phys., 40:2431 (1969).

[54] D. Saint-James, G. Sarma, and E. J. Thomas/Type II Superconductivity", Pergamon, Oxford, 1969.

[55] J. Willbrand, and R. Ebrling, Metall (Berlin), 7:677 (1975).

[56] A. D. Mclnturff, G. G. Chase, C. N. Whetstone, R. W. Room, H. Brechna, and W. Haldemann, J. Appl. Phys., 38:524 (1967).

[57] S. L. Wipf, and M. S. Lubell, Phys. Lett, 16:103 (1965).

[58] J. E. Evetts, A. M. Campbell, and D. Dew-Hughes, Philos. Mag, 10:339 (1964).

[59] R. Hancox, Appl. Phys. Lett, 7:138(1965).

[60] P. S. Swartz, and C. P. Bean, J. Appl. Phys., 39:4991 (1968).

[61] P. F. Chester, Rep. Prog. Phys., 30:561(1967).

[62] M. N. Wilson, C. R. Walters, J. D. Lewin, and P. F. Smith, J. Phys., D3:1518 (1970).

[63] Z. J. J. Stekly, J. Appl. Phys., 37:324 (1966).

[64] D. Dew-Hughes, K. Seshan, and K. A. Wallis, Cryogenics 14:647 (1974).

[65] D. N. Lyon, Adv. Cryog. Eng, 10B:371 (1965).

[66] Z. J. J. Stekly, and J. L. Zar, IEEE Trans. Nucl. Sci., NS-12.367 (1965).

[67] J. W. Ekin, Cryogenics, 20:611 (1980).

[68] B. B. Goodman, and M. R. Wertheimer, Phys. Lett, 18:236 (1965).

[69] S. L. Wipf, and M. S. Lubell, Phys. Lett, 16:103 (1965).

[70] H. Kobayashi, K. Yasukochi, and T. Ogasawara, Jpn. J. Appl. Phys., 9:889 (1970).

[71] Y. B. Kim, C. F. Hempstead, and A. R. Strang Phys. Rev., 131:2486 (1963).

[72] H. R. Jr. Hart, Proc. Brookhaven National Laboratory, Summer Study, pp. 571-

600(1868).

[73] M. N. Wilson, Superconducting Magnets, Clarendon Press, Oxford, p. 146,

1983.

[74] T. Ogasawara, Cryogenics, 27:123 (1987).

[75] Z. J. J. Stekly, and J. L. Zar, IEEE Trans. Nucl, Set, 12:367 (1965).

260

[76] S. L. Wipf, IEEE Trans. Magn., MAG-15:379 (1979).

[77] D. N. Cornish, Proc. 8th Int. Cryogenic Engineering Conf., Geneva, pp. 267-281 (1980).

[78] A. P. Martinelli, and Wipf, S. L., Proc. Applied Superconductivity Conf., Annapolis, Md., pp. 331-340 (1972).

[79] O. Tsukamoto, and S. Kobayashi, J. Appl. Phys., 46:1359 (1975).

[80] W . G. Steward, Int. J. Heat Mass Transfer, 21:863 (1978).

[81] Y. Iwasa, Cryogenics, 19:705 (1979).

[82] M. N. Wilson, IEEE Trans. Magn., MAG-13:440 (1977). (L. Dresner, IEEE Trans. Magn., MAG-13.670, (1977).)

[83] L. Dresner, Cryogenics, 20:558 (1980).

[84] L. Dresner, Cryogenics, 24:283 (1984).

[85] G. Ries, Proc. Workshop on Stability Superconductors in Helium I and Helium II, Saclay, France, pp. 195-207 (1981).

[86] M. Nishi, T. Ando, T. Hiyama, H. Tsuji, Y. Takahashi, K. Okuno, K. Yoshida, E. Tada, K. Koi;_umi, T. Kato, H. Nakajima, O. Takahashi, M. Oshikiri, S. Shimamoto, andK. Yasukochi,./. Phys., 45:C1-131 (1984).

[87] G. Bon Mardion, G. Claudet, and P. Seyfert, Cryogenics, 19:45 (1979).

[88] S. W . Van Sciver, Helium Cryogenics, Plenum Press, New Yory, p. 141 (1986).

[89] C. Schmidt, Appl. Phys, Lett, 32:827 (1978).

[90] Z. J. J. Stekly, R. Thome, and B. Strauss, J. Appl, Phys., 40:2238 (1969).

[91] W . J. Jr. Carr, AC Loss and Macroscopic Theory of Superconductors, Gordon and Breach, New York, p. 62 (1983).

[92] M. N. Wilson, Superconducting Magnets, Clarenndon Press, p. 165 (1983).

[93] V. B. Zenkevitch, V. V. Zheltov, and A. S. Romanyuk, Cryogenics, 18:93 (1978).

[94] T. Ogasawara, Y. Takahashi, K. Kanbara, Y. Kubota, K. Yasohama, and K. Yasukochi, Cryogenics, 19:736 (1879).

[95] K. Kanbara, Cryogenics, 27:621 (1987).

[96] W. J. Jr. Carr, J. Appl. Phys., 46:4043 (1975).

[97] W. J. Jr. Carr, IEEE Trans. Magn., MAG-13:192 (1977).

[98] W. J. Jr. Carr, IEEE Trans, Magn., MAG-15:240 (1979).

[99] G. Ries, IEEE Trans. Magn., MAG-13:524 (1977).

[100] F. Sumiyoshi, K. Koga, H. Hori, F. Irie, T. Kawashima, and K. Yamafuji, Cryogenics, 23:619 (1Y83).

[101] I. HJasnik, Proc. 6th Int. Conf. on Magnet Technology, Bratislava, ALFA,

Bratislava, Czechoslovakia, pp. 575-596 (1978).

261

102] T. Ogasawara, Y. Takahashi, K. Kanbara, Y. Kubota, K. Yasohama, and K. Yasukochi, Cryogenics, 20:216 (1980).

103] T. Ogasawara, Y. Takahashi, K. Kanbara, Y. Kubota, K. Yasohama, and K. Yasukochi, Cryogenics, 21:97 (1981).

T04] T. Ogasawara, Y. Takahashi, K. Kanbara, Y. Kubota, K. Yasohama, and K. Yasukochi, IEEE Trans. Magn., MAG-17:967 (1981).

105] T. Kawashima, F. Sumiyoshi, F. Irie, and K. Yamafuji, Cryogenics, 24:313 (1984).

T06] B. Turck, IEEE Trans. Magn.,MAG-13:248 (1977).

T07] K. P. Jungst, and G. Ries, J. Appl. Phys., 48:3959 (1977).

T08] S. Murase, H. Shiraki, M. Tanaka, M. Koizumi, H. Maeda, I. Takano, N. Aoki, M. Ichihara, E. Suzuki, K. Noto, K. Watanabe, and Y. Muto, IEEE Trans. Magn., MAG-2L316 (1985).

T09] D. S. Beard, W. Klose, S. Shimamoto, and G. Vecsey, Fusion Eng. Des., 7(1-2) (1988).

110] V. V. Sytchev, V. B. Zenkevitch, V. V. Andrianov, and V. P. Baev, Sov. Phys. JETPLett, 9:31 (1969).

Til] K. Kaiho, K. Koyama, and I. Todoriki, Preprint, Int. Institute of Frigenation, Commission I Meeting, Tokyo (1970).

112] H. Tsuji, Y. Takahashi, K. Okuno, L. Dresner, E. Tada, T. Ando, R. S. Kensley, T. Ogasawara, K. Kuroda, T. Sato, T. Hamajima, S. Shimamoto, and K. Yasukochi, Proc. 9th Symp. Engineering Problems Fusion Research, IEEE, New York, pp. 2035-2038 (1981).

113] S. Forster, G. Friesinger, U. Jeske, H. Katheder, G. Nother, A. Nyias, G. Schenk, C. Schmidt, L. Siewerdt, M. Susser, A. Ulbricht, and F. Wuchner, Proc. 9th Int. Conf. on Magnet Technology, Zurich, pp. 32-35 (1985).

114] T. Ogasawara, J. Phys., 45:C 1-443 (1984).

115] Y. Dubots, A. Fevrier, J. C. Renard, J. P. Tavergnier, J. Goyer, and H. G. Ky,

IEEE Trans. Magn., MAG-21:117 (1985).

116] I. Hlasnik, Proc, Int. Symp. on Flux Pinning and Electromagnetic Properties

in Superconductors, Fukuoka, Japan, pp. 247-253 (1985).

117] K. Yamafuji, M. Iwakuma, Y. Tomita, K. Funaki, F. Sumiyoshi, T. Fukutsuka,

R. Ogawa, T. Miyatake, and K. Matsumoto, IEEE Trans. Magn., MAG-

23:1355. (1986).

118] Y. Kubota, H. Okon, and T. Ogasawara, Cryogenics, 26:654 (1986).

119] S. Foner, and B. Schwartz (eds), "Superconducting Machines and Devices,

Large Scale Applications", Plenum, N.Y., 1974.

120] P. R. Critchlow, and R. Roberge, "Applied Superconductivity Past Present And

Future", Superconductivity and Superconducting Materials Technologies, P.

Vincenzini (Editor), p. 613, Techna Sri, 1995.

121] R. E. Jones, et al, Adv. Cryog. Eng, 15:141 (1970).

262

[122] S. G. Wojcicki, "The Superconducting Supercollider", Physics Today, 38(1), 1985, S32-S33.

[123] B. Schwarzschild, "Panel Ceaffirms High-Field Magnet Choice for Supcollider", Physics Today, 39(7), July 1986, p.21.

[124] D. Larbalestier, et al., "High-Field Superconductivity", Physics Today, 39(3), 28-29, 1986.

[125] D. Lambrecht, IEEE Trans. Magn., MAG-17, 5, Sept., 1981.

[126] N. Maki, T. Sanematsu, and H. Ogata, IEEE Trans. Power Appa. Sys., PAS-99, 1 Jan/Feb., 1980.

[127] J. L. Smith, J. L. Kirtley, and P. Thullen, IEEE Trans. Magn., MAG-11, 128 (1975).

[128] H. H. Woodson, A. J. J. Stekly, and E. Halas, IEEE Trans. Power Appar.

Syst, PAS-85, 264 (1966). (A. D. Appleton, J. S. H. Ross, J. Bumby, and A. J. Mitcham, IEEE Trans. Magn., MAG-13, 770 (1977).)

[129] R. W. Boom, B. C. Haimson, G. E. Mcintosh, H. A. Peterson, and W. C. Young, IEEE Trans. Magn., MAG-11(2):475 (1975).

[130] W. V. Hassenzahl, IEEE Trans. Magn., MAG-11(2).482 (1975).

[131] S. M. Schoenung, W. V. Hassenzahl, and P. G. Filios, "U.S. program to develop superconducting magnetic energy storage", Intersoc. Energy Conv. Engng. Conf, ASME, Denver (1988) p. 537. M. Masuda, Moden Power Systems, (March), p. 58 (1987).

[132] R. J. Loyd. T. E. Walsh, and E. R. Kimmy, "Key design selections for the 20.4 M W h SMES/ETM', IEEE Trans. Magn., 27:1712 (1991). M. K. Abdelsalam, and Y. M. Eyssa, "Quench protection for the 21 M W h ETM coil", IEEE Trans. Magn., 27:2316 (1991).

[133] H. J. Boenig, and J. F. Hauer, "Commissioning tests of the Bonneville Power Administration 30 MJ Superconducting Magnetic Energy Storage Unit", IEEE Trans. Power Apparatus and Systems, Vol. PAS-104, No. 2, pp. 302-312, February, 1985. (Y. Mitani, K. Tsuji, and Y. Murakami, "Application of superconducting magnetic energy storage to improve power system dynamic

performance", IEEE Trans. Power Systems, Vol. 3, No. 4, pp. 1418-1425,

November, 1988.)

[134] J. R. Powell, and G. T. Damby, Mech. Eng, 89(11):30 (1967).

[135] T. Ohtsuka, and Y. Kyotani, IEEE Trans. Magn., MAG-15(6):1416 (1979).

[136] Special M A G L E V issue, Cryogenics, 15 (1975).

[137] M. Wood, "Trials and triumphs of superconductivity: The making of Oxford

Instruments", IEEE Trans. Appl. Supercond, Vol. 5, No. 2, pp. 152-162, June

1995.

[138] E. T. Laskaris, R. Ackermann, B. Dorri, D. Gross, K. Herd, and C. Minas, "A

Cryogen-Free Open Superconducting Magnet For Interventional MRI

Applications", IEEE Trans. Appl. Supercond, Vol. 5, No. 2, pp. 163-168, June

1995.

263

139] R. C. Ferguson, and W . D. Phillip, Science, 157 (1967) 257.

140] U. Ts. Andres, "New Trends in Technique for the Magnetic Separation and

Cleaning of Granular Mixtures", 6th Inter. Conf. on Mag. Tech., (MT-6), pp. 291-300, 1977.

141] J. A. Oberteuffer, "High Gradient Magnetic Separation: Basic Principles, Devices and Applications", Proc. Conf. USA Industrial Applications of Magnetic Separator, p. 3-7, 1979.

142] H. Kolm, "High Gradient Magnetic Separation", Scientifc American, p. 47-54, 1975.

143] R. Gerber, "Some Aspects of the Present Status of H G M S " , IEEE Trans, on Magnetics, MAG-18(3): 812-816, May, 1982. (P. G. Marston, The Use of

Electromagnetic Fields for the Separation of Materials, 1978, p. 1-14.)

144] Mostafa M. Abdelsalam, "A Centrifugal channel-straigh wire supercconducting magnetic ore separator", A Preliminary Proposal, 1980, p. 1-64.

145] J. M . Pastrana, "Magnetic Properties of Natural Hematite and Goethite", The Society of Min. Eng. of AJJVIE of Trans., 1977, 262(1), p. 1-5.

146] C. P. Bean, "Theory of Magnetic Filtration", Magnetic Filtration American Physical Society, Bulletin, 1971, 16, p.3 50.

147] J. A. Oberteuffer, "Magnetic Separation: A Review of Principles, Devices and

Applications", 1974, p.223-238.

T48] E. Maxwell, "Magnetie Separation-The Prospects for Superconductivity",

Cryogenics, 1975, 15(4), p. 179-184.

149] Z. J. J. Stekly, "High Intensity Magnetic Minerals Separator", IEEE Trans.

Mag., 1976, MAG-12, 5, p.493-497.

T50] R. Houot, "Separateur Magnetiqne a Solenoide Supraconducteur", Industrie

Minerale-Les Techniques, 1982, Fevrier, p.51-61.

T51] Z. J. J. Stekly, "Coal Preparation Using Magnetic Separation", EPRICS-1517,

1981, 2, p. 1-94.

152] H. K. Collan, "Superconducting Open-Gradient Magnetic Separator", Trans-

Inst. Min. Metallur. Sect. C, 1982, No. 91, March, p. c5-cll.

153] J. Svoboda, Perspektivy Supravodivosti V Magnetickem Rozdruzoani

Mineralu, Rudy, 1978, p. 62.

154] H. Cohen, and J. A. Good, "The Principles and Operation of a Very High

Intensity Magnetic Minerals Separator", IEEE Trans. Mag, 1976, Mag-12,

No. 5, p. 552.

155] D. A. Buck, Proc. IRE, 44:482 (1956).

156] R. C. Jaklevic, J. Lambo, A. H. Silver, and J. E. Mercereau, Phys. Rev, Lett,

12:159(1964).

T57] D. N. Langenberg, W . H. Parker, and B. N. Taylor, Phys. Rev., 150:186

(1966).

158] J. Matisoo, Proc. IEEE, 55:172 (1967).

264

[159] Z. Z. Sheng, and A. M. Hermann, Nature, 332 (1988) 138.

[160] A. Schilling, et al., Nature, 363 (1993) 56.

[161] M. K. Wu, J. R. Ashburn, C. J. Torng. P. H. Hor, R. L. Meng, L. Gao, Z. J. Huang, Y. Q. Wang, and C. W. Chu, Phys. Rev. Lett, 58 (1987) 908.

[162] H. Maeda, Y. Tanaka, M. Fukutomi, and T. Asano, Jpn. J. Appl. Phys., 27 (1988) L209.

[163] T. Hikata, K. Sato, and H. Hitotsuyanagi, Jpn. J. Appl. Phys., 28 (1989) L82.

[164] T. Hikata et al., Proceedings of the International Workshop on Superconductivity, Jun. 6-9, 1994, Kyoto, Japan, p. 69.

[165] R. B. Poeppel, B. K. Flandermeyer, J. T. Dusek, and I. D. Bloom, "Chemistry of High Temperature Superconductors", ACS Series 351, p.261.

[166] S. Tin, T. H. Tiefel, R. B. van Dover, and D. W. Johnson, Jr., Appl. Phys. Lett., 51, 203(1987).

[167] R. Haider, Y. Z. Lu, and B. G. Giessen, Appl. Phys. Lett., 51, 538 (1987).

[168] S. Tin, T. H. Tiefel, R. B. van Dover, and D. W. Johnson, Jr., Appl. Phys. Lett., 51, 203(1987).

[169] H. Krauth, K. Heine, J. Tenbrink, M. Wilhelm, and H.-W. Neumuller, Proc. ISS'90, Sendai, Japan (1990).

[170] K. Sato, T. Hikate, H. Mukai, M. Ueyama, N. Shibuta, T. Kato, T. Masuda, M. Nagata, K. Iwata, and T. Mitsui, IEEE Trans. Magn., MAG-27T231

(1991).

[171] K. H. Sandhage, G. N. Riley, Jr., and W. L. Carter, "Critical Issues in the OPIT Processing of High-Jc B S C C O Superconductors", JOM, Vol. 43, pp. 21-24

(1991).

[172] J. Tenbrink, M. Wilhelm, K. Heine, and H. Krauth, "Development of High-Tc Sperconductor Wires for Magnet Applications", IEEE Trans. Magn., Vol. 27, No. 2 (1991). (P. Haldar and L. R. Motowidlo, "Processing High Critical Current Density Bi-2223 Wires and Tapes", JOM, Vol. 44, pp. 54-58 (1992).)

[174] C. J. Christopherson, and G. N. Riley, Jr, "Development of twisted high-temperature superconductor composite conductors", Appl. Phys. Lett., 66 (17),

1995, pp 2277-2279.

[175] S. X. Dou, H. K. Liu, M. H. Apperley, K. H. Song, and C. C. Sorrell,

Supercond. Sci. Technol., 3, 1990, 138.

*[176] S. X. Dou, H. K. Liu, Y. C. Guo, and J. X. Jin, "Enhanced Flux Pinning Through A Controlled Melting Process In Ag-Sheathed Bi-Pb-Sr-Ca-Cu-0 Systems",

Proceedings of the International Ceramic Conference, Australia 1992

(AUSTCERAM 92), Vol.2, p.635-640.

[177] R. Bhasale, H. K. Liu, and S. X. Dou, "Effect Of Degree Of Reduction In

Thermomechanical Processing On The Transport Characteristics Of Ag-Clad

(Bi,Pb)2Sr2Ca2Cu30io+x High-Tc Superconducting Wires And Tapes", Proceedings Of The International Ceramics Conference, Sydney, Australia, 25-27

July, 1994, Vol. 1, No.2, pp. 1177-1180.

265

[178] S. X. Dou, H. K. Liu, R. Bhasale, M. Ionescu, Q. Y. Hu, E. Babic, and I.

Kusevic, "Fabrication And Properties of Ag-Sheathed Bi-2223 Superconducting Tapes", Superconductivity And Superconducting Materials Technologies, ed., P. Vincenzini, Techna Sri, 1995, pp. 227-234.

[179] J. Yau, N. Sawides, and C. C. Sorrell, "Metal-Ceramic Intergrowths In Ag-Clad Bi-Pb-Sr-Ca-Cu-0 Superconductor", Proceedings of the International Ceramics Conference, Sydney, Australia, 25-27 July, 1994, Vol.1, No.2, pp. 1129-1134.

*[180] J. X. Tin, C. Grantham, H. K. Liu, and S. X. Dou, "The electrical behaviour of high Tc superconducting ceramic bulks and their application in current limiting devices", Proceedings of the International Ceramics Conference, Sydney,

Australia, 25-27 July, 1994, Vol.1, No.2, pp. 1205-1209.

*[181] H. K. Liu, J. Yau, Y. C. Guo, Q. Y. Hu, J. X. Jin, N. Sawides, and S. X. Dou, "Long Ag-clad Bi-based superconducting tapes and coils", Physics B, 194-196 (1994) 2213-2214.

[182] T. Hikata, T. Nichikawa, H. Mukai, K. Sato, and M. Hiotsuyanayi, Jpn. J. Appl. Phys., 1989, 28:2881. (J. B. Goodenough, and A. Manthiram, J. Solid State Chem., 1990, 88, 115.)

[183] S. X. Dou, H. K. Liu, J. Wang, M. H. Apperley, C. C. Sorrell, S. J. Guo, B. Loberg, and K. E. Easterling, Physica C, 1990, 172, 63. (Raveau, B., Michel,

C , and Hervieu, M., J Solid State Chem., 1990, 88, 140.)

[184] K. Heine, J. Tenbrink, and M. Thoner, Appl. Phys. Lett, 1989, 55:2441. (N. Enomoto, H. Kikuchi, N. Uno, H. Kunakura, K. Togano, and K. Watanabe, Jpn. J. Appl. Phys., 1990, 29, L447.)

[185] N. Sawides, A. Katsaros, and S.X. Dou, Physica C, 1991, 179:361.

[186] S. X. Dou, H. K. Liu, K. E. Easterling, S. J. Guo, and X. G. Li, J. Materials

Research, 6, 2287 (1990).

[187] Y. Yamada, B. Oberst, and R. Flukiger, Supercond. Sci. Technol., 1991, 4,

165.

[188] S. X. Dou, et al., Phys. Rev. B, 40 (1989), p. 5266. (M. Takano, et al., Jpn.J Appl. Phys., 27 (1988), p. 1041.) (L. Pierre, et al., J. Appl. Phps., 68 (1990), p.

2296.)

[189] P. Krishnaraj, et al., Pysica C, 215 (1993), p. 305. (B Sarkar, et al., Mat. Res. Bull., 28 (1993), p. 263.) (S.E. Dorris, et al., Physica C, 212 (1993), p. 66.)

[190] K. Aota, et al., Jpn. J. Appl. Phys., 28 (1989), p. L2196. (J. C. Grivel, et al.,

Supercond. Sci. Technol, 6 (1993), p 725. (A. Oota, et al., Jpn. J. Appl. Phys.,

28(1989), p. LI 171.)

[191] Q. Li, K. Brodersen, H. A. Hjuler, and T. Freltoft, "Critical current density

enhancement in Ag-sheathed Bi-2223 superconducting tapes", Physica C, 111,

1993, pp 360-366.

[192] K. Sato, N. Shibuta, H. Mukai, T. Hikata, et al., Adv. Superconductivity IV:

proc. ISS"9I (Eds Hayakawa, H. and Koshizuka, N.) Springer, Tokyo, Japan

(1992) 559.

266

[193] K. Sato, N. Shibuta, H. Mukai, T. Hikata, M. Ueyama, T. Kato, and J.

Fujikami, "Bismuth superconducting wires and their application", Cryogenics, 33 (1993) pp. 243-246.

[194] U. Balachandran, A. N. Iyer, J. Y. Huang, R. Jammy, P. Haidar, J. G. Hoehn,

Jr., G. Galinski, and L. R. Motowidlo, "Recent Advances in Bismuth-based superconductors", JOM, 46 (12), Dec, 1994, pp 23-25.

[195] R. Hawsey, and J. Daley, DOE's Program on the Application of High-Tc Wire, JOM, 47 (8), Aug., 1995, pp 56-58.

[196] U. Balachandran, A. N. Iyer, et al, Recent Advances in Bismuth-based superconductors, JOM, 46 (12), Dec, 1994, pp 23-25.

[197] U. Balachandran, et al., Appl. Supercond, 2 (1994), p. 251.

[198] M. Okada, A. Okayama, T. Morimoto, T. Matsumoto, K. Aihara, and S. Matsuda, Proc. 5th U.S.-Japan Workshop, Fukuoka, Japan, pl50, 1987.

[199] K. Osamura, T. Takayama, and S. Ochiai, Appl. Phys. Lett, 55:396, 1989.

[200] H. Krauth, K. Heine, J. Tenbrink, M. Wilhelm, and H. W. Neumuller, Proc. ISS'90, Sendai, Japan, 1990.

[201] N. Enomoto, Proc. ISS'90, Sendai, Japan, 1990.

[202] K. Togano, H. Kumakura, D. R. Dietrich, H. Maeda, J. Kase, and K. Nomura, Proc. ISS'90, Sendai, Japan, 1990.

[203] K. Sato, et al., Proc. ISS'90, Sendai, Japan, 1990; IEEE Trans. Magn., MAG-27:1231 (1991).

[204] K. Aihara, T. Matsumoto, M. Okada, S. Matsuda, and M. Seidou, IEEE Trans.

Magn., MAG-27:894 (1990).

[205] G. Reis, "Magnet Technology and Conductor Design with High Temperature Superconductors", Cryogenics, 33 (1993), 609-614.

[206] J. P. Singh, J. Joo, N. Vasanthamohan, and R. B. Poeppel, "Role of Ag Additions in the Microstructural Development, Strain Tolerance, and Critical Current Density of Ag-Sheathed BSCCO Superconducting Tapes", J. Matrr,

Res., 8 (1993), 2458-2464.

*[207] S. X. Dou, Y. C. Guo, J. Yau, J. X. Tin, and H. K. Liu, "Mechanical properties of Ag-clad Bi-based superconducting composite tapes", Physics B, 194-196 (1994)

1577-1578.

[208] K. Sato, T. Hikata, H. Mukai, M. Ueyama, N. Shibuta, T. Kato, T. Masuda, M. Nagata, K. Iwata, and T. Mitsui, "High-Jc Silver-Sheathed Bi-Based

Superconducting Wires", IEEE Trans. Mag, 27 (1991), 1231-1238.

[209] R. Flukiger, B. Hensel, A. Jeremie, A. Perm, and J. C. Grivel, "Processing of

Classical and High Tc Superconducting Wires", Appl. Supercond, 1 (1993),

709-723.

[210] H. Mukai, N. Shibuta, T. Masuda, T. Hikata, et al., Adv. Superconductivity III:

proc. ISS"90, eds K. Kajimura, and H. Hayakawa, Springer, Tokyo, Japan

(1992) 559.

267

[211] K^Sato, N. Shibuta, H. Mukai, T. Hikata, M. Ueyama, T. Kato, and J.

Fujikami, "Bismuth superconducting wires and their application", Cryogenics (1993) 33:243.

[212] U. Balachandran, A. N. Iyer, J. Y. Huang, R. Jammy, P. Haidar, J. G. Hoehn, Jr., G. Galinski, and L. R. Motowidlo, "Recent advances in bismuth-based superconductors", JOM, (1994) 46:23.

[213] Y. Enomoto, T. Murakami, M. Suzuki, and K. Moriwaki, Jpn. J, Appl. Phys., (1989) 26:L1248. (S. Martin, A. T. Fiory, P. M. Fleming, G. P. Espinosa, and A. S. Cooper, Appl. Phys. Lett, (1989) 54:72.) (Y. lye, T. Tamegai, T. Sakakibara, T. Goto, N. Miura, H. Takeya, and H. Takei, Physica C, (1988) 153-155:26.) (S. Martin, A. T. Fiory, P. M. Fleming, L. F. Schneemeyer, and J. V. Waszczak, Phys. Rev. Lett, (1988) 60:2194.)

[214] T. T. M. Palstra, B. Batlogg, L. F. Schneemeyer, and J. V. Waszczak, Phys. Rev. Lett, (1988)61:1662.

[215] Y. Yeshurum, and A. P. Malozamoff, Phys. Rev. Lett, (1988) 60:2202.

[216] P. L. Gammel, L. F. Schneemeyer, J. V. Waszczak, and D. J. Bishop, Phys. Rev. Lett, (19SS) 61:1666.

[217] E. W . Collings, Cryogenics 28, 724-733 (1988).

[218] T. Ogasawara, Cryogenics 29, 3-5 (1989).

[219] C. L. Tien, M. I. Flik, and P. E. Phelan, Cryogenics 29, 602-609 (1989).

[220] D. Ito, Cryog Eng. Jpn., 22, 383-385 (1987).

[221] E. W . Collings, Adv. Cryo. Eng, (Materials) 34, 639-646 (1988).

[222] Y. Iwasa, IEEE Trans. Magn., 24, 1211-1214 (1988).

[223] S. Zannella, etal., "AC self-field losses in BiSrCaCuO superconducting wires",

Physics and Materials Science of High Temperature Superconductors I, Eds. R. Kossowski et al., Kluwer Academic, The Netherlands, 1992, pp. 633-642.

[224] E. W. Collings, et al., "AC loss and dynamic resistance of a high-Tc strand carrying a direct current in a transverse A C magnetic field", Adv. in Cryogenics Eng, (Materials), Vol. 38, pp.883-891, eds. F. R. Fickett, and R. P. Reed, Plenum Press, N e w York, 1992.

[225] J. Gannon, et al., "AC losses in Bi-2223 composite tapes", Adv. in Cryogenic

Eng., Vol. 40, p.45, R. P. Reed et al., Plenum Press, New York, 1994.

[226] T. Kato, et al., Adv. in Superconductivity VI, (Tokyo: Springer, 1994), pp.

1273-1276.

[227] H. Fujishiro, et al., IEEE Trans. Magn., 30 (1994), p. 1645.

[228] P. F. Herrmann, C. Cottevieille, G. Duperray, A. Leriche, T. Verhaege, C.

Albrecht, and J. Bock, "Cryogenic load calculation of high Tc current lead,"

Cryogenics (1993) 33(5):555-562.

[229] P. F. Herrmann, E. Beghin, G. Bottini, C. Cottevieille, A. Leriche, T.

Verhaege, and J. Bock, "Test results of a 1 kA (2 kA)-20 kV H T S C current

lead model", Cryogenics, (1994) 34(6):543-548.

268

[230] T. Kato, N. Shibuta, K. Sato, T. Masuda, Y. Hosoda, S. Terai, and E. Haraguchi, (1993) Adv. in Superconductiviy, (Proc ISS'92), p. 1239.

[231] T. Kato, K. Sato, T. Ando, T. Isono, and H. Tsuji, "Development of 5m long

and 1 kA go & return superconducting bus bar," (Submitted to the Proceedings of IS S'94, Adv. in Superconductiviy, Japan)

[232] T. P. Beales, C. M. Friend, L. Le Lay, M. Molgg, C. Dineen, D. M. Jacobson,

S. R. Hall, M. R. Harrison, P. F. Hermann, A. Petitbon, P. Caracino, L. Gherardi, P. Metra, G. Bogner, and H.-W. Neumuller, "Conductor development suitable for H T S C cables," Supercond. Sci. Technol., 8 (1995) 909-913.

[233] R. Hawsey, and J. Daley, DOE's Program on the Application of High-Tc Wire, JOM, 47 (8), Aug., 1995, pp 56-58.

[234] K. Funaki, et al., "Design and Construction of a 500 kVA-Class Oxide Superconducting Power Transformer Cooled by Liquid Nitrogen", prepared for Proc. ICEC16/ICMC, Kitakyushu, Japan, 1996.

[235] G. B. Lubkin, "Power Applications of High-Temperature Superonductors", Phys. Today, Vol. 49, No. 3, pp 48-51 (1996).

[236] E. M. Leung, .and A. Rodriquez, "Developing the H T S SC Fault Current Limiter," Superconductor Industry, Vol. 9, No. 1, pp. 20-23 (1996).

[237] 2.4 kV Current Limiter Prototype, News Release, American Superconductor Corporation, Westborough, M A , November 1995.

[238] J. D. Edick, R. F. Schiferl, and H. E. Jordan, "High Temperature Superconductivity Applied to Electric Motors", IEEE Trans. Appl. Superc,

Vol. 2, No. 4, December, 1992.

[239] D. J. Waltman, and M. J. Superczynski. Jr., "High-Temperature Superconducting Magnet Motor Demonstration", IEEE Trans. Appl. Superc,

Vol. 5. No. 4., Dec, 1995.

[240] R. E. Dietrich, F. Berg, and E. Thuries, "Solution Brought by the One-Cycle Circuit Breaker to Some Network Stresses", International Conference on Large

High Voltage Electric Systems, (Paris), No. 13-09, August 30-September 7, 1978.

[241] E. Thuries, V. D. Pham, Y. Laumond, T. Verhaege, A. Fevrier, M. Collet, and M. Bekhaled, "Towards the Superconducting Fault Current Limiter", IEEE

Transactions on Power Delivery, Vol. 6, No. 2, April 1991.

[242] C. A. Falcone, J. E. Beehler, W . E. Mekolites, and J. Grzan, "Current Limiting Devices", A Utility's need, IEE - PAS, Nov/Dec, 1974, pp 1768-1775. (V. H.

Tahiliani and J. W . Porter, "Fault Current Limiters - An overview of EPRI

Research", IEE - PAS, Sept/Oct, 1980, pp. 964-1969.)

[243] B. P. Raju, K. C. Parton, and T. C. Bartram, "Fault current limiting reactor with

superconducting D C bias winding", CIGER 1982, 1-9 September, No. 23-03.

[244] K. E. Gray, and D. E. Fowler, "A superconducting fault-current limiter", J. Appl.

Phys., 49 (4) (1978) 2546-2550.

[245] J. R. Gavaler, M. A. Janocko, A. Patterson, and C. K. Jones, J. Appl. Phys., 42

(1971) 54.

269

[246] V. L. Newhouse, Applied Superconductivity, Wiley, N e w York, 1964, p. 132. (O. K. Mawardi, U.S. patent 3,384,762.)

[247] P. Tixador, Y. Brunet, J. Leveque, and V. D. Pham, "Hybrid superconducting a.c. fault current limiter principle and previous studies", IEEE Transactions on Magnetics, Vol. 28, No. 1, pp. 446-449, January 1992.

[248] D. Ito, K. Tsurunaga, E. S. Yoneda, Y. Sugiyama, T. Hara, K. Okaniwa, H.

Hoshino, and T. Yamamoto, "Superconducting Fault Current Limiter Development", IEEE Transactions on Magnetics, Vol. 27, No. 2, pp. 2345-2348, March, 1991. (D. Ito, E. S. Yoneda, K. Tsurunaga, T. Tada, T. Hara, T. Ohkuma, and T. Yamamoto, "6.6 kV/1.5 kA-Class Superconducting Fault Current Limiter Development", IEEE Transactions on Magnetics, Vol. 28, No. 1, pp. 438-441, January, 1992.)

[249] H. J. Boenig, and D. A. Paice, "Fault Current Limiter Using a Supercoducting Coil", IEEE Transaction on Magnetics, Vol. MAG-19, No.3, pp. 1051-1053, May, 1983.

[250] S. Shimizu, et al, "Single-Line-to-Ground Fault Test of a 3-Phase Superconducting Fault Current Limiting Reactor", IEEE Transactions on Magnetics, Vol. 28, No. 1, January 1992.

[251] T. Ishigohka, and N. Sasaki, "Fundamental Test of N e w D C Superconducting Fault Current Limiter", IEEE Transactions on Magnetics, Vol. 27, No. 2, March

1991.

[252] R. W . Boom, and H. A Peterson, "Superconductive Energy Storage for Power Systems", IEEE Trans. Magnetics, Vol. MAG-8, No. 3, pp. 701-703, September,

1972.

[253] Y Matani, et al., "Fundamental study of the system capacities of superconducting magnetic energy storage for the stabilisation of power system", Electrical Engineering in Japan, Vol 108, pp 39-45, 1989.

[254] M. K. Abdelsalam, R. W . Boom, H. A. Peterson, D. J. Helfrecht, and W . W. Bratley "Operational Aspects of Superconductive Magnetic Energy Storage (SMES)", IEEE Trans. Magnetics, Vol. MAG-23, No. 5, pp. 3275-3277,

September, 1987.

[255] C. Chen, and W . J. Bonwick, "Design Principles For S M E S Magnet For Power System Stabiliser", Proc. AUPCEC9I, pp. 66-71, Monash University,

Australia, 3-4 Oct., 1991.

[256] K Koyama, et al., "The relationship between stored energy and current density of superconducting magnets", Cryogenic Engineering, Japan, Vol 14, 1979, pp

32-37.

[257] W . Paul, Th. Baumann, J. Rhyner, and C. F. Platter, "Tests of 100 k W High-Tc Superconducting Fault Current Limiter", IEEE Trans, on Applied

Superconductivity, Vol. 5, No. 2, pp. 1059-1062, June, 1995.

[258] K. Ebihara, T. Dcegami, M. Akazaki, and H. Nakajima, 3rd FED Workshop on HiTcSc-ED, Kumamoto, Japan, 44 (1991). (T. Ikegami, H. Nakajima, et al.,

"Application of High-Tc Superconductor to Current Limiting Devices", manuscript

received August, 1992.)

270

[259] J. Acero, L. Garcia-Tabares, M. Bajko, and J. Calero, "Current Limiter Based

On Melt Processed Y B C O Bulk Superconductors", IEEE Trans. Appl. Supercond, Vol. 5, No. 2, June, 1995, pp. 1071-1074.

[260] V. Meerovich, V. Sokolovsky, and J. Bock, "Quenching in a high-Tc superconducting ring", IEEE Trans. Appl. Supercond, Vol. 5, No. 1, March, 1995, pp. 22-25.

[261] E. M. W. Leung, G. W. Albert, M. Dew, P. Gurrola, K. Muehleman, B.

Gamble, C. Russo, G. Dishaw, H Boenig, D. Peterson, and A. Rodriguez, "High temperature superconducting fault current limiter for utility applications," Proceedings of the International Workshop on Fault Current Limiters, pp. 155-163, Jerusalem, Israel, April 5, 1995.

[262] J. Tenbrink, M. Wilhelm, K. Heine, and H. Krauth, "Development of High-Tc Superconductor Wires for Magnet Applicatians", IEEE Trans. Mag, Vol. 27, No. 2 (1991).

[263] J. W . Lue, L. Dresner, S. W . Schwenterly, D. Aized, J. M. Campbell, and R. E. Schwall, "Stability Measurements on a 1-T High Temperature Superconducting Magent", IEEE Trans. Appl. Supercond, Vol. 5, No. 2, June, 1995.

[264] N. Schiff, and E. Schempp, "HTS Conductor Developments for Power Applications", Superconductor Industry, Vol, 7, No. 2 (1994).

[265] G. B. Lubkin, "Applications of High-Tempreature Superconductors: Approach the Marketplace", Phys. Today, Vol. 48, No. 3, pp. 20-23 (1995).

[266] R. B. Stephens, "High Tc Wire Design and Power Losses on Cycling SMES Devices", IEEE Trans. Appl. Supercond, Vol. 1, No. 2, pp. 122-129, Sept.,

1992.

[267] D. J. Waltman, and M. J. Superczynski. Jr., "High-Temperature Superconducting Magnet Motor Demonstration", IEEE Trans. Appl.

Supercond., Vol. 5, No. 2, Dec, 1995. (D. U. Gubser, "Superconducting Coil Development and Motor Demonstration: Overview", J. Electronic Materials,

Vol. 24, No. 12, 1995.)

[268] Y. Kamisada, T. Koizumi, M. Satou, and Y.Yamada, "R&D of Ag-Sheathed Bi-2223 Superconducting Tapes", IEEE Trans. Mag, Vol. 30, No. 4, pp.

1675-1679, July, 1994.

[269] H. Ogiwara, Y. Yamada, M. Satou, T. Kitamura, and T. Hasegawa, "Critical current density and best transport current of an Ag sheathed Bi-2223 Tape conductor", Superconductivity and Superconducting Materials Technologies,

ed., P. Vincenzini, Techna Sri, 1995, pp. 149-156.

[270] U. Balachandran, A. N. Iyer, C. A. Youngdahl, L. R. Motowidlo, J. G. Hoehn, Jr., and P. Haldar, "Fabrication and Characterization of High-Tc Tapes and

Coils Made from Ag-Clad Bi-2223 Superconductors", Advanced Materials VI: Superconductors, Surfaces and Superlattices, (eds. H. Sakaki et al.), Vol. 19A,

pp. 385-389, 1994.

[271] K. H. Sandhage, G. N. Riley, Jr., and W . L. Carter, "Critical Issues in the OPIT Processing of High-Jc B S C C O Superconductors", JOM, Vol. 43, 1, 21-24

(1991).

271

[272] P. Haldar, and L. R. Motowidlo, "Processing High Critical Current Density Bi-2223 Wires and Tapes", JOM, Vol. 44, pp. 54-58 (1992).

[273] R. A. Hawsey, R. S. Sokolowski, P. Haldar, and L. R. Motowidlo,

"Development of Practical High Temperature Superconducting Wire for Electric Power Applications", Proceedings of the Fourth International

Conference and Exhibition: World Congress on Superconductivity, Vol. II, pp. 624-631, Orlando, Florida, USA, June 27 -July 1, 1994.

[274] M. J. Minot, W . L. Carter, J. J. Gannon, Jr., R. S. Hamilton, P. K. Miles, D. R. Parker, G. N. Riley, Jr., M. Rupich, M. Teplitsky, E. D. Thompson, and K.

Zafar, "Advances in long-length B S C C O H T S Multifilamentary composit wire development", Adv. Cryog. Eng, Vol. 40, pp. 131-137, 1994. (U. Balachandran, A. N. Iyer, P. Haldar, J. G. Hoehn, Jr., and L. R. Motowidlo, "Progress in Development of Tapes and Magnets Made from Bi-2223 Superconductors", Proceedings of the Fourth International Conference and Exhibition: World Congress on Superconductivity, Vol. II, pp. 639-649, Orlando, Florida, USA, June 27 -July 1, 1994.)

[275] Y. C. Guo, H. K Liu, M. Ionescu, and S.X. Dou, "Ag-clad

(Bi,Pb)2Sr2Ca2Cu30io+x high Tc superconducting wires and tapes", Proceedings of the international ceramics conference, Sydney, Australia, 25-27 July, 1994, Vol. 1, No. 2, pp. 1228-1233. (S. X. Dou and H. K. Liu, "Ag-sheathed Bi(Pb)SrCaCuO superconducting tapes", Supercond. Sci. Technol., Vol. 6, pp. 297-314, 1993.)

[276] M . N. Cuthbert, M. Dhalle, J. Thomas, A. D. Caplin, S. X. Dou, Y. C. Guo, H. K. Liu, R. Flukiger, G. Grasso, W . Goldacker, and J. Kessler, "Transport and Magnetisation Measurements of Bi2223/Ag Tapes and the Role of Granularity on Critical Current Limitation", IEEE Trans. Appl. Supercond, Vol. 5, No. 2,

pp. 1391-1394, June, 1995.

[277] U. Balachandran, A. N. Iyer, P. Haldar, J. G. Hoehn, Jr., and L. R. Motowidlo "Progress in development of tapes and magnets made from Bi-2223 superconductors", Proceedings of the Fourth International Conference and Exhibition: World Congress on Superconductivity, Vol. II, pp. 639-649,

Orlando, Florida, USA, June 27 -July 1, 1994.

[278] K. C. Goretta, J. A. Cluff, J. Joo, M. T. Lanagan, J. P. Singh, N. Vasanthamohan, Y. Xin, and K. W . Wong, "Mechanical Properties of High-Temperature Superconducting Wires", Proceedings of the Fourth International Conference and Exhibition: World Congress on

Superconductivity, Vol. II, pp. 633-637, Orlando, Florida, USA, June 27 -July

1, 1994.

[279] J. P. Singh, J. Joo, N. Vasanthamohan, and R. B. Poeppel, "Role of Ag Additions in the Microstructural Development, Strain Tolerance, and Critical Current Density of Ag-Sheathed B S C C O Superconducting Tapes", J. Mater,

Res., Vol. 8, No. 10, pp. 2458-2464 (1993).

[280] J. Schwartz, J. K. Heuer, K. C. Goretta, R. B. Poeppel, J. Guo, and G. W . Raban, Jr., "High Temperature Mechanical Properties and High-Sheaths for

Powder-In-Tube Tapes", Appl. Supercond, Vol. 2, No. 3/4, 271-280 (1994).

272

[281] S. Patel, S. Chen. T. Haugan. F. Wong, and D. T. Shaw, "Effect of bending on

Jc-8 characteristic of silver-sheathed oxide superconducting tape with Sausaging", Cryogenics, 35 (1995) 257-262.

[282] U. Balachandran, A. N. Iyer, R. Jammy, P. Haldar, J. G. Hoehn, Jr., and M.

Suenaga, "Processing and Properties of Long Lengths of Ag-Clad B S C C O Superconductors and High-Tc Magnets", Advances in Superconductivity VII, (eds. K. Yamafuji and T. Morishita), pp. 781-784, 1995.

[283] K. Sato, "The Bulk Application of High-Tc Superconductors in Japan", JOM, 47 (8), Aug., 1995, pp 65-67.

[284] E.W. Collings, "Conductor Design with High-Tc Ceramics: A Review", Advances in Superconductivity II, (Proceedings of the 2nd International Symposium on Superconductivity (ISS '89), November 14-17, 1989, Tsukuba, Japan.), eds., T. Ishiguro and K. Kajimura, pp. 327-333, Springer-Verlagm, Tokyo, 1990.

[285] M . D. Sumption, and E. W . Collings, "Stability, Mechanical Considerations, and A C Loss in H T S C Monoliths, coils, and wires", Proceedings of the Fourth International Conference and Exhibition: World Congress on

Superconductivity, Vol. 1, pp. 138-147, Orkindo, Florida, USA, June 27 -July 1, 1994.

[286] E. W . Collings, "Design of High-Tc Superconductors for D C and A C Applications", 2nd Japan International S A M P E Symposium and Exhibition, Chiba,

Japan, Dec. 11-14, 1991.

[287] E. W . Collings, K. R. Marken, M. D. Sumption, John R. Clem, S. A. Boggs, and M . V. Parish, "AC Loss and Dynamic Resistance of a High-Tc Strand Carrying a Direct Current in a Transverse A C Magnetic Field", Prepared for the

CEC/ICMC Conference, Huntsville AL, June 10-14,1991.

[288] J. Horvat, S.X. Dou, H.K. Liu, and R. Bhasale, "Critical current through strong links in Ag/Bi-Sr-Ca-Cu-0 superconducting tapes", Physic C, 271 (1996) 51.

*[289] S. X. Dou, H. K. Liu, Y. C. Guo, J. X. Jin, Q. Y. Hu, D. L. Shi, S. Salem-Sugui, and Z. Wang, "Critical current density and irreversibility behaviour in Ag-sheathed Bi-based superconducting wires fabricated using a controlled melt

procedure," ,4/7/?/. Supercond, Vol. l,Nos. 10-12, pp. 1515-1522, 1993.

[290] M . D. Johnston, J. Everett, M. Dhalle, A. D. Caplin, J. C. Moore, S. Fox, C. R. M . Grovenor, G. Grasso, B. Hensel, and R. Flukiger, "Non-Destructive Measurement of Current Distribution in H T S Conductors", High Temperature

Superconductors: Synthesis, Processing, And Large-Scale Applications, Edited by U. Balachandran, P.J, McGinn, and J.S. Abell, pp. 213-222, Minerals-

Metals-Materials (TMS) 1996 Annual Meeting, California, Feb. 4-8, 1996.

[291] J. Paasi, M . Lahtinen, J. T. Eriksson, and J. Sarkaniemi, "Magnetic Flux

Distribution Dynamics", IEEE Trans. Mag, Vol. 30, No. 4, pp. 1855-1858,

July, 1994.

[292] J. Paasi, and M . Lahtinen, "Characterization of high-Tc superconducting tapes

using Halll sensors", Physica C, Vol. 216, pp. 382-390, 1993.

273

[293] G. Grasso, and R. Flukiger, "Self-field effect and current distribution in Bi(2223) tapes at 77 K", Physica C, 253 (1995) 292-296.

[294] E. W . Collings, and M. D. Sumption, "A Studty of A C Losses in HTSC/Ag

Multifilamentary Strands", Prepared for the 7th US-Japan Workshop on High-

Tc Superconductors, National Research Institute for Metals, Tsukuba, Japan, October 24-25, 1995.

[295] J. J. Gannon Jr., A. P. Malozemoff, M. J. Minot, F. Barenghi, P. Metra, G. Vellego, J, Orehotsky, and M. Suenaga, "AC Losses in Bi-2223 Composite Tapes", Advances in Cryogenic Engineering, Vol. 40, pp. 45-52, 1994.

[296] K. Kwasnitza, and St. Clerc, "Specific Aspects Of Ac Losses In High-Tc Superconductors", Advances in Cryogenic Engineering, Vol. 40, pp. 53-61, 1994.

[297] P. Fabbricatore, G. Gemme, R. Musenich, M. Occhetto, and R. Parodi, "Simple numerical model to interpret the A C measurements on type-II superconductors", Cryogenics, Vol. 32, No. 6, pp. 559-568, 1992.

[298] Y. Yang, T. Hughes, D. M. Spiller, C. Beduz, M. Penny, R. G. Scurlock, P. Haldar, and R. S. Sokolowski, "Measurements of the self-field a.c losses in PbBi2223 tapes with different core/sheath configuration", Supercond. Sci.

Technol., 9 (1996) 801-804.

[299] M . Ciszek, A. M. Campbell, and B. A. Glowacki, "The effect of potential contact on A C loss measuurement in superconducting B S C C O tape", Physica C, 233 (1994) 203-208.

[300] J. R. Clem, and A. Sanchez, "Hysteretic A C losses and susceptibility of thin superconducting disks", Physical Review B, Vol. 50, No. 13, pp.9355-9362,

1994.

[301] A. M . Campbell, "AC Losses in High Tc Superconductors", Applied superconductivity conference, Boston, M A , U S A Oct. 16-21, 1994. (IEEE Transactions on Applied Superconductivity, Vol. 5, June 1995.)

[302] M . Ciszek, B. A.Glowacki, S. P.Ashworth, A. M. Campbell, and J. E. Evetts, "AC losses of Ag-(Bi,Pb)SrCaCuO-2223 Tapes in Combination of Transverse

External Msgnetic Field and Transport Current", Applied superconductivity

conference, Boston, M A , USA, Oct. 16-21, 1994. (IEEE Transactions on Applied

Superconductivity, Vol. 5, June 1995.)

[303] A C Loss and Macroscopic Theory of Superconductors, N e w York: Gordon

and Beach Science Publishers, 1983, pp. 70-73.

[304] S. A. Boggs, E. W . Collings, and M. V. Parish, "AC Losses in H T S C Conductor Elements", IEEE Trans. Appl. Supercond, Vol. 2, No. 3, pp. 117-

121, Sept., 1992.

*[305] J. X. Jin, "High Tc Superconducting Ceramics and Their Application in the

Design and Construction of Electrical Fault Current Limiters", Thesis for Master

of Science, from the School of Materials Science and Engineering, the

University of N e w South Wales, Australia, 1995.

274

[306] D. M . Kroeger, and A. Goyal, "Models for long-range current flow in bulk oxide superconductors", JOM, 46 (12), Dec, 1994, pp 14-19.

*[307] H. K. Liu, P. Bain, R. Brasale, J. X. Jin, Y. C. Guo, and S. X. Dou, "Investigation

of Different Angle Joining of Superconducting Tapes", The 1995 International

Workshop on Superconductivity Co-Sponsored by ISTEC and M R S , Program and Extended Abstracts, pp. 375-376, June 18-21,1995, Maui, Hawaii, USA.

[308] R. Jarnmy, A. N. Iyer, J. Y. Huang, M. Chudzik, U. Balachandran, and P. Haldar, "Recent Developments In Fabrlcatton And Properties Of Ag-Clad Bscco Conductors", High Temperature Superconductors: Synthesis, Processing, And Large-Scale Applications, Edited by U. Balachandran, P.J, McGinn and

J.S. Abell, pp. 1-11, Minerals-Metals-Materials (TMS) 1996 Annual Meeting, California, Feb. 4-8, 1996.

[309] C. M . Friend, and D. P. Hampshire, "Direct evidence for a change at 50 K in

the critical current properties of a Bi2.xPbxSr2Ca2Cu308 tape in high magnetic fields", Physica C, 252 (1995) 107-116.

[310] J. Horvat, S. X. Dou, H. K. Liu, and R. Bhasale, "Critical current through strong links in Ag/Bi-Sr-Ca-Cu-0 superconducting tapes", Physica C, 271 (1996)51-58.

[311] L. J. Giacoletto, "Electronics designers' handbook", McGraw-Hill, Inc., 1977.

*[312] J. X. Jin, C. Grantham, H. K. Liu, and S. X. Dou, "(Bi, Pb)2Sr2Ca2Cu3O10+x Ag-clad high-Tc superconducting coil and its magnetic field properties", Philosophical Magazine B, Vol. 75, No. 6, pp. 81-826, 1997.

[313] E. M . W . Leung, G. W . Albert, M. Dew, P. Gurrola, K. Muehleman, B. Gamble, C. Russo, G. Dishaw, H Boenig, D. Peterson, and A. Rodriguez, "High temperature superconducting fault current limiter for utility applications,"^. Cryog. Eng, Vol. 42, pp. 961-968, 1996.

[314] R. Giese, "Overview/Introduction to superconducting-fault-current limiters", Proceedings of the International Workshop on Fault Current Limiters, pp. 18-

90, Jerusalem, Israel, April 5, 1995.

[315] R. F. Giese, "Introduction to and overview superconducting-fault-current limiters," Proceedings of the International Workshop on Fault Current

Limiters, pp. 13-17, Jerusalem, Israel, April 5, 1995.

*[316] J. X. Tin, C. Grantham, S. X. Dou, and H. K. Liu, "Prototype fault current limiter built with high Tc superconducting coils", Journal of Electrical and Electronics

Engineering, Australia, 15 (1), March 1995, pp 117-124.

*[317] J. X. Tin, S. X. Dou, R. Bhasale, H. K. Liu, C. Grantham, N. Sawides, and M. Ionescu, "Consideration of Using H T S Coils as a Magnetic Core Bias Winding",

Advances in Cryogenic Engineering, Vol. 41, pp. 427-432,1996.

*[318] J. X. Tin, C. Grantham, X. Y. Li, Z. Y. Liu, Y. C. Guo, T. R. Blackburn, H. L. Liu, J. N. Li, H. K. Liu, Z. J. Zeng, J. Y. Liu, and S. X. Dou, "Electrical

Applications of a H T S Saturable Magnetic Core Fault Current Limiter",

Proceedings of ICEC16/ICMC, pp. 1037-1040, Kitakyushu, Japan, May 20-24,

1996.

275

*[319] J. X. Jin, C. Grantham, X. Y. Li, D. Sutanto, Z. Y. Liu, Z. J. Zeng, H. L. Liu,

H. K. Liu, Y. C. Guo, J. Y. Liu, T. R. Blackburn, and S. X. Dou, "EMTP

Analysis of a High Te Superconducting Fault Current Limiter for Electrical Application", Proceedings of the 3rd International Conference Electrical Contacts, Arcs, Apparatus and their Applications (IC-ECAAA), pp. 374-381, Xi'an, China, May, 19-22, 1997. (J. X. Jin, C. Grantham, H. K. Liu, Z. J. Zeng, T. R. Blackburn, and S. X. Dou, "Electrical Application of High Tc Superconducting Saturable Magnetic Core Fault Current Limiter", LVA-11, Conference Handbook, p. 243, ASC'96, Pittsburgh, USA, Aug. 25-28, 1996. Accepted for IEEE on Superconductivity, June, 1997.)

[320] William M. Flanagan, "Handbook of Transformer Design and Applications", Second Edition, McGRAW-HILL, INC., USA, 1993.

[321] Department of Electrical Engineering, Massachusetts institute of Technology, "Magnetic Circuits and Transformers", Massachusetts Institute of Technology, J O H N W I L E Y & SONS, INC., USA, 1957.

[322] M . G. Say, P. D. Atkinson, A. Baracz, E. H. F. Smith, J. A. Lunt, and W. S. Melville, "Magnetic Amplifiers And Saturable Reactors", G E O R G E N E W N E S

LIMITED, Printed in Britain by the Whitefriars Press Ltd., 1954.

*[323] J. X. Tin, S. X. Dou, C. Grantham, and H. K. Liu, "Preparation of high Tc

superconducting coils for consideration of their use in a prototype fault current limiter'', IEEE Transactions on Applied Superconductivity, Vol. 5, No. 2, pp.

1051-1054, June 1995.

[324] R. Roeper, "Short-circuit Currents in Three-phase systems", Siemens Aktiengesellschaft, Berlin and Munchen; John Wiley and Sons, 1985.

[325] Y. S. Wang, S. Qi, and W. J. Bonwick, "Development of VLF H V Power Supply", Proceedings Of The Australian Universities Power Engineering Conference, (AUPEC96), pp. 395-398, The University of Melbourne,

Melbourne, Australia, 2-4 October, 1996.

[326] T. Czaszejko, R. P. Lisner, F. Ahmar, and D. Rinarelli, "A Cycloconverter-Derived High-Voltage VLF Source For Cable Testing", Proceedings Of The Australian Universities Power Engineering Conference, (AUPEC'96), pp. 583-

587, The University of Melbourne, Melbourne, Australia, 2-4 October, 1996.

[327] A. H. Baghurst, and A. P. Jimenea, "A New Portable A C Power Source for Testing High Voltage Power Cables", Proceedings of the Electrical Engineering Congress, pp. 129-133, Sydney, Australia, 24-30 November,

1994.

[328] R. C. Tychsen, and Q. Su, "Evaluation of Partial Discharge Detection Techniques on Hydro-Generators, Proceedings of the Electrical Engineering

Congress, pp. 205-209, Sydney, Australia, 24-30 November, 1994.

[329] T. R. Blackburn, B. T. Phung, and R. C. Sheehy, "The Use of Computer-Based Partial Discharge Measurement for Plant Condition Monitoring", Proceedings of the Electrical Engineering Congress, pp. 211-216, Sydney, Australia, 24-30

November, 1994.

276

*[330] J. X. Tin, C. Grantham, H. K. Liu, M. Apperley, and S. X. Dou, "Electronic High

Voltage Generator with a High Temperature Superconducting Coil", Advances in Cryogenic Engineering, Vol. 41, pp. 991-996, 1996.

*[331] J. X. Tin, S. X. Dou, H. K. Liu, and C. Grantham, "A high temperature superconductor high voltage generator", High Temperature Superconductors: Synthesis, Processing, And Large-Scale Applications, Edited by U. Balachandran, P. J. McGinn and J. S. Abell, pp. 233-239, Minerals-Metals-Materials (TMS) 1996 Annual Meeting, section of H T S synthesis, processing, and large scale applications V: applications and conductor development, Feb. 7,1996.

*[332] J. X. Tin, C. Grantham, H. K. Liu, and S. X. Dou, "High Voltage Generation with

a High Tc Superconducting Resonant Circuit", LSB-4, Conference Handbook, p. 235, ASC'96, Pittsburgh, USA, Aug. 25-28, 1996. Reviewed for IEEE on Superconductivity, June, 1997.

*[333] C. Grantham, J. X. Jin, S. X. Dou, H. K. Liu, and K. J. S. Holder, "Novel electronic high voltage generator with high temperature superconducting coils", Proceedings of the International Power Engineering Conference, Singapore, 27 Feb.-l March, 1995, Vol. 2, pp. 234-238.

[334] C. Grantham, F. N. Kong, and H. Jiansheng, "Novel electronic high voltage generator", Journal of Electrical and Electronics Engineering, Australia. Vol.

11, No. 2, 1991, pp. 117-119.

[335] F. W . Grover, "Inductance Calculations", Dover Publications, Inc., N e w York,

1962.

277

PUBLICATIONS

(HTS)

J. X. Jin, C. Grantham, Y. C. Guo, J. N. Li, R. Bhasale, H. K. Liu, and S. X. Dou, "Magnetic Field Properties of Bi-2223/Ag HTS Coil at 77 K", Physica C, Vol. 278, pp. 85-93, 1997.

(FCL)

J. X. Jin, S. X. Dou, C. Grantham, and H. K. Liu, "Preparation of High Tc Superconducting Coils for Consideration of Their Use in a Prototype Fault Current Limiter", IEEE Transactions on Applied Superconductivity, Vol. 5, No. 2, pp. 1051-1054,1995.

J. X. Jin, C. Grantham, S. X. Dou, H. K. Liu, Z. J. Zeng, Z. Y. Liu, T. R. Blackburn, X. Y. Li, H. L. Liu, and J. Y. Liu, "Electrical Application of High Tc Superconducting Saturable Magnetic Core Fault Current Limiter", IEEE Transactions on Applied Superconductivity, Vol. 7, No. 2, pp. 1009-1012, 1997.

(HVG)

J. X. Jin, S. X. Dou, H. K. Liu, and C. Grantham, "High Voltage Generation with a High Tc Superconducting Resonant Circuit", IEEE Transactions on Applied

Superconductivity, Vol. 7, No. 2, pp. 881-884, 1997.

278