IAEA-TECDOC-594
Superconducting materials and magnets
Proceedings of a Specialists Meeting held Tokyo,in September4-6 1989
INTERNATIONAL ATOMIC ENERGY AGENCY IAEe Th A doe t normallsno y maintain stock f reportso thin si s series. However, microfiche copies of these reports can be obtained from
INIS Clearinghouse International Atomic Energy Agency Wagramerstrasse5 0 10 P.Ox Bo . A-1400 Vienna, Austria
Orders shoul accompaniee db prepaymeny db f Austriao t n Schillings 100, for e for e chequa th f m th IAEf m o n i o n i r eAo microfiche service coupons orderee whicb y hdma separately fro INIe mth S Clearinghouse. SUPERCONDUCTING MATERIAL MAGNETD SAN S IAEA, VIENNA, 1991 IAEA-TECDOC-594 ISSN 1011-4289
Printed by the IAEA in Austria April 1991 FOREWORD
e productioTh ultrf o e energf ano us hig y yb h temperature plasma comprised through magnetic confinemen economicalls ti y feasible only when superconducting magnets produced of superconducting materials with zero electrical resistanc successfulln eca appliee yb fusioo t d n power systems. Comparison energf so y consumption between normal conductind an g superconducting magnets have shown that only by using superconducting materials a profit could be realized in the energy balance of the system. In particular, even the question of the behaviour of superconducting materials magnetd an s expose severo t d e neutron irridiatio hugd nan e electromagnetic forces, which may impact the superconducting materials and deteriorate their properties importanf o s i , t concern. Therefore developmene ,th f o t superconducting materials and magnets is the key of the realization of the syste d intensivman e worldwide efforts have been devoted bringing these technologie practicen i s . Oxide superconducting materials, recently discovered, which show zero electrical resistance above the liquid nitrogen temperature have initiated new activities in research and development areas. These new materials might provide an additional option for superconducting magnets. But also in the field of superconducting wire manufacturing, successful investigations have been performed. For this reason, the Agency convened a Specialists' Meeting to provide an opportunity for experts to present and discuss their recent work of research, developmen d applicatiotan thin i n revieo st ared wan a gained experienced san to show next steps for further development directions. e followinTh g report contain paperl sal s presented durin meetinge th g . EDITORIAL NOTE
In preparing this material for the press, staff of the International Atomic Energy Agency have mounted and paginated the original manuscripts as submitted by the authors and given some attention to the presentation. The views expressed papers,the statementsin the general the made and style adoptedthe are responsibility namedofthe authors. necessarilyviewsnot The do reflect those governments ofthe of the Member States or organizations under whose auspices the manuscripts were produced. thisin The bookuse of particular designations countriesof territoriesor does implynot any judgement by the publisher, the IAEA, as to the legal status of such countries or territories, of their authorities institutions delimitationand the of or theirof boundaries. The mention specificof companies theirof or products brandor names does implynot any endorsement or recommendation on the part of the IAEA. Authors are themselves responsible for obtaining the necessary permission to reproduce copyright material from other sources. This text was compiled before the unification of Germany in October 1990. Therefore the names German Democratic Republic Federaland Republic Germanyof have been retained. CONTENTS
Summar Meetine th f yo g ...... 7 . H. Wada, J. Kupitz
PRACTICAL HIGH-FIELD SUPERCONDUCTORS (Session A)
Developments of advanced superconductors for fusion in Japan (A.bstract) ...... 15 TachikawaK. Technical superconducting materials ...... 17 A.D. Nikulin High-field superconducting properties and electrical resistivity of (Nb,Ti)Sn composite
conductors by the tube method ...... 3 . 27 S. Murase, Shirdki,H. Nakayama,S. Aoki,N. Ichihara,M. SuzukiE. Progress in superconducting wires in Furukawa, Japan ...... 37 Y. Tanaka Flux pinnin thiy gb n or-Ti ribbon superconductinn i s g Nb-Ti ...... 7 4 . T. Matsushita, S. Ota.be, T. Matsuno, K. Yamafuji
Strain dependenc f criticaeo l curren advancen i t d Nb3Sn superconductors ...... 5 5 . T. Okada, K. Katagiri, H. Yoshida
High fiel c characteristicdJ f in-sitso u processe processeE d VCC 3d Gaan d Nb3Al wires ...... 7 6 . Noto,K. Sato,T. Watanabe,K. Noguchi,T. Iwabuchi,A. Yamazaki,K. Saito,S. Ikeda,S. IkedaK.
Nb3Al conductor developmen t ENEa t EM/LMId Aan , Italy ...... 1 7 . R. Bruzzese, N. Sacchetti, M. Spadoni, G. Barani, G. Donati, S. Ceresara
Development of Nb3Al multifilamentary wire by the Jelly-Roll process ...... 75 Ohmatsu,K. Oku,G. Takei,H. Nagata,M. Ando,T. Nishi,M. Takahashi,Y. S. Shimamoto
Nb3Al multifilamentary superconductor ...... 81 T. Takeuchi, Kosuge,M. lijima,Y. Kiyoshi,T. Inoue,K. Kuroda,T. Itoh,K. WadaH.
OXIDE SUPERCONDUCTORS (Sessio) nB
Potential methods for the fabrication of high-T superconductors for wires and cables
(Abstract) ...... c 1 9 . K. Tachikawa, K. Togano
High-Jc silver-sheated BiPbSrCaCuO superconducting wires ...... 93 Sato,K. Mukai,H. HikataT. Superconducting propertie Bi-base th f o s e oxide multifilamentary tape magnetin i s c fields ....1 10 . H. Sekine, Inoue,K. Maeda,H. NumataK. Application of superconducting lens electron microscope to observation of microstructures in superconductors ...... 107 YoshidaH. High-resolution transmission electron microscop f crystayo l structures, defects, surfacd ean interface bismuth-basen i s d superconductors ...... 5 11 . Y. Matsui SUPERCONDUCTING MAGNET APPLICATIOND SAN S (Sessio) nC
Magnetization studies of multifilamentary strands for superconducting supercollider (SSC) application Methods— f controllinso g proximity-effect couplin residuad gan l magnetization ...... 129 Callings,W. E. K.R. Marken, Jr., M.D. Sumption Effec f heao t t capacit matrid yan x resistivit stabilitn yo f superconductoryo n si fast changing fields ...... 3 14 . E.Yu. Klimenko, N.N. Martovetskij, S.I. Novikov Rational desig high-currenf no t cable-in-conduit superconductors ...... 149 L. Dresner Development of pulsed superconducting magnets and their operating characteristics ...... 165 T. Onishi ITER magnet development program at the Lawrence Livermore National Laboratory, USA ... 175 S.S. Shen Researc developmend han t projec f superconductino t g helica Nationae l th coi t a l l Institute for Fusion Science, Japan ...... 181 YamamotoJ. Model coil testing to complete the database for the NET fusion device ...... 189 J. V. Minervini, N. Mitchell Developmen f superconductino t g magnet tokamaa r fo s k fusion machin t JAERIea , Japan (Abstract) ...... 201 T. Ando, ShimamotoS. The Sultan-in project ...... 203 dellaA. Cone, Pasotti,G. Ricci,M. Sachetti,N. Spadoni,Mut,M. Dal Spigo, G. G. G. Veardo, J.A. Roeterdink, J.D. Elen, A.C. Gijze, W.M.P. Franken, E. Aebli, I. Horvath, B. Jakob, C. Marinucci, P. Ming, G. Pastor, V. Vécsey, P. Weymuth Constructio high-fielw ne a f no d laborator Nationae th t ya l Research Institute for Metals, Japan ...... 209 Wada,H. Inoue,K. Kiyoshi,T. Asano,T. Itoh,K. AokiH.
Lis f Participanto t s ...... 5 21 . SUMMARY OF THE MEETING
. WadH a National Research Institut Metalsr efo , Tsukuba City, Ibaraki, Japan J. Kupitz Divisio f Nucleano r Power, International Atomic Energy Agency, Vienna
invitation O Governmene th f no Japaf to IAEe nth A convene Technicae th d l Committee Meeting on Superconducting Materials and Magnets was held in Tokyo on September 4-6, 1989. The meeting was hosted by the National Research Institute for Metals, and included a lab tour to the host institute. The meetin s attende gwa participant3 3 y db s from CEC, Italy, Japane ,th
Netherlands IAEAe th USSe .USAe d th ,th Ran ,
The purpose of the meeting was to provide an international forum for the d recenreviean t d discussiotar an w progres e state th th f eo f n o se madth n i e developmen sperconductinf o t g material magnetd san d theisan r applicability for fusion devices. Thus meetine scope th , th f eo g include topice th d n so practical high-field superconductors, oxide superconductors showing superconductivity above liquid nitrogen temperature, and superconducting
magnet d applicationssan meetine Th . endes gwa witp du rounha d table discussion on the subjects proposed by the participants.
1) Practical high-field superconductors 3 sessions were devoted to the presentation and discussion on practical high-field superconductors. At the beginning a general review was developmene giveth n no advancef o t d superconductor r fusionfo s , referrino t g ac/pulsed-field superconductor higd san h de-field superconductors. Thene ,th
sessions covered various topics mainly on alloy and intermetallic compound
superconductors, such as NbZr, NbTi, Nb,Sn, V0Ga, Nb_Ge and PbMo,S0. J J J Do Italy e fabricatioTh : multifilamentarf no Jelly e wirel th A b y ysN -b Roll method was reported, together with superconducting properties of the short sample wires problemd ,an s relateproductioe th o t d f industriano l scale lengths of wires and possible ways to overcome such problems were discussed.
Japanl A e researcb Th N : d d developmenan h an a G V , NbTif o tSn b ,N
in universities, national laboratories and industries was reported in terms of fabrication technique propertiesd san . Fundamental studieflue th x n so pinning and the strain effects in superconductors were also presented. USSR: The development of technical superconductors, such as NbZr, NbTi, designe, Ge b N y som db d ean researcn NS b h institute d commerciasan l plant s reportedswa . This report also emphasize importance th d e th f o e researc d developmenhan oxidf o t e superconductor connection si n wite th h progres applief so d superconductivity.
It emerged in the sessions that Nb Al might be a new practical
superconductor for high-field as well as ac use.
2) Oxide Superconductors Due to their potential importance on the development of the fusion power system, the presentation and discussion of oxide superconductors was speciall ye topic addeth o n thit si d s meetingpapere th l s Al .wer e submitted from Japan.
Fabricatio wees na s fundamenta la l studies were discusse sessions2 n i d .
A review developmene tal th s mad kwa n eo fabricatiof o t n processe high-Tr fo s c
oxide superconductor Japann i s . Then, efforts were reporte n fabricatini d g
Pb-dope oxidi dB e superconductin gs pointe tapewa wiresd t that i san ou d td ,an
oxide superconductors migh use e e b ablvert b a do t ey high field thef i s y cooled dow liquio nt d helium temperature where metallic superconductors were 3 used exampler Fo . ,criticaa l current densit 7x1f yo s reporte0wa a r fo d tap 23Tt ea . High-resolution electron microscop presentes ywa e th n o d characterization of microstructures in Bi oxide superconductors. The application of superconducting lens electron microscope was also discussed.
3) Superconducting magnets and applications 3 sessions were assigned to the presentation on superconducting magnets and applications, and various projects involving superconducting magnets were reported. e modeCECTh :l coil testing progra completo t mT NE e database th eth r fo e fusion device was presented, and the definition and test requirements for the model solenoid test facility were reported e subsizTh . d fulean l size conductors developed wil SULTAe testee lb th t Na d facilit Switzerlann i y r o d the FENI Xwhil, US e fulfacilit e th l th sizn i ye conductor modee th ld san coils will be tested at the TOSKA facility in FRG. Italy A collaborativ: e project among Italy Netherlande ,th d san
Switzerland SULTAe ,th ProjectI NII s reported,wa e SULTA majoTh a . s Ni r test
facilitry in the EC for the development of high-current superconductors, and
the SULTAN III will be remodeled out of the SULTAN II and used for the NET
Project. Netherlandse Th e researc developmenTh e : th n hi high-currenf o t t
conductors for large magnet systems, such as accelerators like LHC-CERN and
HERA-DES r fusioYo n tokamaks like NET-ITER s reported,wa . Measurementc d f so and ac properties and losses, theoretical models and calculations, and the
constructio ndipoln S programm b N ea magne f o e t were presented.
Japan e developmenTh : t program superconductinf so g magnets reported
could be classified into 2 types. One involved the development of a variety of superconductin d pulsan ec gd magnet appliee b o t sfusioo t d n machines, such as tokamaks and heliotrons. The other referred to the construction of
high-field facilities where superconducting magnets and hybrid magnets as well pulss a e copper-magnets were use evaluato t d e developed superconducting materials including oxide superconductors. USA: SSC-related magnetization studie NbTi-basen so d multifilament composite strands demonstrated positive effects of Ni additons into the strand formulatio n fielno d stabilit homogeneityd an y . ITER magnet development progra s presentedmwa , includin developmene th g T clas12 Sn-baseb sf N o t d toroida d poloidaan l l field coil conductor constructioe th wels s a s la a f no magnet test facility, FENIX. A design study of high-current cable-inconduit superconductors were presented, indicating the high capability of such superconductors for stable operation at quite high current densities. USSR A stabilit: y stud s presenteywa d which analyze stabilite th d f o y superconducting wire n fassi t changing field y takinsb g into accoun heae th tt capacitmatrie th d x an yresistivity s showwa t nI . tha thir fo tn wiree th s permissible field changing rates wer 0 time1 e s large a s th s thos ea r fo e frozen flux model. This may explain the contradictions between the theory and
experimental results d allo,an estimato wt e stabilit d pulseyan limitc da n i s magnets.
4. Round Table Discussion
The IAEA planned a round table discussion at the end of the meeting an informasa l opportunit directlf yo y exchanging opinion d commentsan s among participants workin varioun i g s projects subject2 . s wer facn i e t proposed an d discussed by the participants.
Cable-I a Whe) s (a nha n Conduit Conductor (CICC preferrede b o )t ? It turne t frodiscussioe ou dth m n that some concern existed aboue th t fixation of the cables in a CICC, allowing conductor motion and high local
heat generation. It was then suggested to divide a CICC coil into parallel
subcoil d chargsan e the y separatmb e power supplies, resultin betten i g r
mechanical fixation, lower ac losses and more effective heat removal.
Expectation was presented that ac losses be lower than in Rutherford type conductors, whereby use of a CuNi alloy and absence of solder would further reduc e lossesth e , however rathet ,a r higher production cost.
10 (b) Is there any need for international coordination in the utilization
of large test facilitie superconductinr fo s g magnets? A potential need for international coordination was suggested in terms of collaborative utilization of existing large superconducting magnet facilities developed in fusion, accelerator or many other large projects. It was then pointed out that each magnet in a particular project had been designed for a particular purpose and would be rather difficult to be used for other purpose, while some facilities suc largs ha e cryostats coul usefue db l for different experiments, too. Finally, future activities were discussed, and recommendation was made for convening a meeting on engineering problems of superconducting materials. Such meeting shoul organizee db n internationaa s a d l seminaoccasioe th n ro n olargfa e conference, e.g. Applied Superconductivity Conference, August 1990.
Next page(s) 1 lef1 t blank PRACTICAL HIGH-FIELD SUPERCONDUCTORS
(Session A) DEVELOPMENTS OF ADVANCED SUPERCONDUCTORS FOR FUSIO JAPAN I N (Abstract)
K. TACHIKAWA Faculty of Engineering, Tokai University, Hiratsuka, Kanagawa, Japan
AC and Pulsed Field Superconductors Nb-Ti conductors for pulsed field use upto several Hz, e.g. poloidal field magnet, with filament diameter of 1-20 d larguran n e current capacities have been developed. Con- ductors for AC power applications with filament size of 0.1 e beinar n gur A developedsuperconductin1 kV - 0 50 A . g coil was successfully operated at 50-60 Hz. The development of superconductor e wilus lC A brin r fo sg abou a largt e impact in the electric engineering. However, there are »till a few parameters to be optimized, e.g. barrier material, fila- aent fabrication, wire configuration design etc., for AC superconductors. High-Field Superconductors e high-fielTh d performanc f NbjSo es beeha n n significant- ly improved by the Ti doping. The aultifilamentary (Nb,Ti}3 Sn wire wit i bronzT h e beinar e g widely user generatinfo d g fields 12-17 T at 4.2 K, The conductor was used also for the construction of axicell coils of MPTF-B at LLNL. The strain tolerance of HbaSn wire was also improved by the Ti doping. The fabrication of NbaAl-based superconductors with high- field performance superior than that of (Nb.Ti)gSn is being studied through different methods. However, stoichiometric Nb3A s stabli l e onl t elevatea y d temperatures d thenan , a rapid cooling technique, e.g. continuous beam irradiation process is required to attain good high-field properties. A critical current densit f 1.6x10o c J y ^ A/cm s bee- ha 2 ob n e EB-irradiateth r fo K 2 4. d d NbafAl.Gean T taine0 3 t )a d tape. Moreover, superconductors with crystal structures other than A-J15 type beinar e g developed S V2(Hf,ZrCI . s fab)wa - ricated by a composite method, and B-I type NbN was process- ed by a continuous sputtering. Recently, the Jc of Chevrel phase PbMogSg wire prepared by a powder metallurgy process s gaineha d appreciable improvemen n substitutioS y b t r fo n Pb, HIP treatment, reduction of strain effect etc. A Jc oe f Th Ixlo . K 4 2 A/cm 4. s bee d ha 2 nan obtaineT 5 2 t a d possibilit f usino y g high-T r genec fo oxid K - 2 e 4. tap t a e rating high magnetic field s alsi s o under consideration.
Next page(s) left blank 15 TECHNICAL SUPERCONDUCTING MATERIALS
A.D. NIKULIN A.A. Bochvar's Ail-Union Scientific Research Institute of Inorganic Materials, Moscow, Union of Soviet Socialist Republics
Abstract
e maiTh n design characteristic d propertiean s f superconductino s g wires of roun d rectangulaan d r section e bas th f niobium-titaniuo e n o s m alloys, intermetal1ic compounds, develope a numbe y b df scientifi o r d an c research institutes and produced by the Soviet industry, are given, The superconducting wires of round section in a copper matrix with the number of Nb-Ti filament o 891t p u 0s have Nb-Tth e i packing factor d froan m5 0.10. o 6t e 0.5-2.a wireTh thdi . i es mm 0ma filamentwit 0 a twis19 h o a frot t di s6 m m dependinpitcm 5 2 h o n frotheit o g 8 m r purpos d constructioan e n have 2 jc= (1.4-2.75)x 10° ft/cm in the field of 5 T. The wires of rectangular section 7-28 mm2 with a number of filaments up to 3000 and the packing facto7 hav0. e. Supero T rt Ie fiel 5 cfro3 =5.8-1 th f - 0. o mn d i A 8k conductors on the base of Nb3Sn, V3Sa, V3Si , Nb3Al intermetal1ics, obtained by the method of diffusion interaction in solid or liquid states are investigated in detail f singlo c J .e sectio ee basth wire f th Nbn o e o n 3 nSo s( n without coppes reacheha ) r d (2,7-5.0)xlO , thaT 4 te *fiel1 th A/cmf o n di 2 madn analogou a t possibli ee us o t se wir e Tokamak-1eth typr fo e 5 installati- on. The V3ßa base wires with the number of filaments up to 15000 have jc=3xlO* e maiTh n. T e trenddevelopmen 4 th e fiel1 A/cmth f f o s o n di =d properan t - tie f higo s h temperature superconductors, produce e USSRth e n ar ,i d consi dered.
INTRODUCTION
e SovieIth n t Uniod othean n r countrie e developmentth s e spherth n i se of technical superconductivity were primarily stimulated by the needs f nucleao r scienc d engineerinan e e fielth f higo n di g h energy physics, plasma, nuclear and solid-state physics. Research and developments to produce technical superconductors greatly progresse e seconth n e i 60dd th hal1 "f o "f /!/. During a short period of time helium temperature level superconductors were designe r differenfo d t purposes first odium-zirconiu i basen n o d d an m 1 niodiurn-titanium alloy e beginnin th e 70 st th a (T "f d *o cg=8-lan ) K l e basioth nf intermetal1io s c B typ3 compoundA ee th (T f e=14.5-23.o s ) I2I.2K e samAth t e time throughou e wholth t e world higher transition tempera- ture superconductors were searched for. Beginning from 1973 Nb36e was a record in relation to the critical temperature equal to 23.2 K and this gave hope that it could be used at liquid hydrogen temperatures. Recent discovery of high temperature superconducting oxide compounds LaSrCuO /3/ and later YBaCuO, BiSrCaCuD, TIBaCaCuO and others having the transition tempera- ture of 33-130 K was a revolutionary jump in the progress of applied super- conductivity.
The analysis of the problem status shows that the promising search for way f higo s h curren d higan t h temperature superconductor production will proceed simultaneously wite desigth hd operatioan n f traditionao n w lo l temperature superconductor devices including large magnetic systemr fo s charged particle accelerators, fusion plant d othersan s .
17 PRACTICAL SUPERCONDUCTING WIRES This paper briefly discusses the Soviet developments of high current superconductors for technical purposes. Superconducting alloys and compounds cannot be used directly for practical purposes. Mechanical and electromagnetic perturbations can destroy superconductivity condition; to prevent this a superconductor must be stabilized. Procedures used for superconductor production and the operational conditions of a magnetic system sophisticate their structure. In this relation an engineered superconductor is a complex composite system comprising dissimilar materials /4/. It is usually a oiulti *latienlar< y wir n whici e h fine superconducting filament e placear sa n i d copper or bronze matrix with diffusion barrier layers, reinforcing elements and partitions a higmad f o he electrical resistance material. The work on superconductor materials comprises a large complex of physical examinations of the current carrying capacity nature of supercondu- ctors e electrophysicath , l characteristic f wireso s , superconductivity properties of alloys and compounds. One IB required to solve fundamental and engineering problems suc s developmena h f methodo t r physicafo s d procesan l s desig f superconductino n g wires, calculatio f theio n r characteristicse th r Fo . commercial productio f superconductoro n s high quality metalli d nonan c - metallic materials as well as technically elaborate equipment and processes e neededar e quicTh . k developmen e commerciath f o t l productio f supercono n - ductor s encouragewa s e Sovieth y b td experience gained froe worth mk with rare, refractor d non-ferrouan y s metal d theian s r base alloys. Nevertheless, processes were required for the production of special grade high purity titanium, niobium, vanadium, copper as well as bronze, high electrical resistance alloys and various semi-products such as tubes, rods, shapes, wire and sheets. The use of vacuum metallurgy including electron beam, arc (conventional and skull), induction meltings provided for the production of high purity metals and alloys having adequate chemical and structural mityunr ifo . e studieTh s inte rhéologith o e regularitie d colan dt ho f o s multicomponent composite deformations mad t possibli e o fine t wantee th d d engineering solutions for the commercial production of technical long-size superconductors e processeTh . s suc s extrusiona h , drawing, rolling, welding, soldering, electroplating, vacuum or protective atmosphere heat treatments are widely used for the production of superconductors. The major operations for conditioning composite precursors, their assemblage and sealing require e technologth e "whitth f o ey gloves" type. Unique procedure d equipmenan s t have been develope r manfo d y operations, among them twistin f multifilameo g - ntary conductors. Method d devicean s s have been designe o controt d e th l superconductivity characteristics, mechanical properties, composition and structure of superconductors. Comprehensive studies were required to improve the critical current density of superconductors having different compositions and to establish the relationship between the critical current density and process parameters. Investigations inte finth o e structure formatiod an n diffusion barriers using electron microscope, X-ray diffraction, X-ray microspectrometry methods, Auger analysers together with the process development mad t possibli e o establist e e scientificallth h y documented approac e technologyth o t h . Different magnetic systems require different superconducting materials. e samAth t e time under commercial production conditions excess diversitf o y wire gauges is not economical. This discrepancy is solved by the metallurgical productio f modulao n r element e directlsb than ca t y employed for winding small magnetic systems; at the same time with the cable produc- tion methods thee use o ar producyt d e high current carrying elementr fo s large magnetic systems /5/.
18 NIOBIUM-TITANIUM ALLOY BASE SUPERCONDUCTORS Niobium-titanium composite superconductors in a high purity copper e majomatrith e r ar materiax f superconductino l g magnetic systems. This i s primarily accounted -for by the adequate mechanical properties of the alloys. To-day, commercial plants produce a wide range of superconducting wires and modular elements on the HT-50 (507. Nb,50X Ti) alloys base. Table 1 lists the principal characteristics of Nb-Ti technical conductors produced in the USSR.
Table 1, List of Soviet HT-50 Alloy Base Superconducting Materials
Diameter Filament, PF* Fil ament Twist Current carrying section, number dia i pitch, capacity H=5 T, T=4.2 K mm ßm mm le, A Jc**.10-"A/cm2
2.0x3. 5 2970 0.42 30 150 5800 2.0 3 x 5 2970 0.42 42 150 10000 1.7 4 x 7 3000 0.5 70 200 18000 1.3 0.85 8910 0.42 6 8 550 2.5-2 .75 0.5;0. 85 2970 0.42 6;10 8-10 - 2.5 1.0; 1.2 2970 0.42 12; 14.5 15-20 - 2.2 1 .5;!. 75;2.0 2970 0.42 18;21;24 20-25 - 1.8 0.38;0 .5;0.7 6 0.4 104; 129; 182 8-10 - 1.5 0.5;0. 7 60 0.5 45;63 15-20 - 2.6-2 .8 0.85;! .0 60 0.5 76;90 15-20 - 2.2-2 .5 0.5jO. 7;0.85 210 0.5 24;34;4i 10-25 - 2.5-2 .8 1.0; 1.2;1.5 210 0.5 48;58;72 10-25 - 2.0-2 .2 1,75;2 ,0 210 0.5 84;96 10-25 - 1.4-1 .6 0.85 IS 0. 16 71 10-25 - 2.5 0.2-0. 5*## 75000 0.3 0.4 1.3-3.3 - 1.6-2 .0 - rati F f Nb-To #P o i alloy are o totat a l superconductor area. #*J - criticac l current densit f Nb-To y i alloy. #** - highly resistive matrix.
The superconductor range comprises rectangular wires for medium size magnetic systems and excitation windings for crycelectromachines. Conductors 7 to 28 mmz in section have about 3000 filaments 30-70 mm in dia and the factor of packing with Nb-Ti alloy equal to 30-50X. The high unit face- to-fac m witk e a relativelh5 lengt 1. o t h y large wire section requires largesize precursors in case of the metallurgical fabrication or the soldering of a small diameter fine wire into a rectangular section shape. The need for high critical current density requires high cold work. Cold working processes have been developed for large section precursors which enabled the critical current density of 2.0 and 1.3x10° A/cm2 in the field of e attaineb o t n 2.0x3. T i d d4.0x7. 5 an m 5m m sectio m 0 n wires, respectively. Fig 1.
For some years the design and technology have been developed to produce a superconducting wire for the Acceleration-storage comples (UNK) in Serpukhov /6-8/. Stringent requirement e placear s d upon this conductor s criticait , l current, copper stabilizer quality, twist pitch, losses and other
19 Fig 1, Cross section o-f CKHT-0.05-8910-0.42 (A), CnHT-3x5-2970-0.42 (B)
characteristics. The large-scale production of conductors for UNK required large-scale operations under commercial conditions. The developed commercial process using composit m diae m e selecte precursor5 th , 31 o t d p u s thermomechanical treatment conditions mad t possibli e o mastet e e th r productio f a conductor0.8o ndi 5m m s with 8910 filaments e packinth , g factor of 0.42 e filamenth , t diamete e criticarth lesd an s l m thacurrenjn 6 n t density more than 2.5x10° ft/cm3 and more than 2.75x10° A/cma in the field of 5 T depending on the precursor sizes and process conditions. Fig 1. The introductio a diffusio f o n n niobium e uniforbarrieth d an mr optimized structure mad t possibli e t onl no eo realiz t y e higth e h critical current bu t also to increase the length of single wire up to 5-6 km.
Assemblages more than 300 kg in weight were used for the fabrication of wires 0.85 mm dia containing 14190 filaments 4.5jnm dia in a copper matrix in whic e criticath h l current densit f 3.4x10o y s wa °T A/cm5 t a 2 reached. To-day superconductin gr combineo wire a higm n oh i sh d matrix with 75000 filaments 0.4-1.0 mm dia are being successfully designed, they will be use t alternata d e currents.
INTERMETALLICS BASE SUPERCONDUCTORS
Comprehensive investigations into superconductor n intermetallio s e Nb3Sn, V36a, V3Si and oth. base produced by the methods of solid and liquid diffusion interaction permitted the design and commercial production of a rather wide Nb3Sn base wire gauge. Table 2 gives the list of the wires produced and their principal characteristics. Despite the fact that some specialists are sceptical about Nb3Sn conductors due to e inherenth t brittlenes f Nbo s 3Sn compound, scientist d planan s t engineers have found technical solutions that furnis e commerciath h l produ- ction of single conductors and current carrying elements for the large magnetic system of the fusion facility Tokamak-15 /9/.
20 Table 2. List of Soviet Nb3Sn Intermetallics Base Superconducting Materials
Hire dia, Filaient Filaient Twist Current carrying capacity Notes nn number dia, It* H'=8 T H=14 T
le, A Jc. 10"* fl/ci :' Ic,. AJc 10-* fl/c«*
0.5;0.8 . 1 14641 7i2.7 10-25 200:450 10.2(9.00 10 ; 40 2.0;2.0 Non-stabilized
1.Oil; .2 . 3 14641 5;4.0; 25 600; 800;7.U7.Il 140; 190; 1 .8il.7i . • - 1.5 5. 1 1000 6.3 280 1.6
1.0; 1.5 44521 2.2j3.3 25 700; 1200a.9;6.8 150; 300 2.0,1.7 _ » - 1. 0;1.2; 14641 3.4;5.1; 5 2 550;750; 10 .0;9.4; 135; 180; 2.4;2.3; Stabilized 1.5 4. 0 1080 8.7 265 2.2
1.Oil-2; . 2 44521 2;2.6i 25 5£0;900; 10.6;10.4; 140; 210 2.5j2.5 _ * _ 1.5 3. 3 1180 9,t>
J- criticac l current densit r sectionpe y , without stabilization.
Nb3Sn superconductor e producear s d bot s stabilizea h d withouan d a t copper stabilizer. In the latter case niobium-tin composites are twisted, heat treated to form a superconducting compound and then electroplated with copper. The critical current density of single wires is (7-il)xlO* A/cmz (for the section without copper) in the field of 8 T and (2-2.7)xlO* A/cma in the . NbT 34 Sfiel1 n f baso d e conductors 0.5-1. a werdi e m 0m use o manufacturt d e a great number of magnetic coils with the induction of 11-12 T. To-day niobium-tin superconductors are being designed with the alloyed bronze matri d niobiuan x m filaments n thii ; e sfield th cas n f 8-1e i eo s th 4T critical current density increases by a factor of 1.3-1.5. Conductors with an internal tin core have been commercially fabricated; conductors 0.5-1.0 mm dia (1938 filaments 5-10 mm dia) have the overall current density more than 14 and 5x10* A/cm* in the fields of 8 and 14 T, respectively. Fig 2. A wide range of examinations have been carried out comprising mechanical properties, irradiation of superconductors and methods of producing in situ niobium-tin microcomposite s wel a ss superconductor a l n Vo s3Si , Nb3Ge, Nb3Al, Vz(Hf,Zr), PbMo^Se and oth. base. Stabilized VBa base wires were designed; their diameter is 0.5-1.0 mm,
the numbe f filamento r 1.5-3.( s 3 o 15000t p e 4criticau jnni th ;s i ) l current density of 0.5 mm dia conductors is 8 and 3x10* A/cm2 in the field of 8 and , 1respectivelyT 4 . Promisin2 g Fi . e glos th dat sn o aleve l were receiver fo d niobium-tin conductors havin a twism m gd an tfilament5 m pitc2. m f 1 o h f o s and designed for operation at commercial frequency currents. To-day, large
section conductor a pilo f e undeo t ar s S fusiorSM developmenn i n e us r fo t reactor a curren o s ; t carrying cable n sectioconductoi 2 mm 4 n 7x havinr g more tha a million n filament e fielth sf o n dcarriei A k e curren 2 th s1 f o t 9 T (4.3x10* A/cm2).
21 Fi. Cros2 g s sectio f internao n n superconductoti l wita di h n a 0 1. r 1938 filanents (A); cross section of V3Ga base stabilized superconducto a wit sdi in h 0 14641. r 1 filanent) (B s
HIGH TEMPERATURE SUPERCONDUCTORS
The work on the method of YBazCu07-„ synthesis showed that its serai-products in the form of powder, strips, pellets can be produced having adequate electrophysical propertie e solith y db s state synthesi f initiao s l oxides, nitrate pyrolysis, co-precipitation of carbonates from aqueous solutions; cryogenic processes gave adequate results /ll/. In different organizations throughout the country the production of YBa2Cu307_H ceramics s beeha n mastered havin e transitioth g n temperature 91-9 5K The results of the experiments /10/ and theoretical analysis showed that production of long high-current HTSC ceramics base superconductors entails some grave difficulties. Methods of powder metallurgy, cold work of metal sheathed ceramics, platin y differenb g t methods, productio f textureo n d materials by melting and directed crystallization as well as by extrusion are of great interest for the solution of the problem of high current HTSC production. In our investigations the methods of metal sheathed powder deformation have progressed most of all /10,12,13/. Fig 3. The work to produce composite superconductors by the particular method is being carried on in three main directions: - examination of structural changes and properties of a ceramic core in the process of conductor deformation, the optimization of working (drawing, rolling and oth.) at different temperatures, the formation of a dense ceramic core in conductors; 22 Fi. Cros3 g s sectio f nultifilaaentaro n y YBaaCu3Q7_,, wire. - optimization o-f the mechanical properties and design of a sheath that permits elimination or minimization o-f reaction with YBaCuQ and the -fit o-f the thermal expansion coefficient of a sheath to that of ceramic to eliminate ceramics cracking during thermal cycling; - optimizatio e conditionth f o n f sinterino s d saturatioan g f ceramio n c constituents with oxyge n finishei n d product o product s e a densthe n i me textured state. The influenc f differeno e t t kindworho f kcolo d s (pressingan d , extrusion), impact loading (hammering n ceramico ) s density changes wa s investigate o establist d h conditions leadine maximuth o t gm densificatiod an n s showi t I n hig tha. e c densitI hth t d densificatioan y n extenf o t ceramic specimens depend on the force and rate of cold pressing as well as on the degree of the initial powder dispersion; the maximum density of as sintered cold pressed ceramic s 5.5-5.i s 8 g/cm3, Investigation t worho kf o s processes showed that although hammering produces a high density material (p=5.85-5.95 g/cm3), hot pressing (p=5.7 g/cm3) and the more so hot extrusion (p=6.0 g/cm3) are of greater interest. The critical current density of specimens hammered and subsequently heat treated was lower than that of hot pressed specimens ((6-8)xl02 A/cm2 at 77 K in the zero field). This is likely to be related to large structural changes effected by impact loading. In is established that specimens hot extruded at 900°C have a favourable texture of the (001) type that forms over the whole specimen section; and after anneals at different temperatures it neither disappears nor alters; the specimen structure is characterized by a directional orientation of grains. Composite materials were used to study the influence of a sheath material strength (steel, copper, nickel, silver and oth.) on changes in the density and superconductivity of HTSC ceramics "123" during cold drawing. It is shown that the maximum attainable density of ceramics grows with the strengte materialth a sheathl f al o h f sO . studied onl n steei y r combineo l d steeg sheath A s possibl + wa l t i s o react ee ceramic th h s densit f 6.0-6.o y 1 g/cm y colb 3 d drawin o 807.t g . Heat treatment require o recovet d e th r superconductivity propertie o sintet d an rs oxides entails additional difficultie o bott e chemica th e h du sd physicaan l l compatibilitf o y YBa=Cus07_x ("123") with metallic sheaths e metalth f O s. 23 ) onlNi y , nickeFe investigate , Y l , showeCu , M a relativeldd, Ta (Mb , Zr , y weak interactio o fort n m CuQ-nNiO boundary layer with some copper replacey b d nickel e resultanTh . t laye s dielectriha r c propertie d inhibitan s e th s current from getting int a specimeo n whica furthe s i h r complicatiod an n e specimenus make e on s s wit a partiallh y removed nickel sheath which 2 2 showed Jc=1.3xi0 A/cm at 77 K. o diffusioN n interaction between yttrium ceramic d silve an ss founwa r d out, besides silver is permeable to oxygen which makes it one of the most promising sheathing materials. However e TEC ,th f phas o s e "123 d silvean " r significantly differ (14xlO~* T~* and 20xlO~A T"1}, respectigely which is one e causeoth f s that inhibit e stablth s e high critical curren f compositeo t s in a silver sheath. Nevertheless, superconductor specimens were produced on yttrium ceramics base in a silver sheath, they are in the form of a band 0.05-0.15 mm thick and 3-5 mm wide; the ceramic core thickness is adequately uniform ovee wirth r e length e criticaTh . l current densit s 7.8xl0i y 3A/cmt a 2 77 K in the zero field which is close to the published values 714,15/. It has been confirmed that the deformation of ceramics increases the critical current density by several times. It is apparent that the high critical transport curren y probletke (10*-10a s mi ) K ° 7 A/cm7 t a 2 that determine e technicasth l feasibilit f higo y h temperature oxide supercon- ducting compounds. Along with this a wide variety of challenging technical problems must be solved fundamentally after which the conclusion might be drawn as to the feasibility of introduction of new superconductors into the technology. CONCLUSION To-day in the USSR Nb~Ti alloy, Nb^Sn intermetal1ics base superconducting wires have been developed and brought to a commercial level, n magnetii thee designe e ar y us cr fo systemd n accelerator-storaga f o s e complex and a fusion facility Tokamak-15. A large complex of scientific researc d technologicaan h l wor s undei k y aimewa r t producina d g wires from high temperature superconductors havin a curreng t carrying capasity appropriate for practical usage. REFERENCES . 1 MeTajuioBeABHM n BMerajiJioüiHSMK a CBepxriposoAHMKOB. CöopHM« CTaretö. MocKBa, "HayKa", 1965. 2. HHKyjiHH A.ft., HoTaHMH B.H., HepHonjiBKOB H.A. n ap. "MHoroKH^bHwe Marepnaji s TexHMHecKoran « o ncnoJib30BaHna. KH B " MOCKB3, ATOHM3A3T, 1977, T.4, C.5~14. 3. Bendors I,, Muller K. Possible high Tc superconductivity in the Ba-La-Cu-0 system . PhysZ . . 64:189-19B . 3 (1986). . 4 SepHon^eKDB H.A. CBepxnpaBOAsiiijMB MaTepnaji B cospeneHHou K TBXHMKB. BBCTHHK AH CCCP, 9, 1978, c.48-61. 5. K^HMBHKD E,K)., KasansaH f.f. M «p. CBepxnpoBQASUine TpaHcnoHHposaHHwe npoBOfla fljifl Kpynnux Mari-WTHWx CHCTen. I-».neKTpoTexHMKa, 1983, 10, c.6-8. 6. AreeB A.M., EaJioeKQB B.H., PpMflacoB B.W. n ap. npenpHHT.MU)B3, 8, c.80-96, CepnyxoB, 1980. . Fil'kin7 , V.F.Sogulya, V.P.Kosenk . Composital t e o e SuperconductorK UN r fo s Magnets. Proceedin f o Workshog n Superconductino p g Magnet d Cryogenicsan s , 12-16y BNLMa , , 1986, p.56-59. 8. SBJIBHCKM^ T.K., UlMJibKMH B.S., MBTTB B.J1. M flp. floxJiaA Ha Kom&epemjMH no KpworeHHHM Majepua/iaHï crpyKtypa, npt-iMeHeHMfl n MwHb 7-10, 1988, r.lOeHbflHr KHP. 24 9. KoHOBa.nDB B. SBMHOB CBBTMJIO. MSBBCTMSI N 365, 29.12.1988. 10. HnKH$QpOB A.C., HMKyjlMH A.A.) IpMJlbKMH B.fl., UlUKOB A.K., HepHOnjtBKOB H.A. > KOMnOSMTHbiAP M e npOBOAHMK a OCHOBH H DBDf Bfip C BPi pX l SL|MX COeflUHBHMW La-Sr-Cu-0 n Y-Ba-Cu-0. AroMHas» >Heprna, T.62, Bwn.6, MMHI. 1987, c.421-422. 11. Kayjib A.P., TpeTbSKoe (O.A. M Ap. CMHTBS CBepxnpoBOflsiuinx CJIDKHWX OKCKAOB "CsepxnpOBOflMMOCTb". Bun.lo McnojibsoBaHMn K P . » aTOMHofi 3Heprnn, CCCP, MocKBa, 1987, c.8-10. 12. AopoitteeB F.J)., KVIHMBHKD E.tfl., HnKy.nMH A.fl., [pmibKMH B.3., HepHonjieKOB H.A. M flp. "IlpOlSjieMH BblCOKOTBMneparypHOW CBBpXnpOBDflMMOCTM" H.2, 1987, crp.228-229. 13. HMK'yjiHH A.A> i (DnjibKHH B. 5). , AaswflOB U.M., tlluxoB A.K., H.A., KJIMMBHKO E.lfl. n Ap. Co. "CBepxnpoBOflMMOCTb". (jMSMKa, XMMMSI, rexHMKa", Bwn.3, 1988, c.49-54. . (JlJiwKMrep$ . P . 14 , Mw^JtBp. MeiajuiyprM n SKPMTHSBCKM B TOK npoBOAaxB M . TpyaH KOH(j)epeHUMM no BucoKOTBMnepaTypHUM CBBpxnpOBOflflujHM Marepna^aM M BbicoKOTBMneparypHofi CBBpxnpoBOAMMOCTM, 29.02-04.03.1988, 15. M.Okada, A.Okayaraa "Fabrication of Ag-Sheathed Ba-Y-Cu Oxids Superconductor Tape" Japanese Journal of Applied Physics 27(2): 185-187 (1988). Next page(s) left blank 25 HIGH-FIELD SUPERCONDUCTING PROPERTIES AND ELECTRICAL RESISTIVITY OF (Nb,Ti)3Sn COMPOSITE CONDUCTOR TUBE TH E Y METHOSB D S. MURASE, H. SHIRAKI, S. NAKAYAMA Toshib CenterD aR& , Kawasaki . AOKIN . ICHIHARAM , . SUZUKE , I Showa Electric Wir Cable& e Company Ltd, Kawasaki Japan Abstract Toshiba Corporation and Showa Electric Wire & Cable Co. , Ltd. have cooperatively devoted great effort o develot s e tubth pe processed (Nb,Ti)„Sn conductors. These efforts have brought about the following O results. (1) High critical current density was obtained at high fields for the conductor with higcontentn hti . (2) The conductors have been widely used for high field applications. INTRODUCTION Ther s bee n ha eincreasina n g deman r developinfo d g high critical current density (Jc) superconductors having superior high-field performanc d R higan (residuaRR eh l resistance ratio r manfo ) y applications. Toshiba Corporatio d Showan n a Electric Wir & Cable, . Co e Ltd. have cooperatively developed (Nb,Ti)_Sn conductors processed by tube th e metho. year2] r , fo d [1 s The tube processed (Nb,Ti)„Sn conductor has many advantages, as compared with the usual bronze method. The advantages include no intermediate annealings durin e reductioth g n procesa hign d ti h an s content in the copper clad tin core composite resulting in high Jc values. This paper reports various aspects on the tube processed conductor; the fabrication process, the dependences of the critical curren n contenttti densite th , n applieo y d tensile strain d appliean , d bending strain, the Jc below 4.2 K, the n value, RRR and also fabricated conductors for practical use. 27 FABRICATION PROCESS e fabricatioTh n process, s followsshowa n s Figi i n , .A .singl1 e core wire consistin b tubN ea witf o ga coppe h r sheathe n corti de inside and a copper tube outside, were used. Ti was added to the elemen r o elementt s which forme a singld e core o T fabricat. e multifilamentary conductors, a number (one to 931) of single core wires were bundled together in a copper tube and extruded by hydrostatic extrusion. Then they were drawn down to the final sizes without any intermediate heat-treatment d finallan , y submitte o reactiot d0 70 t a n C to form a (Nb,Ti)„Sn layer inside the Nb tubular filaments. The present method can vary the equivalent Sn content widely inside the filament, dependin e coppeth n ro g sheath thickness e conductorsTh . , % equivalen 0 9 - contentn 0 ti t2 whic d ha h, were prepare orden i d o t r investigate the Jc values at high-fields in relation to the equivalent tin content. Various fina le conductor sizeth f so s were fabricater fo d research use (one to seven filaments) and practical use (over 100 filaments) .A cross-sectiona lconductoe th vie f wo shows i r Fign i n . 2 CRITICAL CURRENT MEASUREMENT Critical currents (Ic) were measured in a transverse magnetic field up to 23 T, generally using the one micro-V/cm criterion. A superconducting magne f Toshibo t a c CorporatioI e th s user wa nfo d measurement a hybri d an sd , belomagnewateT a 5 d 1 wran t cooled magnet of Tohoku University were used from 15 to 23 T. The Jc was defined as Ic divided by the conductor cross-section area, excep copper fo t r (non copper Jc). Jc - Sn CONTENT n contenS e Th t c valuedependencJ e s e studieth wa sth r f o fo yd (Nb.Ti) Sn layer and for the case of non copper at 14 T [1]. The non O copper Jc values increased as the (Nb.TiKSn layer thickness and the o equivalen n contene copper-titi tth n i t n composite 0 9 increase e th o t d % tin content. On the other hand, Jc values for (Nb.Ti) Sn layer at 14 22 f abouo T t 1500 A/m d 200an m0 A/mm were obtaine r samplefo d f leso s s than 35 % Sn content and for more than 50 % Sn content, respectively. 28 Cu TUBE UNREACTH>Nb(Ti) STACK 8 DRAWING HEAT-TREATMENT Single core Mult ifilamentary wire Fig. 1 Fabrication process for (Nb,Ti)3Sn conductor by tube method Fig. 2 Cross-section view of fabricated conductors Considering that all samples had the same grain size about 200 nm, which was independent of Sn content, it was assumed that a higher Jc (Nb,Ti n laye)S r with mor n contenS e % tha 0 t5 n woul broughe b d t forth O n bincreas a yuppee th rn i ecritica l field. Figur show3 e e relatiosth n e equivalenbetweeth d n coppe an n e no contenCu-Sn ti th c tJ rn i t composite inside the tubular filament The Jc values are seen to increas n conteneti wite tth h increas d shos peaan eit w - k 0 valu5 t a e 70 % tin content. The highest non-copper Jc, for instance about 400 2 A/mm at 18 T, and the Jc decrease was the smallest for the 50 % Sn sample, as the magnetic field increased. Typical Jc - magnetic field curves are shown in Fig. 4 for champion data and practical use data of multifilamentary conductor. Sn s % wit 0 5 h 29 Fig. 3 Jc vs. tin contents for various magnetic fields 20 4O 60 80 1OO Sn} conten% t (w t l03rrV-r E i l ! "Z x Ov r O Champio\ n dat \ a \ o £ A Practicae us l -1.0/iV/cm criterion t 4.2,a K A .1= i i l i i ii ii 0 lO'L-Nr 15 16 17 18 19 20 21 22 23 24 Magnetic fiel) (T d Fig. 4 Jc vs. magnetic field for conductors Jc K BELO 2 4. W The field dependence of Jc was studied below 4.2 K for superfiuid helium use, as shown in Fig. 5 [2]. Ic was increased, as the temperature decreased. The temperature dependence of Ic was larger at higher magnetic fields e normalizeth ; d critical current c (2.1I , / 3) K Ic (4. whic, ) 2K h increased with increasing magnetit a c3 field1, s wa , e uppeTh r. T critica5 1 t a l5 1. fiel d dan alsT 0 o1 increaseT 5 2. y b d from 4.2 K to 2.13 K, which was obtained from those data using Kramer 's plot. These result n conductorsS wer% 5 e 2 highee obtainee Th . th rr fo d 30 — <=.<-> i i i i i |S 1.8 o 1ST * 10T s E1-6 += o - \ " •£ l—t 0 1.4 H - x\ ^ | 1.2 : : . X , z i o 2 4 Temperature, T(K) Fig 5 .Normalize d critical current K belo 2 w4. tin conten ) conducto% t 0 5 (suc s s considerei a rh o havt d a highee r normalized critical current at higher fields around 20 T, because the upper critical fiel mucs i d h highe r highefo r conductors n rti . Jc UNDER TENSILE STRAIN . tensilJvs c e stress-strain characteristis wa c T aroun 5 1 d measured using a newly developed apparatus [3] . The applied strain for the conductor was measured by strain gauges directly attached to the conductor. A stress-strain curve was obtained at 4.2 K using a different tensile test machine. Figure 6 shows normalized critical curren c (tensilI t e strain)/Ic . tensilvs o ee conductor straith r fo n . i | i | r i'i' 0 __ ^ 1-0tf^^^äo^^< x Bending o ^ 0.8: % VV i _— » _ 5 0.6 i Tensile i at 1ST o " "0.4 t T3 T 3 1 o ». 0> N \ A 14 T | 0.2- * \5J - o - a 16 T n i i i i 1 i i i i 0.2 0.4 0.6 0.8 .0 ) Strai(% € n, Fig 6 .Normalize tensil. vs c J de strai bendind an n g strain 31 Ico mean c undeI s r zero applied tensile strain e conductoTh . r showea d small peak effect; the maximum critical current Icm (Icm/Ico = 1.03) was obtained at 0.3 % tensile,strain. Ic degradation occurred at about 2 0.4 % (corresponding 15 kgf/mm tensile stress). Furthermore, it was obtained that Ic was reversible below 0.48 % but irreversible above 0.55 % tensile strain. Jc UNDER BENDING STRAIN t applieNormalizea c e ratiJ th d f o , obendin Jc d t ga straic J o t n zero bending strais wa f Jcoo T n. bendin5 vs ,1 - g 3 1 strai r fo n measured s showa , Fign i n 6 .[4]a result s .A ,% normalize2 0. t a c J d bending strain equaled Jco, and Jc degradation of Jc at 0.4 % bending strai % straic decreass les J 6 f Jco wa ns o e s0. wa n .% Th that a e0 1 n f Jcoo % . 0 Simila2 r results were fielTh e d. obtaineT 6 1 o t d T fro 3 1 m t changno d edependencdi undec J n appliea rf o e d bending strain condition; a 50 % ratio between Jc at 16 T and Jc at 13 T was maitained under any applied bending strain. These values are high enough, compared with the bronze processed conductor; 20 % and 30 -40 % Jc % bendin 6 0. gd strainan % s obtaine 4 wa , 0. T t 6 a d1 decreas- 5 1 t a e bronze th r efo % 0 4 s wa ) T 3 1 t (a c J / ) T 6 1 respectively t (a c J d an , processed conductor. VALUn E n value e Th s obtaine V curvdI- froee th from measuremenc I m t were define voltage: s V=aV a d ( current: l I , constant): a , e effecTh . t of the n values on magnetic field is shown in Fig. 7 for a seven-filamentary Hig. conducto ] n valueh[1 n sS r abou% wit 0 0 4 5 ht for arounT wer 6 e1 d obtained d decrease an e ,magneti th s a d c field increased. The n values changed with the number of filaments of the conductor; smaller n values were obtained for a larger number of 8 filamentar10 a filaments r fo yr instanc6 fo 2 n ,valu f n o a e n valu e eTh unde. T conducton appliea 3 r1 t a dr strain condition decreased as the bending strain increased, as shown in Fig. 8 [4]. As the bending strai s e applieconductorwa th n o t de conductoth , s wa r subjected to compressive and tensile strain simultaneously on the same cross-sectional plane. Therefore, various strains were applieo t d filament e conductorth n i sd thuan , s each filamen a differen d ha t t Thi . valuy havJc sma ef eo cause valuedecreasn e de th th . n i e 32 30i -i———i———i—— ol3T A14T a I5T x I6T 0 (Nb,Ti)3Sn •u Tube method UJ i to 15 20 Magnetic field (T) 0 0.2 0.4 0.6 0.8 BENDING STRAI) N(% valuen Fig7 . .magneti vs s c field Fig. 8 n value vs. applied for seven-filaments conductors bending strain RKR Whes obtainedi a hignR RR he heat-treatmen th , t tim s generalli e y shorter d thu an t bring,i s c valueJ s d vicw fortan ,lo e a hversa n I . view of these facts, an internal oxidation (IOX) process has been newly developed to achieve a higher RRR of over 100 and a higher Jc simultaneousl s e defineratiwa th f y electricao R os a [5]RR d e .Th l resistance at room temperature to that at 20 K. The IOX process oxidizes impurity element a coppe n i s r stabilizer matri o oxidest x , which make a slowe r resistivity matri d increasean x X sIO RRRe Th . process is as follows. First, the condcutor is heat-treated at around 300 C in air and a copper oxide layer is formed on the periphery of the conductor. Second, oxygen diffuses to the inner side (copper stabilizer) of the conductor and then combines the impurity elements in e coppeth r stabilizer whe e reactioth n o fore t n(Nb,Ti).,S th m n layes i r O vacuun r atomospherei A C r carriet arouno ma 0 t 70 dou e dimpurit Th . y oxides dispersed in copper do not contribute to electrical resistivity e coppee puritanth th d f ro y e stabilizematrith n i x r increasesa s A . result, RRR and conductivity above 20 K of the conductor increased extremely, which were higher than those of the raw material for the copper stabilizer. When sample conductors were heat-treatee th n i d 33 conventional heat-treatment conditionR valueswerRR 3 9 ,o t e fro 5 7 m obtained. However, from 355 to 415 RRR values were obtained using the IOX process, without Jc deterioration. A small coil with an 80 mm outer diameter and 100 mm height by the wind and react method was also fabricated using the IOX process. Sample conductors were cut out from s measuredwa eac R hRR e coil s showlayea d ,th Fign an ,f i n o r . Thi.9 s resul l samplesX procestal IO s shower alse wa sfo Th . o R d RR abou 0 40 t applied to a copper clad cupro-nickel core composite wire. The process e wirth e o 440 f t o fro n optica6 A .R 7 improvem RR e l th dmicroscop y stud s carrieo investigatt wa y t ou d e caus th e f this a resulteo s A . a , black layer, nickel oxide, which surrounded the CuNi core, was confirmed at the Cu/CuNi interface for the composite wire. It is considered that the circular oxide layer played a role as a diffusion barrier, and thus a high RRR was obtained. From the above mentioned X resultsprocesIO e s considerei sth , e applicablb o t d e e th als o t o ordinary bronze oriented Nb3Sn conductor, which will not need a Nb or Ta barrier. .o •g 600 0 \ t o o o o o* 0 1 400 JQ.O °A. 0°-- _____ -__ _ -0- in ^° ° ° 00 '55 0 g e o o 200 - P 0) 1 1 1 1 1 1 <£ 1 5 10 15 20 25 Number of turn from the inner side of the coil Fig. 9 RRR for conductors from coil winding CONDUCTOR APPLICATIONS Tube processed conductors have been used for many high-field applications. Som f exampleo e s usin a monolitg h wire 16.ar e 7T coi l for Tohoku University ] coild [6]20.[8 r an fo sT 1 , ] 19.3[7 T 6 Karlsruh y superfluib e d heliun m ow coolin a 15.r ou d 5 T coian r g fo l use [9] and another 15 T class coils by immersing into pool boiling helium cooling o T giv. e example a larg f o es current assembled conductor, a copper housing encased conductor for the hybrid magnet at MIT, as shown in Fig. 10, and a cable-in-conduit typed forced cooling conductors (486 strands in the conduit) for DPC-TJ [10] and proto-type conductors at the test coil for the fusion experimental reactor [11] in 34 Fig. 10 Cross-section view of conductor for MIT JAERI were fabricated. Especially e 20.th , 1 T cois performeha l e th d highest magnetic field in the world made of superconductors only, and the hybrid magnet conductor for MIT has achieved 3980 A Ic and 952 A/mm2 Jc at 12.1 T. CONCLUSION A tube processed (Nb,Ti)„Sn conductor, having a high Jc value and O a hign contenti h t copper-tin composite s beeha ,n develope r higfo dh field applications. The Jc vs. magnetic field relation to tin content, temperature s, tensil K belo 2 4. we strain d bendinan , g strain, together with RRR, and the n value have been studied. The obtained results were 2 n conducto S t e highes 0 a A/m% th 70 ma 0 d sc 5 J tha rfollowse Th ) (1 . 16 T. (2) Ic (2.13 K) / Ic (4.2 K), which increased with increasing degradatioc I T 5 1 t A n ) occurre(3 . T magneti t 5 a d1 t a c5 field1. s wa , about 0.4 % and the irreversible strain was about 0.5 %. (4) At 16 T, 10 % Jc degradation was within 0.4 % bending strain, while over 20 % degradatio s observewa n d % bendinunde 4 0. r g e bronz th strai r efo n processed conductor. (5) The IOX process has been developed for obtainin d higg an simultaneoulyc hig hJ R hRR . Finally, the tube processed conductors have already been widely used for high-field magnets. ACKNOWLEDGEMENTS e authorTh s would liko thant e k Tohoku University, National Research Institute for Metals, JAERI, and MIT for helpful discussions. 35 REFERENCES . Shiraki1H . . NakayamaS , . TanakaM , . MuraseS , . AokiN , . IchiharaM , , K. Watanabe Muto . Y Noto. ,K d ; an , Proc. International Meetinf go Advanced Materials, Materials Research Society (1988, Tokyo). . Murase2S . . ShirakiH , . TanakaM , . KoizumiM , . MaedaH , . TakanoI , , N. Aoki, M. Ichihara, E. Suzuki, K. Watanabe, K. Noto, and Y. Muto; IEEE Trans Magn.n o , , MAG-21, (1985 316-319. )pp . . Koizumi3M . . Murase. MaedaS H , d ;an , Proc. llth Symp n Fusioo . n Engineering, (1986), pp. 474-477. 4. K. Inoue, T. Takeuchi, K. Itoh, S. Murase, H. Shiraki, S. Nakayama, T. Fujioka . Sumiyoshi,Y . HamajimaT d an , ; Proc. llth International Conférenc Magnen eo t Technology (1989, Tsukuba). 5. S. Nakayama, H. Shiraki, & S. Murase; Proc. thé Meeting of Cryogenic Engineerin Japann i g , (1989. 97 . )p 6. K. Noto; Proc. US/Japan Workshop on High Field Superconducting Material Fusionr fo s , (1986). 7. P. Turowski and Th. Schneider; Proc. 10th Int. Conf. Magn. Tech., (1987). 8. R. Fluekiger; Commemorative Symp. ISTEC, (1988). . MaedaH . 9 . UrataM , . TakanoI , . OgiwaraH , . MiyakeS . ,AokiN d ;an , Proc. 10th Int. Conf. Magn. Tech. (1987). 10. H. Mukai, 0. Osaki, T. Hamajima, M. Naganuma, H. Shiraki, T. Fujioka, M. Nishi, Y. Takahashi, H. Tsuji, T. Ando and S. Shimamoto; Proc. llth Int. Conf. Magn. Tech. (1989). 11. K. Yoshida, M.F. Nishi, Y. Takahashi, H. Tsuji, K. Koizumi, K. Okuno . Wachi Y ,. Ando T . Itod I ,; an h Proc. llth Int. Conf. Magn. Tech. (1989). 36 PROGRES SUPERCONDUCTINN SI G WIRES IN FURUKAWA, JAPAN . TANAKY A Yokoham LaboratoriesD aR& , Furukawa Electric Company Ltd, Yokohama, Japan Abstract recend an tt ar Thprogres ee th statn superconductin i sf o e g wired an s demonstrating magnet Furukawn si reviewee aar discussedd dan e contributionth f .O s to superconductivity made by Furukawa, three have been outstanding. The first is to suppl practicae yth l superconducting wires with excellent superconducting properties involving reliability and resonable cost. Secondly, Furukawa has understood critical relationship f processingo s superconductino st g characteristicso t thire .s Th i d challenge developing advanced superconducting materials, demonstratin w highne ga - field magnet. Nb-Ti alloy, Nb.,S V-.Gd nan a compound superconducting wires have been performe practicae th s a d l superconducto applicationsC D r rfo C applicationA r .Fo s these superconducting materials have been also developed a well-controlle n I . d conditio maximue nth m critical curren multifilamentartf o densitT 5 t ya y Nb-Ti wire othee th 0 rA/c exceed1 n handO x m . 8 ,s3. multifilamentary Nb-,Al wires with ultra fine filaments around 0.1 yUm have been developed, resulting in success. The oxide superconducting wires suc YBC s BSCCd a h alsO e an Oar o developed, reaching high transport current densitie ovef so r 1x1 0already, K A/c 7 7 mt .a INTRODUCTION As regard hige sth h field superconductors, manufactuerers rank these candidates with respect to ease of fabrication, reliability in mechanical handling and cost as well as superconducting properties. First of all, we should suply the superconducting wires and /or magnets enough to understand and estimate the superconducting properties. The superconducting properties which are important for appications such as MRI, accelerator dipole magnets and fusion are transition temperature uppe, Tc ,r critical field, Hc2 criticad ,an l current e densityTh . Jc , oftee ar firs no referetw t intrinsi s a o dt c properties because the e determineyar d chemicae byth l compositio stronglt no d nan y dependen microstructuree th n to s i s ,a the case for Jc. However, the practical superconducting wire must have ease of fabricatio handlingd nan , resultin higga witc hJ hreasonabla e costo S . far, only three kind superconductorf so s have been develope r practicafo d l use, namely Nb-Ti alloy, Nb.,S V-Gd nan a compounds above-mentionee th , n whico t hfi d basic propertie practicad san l requirements. Quite recently situation regarding choice of superconducting materials and wires have changed for the more advanced level in the magnetic field, the operating frequency, the mechanical strength and the critical temperature. As a result, the scientific knowledge and the experience have been needed much more in order to understand critical relationship f fabricatino s g processe o superconductint s g properties. This necessit Nb-Te y th pointein i allo t dou y superconducting wires, firstly.(1) In this system, we have also obtained three fundamental limiting factors f non-homogenito chemicae th n i yl composition, formatioe intermetallith f o n c matrix-filamencompoune th t a d t interfac filamene th r eo t non-uniformite th d an y pinning force due to c< -Ti precipitates and drawing strain after heat treatments, coming to the highest critical current density. The new programs of the Superconducting Super Collide rLarge (SSCth ed Hadro)an n Collider (LHC) have been encourage thesy db e approaches. However there are many historical knowledges of the Nb.,Sn and V_.Ga compound superconducting wires, a major difference between Nb-Ti alloy and these compound superconductors is its ease of fabrication or its ease of handling. The other high- field superconductor alse sar o compound deforme e whicb n hca d onln i y % abou 2 t0. tension before they fracture and become useless in practical superconducting wires. On the other hand, the oxide superconducting wires are in more strict situation because experienc s lackini e makinn i g g wire d scientifian s c informatios i n significantly comprex for controlling superconducting properties. 37 e presenTh t pape s beeha rn preparee Septembeth r dfo r 1989 Meeting on Superconducting Material Magnetsd an s , organize IAEy db A Specialists, It briefly reviews recent work relevan high-fiele th o tt d superconducting wires. Detailed arguments have been or will be presented elsewhere.(2)(3)(4)(5) PRACTICAL SUPERCONDUCTING WIRES Nb-Ti ALLOY The industrial fabrication of Nb-Ti alloy superconducting wires is well developed, applicationsI largelMR r fo y additionn .I , valuable results froe th m TEVATRON project TOREe ,th SPRA program TOPAe ,prograC th ZSS projece mth d shotan w that significant improvements in Nb-Ti alloy wires are possible. The effort has been focused on improving the Jc around 5 T to 8 T since 1970. The experimental Jc value was demonstrate fabricatioe th y db commercializef no d product shows sa Fign ni .1 These Jc values were achieved for strands with random filament sizes by a combination of homoginized Nb-Ti alloy ingot and an alternate heat treatment schedule developed since 1970. Nb-Tf o c iJ superconductine Th dominatee gar intrisiy db c factors which depend basie oth nc fluxoid-microstructure interactio extrisid nan c factors which relate with filament non-uniformities such as the sausaging or the sunflower effect. The optimum transport currenfacn i tbalanca s i t e betwee n optimizatioa n e th f o n intrinsic microstructure and a minimization of filament non-uniformities.(6) Garber et al. has shown that the resistive transition index, n, is an excellent diagnostic e presencfoth r extrinsif eo c limitations typicae . Th H characteristin- l c rises steeply with decreasing fields when the wire is large, but as the optimum Jc is reached, the low field slope of the curve declines, as shown in Fig. 2. (4) The 2best Jc results tha have tw e obtaine145 d 8 excee 0r ( an fa ) A/m o dmT ds 3835 ( 0 A/mm 4.2Kt a ) . T I I t I I I I I I q 14.9JJIT1 100 » 10.0pm A 7.3jum 5.Sum c so Tevatron o 0 1 8 6 4 2 0 B,T 1970 1980 1990 enhancemenc J 1 Fig. NbTn ti i super- Fig. 2 Effect of filament size on n- conductinf wires H characteristics in NbTi wires The drawing strain in the thermo-mechanical processes is a significant facto optimizinn ri . FiggJc valuc .showJ 3 e es th oftaine functioa s da f ntotao l drawing strain define, ,Et totas da l cold work froextrusioe mth ne fina sizth o let one. The maximum Jc values at 5T indicated by 4/380°C, 5/380°C and 6/380°C, shift to higher strain with increasin numbee th heaf f gro o t treatment thesd san e values 38 are in the range of Kaniti's empirical relationship between Jc and Et. But the best Jc results deviate fron his estimation, coining to the maximum Je on a small total drawing strain difference Th . e betwee composite nth e wires strongly e depenth n do thermo-mechanical conditions before final drawing process. At five filament sizes les. g s. , e tha n 1Q0m contrai,o t extrinsie lth c factors is significantly importan orden ti achievo rt e high critical current deusitiese Th . principle cause of filament non- uniformities in fine filaments of less than 10/ftn is the presence of intermetallic compounds at the matrix- filament interface. To preven intermetalligrowte e th tth f ho c compound maintaio t d san n filament quality, a diffusion barries was added to each individual filament.(8) This technique had been already developed in a Japanese company. This is also effective for altanative approaches such as the double stacking method for the fine filament conductor and e ultrth a fine filament wire, resultin performance th n gi sub-microf eo n filaments for AC applications.(9) The extrinsic limitation of the Jc values is affected by twisting, final annealing and cabling after controlling the Jc valuces in a strand for the cables. The effects of twisting on the critical current degradation are show n ) FigTesi n(3 . t.4 results, whicD programR& hC obtaineSS ,e th revea n do l tha criticae tth l current degradation depend pite th h n lengtso h after twistine th g composite strand. 4.5 4.0 OJ § 3.0 o 6/375°C 2.5 o © 3/420°C ~3 • 4/380°C 2.0 o 5/380°C\ - * 6/380°C 0 2 5 1 0 1 1.5 TWIST PITCH (mm) 4 6 8 10 12 14 16 Total Drawing Straint ,E Fig. 3 Critical current densities Fig 4 .Critica l current degradation versus total drawing strai NbTn ni i due to twisting of optimized NbTi Wires wires Nb3Sn COMPOUND The recent progress in fabrication process of the compound superconducting wire contributes sha d expansivel develoo yt high-fiele pth d superconducting magnets. The so-called "bronze process", as a predcminant one of several processes, has been develope successfulld an d y high-fielappliee th r dfo d magnets processe th n I b .N , filaments are processed in a bronze matrix and then reacted typically at 650-750" C 5 day 2- o fort Nb.,Sse r mth fo n superconducting layer. This proces s weli s l established in Furukawa and has been used to fabricate about 10 tons of superconductor shows sa Fign ni primaril., 5 fusior yfo n research magnets.(10o Tw ) major limitations of this process are (1) the limited ductility of bronze requires many indermediate anneals during drawing, and (2) the need to co-process with bronze means thae overalth t l current densit reduceds yi , comin o requirt g e higth eh critical current density of compound itself. Aa modifies d bronze proces improvo st e criticath e l current densities, titanium additions which made either to the core or the matrix, or both have been 39 developed.(1 1) Fig.6 shows overal withouc J l t copper versus magnetic field, comparing wit conventionaha l wire optimue Th . m amoun titaniuf to e founb s mi o dt e matrixth cor e 0.3-0.o d t th ,e an % o respectivelyt t 5a % 3-t a 5 . Titanium addition increasinn o t ac growtse gth Nb~She ratth nf e o laye d enhancinran e th g high-fiel becausc dJ increasinf eo uppee th rn g i critica l field ,e titaniuHcTh . m addition also act refininn so g grain matrix e sizeth f so , resultin? a strengthen gi n bronze matrix.(11) 50 \. 10 - u ^ % Twt i Core/Matx. I ' o 0/0 7500,40 o • 0.7/0 750 C, 1.5D 0.5 A 0/0.19 700 C, 80 a 0.7/0.4 700 C, 8D 8 1 6 1 4 1 2 1 0 1 8 0 Magnetic Field, T Fig. 5 Nb3Sn superconducting wires Fig Effec6 . titaniuf to m additionn so e MFTF-th r Bfo project critical current densit Nb3Sf yo n wires Critical current densities on the bronze-processed Nb,Sn wires with various filament size founs showe i filumene ar t sFigdth n I n i .. 7 t siz o obtaiet e th n optimum Jc becomes around one micron and the Jc degradation occures steeply below sub-micron sizes because thanon-uniformitiee tth s before final heat treatmente ar s intermediate carrieth n o d e bronz e anneth f elo matrix procesa s A . s serveye th , Nb.,Sn superconducting wires wit same hth e filamen tmicroe sizon f eno producey b d bronze th e internaprocese th d procesn san lti s wer ec degradatioJ compare e th n do n at the high magnetic fields.(12) It revealed that the bronze processed wire when compared to the internal tin processed one exhibits superior performance at the high e bronze-processeth magneti f o T c12 fieldst a d r examplec fo J ,wir e eth , correspnding with a value of about 3 times for the wire prepared by the internal tin process. An allowable strain in the multifilamentary Nb Sn wires is limited around 0.7% e titaniuth eve n i nm added conductor orden I .o avoirt d this strain effect, different approaches have been developed for applications. The large diameter of laboratory R&D solenoids and magnets for fusion experiments, and the huge design of the conductor permit the use of a reacted Nb,,Sn superconducting wires. Most accelerator dipole or generator designs require that the conductor is bent around a relatively tha d smale conductoan tth d l en s diametee somewhai rth t a rt more delicate conductothae th n fusionr rfo thesr Fo .e delicat e smaleth coil ld an s coils ,choica desigf eo n concepts shoul needee db d from react after windin winr go d after reacting. Especially, to the wind after reacting concept tape shaped wires and stranded wire appliedshowe s sb a Fign n ni ca . .8 40 2.5 2.0 M a> Q g 1.0 o "à (3 0 0.5 1 1 1 l l 1 1 0.5234 0 1 5 Nb3Sn Filament Size, um 7 OveralFig. versue lJ s filament size 8 AlternatFig. e shape Nb3Sf so n wires, of NbSSn wires tape (upper stranded )an e (downdon ) V3Ga COMPOUND We developed multifilamentary stranded V,Ga compound superconducting wires in 1974(13), and demonstrated a 10-Tesla magnet with the stranded wire. Secondly, to increase in current capacity of the wire, a multifilamentary stabilized tape or monolithic wirbees ha ne developed. (14) This tap fabricates e wa bronz th y eb d process composite Th . e tape contains 2,538 vanadium filaments embede a copper n di - gallium matrix, involvin coppega r sheat stabilizationr hfo rollee Th . d composite tape was heat-treated at 650 C for form V,Ga compound layer. The dimentions of the tape fabricated were m wid0.1m e0 6m 5. thicm includin a kan a non-organig c insulation. The coil consisting 6 double pancake module of 30 mm in bore was tested in a Nb-Ti magnet having a bore diameter of 124 mm, producing a magnetic field of 12 Tesla.(15) Recently have ,w e develope in-site dth u V_.Ga compound tape coie .Th l consisting 1 4 double pancake modules made from the in-situ V^Ga tape was tested in a Nb-Ti bias field magnet, generating a magnetic field of 14.4 Tesla, in which the bias field of the Nb-Ti coil was 9.9 Tesla. (16) The cross-sectional view of the tape and the complete coil show in Fig. 9 and Fig. 10, respectively. Although the longitudinal distribution of vanadium for the tape was so good and the filaments were bridging each higo othet he contenrdu vanadiumf to stabilite ,th fieldw lo t ,ya flux jump phenomena, rémanent fiel hysteresisd dan , which were generally observed, wert no e more significant than our expection. (16) As a result, there was no problems for the in-situ tap applo et commerciae yth experimentar lo l superconducting magnets. Fig. 11 shows the typical Jc-B character!stice for V,Ga tape and the bronze- processed Nb-,Sn wire. In the figure it is found that the critical current densities of V.,Ga compounds are promising to the high field and superior to Nb^Sn compounds, criticaane dth l current densitie in-site th f so u V-.Ga / surfacetap/ B e( e als)ar o superior to those of the bronze-processed V.,Ga tape because the in-situ tape has fevorable current anisotropy. For instance, the current anisotropy at 12.5T, Jo? /Jcj. , is about 1.5. The highest overall Jc// of 2.2*10 A/cm at 12T is obtained Cu-42ate th r fo %tapeV , whil overale arouns ei th t dlJo 1.5*10 A/c 12Tt ma . Thus the current anisotropy advanageouseeme b o st pancake th o st e cools. 41 Fig. 9 Cross-sectional view of in-situ Fig. 10 Complete coil using in-situ V3Ga tape V3Ga tape Generall e straith y n e in-siteffecth n i ut wirs supressedi e f o e on , fundamental problems of the compound superconducting wires seems to be avouidable. e straith t n Bu effectsi in-site th n u tape e significantlsar y affectee th y b d vanadium fraction before reaction and the conversion ratio from vanadium filament to V~Ga layer. Fig. 12 show a typical comparison with the bronze-procèssea Nb_Sn wire. In the case, the composite composition was Cu-40 at% V alloy and the conversion ratio was about 70%. When the conversion ratio in the same composite was about 100%, allowable strains were steeply drawn back below 0.1%.(16) 10e I ———— 1 ———— 1 in-situ process B11.= 5T B//S o 105 Q,; o -3 S o • in-sit° u V3Ga 10" 0 * Bronze V3Ga : ~~~ Bronze Nb3Sn 0.4 0.6 0.8 1.0 1.2 1.4 7 1 5 1 3 1 1 1 9 7 0 Bending Strain,% Magnetic Field,T Fig. 12 Effect of bending strain on Fig. 11 Current anisotropies in V3Ga critical current densities for compound tape superconducting wires 42 COMPARISO PERFORMANCF NO COMPOUNF EO D WITH NbT HIGR iFO H FIELD MAGNETS Fig. 13 attempts to provide an evaluation of the perfoumance in the superconducting wire high-fielr sfo d magnets. Five performanc, eJc factor e th f so intrinsic properties (Tc, Hc2), reliability, ease of fabrication and cost are selcted, and each factor is ranked by three stages. In the diagram, it means that the central region is not available for practical use and the outer region is useful e Nb-TTh .i alloy wire e almossar t performed withou e fundamentath t l properties such as Tc or Hc2. Thus the Tc or Hc2 performance for Nb-Ti wires is requireed finally .approacheo Thertw purpose e e th e ar r cotrollin) sfo : (1 g alloying elements and (2) operating the coils below 4.2K. The latter approach has been demonstrated by Saclay group as shown in Fig. 14. (17) A 10-T test facility has been successfully constructed and tested quite satisfactly at 1.8K under atmospheric pressure. On the other hand, the compound superconducting wires are in apparently strict situation. Especially, the strain effect related with reliability is the most keen problem. This effect seems to be unavoidable even if developing on reinforcements. But other factors depen fabrication do n techniques dependind , an choic e th f o n ego application scales large enhancement will be expected. Thus improvements to the compoun fabricatiod. n processes could eventually chang future eth e situationr Fo . instance, the in-situ process can be conformable for the Jc performance, and the internal tin process for ease of fabrication and resonable cost. Jc Performance Large-scale, Ease of Intrinsic Production Property, Reasonable Cost Reliability, Ease of handling, Strain effect Magnetic Field, T Fig. 13 Performance diadram for super- Fig. 14 Critical currents of the TORUS conductorsI I d an , K measureconductin 2 4. t da g wire 1.8 K ADVANCED SUPERCONDUCTING WIRE Nb3Al COMPOUND Several processes have been developed which allow aluminum filamente b o t s fabricated and then converted to Nb,Al after mechanical deformation is complete. In the first, the powder metallurgy process has been demonstrated. There are two technical problems in the powder process, i.e., the filament continuity along the longitudinal directon of the wire and the filament bridging. Recently, the jelly- roll process and the composite process (18) have been developed, enhancing the Jc 43 valv f fino e e filaments. Fig show5 .1 s effect filamenf so te pinnin sizeth n so g force. It seems to be consisted of three regions, i.e., below 70nm, 70-200 run and over 200nm n eacI .h filament region, although coult ,i e predictedb d thae th t formation mechamis Nb-,Af mo l compound pinnine th d gsan mechanism were significantly changed, we should make efforts to investigate in detail. Fig. 16 shows a comparison of Jc-B characteristics for metallic superconducting wires. The Nb^Al compound superconducing wire runkee ,ar d abou middle tth e eve t higa n h magnetic fields, much attempt in the development for fabrication processes and scientific understandin needede gar . 10" 1.0 NRIM AI-Cu NRIM AI-Mg AI-Mg I AI-Mg o H 10 T, 4.2 K o Q. u. Q. at O 0.5 o NbTi "c *-« c a. o •a « N 5 o at 4.2K O —— Champion data ——— Industriae us l 10 i i t i i » i • 0 60 0 50 0 40 0 30 0 1020 0 0 2 6 1 2 1 8 4 0 Magnetic Field,T Filament Size, nm Fig 6 Jc-1 . B characteristic Nb3Af so l Fig. 15 Effect of filament sizes on wires, compared with practical super- normalized pinning force in Nb3Al wires conducting wires OXIDE SUPERCONDUCTOR Oxide superconductors havsame eth e characteristic hardnesf so brittlenesd san s as intermetallic compound superconductors suc Nb,Ss ha V,Gad nan . Thu a direcs t application of the tecniques elaborated for the intermetallic ccnpounds is useful to attempt in making for oxide superconducting wires. Recently, sufficiently long oxide superconducting wires anri tapes have been made by the composite process, process. In the composite process, a metal tube such as silver was filled with oxide powder loadee Th . d tubthes ewa n processe drawiny db rollind gan g into wire and tape, successively heat-treate r undeh t 800°a 0 dr5 Coxyger 900fo °nC atmosphere. Fig.17 shows Jc-B characteristic Ag-sheathef so d BSCOO superconducting wires at 77K. The Jc values remarkably depend on the degree of crystal orientation. coulc K valueJ 77 de t a skeeTh p 0.01T o higt p hu , Howeve c steeplJ r y decreased beyond 0.01T.(19) On the other hand, in order to measure the transport current at helium temperature, two aligned wires with different F value, which is the degree of crystal orientation, were prepared. As shown in2Fig.18 the Jc indicated F95% decreased wit magnetie hth c field from 1.2*1d 5*1an o t 0 T T 0A/c1 A/c10 t a mt a m then kept almost constant up to 23T.(5) With respect to the transpot current of the oxide superconducting wires, these approaches are very premising to develop the high field superconducting wire coild san . (5)(20) 44 10 F : The degree of crystal orientation I«1 o • • * • • •%•*• F 95% (BlCaxis) ' o F 83% 10 - 10' (BJLCaxIs) _L _L 10 20 30 0 0.01 B(T) Fig. 17 Magnetic field dependence of Fig. 18 Magnetic field dependence of Ag-sheather Jcfo d BSCOO wires,measure Ag-sheather fo c J d d BSOCO wires, K a7 t7 measureK 2 4. t da CONCLUSION Furukawa has contributed to superconductivity by to supply the practical wires, o understant fundamentae th d l propertie o develot d san p advanced superconducting wires. A large improvement in the critical current density of NbTi wires has accompanied by reduction n filameni s t sizd speciaan e l thermo-mechanical technologiest I . appears that extrinsic factors such as sausaging have so far produce majoe dth r limitation transpore th o s t . Nb^StJc n would hole advantagdth e over NbTi because of its higher critical field. However, significant requirements for NfcuSn wires would be ease of fabrication, mechanical reliability and resonable cost, under developed alternate processes would also require. V.,Ga superconducting wires woul almose b dsame tth e situatio Nb^Se th ns na wiresa condidat s A . f o e advanced superconducting wires Nb-.Al compound would be developed successively. However, development processingf o s easr sfo fabricatiof e o d improvementnan f o s superconducting properties will require further RSD and few years for production scale-up and demonstration. Oxide superconducting wires have been runked also as an advanced materials even for higher magnetic fields. ACKNOWLEDGEMEMTS We are very grateful to K. Enomoto, T. Sano for much careful experimental work . Matsumot K . IkedM d r an o ahelpfut fo o d an l discussione Th . particular encouragemen Konum . Shig. M S gratefulld e f tao an ar y acknowledged. REFERENCES (1) D. C. Larbalestier : IEEE Trans, on Magnetics, MAG-21, (1985)257 Tanaka. Y ) Konishi. (2 ,K Matsumot. ,K Megur. S d Proco: oan . Int. Symp n o .Flu x Pinning and Electromagnetic Properties in Superconductors, Fukuoka, Japan (1985) 212 Iked. M Proc) a: (3 . ISSSC, USA, (1989) Matsumot. K Tanak. ) Y (4 d Presentea: oan MT-11t da , Tsukuba, Japan, Sept. (1989) (5) N. Enomoto, H. Kikuchi, N. Uno, M. Ikeda, H. Kumakura, K. Togano and K. Watanabe: Presente MT-11t da , Tsukuba, Japan, Aug.(1989 Submitte): Japao dt Appl. nJ . Phys. 45 Larbalestie. C Warna. . H D . d W s an ) Procr: (6 . Int. Symp n Flu.o x Pinnind an g Electromagnetic Properties of Superconductors, Fukuoka, Japan (1985) 156 (7 Matsumot. )K Tanak. Y d a: o Procan h US-Japa6t . n Worksho n HFSCo p , Bouler, USA Feb. (1989) (8) T. S. Kreilick, E. Gregory and J. Wong : Adv. in Cryo. Eugn. 32(1986)739 (9) Y. Tanaka, K. Matsumoto, M. Yamamoto, D. Tsukamoto and T. Ishigohka : Appl Super. Cbnf. Baltimore (1986), LH-28. (10)R Scanlan. .M Zbasnik. P Baldi . . ,PickeringJ W . L . ,R . ,J Furuto. ,Y . ,IkedM a and S. Meguro: IEEE Trans, on Mangetices MAG-21 (1985) 1087 (11 )Y. Tanaka, Y.Furuto, M. Ikeda, I. Inoue, S. Suzuki and S. Meguro : 2nd US-Japan Workshop on HFSC, San Diego.(1983) (12)Y. Tanaka Tanak. IkedH . ,M d IEE: aaan E Trans Nuclean ,o r Science, NS-32(1985) 3714 (13)Y. Furuto, T. Suzuki, K. Tachikawa and Y. Iwasa : Appl Phys. Lett., 24(1974)34 (14)Y. Tanaka, Y. Furuto, M. Ikeda, I. Inoue and J. Tahii : Proc. ICEC-6, Grenoble, (1976)381 (15)M. Ikeda Oishi. ,K Ban. ,M MegurTanak. ,. S Y d IEEo: aan E Trans n Magnetics.O , MAG-17(1981)2003 (16)K.Oishi Takei. ,N YamadTanak. . ,K Y d IEE: aan E Trans n Magnetics,o , MAG-24 (1988)1393 (17)R. Aymer, P. Genevey, S. Palanque, A. Sagnize, J. P. Soubeyrand and B. Turck : Cryogenics, Sept. (1980)521 (18)T. Takeuchi, Y. lijima, M. Kosuge, T. Kuroda, M. Yuyama and K. Inoue : 1988 Appl Super Conf. Franciscon ,Sa , MF-2 (19)K. Enomoto et al : 36th Spring Convention of the Japan Society of Appl.Phys., (1989) 2a-PC-16. (20)Y. Tanaka SanYamad. . ,T K Jpn o: d Appla. an .J . Phys., 27(1988)799. 46 FLUX PINNING BY THIN a-Ti RIBBONS IN SUPERCONDUCTING Nb-Ti . MATSUSHITAT . OTABES , . MATSUNOT , . YAMAFUJK , I Departmen f Electronicso t , Kyushu University, Fukuoka, Japan Abstract It has been found by Lee and Larbalestier that the critical temperature reduces monotonically, while the critical current density first increases and is followe a decrease y b d , accordin e diameteth s a g f superconductino r g multi- filamentary Nb-Ti wire is decreased. This behavior is mainly caused by the proximity effect between superconducting matrix and normal o-Ti ribbons that becomes more remarkabl e thicknesseth s regiono a e tw f o ss become smaller. The critical temperature and the elementary pinning strength of a-Ti ribbons are theoretically estimate solviny db phenomenologicae gth l Ginzburg-Landau equations for multilayered structure with superconducting and normal layers. e criticaTh l current densit s calculateyi d fro mstatisticaa l summatioe th f no elementary pinning forces e obtaineTh . d critical temperature decreases mono- tonically, while the critical current density increases, with decreasing thicknesses of the two layers. The decrease in Jc observed by Lee and Larbalestier is consid- ered to result from constriction of superconducting layers by heavy cold work, associatee whicb y hma d wit sausagine hth superconductinf go g filamente th n si multifilamentary wire. If a development of drawing technology that does not disorde laminae th r r structure insid filamene eth achieveds i t furthea , r increase c wil J realizee b n l i furthea y db r reductio wiree th f .no 1. INTRODUCTION From the view of application, superconducting Nb-Ti wires are required to have much larger critical current densitie spresene thath n i t stage. Main pinning centers Nb-Tn i normae ar i l a-Ti precipitates processine th r fo s f Nb-TgA o . i wiresn a , empirical combination of cold work and heat treatment is employed so as to obtain suitable geometries of normal a-Ti precipitates for flux pinning. The thicknesses of a-Ti ribbon interstitiad san l superconducting layer decreasee sar d durin reductioga n odiametea f multi-filamentarf ro y Nb-Ti wir coly eb d drawing. These thicknessee sar comparable to or smaller than the coherence length in the superconducting matrix (/J-phase).1) Hence, the proximity effect seems to be notable. Meingast et a/.2) and Lee and Larbalestier1^ measured the critical temperature and the critical current density in Nb-Ti at various stages of cold work after heat treatment. During this process, the wire is drawn extensively. They found that the critical temperature reduced monotonically while the critical current density had a maximu thicknese th mt a a-Tf so i precipitate superconductine th f o abouf so % 20 t g coherence length, according as the wire diameter was reduced. The optimization of a-Ti structure in Nb-Ti wires is required so as to get higher critical current density. But the phenomenon is very complex because of the prox- imity effect, and it is hard to find empirically the optimum condition for the critical 47 current density. Hence worths i t i , estimato yt e theoreticall effece y th f proximit o t y on the critical temperature and the critical current density. In this paper,this prob- investigates i m le solviny db phenomenologicae gth l Ginzburg-Landau equationr sfo superconducting-normal multilayers. The result is compared with the experimental results by Meingast et al. and Lee and Larbalestier. 2. THEORY 1 Basi2. c Equations The superconducting properties in semimacroscopic scale can be described by the phenomenological Ginzburg-Landau equations. These equations in the super- conducting region are given by J_ (-JÄV 2m (1) - * W) - — 1 m m (2) vectoe wherth s i r epotentialA superconductine —* th s i ? \ , g order parameters i J d an , —* the current density. In the normal region, on the other hand, the order parameter obey schrödingee sth r equation — 2m (3) instead of Eq.(l), where an is a positive parameter associated with the repulsive potential for superconducting electron pairs. r estimatioFo e criticath f no l temperature treae case w , th te wher currene th e t e magnetith d t an appliedcno fiele ar d. Sinc e geometrth e e normath f o y l a-Ti ribbon precipitate superconductinn i s g Nb-T s complicatedi i approximate w , e eth present system by the idealized multilayered structure with superconducting and normal layers as shown in Fig. 1. In this figure, the thicknesses of normal and su- perconductin 2dd t,an gn respectively layer2d e sar . These thicknesse normalizee sar d by the G-L coherence length Ç in the superconducting layers. We define the rç-axis, which is also normalized by £, along the direction normal to the multilayer. From periodicitye th have w , e onlFign i . conside o 1 ys t . d + regioe n d th r 7< 7< n 0 In this case Eqs.(l) and (3) are reduced to 3) ; 0 = 3 R - +R drf (4) normal superconducting -2d, Figure 1. Idealized periodic multilayer. 48 dn (5) where R is the absolute value of normalized \£. The parameter 0 is defined by (6) wher h/(2ma= „ e£ coherence th n Y^s i e lengtnormae th n hi l layer. Equations (4) and (5) can be solved analytically and we have R = ~ Tô, n )d + s d drj< < n (7) Q l-P c = (9) and TJQ is an integral constant. In Eq.(8), R(Q) is a value of R at rj = 0. These con- stants, c, rj0 and jR(0) can be determined from the boundary conditions, continuity of R and dR/drj at the interface rj — dn and dRfdrj = 0 at r? = dn + ds.^ Examples of obtained R are shown in Figs. 2(a) and 2(b). In these figures, the ratio of normal and superconducting layers is fixed as d : d = 1 : 4. It is seen that s e proximitth y effect becomes remarkable whe thicknessee nth layero tw f sso becomen small. Figure 2. Order parameter in the multilayer for (a) dn — 0.6 and d, = 2.4 0.8= , .d d an 2 0. = n d an) d(b 49 2.2 Critical Temperature First we consider the condensation energy density F in the two regions. This is givey nb ; dn 0 < rj < dn. (11) In the above, HQ is the bulk thermodynamic critical field. The critical temperature Tdetermines Ci minimue th y db m C , poinwhicF f o th correspond maximue th o st m of R. At the center of superconducting layer 77 = dn + ds, R takes the maximum value, R = \/l — c. Thus, we have 2 In the absence of the proximity effect, this takes —1/2. This enables us to define 2 an effective thermodynamic field under the proximity effect: Hc = Hc0^/l — c . If we assume that the electronic state density at the Fermi surface is not appreciably influenced by the proximity effect, the critical temperature is proportional to Hc. Thus obtaine w , 4^ 2 , TTcQ= cc Vl - (13) where Tc0 is the bulk critical temperature under no influence of the proximity effect. 2.3 Elementary Pinning Force normae th n I l corfluxoida f eo superconductine th , g order paramete reduces i r d zeroo t . Hence condensatioe th , n energy changes whe fluxoie nth displaces di e th n di region where normal precipitates are distributed. The elementary pinning force can estimatee b d fro variatioe mth condensatioe th f no n energy durin gdisplacemena f o t the fluxoid. For simplicity, we shall estimate a difference in the energy between the two cases (a) and (b) shown in Fig. 3. When the fluxoid is located in superconducting normad an l layer depictes sa Fign di . 3(a) energe th , periodir ype c length ) s 2(d„d + s givei y nb fluxoid n o-Ti (a) (b) Figur . Idealizee3 d arrangemen f normao t l layer fluxoidd san . 50 fdn+d, Fdrj. (14) Jo Whe fluxoie nth places di normal-layee th n di r free regio shows na Fign ni . 3(b)e ,th energs yi where (R)s means a value of the order parameter averaged in the superconducting layer. Thus e elementarth , y pinnin ge norma forcth f o e l layer s approximateli s y estimated from TT G'-G ( B When the proximity effect does not occur, the above equation is reduce to fpQ= . (17) D C2 In a practical case, the normal precipitates are not always perpendicular to the fluxoids. Whe normae nth l precipitate perpendiculas si fluxoide th o volume t r th , e 2 of elementary interactio s minimuni d takeman s 2TrC e practicadTh n. l interaction volume is larger than this value, and the elementary pinning force is given afp, where a correctio s i a n factor. Typicall e e widtnorma th havth 2.9e — yf w r ho 8a efo l layer of wt; = 30£. 2.4 Critical Current Density In commercial Nb-Ti wires, strong pinning points, i.e., edge a-Tif so ribbons, with very high density distribute randomly. Hence, the critical current density should be calculated in terms of the statistical summation. For strong pinning centers as in the present case, the transverse correlation length of the fluxoid lattice lee is less than the fluxoid spacing a/. Hence, the fluxoids behave incoherently as in the amorphous state. The longitudinal correlation length £44 is expected to be given by a cut off length 4?ra/. Thus numbee th , elasticalle f pinth o r n si y correlated volums ei (18) where N is a concentration of a-Ti ribbons in a transverse cross section and is approximately givey nb numbee Th practicae th varien r n i 0 s1 l frocaseso t m2 t larg. no Sinc es i enoughen , a deviation from the simple statistical behavior is expected. From a numerical simulation, the strength of a superpinning center composed of n pins with the force 4 estimates i af s 1.6n°da a/. -56 Thus, critical current densit obtaines yi d as) p p a56 7 1.6n o/P , N 0 2 — —— A - J= c wher e flueth x 51 3. DISCUSSION 3.1 Critical Temperature The critical temperature of Nb-Ti composites has been measured at various stages of processing by Meingast et al ^ In the first stage, Tc raises from 9.1 to 9.5 K. This comes fro mvariatioa f compositiono n during precipitatio f a-Ti no e ensuin Th . g cold drawing without heat treatment make c falT s l . OnlfroK e mth 7 y 8. 9.4 o t 4 geometr f a-Tyo i precipitate varies theid swa dan r volume fractio abous % wa n 20 t during this process. This variation seem originato t s e fro e proximitmth y effect. Her compare ew presene eth t theoretical result with their experimental data. From the observed volume fractio f a-Tio n wile ,w le cas limif th d o en /d o 0.2t = t sn i 5 this calculation. From Fig. l in Ref. 2, we assume the bulk critical temperature Tc0 as 9.50 K. The obtained theoretical result of the critical temperature and the experimental dat showe aar Fign ni . 4 . 4 0. 3 0. 2 0. 0.1 Figure 4. Calculated critical temperature for dn/ds = 0.25 and the ob- served resultMeingasy b ) (o s t al.e t 2) Both the theoretical and experimental critical temperatures decrease monoton- ically wit e decreasinhth g thicknesses e experimentaTh . l data agree wit cale hth - culated result with 9 = 1.4 in the range of dn > 0.1. For dn < 0.1, however, the deviation betwee experimentae nth theoreticad an l l result larges si c dropT : s sharply wit decreasine hth g thicknesses. This sudden degradatio c suggestT n ni s thae th t proximity effect becomes more remarkabl 0.1< n sucn .d I extensivelr n ha efo y cold drawn wire, the laminar structure inside the filament may be disordered heavily, resulting in appearance of the region where the superconducting layer is constricted and the ratio dn/ds is locally large. The proximity effect in such a region is expected to be much more significant and the superconducting characteristic seems to be de- graded more. In fact, disordered laminar structures are observed in such wires and this fact supports the above assumption. 52 3.2 Critical Current Density In Fig. 5 the calculated critical current density at B = 5 T is compared with experimentae th Larbalestier,d an l e resultLe y sb 1-1 wher assumee ew d tha ratie th t o of dn/ds is constant (0.25) and w is proportional to dn.^ We also used the param- eters= 5.5facto A = (p10.m f . f #20.6 n o : d 8r T) = 0.13s van Hi 1T 2 0 c 0 c0 gooa t t wit multipliedexperimentse fi ge hth o t calculatee s th a o e dt o s Th c .dJ larger theoretical valu direca e o that experimente te nsummatioth du e b y f ma sn o the forces on respective fluxoids. According to reduction of the layer thicknesses, the calculated Jc increases monotonically, while the observed Jc first increases and then drops e suddesuddenl0.1< Th n . d nr c droyseemfo J likee b f po do t s with suddee th n dro T show f po Fign n i degradatio e . Tha.4 th , is t J als n ni o seemo t s c causee strongee b th y db cr proximit yregioe effecth t a nt wher superconductine eth g layer is significantly constricted. The theoretical results for larger values of dn/ds are shown in Fig. 5. This shows that Jc decreases sharply according as dn/ds in- creases. This disorde laminae th f o r r structur relatee sausagine b th y o dt ema f go the superconducting filaments.^ If the laminar structure is not disordered during the drawing, the critical current density is expected not to be degraded. 3.5 0 = 1.4 -3.0 E =0.25 2.5 2.0, 0.1 0.2 0.3 Figur . Calculatee5 d critica observee th l currend an d4 1. t densit= 6 r yfo results (o) by Lee and Larbalestier.^ 4. CONCLUSION The proximity effect by a-Ti precipitates in multifilamentary Nb-Ti wires is in- vestigate solviny db phenomenologicae gth equationL theoreticae G- l th d san l results e comparear oc nJ d dan wit e experimentahth l resultf o d f Meingaso san . al t e t Lee and Larbalestier. The following results are obtained: 1. The decrease in the critical temperature and the increase in the critical current density according to the reduction of the layer thicknesses are explained by the theory for dn > 0.1. 53 e sudde2Th . n degradatio e criticath n ni l temperatur e criticath d lan e current density for dn < 0.1 may be caused by a constriction of superconducting layers. 3. If a development of drawing technology that does not disorder the laminar structure is attained, further increase in the critical current density is expected. References 1) P. J. Lee and D. C. Larbalestier, Acta metall. 35, 2523 (1987). . Meingast2C ) . LarbalestierC . . DaeumlingD M , d an . LeeJ , . ApplP , . Phys. Lett. 51, 688 (1987). 3) T. Matsushita, J. Appl. Phys. 54, 281 (1983). 4) T. Matsushita, S. Otabe and T. Matsuno, to be published in Proc. Int. Cryo. Mater. Conf. Angeless Lo , , 1989. . Meingast5C ). Thesi Dr . s Ph ,(Univ . Wisconsin, 1988). 54 STRAIN DEPENDENCE OF CRITICAL CURRENT IN ADVANCED Nb3Sn SUPERCONDUCTORS . OKADAT . KATAGIRK , I Institut f Scientifieo Industriad can l Research, Osaka University, Ibaraki H. YOSHIDA Research Reactor Institute, Kyoto University, Osaka Japan Abstract Parameters relevant to the strain dependence of critical current Ic have been investigated with emphasi correlatioe th n so n witconductoe th h r constitution and heat treatment. Examinations have been made on several advanced Nb3Sn conductor course th sdevelopmentn f ei whico e har e Th . conductors involved are; bronze processed NbsSn conductors with third element, internally stabilized conductors, internally reinforced conductors, internal tin diffusion processed conductors and in situ processed conductors straie Th . n characteristic specifiee sar termn i d pref so - compression strain, reversible limit strain and strain sensitivity of Ic. Improvement of the strain characteristics in these conductors through proper choice of parameters is briefly discussed. Effec neutrof to n irradiatio straie th n no dependence th f eo conductors is also presented. Changes in the mechanical properties of the constituents induced by irradiation alter the strain characteristics of Ic. Role of the third additive element to the irradiation-strain synergistic effec discusses ti d with emphasi bot specien e samoune e o th hth th d f sto an additive element. INTRODUCTION The developments have been made on fabrication of advanced Nb3Sn conductors which have high critical current density at high fields. The strain effec criticaf to lconductore currenth f o G knowe s tI b si o nt influenced by various parameters associated with constitution of the conducto head ran t treatment behavioe conductoTh e . th f ro r under strais i n characterized by pre-compression strain £m, reversible strain limit eirrev straie anth d n sensitivit c [11I f yo . 55 conductoe th f o alss G i rI oe altereTh neutroy db n irradiations i t I . changcriticae e th th n mainlo ei t le parameterydu conductore th s n i si d san change th strain partlo i et e nydu state inducee irradiationth y db n I . this paper, the results of examination of strain effects in the advanced NbsSn conductors are presented. EXPERIMENTAL Several kinds of wire conductor, the details are to be described in the next section have been studied. The strain dependence of Ic was measured at T (IM1S R Tohoku University T (RRI6 d ,)an Kyoto University) usin e loadingth g apparatus [2,3], The criterion of Ic was 1 yV/cm and the accuracy of strain measuremen 0.0s twa . 5% The neutron irradiatio performes nwa d usin in-core gth e irradiation facility of RRI, Kyoto University at the reactor ambient temperature (estimated specimen temperature; 355K). The fast neutron flux was 1.6xlO-*-3 n/cm^.sec for neutron energy EX).l MeV. The critical temperature Tc at zero magnetic field was also measured by the induction method. RESULTS AND DISCUSSION Bronze Processed Conductors Addition of Ti improved the strain sensitivity of Ic of a multifila- mentary wire conductor (0.31 mm dia.) at high field [4], The em was almost the same as that of without addition and eirrev was slightly small [4], The latte contrars ri resule th o t y previously reported [53. Further studn i y the effect of Ti addition is required on the view points of 1) suppression of martensitic transformation and 2) suppression of Kirkendall void formation. The position of copper stabilizer slightly affects the strain charac- teristics of Ic- Figure 1 shows the comparison of two conductors which have almos e samtth e conductor parameters exceptin e locatioe th g th f no stabilizer [63. In the internally stabilized wire, the £m is lower and the strain sensitivity is higher as compared to that in the externally stabilized one. These are explained in terms of the change in the tri- axial strain state in the (Nb,Ti)3Sn layer induced by stabilizer confined inside of filaments/bronze composite. 56 Comparison between Fig. l(a) with show2 Fig effece . sth heaf to t treatment. The longer heat treatment time results in the larger Ic and an< smaller £m» £Q,irrev 3- the strain sensitivity of Ic. 1001—————————i—————————i———— A-7 InternallyStabilized (Nb,Ti)3Sn 963K(240h) 100 = 15T c Of 3 50 o D o O 50 - Externally Stabilized (Nb.Ti)3Sn 963K(200h) 0.5 1.0 Strain ( 0.5 1.0 Strain( (a) (b) . Strai1 FIG. n dependenc bronzn i c ! f eo processed conductors. Internally Stabilized (Nb,Ti)3Sn 963K(A8h) 15T 0.5 Strain ( FIG. 2. Strain dependence of lc in bronze processed conductors. 57 Internally Reinforced Conductors In order to withstand huge electro-magnetic force as well as to enable winding after reaction heat treatment composite ,th e conductors with rein- forcement have been developed [7] conductoe ,Th r examine bronza s i d e processed multifilamentary (4 \pi x 4620) composite with 7 reinforcing wire of type 405 stainless steel; one in the center and 6 at corners of the composite inside the stabilizing copper sleeve. The conductor, diameter . h 3 17 r heas i fo t, K treatemm bein3 2 92 1. gt da The flow stress of the conductor increased drastically as compared to tha conventionan i t l bronze processed conductors (Fig . Figur3) . showe4 s Ic vs. strain relation at 1ST. Increase in strain tolerance is seen in £m=0.4 %, which is higher than that in the conventional conductors. The intrinsic reversible strain limit £Q,irrev = ^irrev ~ £nu however, is straie th rathe d nw (0. an sensitivitrlo ) 4 % comparabls i c I f yo r eo slightly hig compares ha conventionao t d l conductors. Thes suspectee ear d tcorrelatee ob d witserratioe th h n occurring aroune strai% th 9 n 0. di n stress-strain curve and micro-strain bursts preceding it. These localized deformation in 405 steel (b.c.c. metal) through twining and/or low temperature embrittlemen deterioratn tca e superconducting characteristict sa that portio leadd an ninferioo st r strai nstraie dependencTh n. c I f eo sensitivity will be improved by adding the third element. Proper choice of reinforcing metal and constitution which minimize direct transmission of local strain in the reinforcement to NbsSn layer is required. Another point worthy to be noted in Fig. 4 is that enhanced recovery of Ic in the Ic vs. strain characteristics within the "reversible strain limit". Thi comprehendes si d throug residuae hth l strain state including radial and tangential direction in the conductor induced by reinforcement and its relaxation through the plastic deformation in some of constituents. 500 Internally reinforced Nb3Sn 923K(196h) B=15T A.2K o CL in to a» «+-i». 1/1 i i r i 0 0.5 1.0 1.5 Strain (%) FIG. 3. Stress-strain curve of internally reinforced conductor. 58 0 1. 5 0. Strai) % n( FIG. 4. Strain dependence of lc in internally reinforced conductor. Internal Tin Processed Conductors In orde reduco rt fabricatioe eth n cost, interna procesn lti bees sha n developed e straiTh . n characteristic single th corn n si eS e conductos ri not excellent, but distribution of Sn cores reduced the strain concentration Kirkendale atth l voids and, therefore ee irre ,th fracture equas vth i o lt e strain [83. Figur showe5 e fractursth e surfac internan a f o en lti processed conductor. It has been reported that the addition of Ti make difficult to fabricate a long conductor. The problem, however, has now been overcome throug improvemene hth conducton ti r desig fabricatiod nan n process [9], In situ Processed Conductors Becaus highef eo r strain tolerance [3] sitn ,i u processed conductors have been one of the most promising conductors for large scale high field magnets hige contenb th hN n I . t conducto higr , howeverc rfo I h , difficulty diffusion oS f e uneveth diffusiod n nS nan botn ni h radia d longitudinalan l directio bees nha problemna . Figur showe6 deviatioe th s reactef no d layer cause y impropedb r heat treatment coupled with hig contentb he N th n I . 59 FIG. 5. Fracture surface of internal diffusion processed conductor. FIG. 6. Fracture surface of in situ processed conductor. 60 worst case ballinn ,ti g resulte botn i d h prematurreductiod an c I n eni fractur conductore th f eo . This proble improves mi carefua y db l heat treatment procedure including one prior to Sn plating. Now the blistering has been controlled throug propee hth r desigcladdinu C f no g thickness [10]. In order to scale up the conductor, CaO melting is being developed [11], Multi-core internal diffusion metho uses i dthin di s conductor whicn ,i e hth desig heaf no t treatmen importants ti . Effec Additioni T f to n Irradiatio,o n Effec. I^ f to Figure 7 shows the dependence of Ic on neutron fluence ()>t in a bronze processed multifilamentary NbaSn conducto (Nb,Ti)3Sa d ran n conductoT 4 t ra diametere [12]Th . , bronze-to-Nb rati Cu-to-nod oan e ratiu nth C f oo conductor identicale sar 0.45d an ;,0 0.3 respectively2. , 1mm c valuI e eTh . of . wireA 5 s12 irradiatioe priod th an o rt 8 12 almos s n, i same 4T tth t e a attributee b t straie no Thth n eo nd t ca chang effecc I n thiet i ta s magnetic field, becaus behavioe Icm e eth maximue th , th f ro mI fre e from strain, c with later). The increment of Ic at the peak is less and the (j>t crossing Ic/Ico fairls i addei 1 T y = e dlowe th NbßS n ri n conductor compares sa thao dt n ti the pure NbsSn. Addition of Ti raises Tc and increases BC2 through increase of resistivity at normal state and suppression of martensitic transformation [13,14], If we assume that neutron irradiation changes BC2 through the same mechanism associated with disorde solutr ro e additione th n i ,c I the e nth (Nb,Ti>3Sn conductor will decrease at lower (j>t as compared to that in the _ pur,^ e NbsSn. Nb35n (Nb,Ti)3Sn (Nb,Ti)3SnRef.15— 17 0 10 1018 Neutron fluence (n/cm2) FIG. 7. Dependence of lc on neutron fluence. 61 effece Th irradiatiof to n reporte Webey db r [15 n ](Nb,Ti)3So n conductor (broken lin Fign mors ei i e.) 7 significan compares ta ourso dt . Possible discrepancreasone th difference r sth fo irradiatioe e th yar n ei n temperature (ambient temperature in the present study and low temperature in temperaturw lo e Weber'sTh amoun. e e Ti th thad irradiation )n f an t i o , however, will induce more significant damag compares ea thao dt ambienf to t temperature becaus mobilitw lo f eo poinf yo t defects. Therefore eth discrepancy is not explained by the difference of irradiation temperature. (Nb,Ti)sSn i Th c eT knows ni depeno conteni nt compounT e th n do n ti d layer [13,14]c increaseT e Th , s witconteni T increasn ha e reached th tan f eo sa peak value at 1 at% Ti and then decreases. Therefore, the difference of Ti conten n (Nb,Ti)sSti n conducto n change rca Th . (2.8 at%). The earlier decrease of Ic in Weber's conductor can be attributed to the larger amount of Ti. The dependence of Tc in our specimens on the neutron fluence is shown in Fig. 8. The difference in degradation of Ic in our (Nb,Ti)3Sn conductor from the pure Nb3Sn conductor mentioned before appears to be associated with change in Tc. According to Sekine [14], the peak of Ic occurs in the Ti content range of 0.5-1.5 at% depending on a magnetic field; the higher the field contenthighei e T , th e rth . partl e b Thi n syca ascribee th o dt strain effect caused by pre-compression coupled with the strain sensitivity which depend magnetia n so contenti T ce th fiel .d dan Therefore , optimum Ti content without significant degradation by irradiation should be chosen depending on the magnetic field to be subjected. Recently neutroV Me effece 4 1 ,th n f tirradiatioo conductore th n no f so NbsS beens ha adde n magnetie a examineT th d r o o witt i T cp hT du 0 fiel 2 f do [16,17], The result indicates that the similar effect as mentioned above 1.0 Bronze Tœ(K) • Nb Sn 17.7 o 3 .u a (Nb,Ti)3Sn 17.7 ,0.5 I i I 1019 0t (n/cm2) FIG. 8. Dependence of Tc on neutron fluence. 62 also case exis higth f eo n thi energy neutron irradiatio thad nan a t T addition appears more deleterious [17] orde,n I comparo rt effece eth f to additive elements, however onlt speciee ,no yth alst amoune sbu o th t constituting A-15 phase shoul takee db n into account. Irradiation-Strain Synergistic Effect dependence strain th o No t alsc e tI bu nar onl o f c value eo I yth f eo changed with the irradiation. Figure 9 shows the strain values which characterizes strain effects in the bronze processed conductors as a function of 1018 1019 Neutron fluence (a) Neutron fluence (n/cm2) (b) FIG . Characteristi9 . c strain value functioa s sa fluencef no . 63 3) The ef increased beyond the 4>t of 5xlOl8n/cm2 ancj aid not change by further irradiation. The increment of em is not fully explained by the hardening of the ductile materials in the conductor. This is partly explained by the elastic softening of Nb3Sn with decreasing temperature. The marked increase in Bra. implies that an additional compressive strain in the direction of the axis broughs ha tNb3Se abouth no t filament s durin irradiatioe gth n through some mechanisms which are not known at this time. If it is the case, increase of Em will occur even in the conductor which is irradiated at cryogenic temperature conductoe th case magne e n th th i efusion r s i f rfo to s ,a . The coincidence between £f and Eirrev is interpreted as follows. Since bronze matrix has been hardened and embrittled by irradiation of high SUMMARY Strain characteristics of advanced Nb3Sn conductors depend on the constitution. Internal stabilization result slighn si t increase th n ei strain sensitivity reinforceA . d conducto highes rha r flow stresd san slightlt highebu , £m ry smaller £0,irre slightld van y higher strain sensi- tivity. An internal Sn processed conductor with Ti addition and large scale in situ conductor with good strain characteristics are being fabricated. The Ic of Ti added NbsSn conductors is degraded at lower §t as compared to that of pure NbsSn. The degree of degradation appears to depend on the amoune specieth d additivf to san e elements. 64 The constituents of the conductors are hardened and embrittled by irradiation e pea Th c increase.kI strair fo m £ ns with ACKNOWLEDGEMENTS Sample conductors have been provide many db y manufacturern si Japan and we thank them collectively. The authors are grateful to colleagues in ISIR, Osaka University for their help in Ic measurements, the member HFLSMf so , Tohoku Universit givinr yfo g convenienc 16.5e us , o Tet SM members of RRI, Kyoto University for help in irradiation and measurements. This wor partls i k y supporte r Scientififo d Grany db Ai n ti c Research 60050004 and 01050005, Ministry of Educ. Sei. and Cult., Japan. REFERENCES ] J.W[I . Ekin, Adv. Cryog. Eng. (19840 3 , ) 823. [2] K. Katagiri, et al. Adv. Cryog. Eng., 36 (1990) in press. Fukumot. M Adv, al ] .t [3 o e Cryog. Eng. ,(19840 3 ) 867. [4] K. Katagiri, et al, Adv. Cryog. Eng., 34 (1988) 531. [5] K. Tachikawa et al, Adv. Cryog. Eng., 32 (1986) 947. [6] K. Katagiri, et al, Adv. Cryog. Eng., 36 (1990) in press. [7] K. Katagiri, et al, Proc. MRS Int. Meeting Adv. Maters., Vol. 6, Eds. M. Doyama, MRS Pittsburgh 1989, 31. Katagiri. K Adv , ] al [8 .t ,e Cryog. Eng. ,(19906 3 pressn )i . Gregor. ibidE , al ] .t ye [9 110] J.V. Verhoeve Adv, al .t ne Cryog. Eng. ,(19862 3 ) 985. [IIAdv , Iken. ]al Y .t oe Cryog. Eng. ,(19906 3 pressn )i . [12 Katagir. ] K Jpn , al .Appl. t J ie . Phys., Suppl. 26-3 (1987) 1521. [13] D.O. Welch, Adv. Cryog. Eng., 30 (1984) 671. [14 IEESekin, . ]H al E t Trans.,MAG-1e 9 (1983) 1429. [15] H.W. Weber, Adv. Cryog. Eng., 32 (1986) 853. [16] F. Weiss et al, IEEE Trans., MAG-23 (1987) 976. [17] M.W. Guinan, P.A. Hahn and T. Okada, Summary Kept. RTNS-II Collab. Res.1988, UCID 21298. [18 IEE , Okad. ]al T E t Trans.ae , MAG-23 (1987) 972. Next page(s) left 5 blan6 k HIGH FIELD Jc CHARACTERISTICS OF IN-SITU PROCESSED VGa AND CCE PROCESSED NbaAl WIRES 3 K. NOTO* . SATO**T , . WATANABE**K , , . NOGUCHI***T . IWABUCHI*A , . YAMAZAKI*K , , . SAITO**S . DŒDA+S , . IKEDA*K , * *Facult f Engineeringyo , Iwate University, Morioka **Institut r Materialefo s Research, Tohoku University, Sendai ***Vacuum Metallurgical Company Ltd, Sambu, Chiba +Miyagi National College of Technology, Natori Japan Abstract Critical current densities in in-situ processed V3Ga wires prepared by a CaO crucible melting and CCE (dad-chip extrusion) processed Nb3Al wires were studied up hybrie th n di magneT 2 t2 o t (HM-2 HFLSMt )a . Tohoku University Cu-20îA . V allos wa y melte usiny b d n inductioga nO crucible Ca furnac a d d theean an , n wire drawn dowo t n the final size of -0.3 mm0. Samples were heat treated (450-550°C)x(2-6.5days) for the reaction of V3Ga compound after a Ga plating on the wire surface. The sample heat 2 treate day2 550°t a dr s Cfo showe bese c th dvalueJ t s higher o tha16.A/cm* t 10 np 2 u expectes i t I T. d that highe c valueJ r t highesa r field wil e attaineb l d aftee th r optimization of alloy composition, heat treatment condition and so on. The best sample 1 of Nb3Al wire prepared by the newly developed CCE process showed Jc higher than 10" . K 2 4. o 18.A/cm t t a p 8 u 2T INTRODUCTION High Tc oxide superconductors" have field superconductors, which have been discoverd in 1986, 75 years after recently shown appreciable progress. the discovery of superconductivity. The In this paper r recenou , t workn so impact of this discovery was very large. the high field characteristics of in-situ However seemt i ,neeo t s a ds likel u r fo y processed V3Ga wires prepared in an little longer time than first expected induction furnace and a CaO crucible2' e practicath r fo l applicatio f thesno e and a new process named as clad chip new materials. Therefore, it is very extrusion (CCE) processe th r 3'fo important to continue the effort to preparation of a superconducting Nb3Al improve or develop conventional high wire will be reported. 67 1. VtCt < 15 20 25 30 B (Tesla) Fig. 1 In-situ PROCESSED V3Ga WIRES In-situ process* is very inter- 5) compound. Figure 2 shows the Jc charac- esting frostandpointe th m f simplso e teristic f sampleso s prepare severay b d l production proces d smalsan l strain heat treatment conditions. The sample sensitivity of the sample comparing with heat treate day2 550°t a dr s Cfo showe d the standard bronze process. Especially, the best characteristics, where Jc is 6 2 in-situ processed V3 Gas ver'ha y higc J h higher than 10" A/cm up to 16.2 T. These at high field region as can be seen in characteristics will be improved further Fig sho 2 .e curvee d th wITh 71 an . 1 s a G afte e optimizatiod th an r V e th n i n c meltear n Jc'i ds in-situ processed V3Ga composition d heasan t treatment wires. conditions. We trie moula d d e procesth r sfo more simplification f thiso s procesy sb E PROCESSECC D Nb3Al WIRES O cruciblusin Ca n inductio a ga d ean n furnace. Cu-25 at XV mould was wire There are many binary systems which draw finane th dow lo nt siz abouf eo 3 t0. have superior superconducting behavior 7 mm0. The wire was heat treated at (450- than Nb3Sn in many A15 compounds '. 550°C) for (2-6.5 days) after the Ga Amon mose g th themt f o ,e Nbon 3s Ai l formatioe platin th V5 3r GA1 a fo gf no interesting system which has higher BC2 68 1O6 O Cl»rf tlilfl mtlltoJ (n,«J. „ ,| I iMot.iiM r/ T . n-) ! «(il,l,„„,i,rio j |,,«K.1IM A»I - I.SS . x I0< (W.U.,.1,,,I,l.| lul- 4 ux nmhoj n-A«<«l.i .<«u Ufa SUe 10mm° O.zcnm' Culling Filling Clod Rolling Bundling f J -« — J 1 ^ """""" ~] 1 | X. — t n T-4.ll! than NbsSn. Ther mane ear y effortr sfo developing practical Nb3Al superconduct- ing wires e processeTh . s tried tilw no l 10 15 20 are liquid quench8>, powder metallurgy9*, 8 (Tesla) modified jelly role10>,Nb-tube method11', Fig. 4 and so on. Recently coworkers hi . SaitS , d oan s have develope processw ne da 12' named clad chip extrusion (CCE) methor fo d preparing a superconducting NbsAl wire. the temperature of the first heat treat- e procesTh f thiso s metho schematis i d - men loweres i t d from 1100°C, whica s i h cally showl claddeA Fign i n. b .3 M d little too high from the industrial point plates wert intecu o square chips, filled of view, to 950°C after the optimization into a billet, and then extruded, and for the samples of the Mg added CCE finalle th e wir s drawo th t ywa e p nu processed Nb3Al wires. final areal reduction rati f 10o 6~108 without intermediate annealing. The final wire samples with ~0-3 mm0 were CONCLUSION two-stage heat treated and then measured. Figure 4 is the comparison of Jc- In conclusion, characteristies in CCE processed super- 1morA . e simple procese in-sitth r sfo u conducting wire with samples prepared by V3Ga wirs attemptewa e y usinb d n ga other several processes. Jc is higher inductio O ncrucible Ca furnac a d .ean than 10* A/cm2 up to 18.8 T at 4.2 K for bese Th t sample showe c higheJ d r than the CCE processed Nb3Al wire. o 16.* A/cmt 10 whic, p 2T u 2 h wile b l The effect of Mg addition into the improved further more aftee th r cladded Al plates was also studied in a composiG optimizatiod an -V e th n i n this process. It turned out that Jc e heath tiont d treatmensan t increase sa littl e and, more importantly, conditions. 69 2. Jc in a Nb3Al wire prepared by a newly Muto: Proc. Honda Int. Sympo Hign .o h developed CCE process exceeds 10* Magnetic Field (1989, Sendaie b o )t A/cmz up to 18.8 T at 4.2 K. Moreover, published. it turne t thae additioou d th t f no . SaitoS ) 3 . IkedaK , Noto. K , . Ikeda,S , third elemen g increasetM a littl e J d e Nagata. A d . Matan J : . Engin. (1989) and, more importantly, lowered the to be published. temperatur firse th tf e o heat treat- Tsuei. C . C : ) Scienc4 0 (197318 e . )57 ment from 1100° twoe 950°o Ct th -n i C . Sue J . Verhoeven D . ,J . J ) . 5 D . E , stage heat treatment. Gibson, J. E. Ostenson and D. K. Finnemore: Ac ta Metall. 28 (1981) ACKNOWLEDGMENTS 1791. The authors would like to thank 6) H. Kumakura, K. Togano and K. Prof. e IwatIkedth ef o aUniversit r fo y Tachikawa: Adv. Cryog. Engin. Mat8 2 . valuable assistances. (1982) 515. They also e technicianthanth o t k d san . NotoK . Watanab) K ,7 . Muto Y d : ean Sei. staffHige th h f sFielo d Laboratorr yfo Rep. RIT3 (19863 A U ) 393. Superconducting Materials (HFLSMe th d )an 8) K. Lo, J. Bevk and D. Turnbull: J. Cryogenic Center of Tohoku University for Appl. Phys 8 (1977.4 ) 2597. many assistances. This work was partl . ThiemeH y. L . Pourrahimi S ,. C ) 9 . B . B , supported by the Grant-in-Aid for Special Schwart . FonerS d zan : IEEE Trans. Research on Fusion (01050005) from the Magn. MAG-21 (1985) 756. Ministr f Educationo y , Sciencd ean . ToganTachikawa. K K ) d 10 oan : Adv. Culture, Japan. Cryog. Engin. Mat. 34 (1988) 451. 11) T. Takeuchi, Y. lijima, M. Kosuge, K. REFERENCES Inoue, K. Watanabe and K. Noto: Appl. 1) J. G. Bednorz and K. A. Müller: Z. Phys. Lett. 53 (1988) 2444. Phys. B 64 (1986) 189. 12) S. Saito, K. Ikeda, S. Ikeda, A. 2) K. Noto, K. Watanabe, S. Murase, A. NagatNoto. K d : aan Proc . llth. Int. Sato . IkedaK , . SaitoS , . YamazakiK , , Conf Magnen o . t Technology (1989, . SaitoT . SatoT , . Noguch. ,T Y d an i Tsukuba) to be published. 70 Nb3Al CONDUCTOR DEVELOPMENT AT ENE EM/LMID AAN , ITALY R. BRUZZESE, N. SACCHETTI, M. SPADONI Centre Ricerche Energia Frascati, Associazione EURATOM-ENEA sulla Fusione, Frascati, Rome . BARANIG . DONATIG , . CERES , S ARA Centra Ricerche EM/LMI, Lucca Italy Abstract A method (th calleo es d "jelly-roll method" bees ha ) n develope pase th t n dyeari s which allow obtaio st n multifilamentary NbaAl wires. A short description of this process as well as a review of the experimental results so far obtained on short sample presentee sar discussedd dan maie .Th n aspec thif to s wor thas ki superconductota r wit hvera y high critical current density (1.8-« producee - b 1.9xl09A/m n ca d) whicHT t 2a h makes this material very interestin r practicagfo l purposes. However, problems relate productioe th o dt significanf no t length wirf so e remain still open. Possible way overcomo st e them are presently under consideration and will be briefly discussed. 1. INTRODUCTION As it is well known, the A-15 materials, with the general stoichiometric formula A3B (A = Nb, V; B = Sn, , etc.Ge , ) constitutGa , AlSi , n interestina e g clasf o s superconducting compounds with rather exceptional properties 18°K,H~ c :T c2~20 300kOean0+ d jc(10T,T,2K) > 105A/cm2. Unfortunately their preparation is usually e higth hdifficulo t reactio e du t n temperatures (often exceedin o theit gd r 100an inheren ) 0°C t brittleness (similar to glass). This problem has been circumvented in many cases by adopting peculiar fabrication procedures, like e.g. chemical vacpour deposition (NbsSn), solid state diffusion from bronze into niobium or vanadium (NbsSn, VaGa), liquid stale diffusion, etc. different from the usual metallurgical methods. In this way a variety of conductors (ribbons, multifilamentary cables, etc. bees )ha n produced and is commercially available. In this paper we present a short description of the fabrication method of a NbsAI based multifilamentary cable and results about its physical properties e choose reasonw Th .y e wh sthi s particular compound among others A-15 are that its critical temperatur estimates it ed (18,an ) d7°K critical fiel aboud( t mak) promisina kG t ei 0 32 g candidat hign ei h field large volume magnet realizations. 2. FABRICATION METHOD Several years ago a method for fabricating Fig. 1 - Sketch of the fabrication process of multifilamentary wires based on the A-15 intermetallic multifilamentary wires compound NbaAl was developed successfully (jelly roll method) although the details of this technique have been certain numbe f theso r e a suitablhexagon y b d ean s already published (1,2 usefus i t ]i l her mentioo et n briefly mechanical processing a final multifilamentary wire is basis it c principle illustrates A . Fign di "filamenta .1 s "i obtained. Alternatively a single extrusion billet with a made by wrapping two superimposed foils of niobium and suitable number of drilled holes can be used. At this point aluminum around a copper cylinder. This structure is then hean a t treatmen s performei t t temperaturea d s ranging inserted in a hole drilled in a copper containe rNbaAe th an t orden di lge comounthC o ° rt e0 85 y o dt b C " fro0 m75 containe s broughi r o hexagonat t l shape. Combinina g the direct reaction of the base elements. 71 Two advantages of this fabrication method are quite evident. First there is no need of intermediate annealings - 5 during the mechanical processing or homogeneization heat treatments l elementA d e an rathe .ar sb SeconN r e th d lc(at0.1; 3. EXPERIMENTAL RESULTS a) Critical Current. A critical parameter in this method e finath ls i thicknes e aluminuth f o t s m foil before s reactiorols beeit ha ed n an nrathe r extensively studied (3J: nearly stoichiometric NbsAi with good superconducting propertie formes si d provided that belos i wcertaia otherwis) um n valu2 0. e ~ espuriou( s phases with poor properties are obtained. Similar effects have been observed by other researchers which Commercial have use a ddifferen t techniqu o product e e NbaAl wires (4,5,6). In Figur have w e2 reporte c (noj e n dth coppert s )v dependence for a single filament wire while in Fig. 3 56 7 8 9 10 11 12 th j e(no n copper a functioe applies th a ) f o nd c B(T) magnetic field is shown. The strong influence of the Al Fig. 3 - Critical current density Jc (non-copper) vs thicknes transpore th n so t propertie quits si e evident. magnetic fielmonofilamentara r dfo y jell-roll conductor ) b Mechanical properties serie.A s of bending tests sha having an Al foil thickness t = 35 nm. A range of Commercial NbaSn dat alse aar o reported been performed by measuring the critical current density of short samples at a fixed field as a function of the bending strain defined as the radio d/D between the wire diameter d and the bending diameter D. Sample differing in filament size, external diameter and heat treatment have been studied and the results showe ar Fig n t appearni I . .4 s tha degradatioo tn f no Ic is evident up to d/D ~0.8% and that after this value the bending strain damages very rapidly the superconducting transport properties. This behaviour suggests a better tolerance of the 1.0 uniaxial strain effect with respect to the analogous —— Ö"3~t"öi9 compound NbsSn. In fact it has been shown, in the case of (normalized) \" NbsSn [7], that the bending strain curve can be interpreted i in termuniaxiae th f so l strain dependenc takinI f eo g into 0.8 ~ i c i accoununiforn no e mth t distributio stressef no s acrose sth _ I section of the wire (from compression to tension passing i i throug neutrae hth l axis). Usin experimentae gth l curvf eo 06 i \ P o \ 0.4 \ \ \ \ coppen no c r 0.2 - (Am/0 s 10 t i 0 05 1.0 1 10' 200 400 600 800 1000 1200 1400 t(nm) Fig. 4 - Critical current as a function of the bending strain. (non-copperK 2 4. d a an s a )T 4 6. t a c - PloJ Figf 2 o t. Different symbols refer to samples with different filament function of the Al foil thickness t. Open triangles refer to size, external diameter, and heat treatment. These previous dat Reff ao 4 . parametgers have no induce on the observed behaviour 72 strails ev n for NbsSn (unfortunatel analogoue yth s curvs ei followinw no contene ar g e essentiallW t wayo ytw s first not available for the moment in our case) it has been try to use niobium foils with oxygen content as low as calculated a degradation of about 15% in Ic at 0 8% of possible, tests are in progress with foils containing about bending strain 125 p p m of oxygen and other tests are foreseen in the next Therefore the suggestion is a better strain behaviour future with foils having only about 60 p p m (very difficult of NbaAl and this is in accordance with results reported by to find commercially available) second use foils of niobium Foner et al [5] in the case of NbsAl wires made by the wit titaniue f additioo hth % 1 f mno which a shoul s a t dac powdermetallurgy method sorf o internat l gette r oxygefo r n improvine th g workability 4 DISCUSSION Fro resulte mth s illustrate precedine th n di g sectiot ni REFERENCES appears that NbaAl is a very interesting material as far as [1] S Ceresara, M V Ricci, G Sacerdoti IEEE Trans on practical applications are concerned Of course, as it should Magnetics.MAG 11,2(1975)263265 e welb l know wiry b n e producers e reallb o t y, usablea , [2] R Bruzzese, S Ceresara, G Pasotti, M V Ricci, N superconducting material has to be fabricated in a wire Sacchetti, M Spadoni VU Symp on Engineering form of significant length not less than 1 km at least This Problem Fusiof o s n Research, Knoxville, Sept 1977, maie isth n proble presentle ar e mw y confronted Higc hJ 2,1255 1258 obtainevaluee b n verr sca dfo y thid smalan s t mean f o l sa [3] B Annaratone, R Bruzzese, S Ceresara, V Pencoli large amoun mechanicaf o t l performewore b o e kt th n do Ridolfini G Pitto, N Sacchett, i IEEE Trann o s composites durin e wirgth e fabrication Breakage th f eo Magnetics.MAG 17,1(1981)1000 1001 filaments is a typical drawback of these processes which is (4] J M Larson, T S Luhman, H F Mernck connected to the ability of materials to withstand heavy Manufacture of Superconducting Materials mechanical r cas worke workabilitou e th n I s f niobiuyo m MeyerhoffEditor (1976) 155 163 cruciae foilth s si l point which set maximue slimia th n to m [5] R Akihama, RJ Murphy, S Foner IEEE Trans on lengths of wire which can be produced and this problem Magnetics MAG 17(1981)274 277 becomes even more severe when we go from mono to H Thieme L [6C ] S Pourrahimi, B Schwartz B , S , multifilamentary wires Our present efforts in trying to Foner Appl Phys Lett 44(1984)260 262 overcome these difficulties start froe facmth t thae th t m FilamentarEk W J [75 Superconductor1 ) A y s workability of niobium is strongly decreased by its oxygen Suenag Clard aan k Editors, Plenum Press Yorw k,Ne (1980)183 720 Next page(s) lef3 t 7 blank DEVELOPMENT OF Nb3Al MULTIFILAMENTARY WIRE BY THE JELLY-ROLL PROCESS . OHMATSUK . OKUG , . TAKEI,H . NAGATM , A Sumitomo Electric Industries, Ltd, Osaka T. ANDO, M. NISHI, Y. TAKAHASHI, S. SHIMAMOTO Naka Fusion Research Establishment, Japan Atomic Energy Research Institute, Naka-machi, Naka-gun, Ibaraki-ken, Japan Abstract NbaAl raultifilamentary wire for a fusion magnet has been developed using e jelly-rolth l process e wir.Th e includin 1 Nb9 g 3Al filaments wit u matriC h x was designed and fabricated. The influence of heat treatment temperature on Nb-Al microstructure C losseA s d measureswa an , detailn Jc i d , ,Tc saralA . l coil was made and was tested in the 13T background magnetic field. The coil successfully charged up to 100% of the short sample critical current, when Jc was 183 A/mm2 at 13.34T, 4.2K. INTRODUCTION (JAERI), 12 T-30 KA Cu-stabilized multi- NbAl wirattractivs i e r smalfo e l filamentary superconducting cable with degradation3 s of critical current density Jcè 700 A/mm2 at 12T, 4.2K for non-Cu (Jc) under the stress and high upper a minimu aren i a m lengt f 600s o hi m critical field (Hca) over 20T. Further- required.[4] more, Jc in a magnetic field over 10T is In order to meet these specifica- reporte o exceet d e valur Nbth dfo 3 Sne . tions, instead of Nb3Sn wire, we have starte o t develod p NbAl multi- r thesFo e reasons, this wir expectes i e d 3 e applicabltb o a fusio o t e n research filamentary wire with a Cu matrix using magnet. the jelly-roll process. This process has In recent years, several processeo st been developee ENEA/LMth y b dI group produc a eNb 3Al multifilamentary wire and high Je value exceeding 10A/mm at 2 have been developed [1,2,3 however], , s reported.[211wa T ] This process i s 3 a suitabl beet ye ne t procesno s ha s considered to be the most suitable if constructed for a fusion magnet because the poor workability can be improved. r firsOu o tt develo goas i lp NbAl of the difficulty in producing a wire 3 with sufficient length, and in making a wire having both good workability and Cu-stabilized multifilamentary wire. high Jc. e th Folargr e superconducting This paper reports the preliminary toroidal field coil in the fusion result f wiro s e product ions.observât ions experimental reactor (FER), which has of microstructure and superconducting been researched and developed by the properties, and results of test coil Japan Atomic Energy Research Institute e Nbwounth 3A Iy b dmultifilamentar y wire. 75 EXPERIMENTAL PROCEDURE Tabl Specification. 1 e e wirth ef o s e jelly-rolTh l proces s appliewa s d fire diameter 1.09mm o fabricatt e single filament rods, which Numbe f filamento r 1 9 s were prepared by rolling a 150#m Nb Filament diameter sheet and a 40/(m Al sheet together on a outer 65/ira u core C d insertinan , g u pipeintC a o . Inner 27/im The rods were reduced to a diameter of 6 2. Cu ratio Imrn^ d thean , a raultifilamentarn y wire Twist pitch 50mm s fabricatewa y b stackind 1 singl9 g e Wire length 200m filament rods, inserting into a Cu pipe, and drawing. Table 1 shows the specification e wird Figurth an ef o s1 e show scrosa s sectiowiree e th Th f .o n wire showed good workabilits wa d an y reduced to a diameter of 1.09mm0 in 200m length. The areal reduction ratio was e thicknesseX 10d th 5 an 4d , b an N f so Al were 0.7/fm and 0.2/ua, respectively. The effects of heat treatment from 700"C to 900Ç on the Nb-Al microstructure were observe atomie y SEMb dTh . c ratios werl A ed analyzean b oN f y EDXb d . X-ray diffractions were also observed to confirm the crystal structure of Nb-Al strand phases. Critical temperature (To) and Jc were measured by the conventional four probe method. AC losses were measured for each sample by the electrical method when an external magnetic field (Bra s applie)wa d at±0.5 ±1Td an T, 0.2Hz. A small test coil was made to demonstrate the practical application of long wire. Nb3Al wire covered with quartz fiber insulation in 110m length was wound on an SUS bobbin. Then it was impregnated epoxy a vacuuresi n i nm environment after NbaAl forminfor g filament 800 2houÇ- r heat treatment. Table 2 shows the specifications of the test d coiFiguran l 2 eshow a s Figure 1. Cross section of the photograph of the coil. The coil was »ire 76 Table 2. Specifications of the coil 700U Inner diameter 36min Outer diameter 59mm Coil length 11lira Number of turns 633 Number of layers 8 Self inductance 5.65niH ffire length 110m Figur . Photograp2 e e coith lf o h charge T p backgrounu 13 unded a r d magnetic field. RESULTS Hicrostructure Figur N photograph 3 eSE show e sth s fracturee oth f d surfacsamplee th f o es heat treated from 700 o t r 90QTÇfo C s foun wa 2 hours l t A dI .e tha th t diffuse drasticallb sN inte e th oth s a y temperature increases and pores are generated at the site of Al. Unreacted H FigurphotographSE . 3 ee th f so Nb, corresponding to the white area of fractured surfac Nb-Af eo l the photographs, exists under the heat filament 77 treatmen 750Ïf o t . Grain growts wa h O ZH clearly heae observeth t treatmens a d t B- 8T A 4H 1000 temperature increases above 800t!. EDX o 8H analysis revealed e averagthath t e atomic ratio of Nb to Al is 3:1, however, observation of the raicrostruc- e Nb-Ath tur f lo e surface usin e higth gh resolutio N showSE n s that several Nb-Al phases exist. X-ray diffraction also show n unreactea s phasb N dd Al-ric ean h phase Nba s wel3 a As la l l phasal n i e samples. » KX O 9O O 8O O TO 0 60 Heat tmtmeni teiperature (°C) Superconducting properties Figure 4. Heat treatnent teœperature Figure 4 shows the heat treatment dependence of Jc temperatur, 8T e t dependenca c J f o e 20 increasee J temperature 4.th . 2K s sa e 2H increases and reaches the maximum value at 800t!~850t, the decreasec J n s rapidly. The maximu, 0 A/mm8T 85 m t a valu 2 s wa e 15 0 850A/mmr 4.229 fo 2hours - tKf 2o c J . A/mm3 14 8 d atan 12T T 0 A/mm13 20 , t a 2 at 14T was obtained for the same heat treatment. 10 Figur 5 showee heath st treatment 600 700 800 9OO WOO temperature dependence of Tc . Tc Heat treatment temperature (°C) increasetemperature th s a s e increases Figur . Hea5 e t treatment teaperature to 800t, reaching 16K d remainan , s that dependence of Tc valu t SOOta e . Figur 6 eC show A resule th f so t IT : losses. The value at Bm = 0.5T is almost constant for the heat treatment from 700Ç to SOOt!, however, the dependence of AC losses at Bm= IT shows a different O.ST phenomenon, i.e .a pea k appears around 10' O as in the case of Jc. Test coil Figure 7 shows the load line of the 10' I coil together e witshorth h t sample O 9O 0 85 0 80 0 7075 O critical current. The coil was charged Heat treatment temperature (°CX2H) T backgroun1S unde a r d magnetic field Figur . Bea8 e t treatnent teaperature wit a currenh t 2 A/minspee f A o .d dependenc losseC A f so e 78 complete superconducting state wito n h TOO eoox» ZH voltages from the coil is achieved 6OO - untill the coil current is 40A. The maximum steady state which could keep 500 2OOO the continuous current charge was 47.3A, — 4OO althoug e voltageth h f o 600/is Y were 300 observed t thaA . t time critical current IOC» density of the wire is 183 A/mm2 for ZOO non-C e umaximu th area d ,man magnetic e windinfielth s f i 13.34To dg . From this resultrecognizes wa t i , d thae th t 4 E 8 10 12 14 I« coil successfully charged up to 100% Magnetic field (T) e shorth of t sample critical current. Figur Coi. e7 l load linshord an e t Table 3 summarizes these results. saiple critical current Table 3. Results of the test coil Complete SC Steady State State Central magnetic field (back ground =13T) 13.26T 13.31T Maximum magnetic field of the windings 13.29T 13.34T Coil currents 40A 47 .3A Average current densities of the windings 19.8A/mm2 23.5A/mm2 Critical current densities of the windings 155 A/mm2 183 A/mm2 DISCUSSION Jc but not the Tc. Therefore, grain boundaries are thought to be one of the The value of Jc of our sample is main flux pinning centers in the Nb3Al relatively low compared with that from s wel a s a othel r A-15 type superconduc- other processes.The reaso consideres i n d ting materials. to be that the thickness of Nb and Al is From the above discussions, not fully reduced in our samples, improvemen expectes i c J o resul t df o t t therefore, the complete Nb Al phase was 3 from increasin e fine-graineth g d Nb3Al not produced as observed by SEM, EDX and phase. In order to achieve this X-ray diffraction. w tryincondition no o reduc t e g ar ee w , Moreover c decreaseJ , s rapidly the Nb and Al thicknesses, and research- abov e c remain e maximuT 850t th Çbu t sa m ing the best heat treatment conditions. value. This tendency is much more clear Finally r preliminarou , y coil test for the long heat treatment. This result indicates e Nbthath 3Atl wire produced suggests thagraie th t n growth causey b d by the jelly-roll process is applicable e lonth g heat treatment decreasee th s for high field magnets over 13T. 79 CONCLUSIONS REFERENCES ] C.L.H.Thieme[1 , S.Pourrahimi, B.B. NbAl multifilamentary wire with 3 Schwartz.and S.Foner, "Improved high u C matri s fabricatewa x d e usinth g field performance of Nb-Al powder jelly-roll process. The wire showed metallurgy processed superconducting wires", Appl.Phys.Lett.44,(19840 )28 good workability and was reduced to [2] R.Bruzzese, N.Sacchetti, M.Spadoni, a diameter of 1.09tnm in 200 m length. G.Barani, G.Donatij and S.Ceresara, Microstructur f Nb-Ad o efundamenta an l l "Improved critical current densities in Nb3Al based conductors", IEEE C losseA e d ar san , propertieJe , Te f o s Transaction n Magnetics,vol.23,NOso . clarified. Maximum Je was 290A/mm2 at 2,(19873 )65 12T, 4.2K for non-Cu area. A small test ] T.Takeuchi[3 , Y.Iljiraa, M.Kosuge, T.Kuroda, M.Yuyama and K.Inoue, s coiwa charge lp T u unde13 d r "Effect f o additivs e elementn o s background magnetic field reachean d d continuous ultra-fine NbF M 3Al 100% of the Jc of the short sample. superconductor", IEEE Transactions on Magnetics,vol.25,NO.2,(1989) 2068 The success of wire fabrication and [4] K.Yoshida, M.Nishi, Y.Takahashi, the test coil demonstrated that the H.Tsuji,K.Koizuai,K.Okuno,and T.Ando NbaAl raultifilamentary wire with Cu "Design of the proto-type conductors e fusiofoth r n experimental reactor" matrix manufactured by the jelly-roll JEEE Transaction n Magneticsso , vol. proces s applicabli s r higfo e h field 25,NO.2,(1989) 1488 magnets. 80 Nb3AI MULTIFILAMENTARY SUPERCONDUCTOR T. TAKEUCHI, M. KOSUGE, Y. IDIMA, T. KIYOSHI, K. INOUE, T. KURODA, K. ITOH, H. WADA National Research Institut r Metalsefo , Tsukuba City, Ibaraki, Japan Abstract Nb-Al is known to be one of the most promising A15 superconductors for practical use due to their high T and high H „. We have been successful in fabricating a continuous ultrafine Nb~Al multifilamentary conductor by an improved composite-diffusion process using various Al-based alloy cores and pure Nb matrix. The highest T of 17.2 K, H (4.2K)„of 24.4 T are obtained at least hitherto. J (4.iZK) values of Î.5 x 10 A/m at 10 T and comparablT 0 2 t a e commercia thoso th t em f A/ o e 0 1 l x Nb,,S 5 n multifilamentary wires and the excellent tolerance to large mechanical strains and magnetic fluctuations indicate that the present Nb-Al conductor s promisini n alternativa s a g Nb„So t e n multifilamentary conductor. INTRODUCTION n applicatioa r Fo f superconductino n g magnet n fusioi s n reactorsl al , the magnet materials hav meeo t e numbea t f fairlo r y stringent requirements. For the superconducting materials the desired properties are (1) large J in high fields, (2) a high tolerance to mechanical strain and radiation, and (3) low AC losses. Recently, extremely high J and excellent strain tolerance were demonstrate n Nb»Ai d l fabricatey b d jelly-roll process[l d powde]an r metallurgical process[2], indicating that Nb»Al is promising as a candidate material for the fusion reactor magnets. However, these processes seem to encounter difficulties in the development oa conductof r with multifilamentary structure sf greato whic e ar h advantag o improvint e e stabilit th gmagnee th f to y against various electromagnetic disturbances. An attempt was made to fabricate the multifilamentary Nb„Al by the composite process using pure Al as a core material, which ende n failuri d e becausworkabilitied ba f o e e th f o s composite[3]. This study was carried out to develop the multifilamentary Nb_Al conductor y improvinb s e workabilitth g e compositeth f o y . EXPERIMENTAL PROCEDURES n general I e compositth r ,fo e process, good muc hardnesn i he th f so cord matrian e vers i x y important d thi an ,s alsi scase th oe with Nb/Al composites s showA .n Fig i n .ther, a larg1 s i ee differenc n hardnesi e s between the pure Al core and the Nb matrix, which results in a breaking of the composit reductioe th t a e n rati . f oveHowever o 0 1 r , alloyinl A g with Mg hardens the Al core preferably, diminishes the difference in hardness, and improves the workability of the composite[4,5]. This improvemen workabilite th n i t s alswa y o achieve y alloyinb d witl A g h Cu(- Ge), Ag(-Ge d Zn[6,7])an e abovs encaseTh wa .l allo eA d ro dy b intN a o tube(14/7 mm$) e resultin.Th g single-core Nb/Al composit s cold-drawwa e n 81 200 O S 0 2 0 1 5 2 Reduction Ratio (Sa/S) Fig 1 .Vicker s hardness versus area reduction ratio curve r pur, fo seNb Al-(3,5,7,10)at.%Md an pur, eAl g alloy. by cassette-roller dies into a wire and cut into short pieces. The 121 short, single-core wires were bundled in a Nb tube(20/14 mm0), and the resulting 121-core Nb/Al composite was cold-drawn into a wire and cut into short pieces, again. This procedures were repeate o timetw d s mored ,an finall million(121xl21xl21)-cor8 1. y e Nb/Al composite with continuous ultrafine Al alloy cores was fabricated. No intermediate annealing is necessary in this procedure, which means this process is economical. After heat treatments, superconducting properties were measureds wa J . calculated by dividing a critical current by the cross-sectional area of the 1.8 million Nb/Al single-core composites(i.e., excluding All Nb used b r tubes(20/1bundling)N fo e th f I . 4 replacee b mmjè n ca ) d wit u tubesC h , th J e define e presenth n i dt stud s expectei y e nearlb o t dy e equath o t l so-called non-Cu overal Furthermore. J l , replacin tubeb N e s th gwit e th h resistive materials make presene th s t Nb,Al C losinterestinA sw lo a s a g superconductor, because it can reduce the electric coupling between filaments[8]. Fig. 2 shows cross-sectional view of the (a) unreacted and (b) reacted 1.8 million-core Nb/Al-7at.%Mg composite wires of 0.7 mm in diameter ,e scannin taketh y b n g electron microscopy. Cylindricad an l ultrafine Al cores less than 100 nm in diameter are clearly separated, and its feature remains unchanged drastically after heat treatments. RESULTS AND DISCUSSIONS Superconducting propertie e presenth f o st Nb»Al strongly e depenth n o d Al core diameter, additive elements, heat-treatment conditions, mechanical strains, magnetic fluctuations, etc. Fig. 3 shows the dependence of T and J on the Al core diameter which is a parameter representing the interdiffusion distance between Nb and Al. With decreasin l corA e e th gdiameter J initialld an ,T y increase, reaca h maximum and then decrease. The dependence of £he Al core diameter on the critica lrelative explainee valueth b ) n (1 ca se y stabilitb d Nb-Ae th lf o y compound phase(Nb.,Al, Nb„Al and NbAl~) in relation to the Al supply from thcore(i.e.l eA instabilite ,th exhaustioe th Nb„Al(NbAf o yo t e f no du ) l thl suppleA y froe cor mors th mi e e remarkabl e smalleth t l a ecore)rA d ,an (2) the proximity effect for extremely thin Nb.Al filaments. At least by the single-stage reaction, excellent superconducting properties are 82 Fig. 2 Scanning electron micrographs of transverse cross section for (a) unreacted and (b) reacted 1.8 million-core Nb/Al-7at.%Mg composit n diameteri n ra e7 .0. wir f o e I t St t i 4.2K, 17T 750*C-2Ah 100 1000 Al-core Diameter (nm) Fig. 3 Dependence of (a) T and (b) J (4.2K, 17T) on the Al core diameter for the Nb/Al composite reacteS at 750»C for 24 h. obtained in the following order: Ag-Ge, Ag, Mg, Cu-Ge, Cu and Zn. The highest T of 15.9 K and is obtained for the Nb/Al-1.5at.%Ag-l.5at.%Ge composite at the reaction temperature of 750°C. e maximuth d an m „ Fig pinnin H 4 show ., T sg force densitF y against the Al core diameter. Maximum T and H ~(4.2K) values ari'oDtained e samth et a filamen r tbor e fo sizNb/Al-2at.%C th nf m abouo n e 0 7 e t th d an u Nb/Al-5at.%Mg composites, althoug e core-sizth h e dependence?oH d an T f are week for the Nb/Al-5at.%Mg composite. However, the peak of F xs observe a smalle t a de differenc l corTh rA e e. tn diamete nm n I e0 4 f o r - * r — TT13.X optimum core size between F and T (H „) is also observed in the submicron NbTi multifilamentary conductor[9], which suggests the normal- 83 30 / Fp,max(4.2K) ,520 20 -20 p0Hc2(4.2K) CuM o 10 10 10 100 1000 Al-Core Diameter(nm) AI-allon Figo 4 .Dependenc „(4.2KH y , F corT d f )ean o e diameter for Nb/Al-2at.%8u(cîosed symbols)pant Nb/Al-5at.%Mg(open symbols) composite heat treated at 750°C for 24 h. superconductivity interface acts as a dominant pinning center. Both the difference in the f (normalized pinning force density) versus h(normalized magnetic field) n anisotropcurvea d an sJ observe n i y a tap n i de conductor support the above interface pinningtlO]. Fig show5 . ) heaeffec e (a th st f treatmeno t ) heat(b t timd an e treatment temperatur H, c2 (4.2KT n o e J d c)(4.2K,10Tan compositee th r )fo s with 90 nm0 Al-Mg core. Maximum values at a given reaction temperature in Fig. 5(a plottee )ar d agai Fign i n . 5(b). With increasing reaction temperatures, T and H „ continue to increase gradually at least to 1000°C, whic n approac s a probabîe compositioi h o th ï f e e o hdu yth o t n stoichiometry. However, the optimized J does not decrease drastically eve higt na h temperature contrasn i s NrLsno t . This seem refleco st t that the major pinning centerpresene th graie n th i s t nt Nb^Aboundarno s i l y interface buth t e between Nb matrix(od 3Aan l r core). High field properties can be further improved by the two-stage reaction. Fig. 6 shows the variation in J (4.2K) at 10, 15, 20 T and H (4.2K functioa s )a reactiof o n n temperatur Nb/Al-Ag-Ge th r fo e e composite e besTh .t critical value e show ar sn thi i n s figure, where th e numeral denoted beside each symbol represent optimue sth m reaction timt ea a given reaction temperature. Closed symbol represente th e cas th f so e two-stage reaction where the samples were heat treated again at a low temperatur improvo t h e lon 0 th e5 f 700°g o e r rangCfo e order parameter. e seconTh d reactio higt a t 700°h a n J fields Cd an increase d „ ,an H e th s this improvemen s i mort e remarkabl highet a e r reaction temperatures. However optimue ,th m reaction time becomes shorter with increasine gth reaction temperature. Therefore, the reaction time is required to be shorter tha second1 n , whee reactioth n n temperatur s abovi e e 1200°Co T . do such a short heat treatment, we applied an electric current pulse directly to the sample(0.1 m in length and 0.7 mm in diameter) immersed in liquid nitrogen[11]. 84 16 15 O Nb/Al-10Mj — Nb/AI-lOMg • Nb/AI -5 Mg — Nb/AI-5Mg 4.2 K 20 850 *.2K 9501C , 10T.4.2K 10 10T.4.2K 10 1C>2 103 10* 105 600 700 800 900 1000 Heat Treatment Time (sec) Heat Treatment Temperature(°C) Fig. 5 Effects of (a) heat treatment time and (b) heat treatment temperatur H (4.2K , J (4.2K,10T T d Nb/Aln )e an o e th r - )fo (5,10)at.%Mg composite. Al core diameter is 90 nm. 40s 10s 30m 2s 10" r O Om -—=t==ap5Z7C~*—S.HST 30s 30s Nb/AI -1.5Ag-1.5Ge JE10' 42K alloI A , y cor*:90nm < Open Symbol'.single-stage reaction Closed symbol:two-stage reaction(.70CTC-50h) 20T 10' 23 Nb/AI-1.5Ag-1.5Ge l alloA 4.2Ky, core:90nm '21 0 10090 00 11080 70 0. 0120 0 1300 Reaction Temperature (*C) J) c(4.2K(a 6 Fig,. 10,r5,20T2(4.2KH ) (b )d versu)an s reaction temperature curve r Nb/Al-1.Sat.%Ag-l.5at.ZGfo s e compositese Th . numeral denoted beside each symbol represents the optimum reaction givea tim t a en reaction temperature. 85 sample th r Fig eshow7 fo .reacte e typica K th s2 H curve - 4. d J lt a s single-stage bth y e reactioe two-stagth a an n e reaction(self heatinw lo + g temperature heatin t 700°C)ga . Non-C multifilamentare u th overal f o J l y Nb-Sn are also given in this figure for reference. At 10 T, the highest J s obtaineNb/Al-Me i th m r A/ fo dg 0 composite1 x 5 oe 1. two-stagf Th . e reaction increases T and H _ by 1.5 K and 3 T, respectively, and significantly improves the 5 at high fields. The improved J by the two- stage reaction is higher than that of the commercial Nb~Sn multifilamentary conductor dope t leasda wit , t. hTi T hitherto 7 1 o t p ,u 42K. 10* 10" AI alloy cor «90n: m Single-stag« reaction —O — 5Mg 8OrO3h 1S.6K 20.7 10' — A— 1CU-1G« 8ArOO.5h15.2K20.3T ' — D— 1.5Aa-1.5G885CrC-1h 15.9K 21.3T Two-stag« reaction - 3Ag oment pul«] 16.4K 23-2T» •-*- 1CU-1G« heaflng 17.2K 24.4T --»- 1.SAg-1.5Get* « 23.44 6 J1 T 106 8 10 12 «4 16 18 20 22 24 Fig 7 Typica. J (4.2Kl ) versu s e varioucurveth r fo ss Nb/Al composites reacted by the single-stage reaction and the two-stage reaction. Non-Cu overall J of Nb-Sn and (Nb,Ti) Sn multifilamentary wires e figure showth r ar reference n fo i en . Since the T of the present Nb,Al is lower than that of the (Nb,Ti)-Sn ,reSuctioa e batH hn i n temperature seem enhanco t s e more remarkably the high field properties of the present Nb-Al- . The increments in H due to reducing the He bath temperature from 4..2 K to 1.8 K is 2-2.5 T, which is larger than that of Nb Sn(1.9 T)[12]. Therefore, as shown in magnetie th Fig, 8 . ce Nb_A H curvfielth - lJ f t whico a ede intersect th h s with that of (Nb,Ti)«Sn increases by 2 T when the He bath temperature is reduced from 4.2 K to 1.8 K. Froe practicamth l point s viewimportani t i , o supprest t e th s dégradation of J with mechanical strains. Tensile test revealed the small 1 degradatio J wit f o nh tensilC ee larg straith ed irreversiblan n e tensile strain(about presene 1.3°%th r t)fo Nb„Al conducto . Bendin] 13 r[ g test also indicate e largth s e toleranc mechanicae th o t e l strain[12]. Excellent tolerance to mechanical strain and high J values comparable to (Nb,Ti)„Sn indicate that the present Nb-Al conductor will be very promising for practical high field superconducting cable. 86 109 1.8K (N E 108 NbF D3«M S n o«Nb/AI-10Mg 16.5 K 950°C-0.5nnin 107 /700°C-4day AANb/AI-7Mg 15.6 K 75Q°C-24h 4 2 2 2 0 2 8 1 6 1 4 1 2 1 0 1 8 B ( T) presene th r tfo Nb-AI(corK 8 1. d J Fig_versuan e 8 K . 2 sH curve 4. t a s diameter:°90 nm) and (Nb.TdJ-Sn multifilamentary wires. Finally, magnetization measurements were carried out to estimate the AC losse[8]. Hysteresis losses for the present Nb„Al conductors were relativel yfiele largth dn i erang e from -0. +0.o 5t attribute, 5T e th o t d magnetization by the Nb matrix. If the Nb tubes used for bundling can be replaced with the resistive materials, this hysteresis loss will remarkably be reduced. This replacing seems to be effective in reducing the fabricatio expensive th f no coste e on ,materialss i sinc b N ee Th . magnetization measurement also revealed that the effective superconducting filament diameteryo4 fieldn 5-1d u2- i aroun an m e arounf 2p T ar s d s 8 d d 4 T, showing that there existed some kinds of electric coupling between filaments. It is noted that d iff around 8 T is one of the smallest diamete r5 compound^superconductin A1 amon e th g g cables, indicating thae th t present Nb~Al multifilamentary wire is also promising as a low AC loss conductor. CONCLUSIONS A new process was developed for fabricating Nb,Al multifilamentary superconductors containing more tha millioa n f continuouo n s ultrafine filaments l allon A diameterlesi f u o syyo e thacore1 Us 0. .n s with small amounts of additives including Mg, Ag, Cu, Zn, etc. resulted in successful cold drawing witn are a b tubh N a t a ereductio n rati . Reacteovef o 0 1 r d wires show superior propertie commerciaa o t s l multifilamentary Nb-Sn; (1) a higher J (4.2K)(for example 1.5 x 10 A/m at 10 T and 2.2 x 10 A/m at 17 T), (2) the smaller degradation in J with mechanical strains, and smallee th ) r(3 effective superconducting filament diameter, indicating that the present Nb„Al conductor is very promising for the practical high field superconducting cable. e authorTh muce ar sh indebte e staf th Tohokf o fo t d u Universitr fo y operatin hybriT 3 2 g d magnet. 87 REFERENCES 1. S. Ceresara, M. V. Ricci, N. Sacchetti and G. Sacerdoti, Nb„Al formation at temperatures lower than 1000°C, IEEE Trans. Magnetics, MAG-11:263(1975). 2. L. H. Thieme, S. Pourrahimi, B. B. Scwartz and S. Foner, Nb-Al powder metallurgy processed multifilamentary wire, IEEE Trans. Magnetics, MAG-21:756(1985). . RoseM . R . Eage,W d Improve. an T r3mechanicalln . i J d y fabricated wirel N A bd ribbons an s , IEEE Trans? Magnetics, MAG-11:214(1975). . InoueK . . . Takeuchilijim,4 Y T d an a , Superconducting propertief o s Nb„Al multifilamentary wire, Appl. Phys. Lett., 52:1724(1988). . InoueK 5. . . Takeuchilijim,T Y d an a w superconductin,Ne g Nb„A wirF M l e made by Nb/Al-Mg composite process, Cryogenics, 29:418(1989). 6. T. Takeuchi, Y. lijima and K. Inoue, Fabrication of Nb,,Al multifilamentary wire using ultra-fine Al-based alloy coresS ,MR Int'l Mtg. Adv. Mats., 6:19(1989). . TakeuchiT 7. . lijima,Y Kosuge. ,M . Kuroda,T Yuyam. . Inoue,M K d aan , Effect f additivo s e element n continuouo s s ultra-fine Nb»AF M l superconductor, IEEE Trans. Magnetics, 25:2068(1989). . ItohK . Yuyama ,M . 8 . Kuroda,T . TakeuchiT , . Wada. Kosug,H M d ,an e Magnetizatio f composite-diffusioo n n processed Nb„Al superconductors containing ultrafine filaments publishee b o t , n Proci d . llth Int'l Conf. Magnet Technology, 1989, Eisevier Sei. Publishers LTD. 9. I. Hlasnik, S. Takacs, V. P. Burjak, M. Majoros, J. Krajcik, L. Krempasky, M. Polak, M. Jergel, T. A. Korneeva and I. Ivan, Propertie f superconductino s g NbTi superfine filament composites with diameters ^0.1 jam, Cryogenics, 25:558(1985). 10. T. Takeuchi, Y. lijima, M. Kosuge, K. Inoue, K. Watanabe and K. Noto, Pinning mechanism in a continuous ultrafine Nb_Al multifilamentary superconductor, Appl. Phys. Lett., 53:2444(1988). 11. TakeuchiT . . Kosuge,M . lijima,Y Watanabe. K . Inou, K d an e , Improvements in high high field properties of continuous ultrafine Nb-A F superconductorM l e publisheb o t , n Advi d. Cryo. Engn, Mater., 36:(1990). 12. T. Takeuchi, M. Kosuge, Y. lijima, K. Inoue and K. Watanabe, Developmen f Nb-Ao t l multifilamentary superconductorse b o t , published in Proc. llth Int'l Conf. Magnet Technology, 1989, Eisevier Sei. Publishers Ltd. 13. T. Kuroda, H. Wada, Y. lijima and K. Inoue, Strain effects on superconducting properties in Nb_Al multifilamentary wires, J. Appl. Phys., 65:4445(1989). 88 Next page(s) left blank OXIDE SUPERCONDUCTORS (Sessio) nB POTENTIAL METHOD FABRICATIOE TH R SFO F NO HIGH-T SUPERCONDUCTOR WIRER CABLESD FO SAN S* (Abstract)C K. TACHKAWA Faculty of Engineering, Tokai University, Hiratsuka, Kanagawa K. TOGANO National Research Institute for Metals, Tsukuba City, Ibaraki Japan Studies on practical properties, such as upper critical field and critical-current density of high-Tc oxide superconductors of Y-Ba- Cu-O, Bi-Sr-Ca-Cu-O, TI-Ba-Ca-Cu-Oand systems, reviewed.are theseof Thec2 H materials muchare higher than those conven-of tional metallic superconductors, indicating higha potentialfor practical applications, even when they are used in liquid nitrogen. However, the Hc2 of these materials are highly anisotropic, as can expectedbe from examinationan theirof crystal structure. addi-In tion to this anisotropy, the presence of weak links and a weak pin- ning force in these materials limit the transport Jc to much lower levels than that required for practical applications. Since the oxide superconductors are intrinsically brittle, special techniques must be developed for making tape or wire conductors. In this paper recent developments of fabrication processes that have good potentiality producingfor wire tapeor conductors high-Tof c oxide are then reviewed. Some details are presented for the powder method (which classifiedis organican of into binderuse a and metal sheath)otherfor fabricationand processes using diffusion, solidification, depositionand techniques. Ag-sheathedthe For 2 oxide tapes, Jc values exceeding 10 000 A/cm at 77 K and 0 T have been reported for both Bi and Tl oxide materials. Further develop- ments fabricationin processes overcomethatcan various prob- lems limiting transportthe required.are J c fale l Th tex thif * to s pape publishes rwa Proceeding e th n di IEEEe th (Augus 8 f . so , VolNo , t .198977 ) 1124-1131. Next page(s) left1 blan9 k HIGH-JC SILVER-SHEATHED BiPbSrCaCuO SUPERCONDUCTING WIRES K. SATO, H. MUKAI, T. HIKATA Osaka Research Laboratories, Sumitomo Electric Industries, Ltd, Osaka, Japan Abstract Froa the discovery of high-Tc oxide superconductors in 1986, many kinds of oxides were reported as superconductors. Among them,the Bi compound has nany advantages :n rare-earthigh-Tno , c K h abov 0 elements10 e , mechanically enhanced alignment, and high-Jc with polycrystalline film . In this paper, we report the results of research on silver-sheathed BiPbSrCaCuO superconducting wires. INTRODUCTION From the discovery of high-Tc oxide superconductors in 1986, many kinds of oxides were reporte s superconductorsa d . Among them,th i compoundB e s man1ha ' y advantages: high-Tc above 100K1' 2n 'rare-eart,no h elements, mechanically enhanced alignment, and high-Jc with polycrystalline film3). In this paper, we repor e resultth t f researco s n silver-sheatheo h d BiPbSrCaCuO superconducting wires.4~7) EXPERIMENTAL Appropriate amounts of BiPbSrCaCuO powder were calcined, sintered and grounded.Th et intpowdepu o s silvewa r e rcomposit th tube d an s swaged wa e , draw d madan n e into flat wires wit a thicknesh s ranging from 0.5m o O.lmmt m . o stetw A p sintering metho s c adoptewerJ wa dd e an o increast dc T . Jc e determined using the dc four probe method. Jc was determined with the criterion of l^V/cm, which means 2X10~#U «cm in resistivity for high-Jc wires. The magnetic fiel 5 Teslad2. e dependencstructureo s measuret Th .wa p u c J t a f do es 4 f BiPbSrCaCuo O were investigated through XRD d TEM,an SEM .X ,ED RESULT D DISCUSSIOAN S N Fig. 1 curve1p~ show f silver-sheathee so th s d wires with different cation ratios. Highe a madC r d econtent an highe i B f r o s zero resistance temperatures. Fig.2 shows the relationship of Jc to the thickness of the wires. Higher-Jc properties were obtaine s thincknesa d s decreased e magnetiTh . c field dependence of Jc is shown in Figs.3 and 4 for the wires with various Jc levels in a zero magnetic field. It is clearly shown in Fig.5 that the increasing of Jc in a zero magnetic field improves the Jc-B property, especially for the field direction parallel to the a-b plane. 93 Fig,1 silver-sheathef o c T d BiPbSrCaCuO wires. TEMPERATURE (K) 15000- 10000 - Fig.2 f silver-sheatheJo c d BiPbSrCaCuO »ires. - 5000- 0 0. I 0. 2 0. 3 0. 4 THICKNESS (mm) Fig.3 Magnetic field dependenc. Jc f eo 0.01 O.I I 3 Magnetic field (T) 94 77.3K -I Jg,4QOA/çm? 0.1 u V/cm • Fig.4 Magnetic field dependence of Je. O.O l 0. l Mognetic field (T) I04 2T Fig.5 Jc-B propertie silver-sheathef so d 77.3K BJPbSrCaCuO wires. I03 104 10s 2 (8=0c J , A/cm ) m 5m . 0 0.5mra FILAMEN6 T3 NUMBER FILAMEN2 T 76 NUMBER WIRE SIZE 0. 29mm'4. 9ran x WIRE SIZE 0. SSiwn'xS. 7mm w AVERAGE CROSS 0.054mm2 AVERAGE 82x10- . 2 CROSS4 mmz w SECTION OF FILAMENT SECTIO F FILAMENO N T Fig.6 Fig.7 38-core silver-sheathed BiPbSrCaCuO fires. 762-core silver-sheathed BiPbSrCaCuO wires. 95 Table 1 Suaœary of superconducting properties of silver-sheathed BiPbSrCaCuO *ires. NUMBER OF Ic Jc Tc WIRE SIZE (nun) 2 FIUMENT (A) (A/cm ) (K) 762 9.24 4310 1067 . 3 x 03 .3 Figs.6 and 7 show the cross-sectional view and dimension of 36-core and 762 -core multi-filamentary superconducting wires, respectively. Tabl 1 showe a s summary of the properties of single-core and multi-core wires. At present, s obtainelowewa c J r rmulti-cor fo d e wire e A resultserie. th f o f so s single-core wires suggests thae smalleth t r dimension wires could produce highes showa Fig.2n c i J rn . Considerin r single-cor fo e sam th c gT e d an e multi-core wiress necessari t i , o improvt y e workinth e g characteristicf o s multi-core wires. XRD pattern e BiPbSrCaCuth f o s O inside silver-sheath show (001) peakf o s high-Tc phase dominant. This good alignment of the crystal is also observed M photographTE d througe transversan th M f SE o hs e cross-sectio e wireth f so n show Figs.. n Detail 9 i e ncompositio d th 8an f o s f BiPbSrCaCuo n O inside silver-sheath were investigated through EDX. Three kinds of the nonsuperconduct- ing phase e observedar s : CaSrPbO, SrCaCu d CaCuOan Oe firs d Th secon. an t d phases are finely dispersed with sizes of below l//m, as shown in Fig.10. TEM observa- tions revealed three typical grain boundaries. First is the boundary with no other phas showin, e e rotatioth g t c-planea n . e witSecon on a thihs i dn amor- phous phase along c-planee thir e th wit on d h an ,smal l particles. These results e showar Figs.11,1n i n . Integrowt13 d an 2 f low-To h c phase with c/2=1A 5 frequently observed s showa , Fig.14n i n . Fig.8 SEM photograph of BiPbSrCaCuO inside silver-sheath. 96 Fig.9 TEM photograph of BiPbSrCaCuO tnstde sliver-sheath. Fig.10 Fine dispersion of nonsuperconduct Ing phases. 97 Fig.11 Grain boundary wit o othen h r phase. 50 À Fig.12 Grain boundary wit a thih n amorphous layer. Fig.13 Grain boundary with small particles. 98 A 0 10 TEM Fig.14 Intergrowth of low-Tc phases. Many structures are observed, such as thin layers, small particles, and intergrowth layers, which can be expected to act as the pinning sites. These structures existed paralle b planea- d Jc~o t an ,l B properties wite magnetith h c field direction paralle b plan a- e improve o n ar magnetiet lno s a d c fielc J d increases. e importanOn t applicatio f theso n e powen wirei s ri s cables. 150-Ampere class cable was tested. Table 2 shows the main features of this cable. Ic was measured at every 5 cm along the longitudinal direction. The results are shown in Fig.15 .e firs th Thi s ti s demonstratio o provt ne capacitth e f silvero y - sheathed high-Tc wire o carrt s y high critical current abov 0 ampere10 e s with uniformity. 99 Tabl e2 150-aipere class superconducting cable. 0 A-clas15 s Ag/BiPbSrCaCuO SuperconduclorI 0 A-clas15 s Ag/BiPbSrCaCuO Superconductor i i i i i i 77. 3K STRAND: 200 THICKNESS: 0.5mm t-O-i WIDT 3.6m: mH LENGT 500m: H m 150 Îc=l77 A NUMBER OF STRANDS: ~ 100 20 3 50 CRITICAL CURRENT: 0.5mmx3.6flimx500mm 20 strands 150 A (77.3 X) x CRITERION: I j(V::|0-"Q'm i i i i i i 0 10 20 30 40 50 length (cm) Fig.15 Ic distribution of 150-anpere class superconducting cable. SUMMARY Pb-doped BiSrCaCuO superconducting wires were fabricated througe th h powder in silver sheath method. Single-core, 36-core and 762-core wares were fabricated e maximuTh . m transport current densit a zer t a 77.n i oy magneti3K c fiel s 17,40wa d 0 A/c r single-corefo m 2 , 7,020 A/c r 36-corfo m 2 d an e 4,310 A/cm r 762-corfo 2 e wires. Zero resistivity temperatur f theso e e wires was 106 K. The magnetic field dependence of Jc(0. l#Y/cm criterion) was summarized as follows for the magnetic field direction parallel to the a~b plane: —6,000 A/cm2 atO.lTesla, ~1,500 Tesl1 A/cmd —20 t a an a 2 0 A/cm2 at 2 Tesla for single-core wires. Highly textured structure was observed with c-axis of the high-Tc phase perpendicular to the longitudinal direction of the flat wires. lengthm 5c 0 , 150-Aaps. class cabl s firswa e t madd testean e d successfully. REFERENCES DH.Maeda, Y.Tanaka, M.Fukutom d T.Asanoan i : Jpn.J.Appl.Phys.27(1988)1209. 2)M.Takano, J.Takada, K.Oda, H.Kitaguchi, Y.Ikeda, Y.Tomii and H.Hazaki: Jpn.J.Appl.Phys.27(1988)11041 3)H.Itozaki, K.Higaki, K.Harada, S.Tanaka, N.Fujimor d S.Yazu:Procan i t 1s . Int.Symp.Superconductivity, August 28-31, 1988, Nagaya(Springer~Verlag, Tokyo, 1989) p.599. 4)T.Hikata, K.Sato and H.Hitotsuyanagi: Jpn.J.Appl.Phys.28(1989)182. 5)K.Sato, T.Hikata, H.Mukai and H.Hitotsuyanagi: Proc.1STEC »orkshop on Superconductivity, February 1-3, 1989, Oiso, (1STEC, Tokyo, 1989) p.119. 6)K.Sato, T.Hikata, H.Mukai, N.Shibut d H.Hitotsuyanagian a : Sprin198S 9MR g Meeting, April 24-29, 1989 n Diego,Sa , M8.71. 7)T.Hikata, T.Nishikawa, H.Mukai, K.Sato and H.Hitotsuyanagi: Jpn.J.Appl.Phys. 28(1989)11204. 100 SUPERCONDUCTING PROPERTIE Bi-BASE TH F ESO OXIDE MULTIFILAMENTARY TAPE MAGNETIN SI C FIELDS . SEKINEH . INOUEK , . MAEDAH , . NUMATK , A National Research Institut Metalsr efo , Tsukuba City, Ibaraki Advanced Technology Research Center, Mitsubishi Heavy Industries, Ltd, Yokohama Japan Abstract e high-Th T -phase Bi-base oxide superconducto s fabricaterwa d into single- core tape multifilamentard an s y tapes wit metaa h l sheath. These specimens were prepared by combination and repetition of cold press or cold roll and sintering. In these specimens axi,c s tende aligo t d n well magnetizatioe Th . n measurement revealed that t 4.2K,a , these tape specime excellenn a d H ha (critican- J t l current density versus magnetic field) characteristics while, at 77K, £he flux pinning force e tapth e n i specimen s reduceswa almoso t d t zero abov thermae th e o t 0.3l e activaTdu - tion, i.e. flue th , x resistive creepth n I . J emeasuremen 4.2Kt a t , these tape specimen scarcels ' 0 s A/c~1 J showef e o mytn t 18T a J ddepended e ,an th n o d magnetic field from 20T to 30T. These results indicate that the Bi-(Pb)-Sr-Ca-Cu-0 tapes could be used, for example, for the innermost coil of an extremely high-field superconducting magnet at 4.2K in a near future. INTRODUCTION temperature measuremene th d an sJ f o t in magnetic fields up to 30T were made for The Bi-Sr-Ca-Cu-0 system (BSCCO) has them in order to study J -H (magnetic high-a T (critical temperature, ~107K) field) characteristic e fluth x d pinninsan g phase ana includes neither rare earth mechanism of this material. element r poisonouno s s elements(l). Further- more, the upper critical field,H „, of these phases are reported to be over 100T at 4.2K(2). These facts indicate that this EXPERIMENTAL PROCEDURE g potentiamateriabi a practicar s fo lha l l n extremeli e us y high magnetic fieldt a s The BPSCCO powder prepared by a co- t 77Ka e .us e th 4.2kr fo wels a ,s a l precipitation method was packed in a metal Recently, the formation of the single high- (Cu and Ag) tube of 10 mm o.d. (outer diam- T phase of this material has been made eter) and 6.7 mm i.d. (inner diameter), and possible by partial substitution of Bi by s colwa d worke m o.ddm int5 . 2. wira o f o e Bi-Sr-Ca-Cu-e th n Pbi 0 system(3-7). (Cu sheath) and 0.5-1.0 mm o.d. (Ag sheath). Therefore, development of the wire (or o.dm m sheathu Th5 .(C e2. ) wirs theewa n tape) fabrication process, study of the rolled into a tape of 4 mm x 1.5 nm in size. magnetic behaviou d improvemenan r f o t nominae Th l catio npowdee ratith f ro used critical current density, magnetin i , J c for the Ag-sheath tape was Bi:Pb:Sr:Ca:Cu fields would be the most important subjects =0.8:0.2:0.8:1.0:1.4 , (accordin6) f re o t g for applicational researches on this mate- nominae anth d l catio ne powderatith f ro rial. Cu-sheate th user fo d h taps Bi:Pb:Srewa : In this study ,Bi-(Pb)e tapeth f so - Ca:Cu=0.92:0.17:1.05:1.12:1.5 (ref 7). The Sr-Ca-Cu-0 superconductor (BPSCCO) wita h powders were calcined at 800C for 14h in Cu sheath or a Ag sheath were fabricated. air. The powder used for the Ag-sheath Magnetization measuremen variout a t s tape was sintered at 845C for 40h in a 101 mixe s (Ar:0_=12:lga d ) before being packed the sample, Boltzmann constant, temperature into the Ag sheath. and the flux pinning potential, respec- For fabrication of a 46-filament Ag- tively. sheath tape, 46 pieces were cut from the monofilaroentary Ag-sheath wire of 0.7 tan o.d-, packe o.dm m 5 d. 7. g inttub A f a o e and 5 ran i.d., and cold worked into a tape powder r fabricatioFo 0.1x m . 6ra mm a 8 f o n1. f o 1330-filament wire, 70 pieces were cut from the monofilamentary wire of 0.5 ran o.d., packed into a Ag sheath of 7.5 mm o.d. and i.d.m m d col5 ,an d worked int wira o f eo 0.7 mm o.d.. 19 pieces were cut from the 0 filamen7 t wire, packed again intoa silvem i.dm 5 . r3. m o.d m sheatd 5 .an f o h and cold worked again int5 1. wira o f o e mm o.d.. ui c tape T was measured by both the stand- 01 ard inductio ne resistiv methoth d an d e method. X-ray analysi s alsswa o made.. e magnetizatioTh n measurements were per- formed wit vibratina h g sample magnetometer in magnetic fields up to 8T at various temperatures. Rectangular specimens of 0.5x3x dimensionn i m 4m s werfrot cu em the sintered tape. Resistive J measure- ments were also madr thesfo e e speciment a s 4.2K in magnetic fields up to 30T with a 20 30 40 50 hybrid magnet of Tohoku University. 2S(deg) 1 X-raFig. y diffraction patterne th r fo s BPSCCO tape specimen and for the powder specimen into whic e tapth he RESULTS AND DISCUSSION specimen has been ground. Figur show1 e n X-raa s y diffraction BPSCCe patterth r O fo n tape specimen which has been sintered at 845C, pressed and sintered agai 845°t na r 36h Fign Cfo I . , .1 an X-ray diffraction pattere powdeth r rnfo specimen into whic e tapth he specimes nha been groun alss i d o shown. Comparisof o n (a) these patterns indicates thaprocese th t s whicn i h pressin d sinterinan g- re e gar i.s 4.2K - peated produces the strong orientation of the c axis in the tape specimen. Figure 2 shows magnetization versus * o magnetic field (M-H) 4.2t curvea K) (a s and (b) at 77K for the BPSCCO tape specimen. -1.5- At 4.2k, the width of the hysteresis in the M-H curve was broad and did not depend 0 magnetie th n o c field shows ,a Fign i n .2 HlkOt] (a). This mean4.2t a KsJ thae th t produceflue th x y b pinnind s sufficii g - entlt dependenno e ys th i hig d n o than magnetic field. At 77K, however, the 0.05 width of the hysteresis became almost zero 77K above 0.3T. This indicates thae th t X' mechanis fluf mo x pinning doe wort sno k above e reason0.3 on t 77K Ta r Fo ,. this clear differenc e hysteresith K f eo 77 t a s from that at 4.2K is considered to result -0.05 - from the thermal activation of fluxoids, -10 0 i.e., the flux creep. HtkOt) Accordin e e flutheorth th x o f t o yg Fig 2 .Magnetizatio n versus magnetic creep, the relaxation of the magnetization field curves at (a) 4.2K and (b) 77K for the non-sheath BPSCCO tape dM/d(lnt) is shown as follows, specimen whic s sinteredwa h , pressed dM/d(lnt) = (d.J /30)(kT/U ) (1) and sintered after the Cu sheath was where d, k, T and U are the dimension of removed. 102 Figure 3 shows the relaxation of the normalized magnetization at various temperatures Fign I magnetizatio, .3 n decays almost linearl e logaristh o t yf mo time,lnt, indicating that this decay of the magnetization is caused by the flux creep t 4.2KA . relaxatio e th , e th f no magnetization was not observable. Figure 3 shows that the relaxation rate at 20K is much smaller than that at 50K or that at 60K. We obtained the pinning potential U with the results shown in Fig. 3 and wite formulth h a (1)20Kt a K .50 ,U K weran60 de 0.04eV, O.OsSe 0.054eVd an V , respectively s i noteworth t I . y thaU t s almoswa t constan n thii t s temperature range. This result contradicts the theory flue th x f melo t bees (8)ha nt .I reporte d A cros Figt I .s sectio e 1330th f -no tha8 flue f tth x re linn i e lattice,FLL, filament Bi-Fb-Sr-Ca-Cu-O wire. become s irregulaa s n amorphoua s a a r t a s temperature T (T = ~40K in the BSCCO system). If tße fTux melt really occurs at a temperature arround 40K, Ü should 1.0 undergo suc suddera h n change lik trana e - resultsr sitionou n I ., however, suca h sudder beet no n s nobserved ha chang U n ei . y ratean ,t A U must be enhanced by o c dopinpr r r producino o g strongee th g r pinning o center n ordesi preveno t r e flutth x creep. o.s 77K o-i 5 0- 0-2 4 0- 0-3 £ Fig, 5 Dependence of J on the bending strain degradates i Th eJ d dow -70o nt y 0.3 b % % bending strain woult y I .sa e fai b do t r to1 »' Time (sec) that this dependence of J on the bending strain is not worse than that of the Fig. 3 The relaxation of the normalized metalic compound superconductor because magnetization at various temperatures. th e strai th dependenc n no improveJ f o e s with decreasing thickness . Figur show4 e pictursa crosa f o es Figur curvH - show6 e J ea s measured section of the 1330-filament Bi-Pb-Sr-Ca- BPSCCe at th 4.2r O Kfo £ap e specimena y sb Cu-0 wire intermediate Th . e annealings standard resistive metho magnetin i d c at 150C which were performed every time whe areae th n l rediuction ratio reache0 1 s gave appropriate hardness to the silver sheat madd han e cold work fabricatiof no this multifilamentary wire possible. ID* Although Bi-Pb-Sr-Ca-Cu-e grainth f so O powder canno e plasticallb t y deformede ,th filaments of the oxide powder in Ag matrix n virtuallca y elongate probably because th e Bi-Pb-Sr-Ca-Cu-O grains can be easily crackestrese th y sdb produced durine th g MO1 cold workresistive Th . eT measuremen t showed that this multifilamentary wire shows zero resistance at~105K. Figur show5 e normalizea s J dversu s B 12 16 20 24 28 32 bendin ge single th strai r -fo n K curv77 t a e H m core BPSCCO tape. The thickness of the Fig. 6 J -H curves measured at 4.2K for oxide superconducting portion is ~100utn. tfie BPSCCO tape specimens. 103 T usin 30 hybrie fieldo th gt dp su magnet Result Figd indicatf Figan 7 .o s 6 . e of Tohoku university. In Fig. 6, speci- that the BPSCCO superconducting tapes have men (1) is the non-sheath tape specimen a big potential for the practical use in sintered at 845°C for 36h, cold pressed extremely high magnetic field 4.2Kt a s . and sintered agai 845t a r n36h Cfo . For example, they could be used as the Specimen (2) is the monofilamentary tape innermost coil of a superconducting magnet (with a Ag sheath) sintered at 840°C fonean i r ro s future r o T o.25 f Althouge th h 36h, cold worked with a flat roll and sin- the presen t higJ tvalu no hs i eenough , tered agai 840* t a n36hr Cfo . Specime) (1 n the practica wile us ll become possiblf i e s measurewa d wit brasa h s shun r protecfo t - the J is more improved by optimizing such tion for the thermal heating. It is note- conditions as composition, fabrication worthy J bote value thae th h tth f so process, sintering, . graion o ns sizd an e specimen almose sar t constan range th en i t of magnetic field from 20T to 30T. This tendency in the J -H curve over 20T is n quiti contrasa e e th o tha t f o t CONCLUSIONS conventional intermetallic superconductors. e intermetalligeneraln I th f o ,J c measurements of magnetization and J compound superconductors decreases very in magnetic fields up to 30T were made for rapidly over 20T. On the other hand, the the BPSCCO tape specimens. The magne- upper critical field,Hc-, of the BPSCCO is tization measurement revealed that the over 100T at 4.2K when the magnetic field tape specimens had an excellent J -H is applied in pararell to the a-b plane. characteristics at 4.2K while the flux Therefore this material is considered to pinning force was reduced to zero above have an excellent J -H characteristic e fluth x o st creep ue e pdu Th .K 77 0.3t a T ° 100T- T .t 80 o degradation of J due to the 0.3% bending Figure 7 shows J -H curves measured strain was -~70%. This dependence of J BPSCCe ath t r 4.2O fo Kmonofilamentar d an y would be comparable to that of the metalic multifilamentary tape specimena y sb compound superconductors. In the standard resistive method in magnetic resistive J measurement at 4.2K, the T usin 23 hybrie fieldo th gt dp su magnet tape specimens.(with high T phase) showed of Tohoku University Fign specime, I 7 .. n scarcels ' J 0 l A/c1 t 1STya Th f m.o J (a)o is the non-sheath tape sintered at depended on the magnetic field from 20T to 84536hr Cfo , cold presse sintered an d d 30T. This indicates that the practical again at 845°C for 36h. Specimen (b) is application of the BPSCCO superconductors the non-sheath tape for which cold press in extremely high magnetic fields at 4.2K and sintering have been repeated twice will become possibl f mori e e improveJ d after the first sintering, and specimen in fields above 20T is obtained for a (c) is the 46-filament tape sintered long length of the tape by optimizing at 840C for 36h, cold rolled and sintered various factors e thanW . k prof. Watanabe again at 840C for 36h. In the specimen of Tohoku Univ. for the use of the hybrid (a) and £b), the J values are as high as magnet. 10 A/cm at 17T and 7x10 A/cm at 23T, respectively. The J -H curve of specimen (b) shows a slight peak effect. The curve (c) shows that the Jc of multifilamentary REFERENCES tape is also independent of the magnetic field in high fields. 1. Maeda , Tanaka,H. Fukutomi, . ,Y d an . ,M high-w ne Asano A T , ,oxidT. e super- conductor without rare earth element, Jpn Appl. J . . Phys., 1988L209, 27 , . 2. Togano Kumakura, ,K. d Dietderichan . ,H , D.R., Critical current magnetid an s c properties of Bi-and Tl-based new high T superconductors, Cryogenics, 1989, 2§, 286. 3. Cava, R.J., Batlogg, B., Sunshine, S.A., Siegrist, T., Fleming, R.M., Rabe, K., Schneemyer, L.F., Murphy, D.W., vanüover, R.B.,, Gallaghers, P.K., Glarum, S.H., Nakahara , Farrow,S. , R.C., Krajewski, 4 2 0 2 6 1 2 1 8 J.J., Zahurak, S.M., Waszczak, J.V., H(T) Marshall, J.H., Marsh Rupp, ,P. , L.W., Jr., Peck, W.F. and Rietman, E.A., H curve- J Fig s7 .measure r 4.2t a dfo K tfie BPSCCO mono- and multifilamentary Physic 1988, aC , 560, 153. tape specimens. 104 Green. 4 Jiang, ,S. Mei, Luo, C. ,Y. , H.L. Koyama. 7 , Kawai d Endo, an , S. , . ,T. U and Politis, C., Zero resistence at 107K Preparation of single 110K phase of in the (Bi,Pb)-Ca-Sr-Cu oxide system, Bi-Sr-Ca-Cu-0 superconductor, Jpn. J. Phys. Rev. B, 1988, 28, 5016. appl. Phys., 27, 1988, L1861. 5. Takano, M., Takada, J., Oda, K., 8- Gammeif p.L>> Schneemyer, L.F., Waszczak, Kitaguchi Miura, Ikeda, ,H. , ,Y. ,Y. J Ca-Cu-0 system, Jpn. J. Appl. Phys., 9> Sekine> H>> Inoue> K., Kuroda, T., and 1988, 27, L1041. Tachikawa, K., Cryogenics, 1989,^9, 96. 6. Endo Koyama, Kawaid ,U. an , . ,T. S Preparatio high-e th f To n phasf o e Bi-Sr-Ca-Cu-0 superconductor, Jpn. J . Appl. Phys., 1988, 27, L1476. ——————————— Next page(s) lef5 t 10 blank APPLICATION OF SUPERCONDUCTING LENS ELECTRON MICROSCOP OBSERVATIOO ET MICROSTRUCTUREF NO S IN SUPERCONDUCTORS H. YOSHIDA Research Reactor Institute, Kyoto University, Osaka, Japan Abstract High resolution electron microscopy has been applied to study of microstructures in A15 and oxide high TC superconductors using the superconducting lens electron microscope which is enable us to observe in atomi 5 typc A1 scal er superconductorsFo e resolutio. K 2 4. t a n , i.e. NboSn, NbgAl, NbgCAl.Ge), structure images with atomic resolution of e clearlar A y5 observed2. . Contrast anomaly sometimes appeart a s grain-boundaries and in grains of these A15 superconductors. Atomic resolution images showin f twio g n structur e clearlear y observer fo d the (a,b) planes of ErBapCuoO . A bright contrast often appears at twin boundaries. Layered structure consistino tw e Er-d on an 0 f o g Ba-0 plane recognizeds i s , where some lattice with different contrast sometimes appears suggesting extra stackin e layereth f o gd structure. Tow kinds of layered structure consisting with two or three lattices betwee e latticth n e image f Bi-o s 0 plane e observedar s . Both lattice image e understooar s BipSrpCaCups a d O ancj Bi2SrpCapCuoOx structure, which correspond to the 85 and 110 K phases, respectively. 1. INTRODUCTION Since the discovery of high Tc superconductors [1,2] many efforts have been devoted to clarify the origin of the superconductivity. The high resolution electron microscopy has effectively been applied to determination of the crystal structures . The structure of oxygen deficient perovskite is formed e transitioth n from tetragona o orthorhombit l c structure introducing many twins in LnBapCugO^c (Ln = V and rare earth elements) superconductors with Tc=90 K. The twin structure has been discussed in a relation with their superconductivity. It has been reported that the layer structure of these oxides produces anisotropy of their superconducting properties. For Bi-Sr-Ca-Cu-0 superconductors the two phases have been observed in connection with two-stape f superconductinso K [3] 0 ge .11 transitionTh d an 5 8 t a s crystal models have been presented e studiebaseth n o dy electro b s d an n neutron diffractio e structurth d an n e modulations wer ee hig founth hy b d resolution electron microscope observations. d higan High c Hc£T h s weln singl ,a i s hig a lc J he crystal, introduca e possibilit o observt y e crystae superconductineth th n i l g state [4,5] orden .I r to clarify the relation between superconducting properties and microstructures, the cryogenic temperature observation of these superconductors is particularly important thin I .s repor e observation tth d oxid an 5 e A1 superconductor n o s s using superconducting lens electron microscope which were already published will be reviewed. 107 . EXPERIMENTAI2 S Superconducting lens electron microscope was originally developed by Siemens group for observation of biological specimens at low temperatures avoiding radiation damage [6]. Recentl e shieldinth y g type superconducting lens ha s been equipped wit higV hk JEO h0 resolutioL20 n electron microscope [7]t i .s A enable us to observe thin foil crystals at atomic resolution at 4.2 K, the multi-beam imaging has been applied to cryogenic temperature observations of microstructure oxidd an 5 e A1 superconductor f so 5,8-10], s[4 . Electron microscope observations were usinV performe k e superconductin 0 th g 16 t a d g lens electron microscope JEM-SCM2000 without tilting goniomete d heatinan r g device. During observatio e specimeth n s welwithouK wa n l2 4. kep tt a ttherma l drifd an t vibration, becaus s locatee i liqui th t i ee n cryostat H i d e magnetiTh . c field at the specimen position is 1.4 T. Thin foil crystals of AI 5 superconductors were prepared by electropolishing NboSC IG frone tap th md froan e m disc f Nb^A o sd Nb^CAl.Ge an l ) pre-thiner A y b d ions. ErBaCuoO and Bi-Sr-Ca-Cu-0 oxides were crushed into thin crystals for observations thesf o C eT . superconductor previousls swa y measured. . RESULT3 DISCUSSIOND SAN S Examples of the high resolution (HR) images of A15 crystal are shown in Figs. 1 - 3 [4,5]. Fig. 1 is a HR image of Nb3Al when the incident beam was paralle <001e th >o t ldirection ,0 lattic wher02 e f 0.2th eo e 6spacinm n s i g resolved. Atomic resolution images of Al 5 structure have been observed as bright dots with similar atomic arrangement of Nb atom chains along the <001> direction s illustratea , e figurth n i e d [11-14]. Fig demonstrate1 . s that almost the same quality of the multi-beam image obtained at 4.2 K as that obtaine t rooa dm temperature. Fig 2 show. s another example imagth ef o e tapeC IG ,e wherth r photographeNb^Sa efo K n 2 crysta4. t s formea d wa l n o d the Nb substrate those 111 lattice of 0.19 nm is resolved. In Fig. 3 a HR lattice images of slightly thick area of NbgSn crystal a contrast anomaly appeared at a grain boundary and in a grain as indicated by arrows, which may arise from in a relation to the flux distribution in the superconducting state [4,5]. Fig.1 HR images at 4.2K for the (001) crystals of Nb3(Al,Ge). 108 Fig 2 . Lattic thir fo n K Nbe 2 3imagS4. n t crystaa e l substrateb growN n no . Fig.3 Lattice imag f NbgSo e n showing contrast anomaly (arrows). 109 show4 imagR FigH . f ErBa2CuoOa so e x, wher o thitw en crystals with different orientations occasionally coexist. The upper crystal was photographed under the condition at which the incident electron beam was normal to the c-axis. There ia layes r structure with c=l.1 whic, 7nm h consist periodicitsa o strontw f go y dar ke wea on line kd daran s k estimates linei e contrast I .th y b d t calculation that heavier atoms give strong darker contrast in these oxide crystals [16]. The layer structure of BaO/ErO/BaO can be understood as the structure model as shown in the figure. A strong contrast layer with wider distance than the regular periodicit s indicatearrog a bi , 4 a we Fig .th y suggestinsees b i dyn i n r E a g and Cu rich plane [15]. The bottom crystal was oriented at the (001) plane being parallel to the electron beam, where twin structure was recognized because the contrast slightly differen r eacfo th twin. Bot(100e th h d (010)an ) lattices wit spacinm hn 0.3 e 9resolved ar g . Arrows indicate twin boundaries, where broad bright contrast appeared sometimes followe broay b d d dark lines. Twin structure in YBaCu0x has been reported by many authors. Fig. 4 HR image at 4.2 K for two thin crystals with different orientations. e twiTh n structur alss i e oe multi-bea th see n i n m imag f ErBa2Cu3Uo e x shown Fig. 5. The structure images corresponding to the atomic lattice spacing are clearly resolved. Atomic arrangement at the twin boundary is in principle, but an irregularity of atomic arrangement and strong strain field appears somewhere s indicatea figuren i d e broaTh . d brigh d daran t k line contrast often appeared along twin boundaries. Althoug t clea ye e contrase origi th t th rh no f now o n,s i t a possibilit e b t ther e contrasbu y th ma e f o y t raise a relatio n i d n wite th h magnetic flux distribution in the superconducting state [8]. 110 Fig. 5 Twin structures of showing contrast anomaly at boundaries. As the crystal structure of LnBaCuqO superconductors (Ln=Y and rear earth x elements) is an oxygen deficient perovskit2 e with the orthorhombic structure, the twin structur believes i e o fort d m durin e transitioth g n from tetragonao t l orthorhombic structure in the annealing process of the sample preparation. The contribution of twin structure to superconducting properties has been discussed e boundarieth r fo s e pathactin r th superconductin fo s a g g e electronth s a d san pinning centres for preventing the movement of magnetic flux in a magnetic field. Among the mixed twin structures a part containing fine micro-twins was reported o shot w highe C thaT r n that containing large twins twie [19]th n f I boundarie, s contain local strain field, the local strain may effectively assist to form the orderin f oxygeo g n atom d vacanciesan s , whic essentias i he boundarie th r fo l o t s act as the superconducting paths. For A15 superconductors the grain boundaries are believed to act as the pinning centres, where many lattice imperfections at grain boundaries are effective for contribution to their strong pinning force. Althoug e twith hn boundary contains less misfit imperfectioe goob dy ma t i , n enough to contribute for pinning force in the oxide superconductors because of their short coherent length closed to the lattice parameter. These speculations introduce reasons to originate the bright and/or dark contrast observed at the twin boundarie Erf^CugCn i s t cryogeni^a c temperature. In Fig .example6 multi-beae th f o s m image f Bi-Sr-Ca-Cu-o s 0 specimee ar n shown, which were photographed whee electroth n n beae c-axiss normath wa m o t l. The layer structure with c/2 = 1.54 nm consists of two strong dark lines and three weak dark lines, which correspond periodicita o st f BiO/SrO/Ca/SrO/Bio y O planes, as indicated in Fig 6(a). A crystal model of BioSroCaCuoO (85 K phase) is illustrated in Fig. 6(c). The layer of Cu0 is too weak in contrast. In a multi-beam image obtained at room temperature 2 there is another layer structure o coexistt , widea whic s ha rh separation thae abovth n e structure [8]. This layer structure consists of four weak lines of SrO/Ca/Ca/SrO layers between the strong dar ke Bi0 lineth 2 f o layerss shows a , Fign i n .6 (b) . This structure is recognized as BiSrCaCu30 for the higher T (110 K) phase. Coexistence of 2 2 x C these two phases correspond2 s to the two-step transition appeared in resistivity. Ill Fig. 6 HR images at 4.2 K for Bi-Sr-Ca-Cu-0. (a) Bi?SroCaCuoO (85 K phase), (b) Bi2Sr2Ca2Cu3Ox (110 K phase). Fig. 7 Structure modulation observed at room temperature (a) and 4.2 K (b) 112 Structure modulation in Bi-Sr-Ca-Cu-0 reported by other authors [17,18] was also observed in the multi-beam images obtained at 4.2 K when the electron beam was parallel to the c-axis. The atomic lattice spacings with a*b=0.54 nm are clearly resolved in the images obtained at room temperature [8,9], however broad and wavy lines strongly appeared along the a-direction. The modulation appeared as a periodicity of five atomic planes but sometimes as that of four planes. The line contrast disappears at a border as indicated by an arrow, where some domain waly presentma l e modulatee imageth Th .f o s d structure seem o appeat s s weaa r k contrast at 4.2 K in comparison with those at room temperature, which may be related to a reduced strain in the superconducting state [8,10]. Origin of the structure modulatio s beeha nn discusse d e periodibaseth n o d c arrangemenf o t oxygen vacancies [20] or a superlattice structure [8,18]. In order to clarify the relation between superconducting properties and microstructure cryogenic temperature observation wil e essentiab l e futureth n i l. 4. SUMMARY The results obtained in the previous experiments are summarized as follows. 1) For AI5 superconductors high resolution images appeared as bright dots just corresponding to the Nb atom chains along the 001 direction. Contrast anomaly s sometimewa s observe t graia d n boundarie grainn i d san s suggesting magnetic flux distribution in the superconducting state. ) ErBapCu32 0 appeare e layeth rs a dstructur e witd o daran tw h k O lineBa f o s one weak dark line of ErO when it was observed normal to the c-axis. The high resolution lattice images of the (001) crystals clearly show twin structure and sometimes shows contrast anomal t twia y n boundaries. 3) The images of BipS^CaC^Ojc (85 K phase) and Bi2Sr2Ca2CuoO (110 K phase) were recognized as the layer structure with a periodicity or threx e and four weak dark lines, respectively, between strong dark lines of BiO planes. The structure modulatio e higth hn i nresolutio n images also observed. REFERENCES G.Bednor) 1 K.A.Mullerd zan . PhysZ , . B649 (1987)18 , . 2) M.K.Wu, J.R.Ashuburn, C.J.Trong, P.H.Hor, R.L.Meng, L.Gao, Z.J.Huang, Y.A. WanC.W.Chud an g , Phys.Rev.Lett 8 (1987)90 , 58 . ) H.Maeda3 , Y.Tanaka, M.Fukutom T.Asanod an . AppliiJ n Jp ,. Phys , L2027 . 9 (1988). 4) H.Yoshida, H.Hashimoto, Y.Yokota and M.Iwatsuki, Jpn J. Aplli. Phys. 26 Suppl. 26-3 3 (1987),94 . 5) H.Yoshida, Y.Yokota, H.Hashimoto, M.Iwatsuki and Y.Harada, J. Electron Microsc. 36, 228 (1987). ) I.Dietrich6 , Proc h Int.9t . Cong Electron o . n Microscopy, Tronto4 17 . ,p (1978). 7) M.Iwatsuki, H.Kitahara, K.Nakanishi and Y.Harada, Proc. 11th Int. Cong. on Electron Microscopy, 1 Kyoto(1986)25 . p ,. 8) H.Yoshida, Y.Yokota, H.Hashimoto, M.Iwatsuki and Y.Harada, New Developments in Appli. Superconductivity, World Sei. Pub. p. 261 (1989). 9) H.Yoshida, Y.Yokota, H.Hashimoto, M. Iwatsuki and Y.Harada, J. Less Common Metals, 151 3 (1989),24 . 10) H.Yoshida, Y.Yokota, H.Hashimoto, M.Takeda, M.Iwatsuki and Y.Harada, Proc. EMAG-MICRO'8 press)n (i 9 . H.Yoshida) 11 , M.Taked d H.Hashimotoan a , Proc h Euro8t . . Cong. Electron Micros. Budapest, p. 231; Phys. Stat. Soli, (a), 87, 473 (1985). 12) H.Yoshida, M.Takada, H.Hashimoto and H.Endoh, In Situ Experiments with High Voltage Electron Microscope, Osaka Univ., p. 229 (1985). 113 13) H.Yoshida, M.Takeda and H.Hashimoto, Proc. Int. Sympo. Flux Pinning and Electromag netic Properties in Supercon., Fukuoka, p. 160 (1985). ) M.Takeda14 , H.Yoshida, H.Endo H.Hashimotod an h . Microscopy0 , , 151 7 (198814 , ) ) D.H.Li15 , H.Shibahara, J.P.Zhang, L.D.Mark, H.C.Marc S.Songd an y , Phisic, C a 156(1988)1 20 , . 16) K.Hiraga, D.Shindo, M.Hirabayashi, M.kikuchi and Y.Shono, 0. Electron Micros.1 (1987)26 , .36 , ) J.M17 . Tarascon Pagee L . Y ,, P.Barboux, B.G.Bagley, L.H.Green, W.R.McKinnon, G.W.Hull, M.Giroud and D.M.Hwang, Phys Rev. B37 9382 (1988). 18) H.W.Zandbergen, W.A.Groen, F.C.Mijlhoff, G.Van Tendeloo and S.amelincks, Physica c, 156, 325 (1988). 19) H.Hayashi, Y.Yokota, S.Morita, N.Horiuchi and S.Hayashi, Jpn J. Appli. Phys. 27, 1856 (1988). 20) T.Kajitani, K.Kusaba, M.Kikuchi, N.Kobayashi, Y.Shono, T.B.Williams and M.Hirabayashi, Jpn J. Appli. Phys. 27, L 587 (1988). 114 HIGH-RESOLUTION TRANSMISSION ELECTRON MICROSCOPY OF CRYSTAL STRUCTURES, DEFECTS, SURFACD EAN INTERFACE BISMUTH-BASEN SI D SUPERCONDUCTORS Y. MATSUI National Institute for Research in Inorganic Materials, Tsukuba-shi, Ibaraki, Japan Abstract Sinc discovere eth nighf yo superconductorC tT investigationf o t lo sa s have been performed to clarify the origin of the superconductivity. The High-Resolution Transmission Electron Microscopy is one of the most effective instruments to get more information and understanding in the superconducting mechanism of oxide superconductors. This paper reviews and discusses the resul studief to Bi-Sr-Ca-Cu-n so 0 superconductors which have been performed at the Japanese National Institute for Research in Inorganic Materials. Special attention has been paid to crystal structures, defects, surfaces and interface examinef so d samples. 1. INTRODUCTION In early 1988, Maeda et al. [1] discovered a new series of high-T_ supercon- Cj ductors in a Bi-Sr-Ca-Cu-0 (BSCCO) system. The two superconducting phases with T„ about 80K and 1 1 0K claimed by them were later identified to be O r a ( 115 not completely indexed by the tetrago- nal substructure of Fig. 1. Such dis- crepancy was later solved by the direct high-resolution transmission electron microscope (HRTEM) observa- tions of the structure by Matsui et . [5,6]al . They showed thae 221th t 2 phas s stronglha e y distorte r moduo d - lated structure with an incommensurate 4.8 times long-periodicity in the b- o direction. The similar modulated structure e alsar so observee th n i d high- TK (220120 (2223d ) an ) O phases [7, 8] . In this way, it was revealed that the modulated structures e commoth e nar feature e serieth f o s f o Bi-based superconductors. Recent HRTEM study also showed that the cleavage surface of the 2212 phase suffers from the wavy atom fluctua- tions due to the effects of the modu- e lationbulth k n i structurs . ] e[9 Althoug e detaileth h d analysie th f o s modulated structures are still in progress s considerei t i , d thae th t modulations are associated with inter- stitial oxygen atoms and/or cation substitutions which are possibly responsible to the formations of Fig. 1. charge carriers, holes or electrons. Average structure of the low- Informatio e modulatioth f o n n struc- TC superconductor, E^S^CaCk^O tures are, therefore, considered to be (2212), after Tarascon et very importan o understant t e th d al. [4]. superconducting mechanisms of the series of Bi-based superconductors. In this paper, the results of the 116 HRTEM studies of crystal structures, defects, surface and interfaces per- r formelaboratorou n i d e brieflar y y reviewed [10, 11]. Most of the samples examined in this study are prepared by the solid state reaction methods using starting powders of 61203, SrCOg, CaCOg and CuO. They e pre-firedar , groun d finallan d y anneale e appropriatth t a d e temperaturef o s 0 °C[1,2] 90 e obtaine - Th 0 . 80 d specimen e grounar s CCln i d ^ using agate mortar and then dispersed onto perforated carbon films. HRTEM observations are performe ) H-900(2 d JEM-4000E) d mainl0(1 an typ 0 y M b y30 eTE V Xk typ0 40 e kV TEM, both enable us to get HRTEM images under the point-to-point resolution of 0. 17 - 0. 19 nm. 2. RESULTS AND DISCUSSIONS 2-1. Modulated Structure Bi-Basen i s d Superconductors As already mentioned, electron diffraction patterns of various reciprocal lattice sections of the low-T_ 2212 phase include many weak satellite reflec- LI tions which are not explained by the simple tetragonal substructure shown in Fig.1 [5,6]. Detailed analysis of the electron diffraction patterns showed thae exacth t t structure shoul e describeb d n orthorhombia y b d c lattice with . ] 6 Thi, [5 sm C =c=3 n mean 8b=2d . 8 .an 0 =4 s .6 B A =a=0 thae crysta, th .t54 l So o has a long-periodicity in the b direction, which is incommensurate with the b parameter of the tetragonal substructure. In order to examine the origin of the incommensurate superstructure, we performed a direct HRTEM observation with an electron beam along the a-axis. In the resultant image shown in Fig.2 e doublth , e line f strono s g dark contras e considerear t o correspont d d e (BiO)th o t 2 double layers e strikinTh . g feature (BiO)th f o 2e layers i s that the horizontal Bi-Bi distance is not constant but fluctuates in the fa- direction. The regions of darker contrast where Bi atoms are compressed to each other was called the "Bi-concentrated bands", as indicated by the symbol e micrographth n i " "B . Another region f weakeo s r contrast where Bi-Bi dis- tance is relatively large are, on the other hand, called the "Bi-deficient s alsi bands" ot I foun. d thae verticath t l Bi-Bi distanc e s largei eth n i r Bi-concentrated bands as compared to the corresponding distance in the Bi- deficient bands. In other words, horizontal compressions and the vertical expansions of Bi-Bi distances are induced in the Bi-concentrated bands. Strong lattice distortions are also induced in the perovskite layers to form zigzag vertical lattice planes. The horizontal periodicity of the two Bi- concentrated , bandwhic nm s abou i 6 sh . 2 tinclude n s(020 te nin r o )e lattice 117 Fig. 2. HRTEM image of the 2212 phase showing the modulated structure [6]. Symbol B stands for the Bi-concentrated band. •mtttWMiiiutf e reasoth planee nb d thi thay e averagan s ma th st e periodicity reat froou dm the electron diffraction patterns is 4.8 times of the b-parameter of the tetragonal substructure. Arrangement f cationo s s proposede th base n o d computer simulations of image contrasts is shown in Fig. 3 [12], Here, we assume that some of the Bi-sites are substituted by the Sr atoms and, on the n thiotheI s. rwayon e hando th ,s i atom B ,e Sr-sited e th soman s th f o ey b s characteristic features of the modulated structures in the low-T_o 2212 phase e obtainedar . Determination e oxygeth f no s arrangements are, however, diffi- cult based on the HRTEM methods because of a weak interactions of electron with light atoms suc s oxygena h . Recently, Yamamot [13. al ] t performee o e th d Rietvelt analysi e powdeth f ro s X-ra d neutroan y n diffraction datd showean a d that the Bi-deficient bands contain interstitial oxygen atoms, while the oxygen atoms in the Bi-concentrated bands have rock-salt type of structure as e tetragonaith s l substructure propose y Tarascob d[4]. al . t e n They however insist that no substitution occurs at the Bi sites, which is different from our mode le HRTE baseth n M o d images. Detailed analysi e imagth f e o s con- e modulatetrastth f o s d structure e stil sar progressn i l . Similar structural modulation e alsar so observe n bote high-Ti d th h _ 2223 c phase wit d als 220e T =110han th o1T =20] K[7 K phas en examplA e [8]th . f o e L« \jf HRTE Me modulateimagth f o e d structur e 222th 3n i ephas s showi e n Figi n . .4 Most of the perovskite layers formed in the 2223 phase have thickness of 1.8 nm which contains three Cu-planes. The crystal however contains various perovskite layer f differeno s t number f Cu-planeso s s indicatea , e righth n ti d hand sid f eaco e h perovskite layern thiI s.s reveale i way t i , d thae th t 118 BiO 1 , î ] 1 » r i i---i BiO » 4 4 • * t f é < l i i | i 1 1 • f f é , 1 ? f i • Bi o i o 0 f 0 • , < ° A f ° • Bi(Sr).O 's 1 « • 4 Öd Sr(Bi).O 0 f ° 4 0 f 0 t o \ o 1 t ° r o I „ ' O C*.0 1 , 1 1 1 i Ci » Ca(Sr) 1 « t î 9 1 1 î * • Cu T ; 1 t BiO i 4 • 1 0 o t i t T t — J _„t..„ 1 — --t— -4 4-— 4- •~1 B 1 t BiO 4 • 4 t 1 î t ? i ^ 1 « I f ? i ai • 4 1 ; 0 o * ° T 0 i 0 o 4 ° Fig. 3. 6 & i 8 I t j 0 T » i o ; t "i »1 0 • ° j 0 f 0 o^o 1 Fig. 4. HRTEM image of the modulated structure and intergrowth defecte 222th 3 n i shigh-T c phase [7]. high-Tc 2223 phase contains many intergrowth defects, as predicted by the 2 broadeninpowde00 e rth diffractiof o g n peak. Arrangement e Bi-conth f o s- centrated bands are very similar to those observed in the low-T c 2212 phase in e thereforW . 2 Fige. concluded thae high-Tth t 2223 - phasor n ea alss ha o c thorhombic superstructure with an incommensurate modulation in the b-direc- wels i lt tion I know. n thae formatioth te intergrowtth f o n h defecte th n i s high-T 2223 phas s considerabli e y reduce e structureth y dopinn b di b P g. c 119 [14. Ikedal ] t showee a d thae modulatioth t n structur e Pb-dopeth n i e d 2223 phas quits i e e different from tha f non-dopeo t d 2223 phase. Similar effects f Pb-dopino e alsar g oe 221founth n 2i d phas [15. al y Che]b e t e n The modulated structure formed in the 2201 phase (T =20K) is also examined o by HRTEM as shown in Fig. 5 [8]. In this case the distance of the two (BiO)2 layer abous i s t c/2=1. . 2nm Formation Bi-concentratee th f so d band e alsar s o observe t thesbu d et for bandno n orthorhombia mo d s c lattic a monoclini t bu e c one with A^O. 54, Bm=2. 6, Cm=2. 7 and alpha=113°. Reason of such a monoclinic distortion is unknown in the present stage. Effects of the Pb-doping in the 2201 phas e recentlar e y examine e e chang th [16th . y Matsu b dal f d o ] et an e i modulation features, similar to that observed in the Pb-doped 2223 and 2212 phases, are observed. . Fig5 . HRTEM image of the monoclinic modulation in the 2201 phase [8]. Both monoclinic and pseudo-orthorhombic unit cells are drawn in the micrograph. 2-2 Intergrowth Defects and Interfaces Examples of the intergrowth defects in the high-T^ 2223 phase are already U show n Fig.4i n . Such intergrowth defect e seldoar s m observe e 220th 1n i d TC=20K phases, although small amoun f o intergrowtt e perovskitth f o h e layers of three and four Cu-layers are observed in the low-Tc„ 2212 phase [17]. Another typ f defeco e t structure observe e Bi-baseth n i d d superconductors i s the discontinuit e (BiO)th f o y2 layerss reportewa s a ,[11]. y Matsub dal .t e i 120 Various types of interface structures are also observed in the Bi-based superconductors. Two kinds of twin structures, the rotation-type twinning e mirror-typth d [17an ] e twinning e observear , , 7 s showa d d n Figsi an n 6 . respectively e firsth n tI exampl. e [17]uppe) n Figth i (a e , 6 . r part shows typica le orthorhombiimagth f o e c modulated structur e 221th 2n i ephase , while the lower part seemingly shows no modulation. This indicates that the upper e loweth rd an parts show e HRTEth s M images projected alone e th [100th gd an ] [010] directions, respectively. In other words, the upper and the lower regions are mutually related by the 90 degree rotation around the common c- axis. One of the interesting conclusions obtained from Fig. 6 is that the twin boundary exist o n betwee(BiOi stw e )th nplane n boto s he sideth f o s arrowhead in the micrograph. This indicates that the upper and the lower (BiO) planes suffer from the structural modulations in the different direc- tions whic e perpendiculaar h o eact r h other s schematicalla , y drawn i n Fig. 6(b). This may be the results of the weak chemical bonding between the o (BiOtw e mirror-typ)th planesn I .e twinnin ge lowefounth n ri d parf o t Fig.7 [10], two domains B and B' have 2201 structure with a monoclinic symme- try [8] . The two regions have common b-axis but the c-axes are mutually related by the mirror operation with mirror plane parallel to the (001) plane. Fig. 6(b). Schematic representation of the rotational twin structure of the 2212 phase [10]. Fig(a)6 .. HRTEM e rotationimagth f o e - type twinning observed in the 2212 phas ) [17]e(a . 121 Figure 7 also contains a phase boundary (PB) between the 2212 phase in the e 220th uppe 1e d lowe phasth ran d n B')parre ) i phas ean Th (A par tB . ( e t boundary plan s almosi e t paralle e commoth b plano a- nt l f boto e h phases. e numberTh f Cu-planeo s eacn i s h perovskite layers neae phasth r e boundars i y indicate e centrath n i dmicrographe l th par f o t . e numberChangth f o es from (-2-2-2- e uppeth n ri ) par o (-1-1-1-t t e loweth n ri ) par s observei te th t a d phase boundary. Some irregularit e sequenceth f o y , suc s (-2-2-1-2-1-1-a h ) is, however observed s indicatea , e centrath n i de micrographl th par f o tt I . is also observed in the micrograph that some of the (BiO) layers are not the 2 complete sheet t suffebu s r froe displacementth m c directio e th n o i forst n m stepped (BiO)2 layers s indicatea , y arrowheadb d micrographe th n i s . Obser- vation f suco s h stepped (BiO)2 layer e alsar s o reportee thith n r filfo df o m Bi-based superconductors. Fig. 7. HRTEM image showing the phase boundary (PB) betweee th n uppe e loweth r r221 220d 2an 1 structures [11]. The lower 2201 domai na mirror-typ alss ha o e twinning with twin boundary parallel to the (001) plane. e interfaceth Mos f o t r boundarieo s n Bi-basei s d superconductor e thuar ss paralle e basath lo t l (001 )e layere planth f o ed structures e exceptionaOn . l e boundarth cas f o e y introduced almost e (001normath o )t l plan s showi e n i n Fig. 8, which is obtained from the fluorine doped BSCCO superconductor [18]. In this case, the right and the left hand sides of the crystal have mutually different layer sequences and these two regions meet at the central part. e (BiOth Som ) 2 f layereo s suffer from complete discontinuity. This means thae (BiOth t ) e 2 changeb layer n ca sd intperovskite-1ike th o e structures without causing large structural instability. 122 . Fig8 . HRTE Me F-dope imagth f o ed BSCCO superconductor. showing the boundary which is associated with discontinuit e (BiO)th f 2o y layers [11]. 3 Profil2- e Imagin f Wavo g y Surface Characterization of surface structure is basically important to understand and control the properties of thin films of the high-Tc superconducting oxides. One of the interesting problem of the surface in Bi-based supercon- ductor whethes i s e bulth r k modulations mentione 1 mak3- en i dsom e effectn o s the surface structures. Because of the weak chemical interactions between o th(BiOtw e ) planess considerei t i , d thae crystae cleaveb th t y o ma lt d separato (BiO tw e interestin b e y )th e e casema planesth t i f thi,I s gwa s. to know whether the concentrations and expansions of the Bi atoms in the (BiO)2 layer e stilar s l inducee singlth n i de (BiO p surfaceto ) e planth . n o e o examinT e surfacth e e structur e 221th 2f o phasee e triew , o perfort d e th m so-called "profile-imaging e cleaveth f o " d (001 n thiI ) s plan tech. ] e[9 - nique e surfac th e examine, b t paralle o e electrose t e th s i do t le nth bead an m surface profile projected in the beam direction is obtained in the HRTEM image. In the present case, the crystal with sharp (001) edge is selected wit[100set its h ]and direction paralle electroan to l n resultanbeamThe . t HRTEM imag s showi e Fign i n . 9 Existence f boto s h Bi-concentrated band) (B s e Bi-deficienth d an e observetar one) (D sd insid e crystal th e edg Th f o e . the crystal showe (001e imagth th s f o )e plane projecte e [100th n i ]d direc- tion. First we can conclude from this image that the top surface consists only of the single (BiO) plane. It is also observed that the top (BiO) plane 123 Fig. 9. Profile imaging of the wavy cleavage surface of the 2212 phase. Symbols B and D stand for the Bi-concentrated and the Bi-deficient bands, respectively [9]. consist e Bi-concentrateth f o s d regions (B') with darke- Bi r contrase th d an t deficient ones (D') with weaker contrast. The B' bands are almost above the D e nearesbandth f o s t (BiO) layers e otheandth n ro , hand ' bandD , s above 2 the B bands of the (BiO)2 layers. The horizontal distance of the two B' band, almos nm s i abou se sam 6 th . t e2 t B parametewit e e orthorhomhth th f o r - 0 S bic superlattic e modulateth f o e d structure. This strongly suggests thae th t B' and the D* bands are the lower half of the original B and D bands, respec- tivelye (BiO)th f go , layers thin I s . n conclud wayca e ,w e thacleavage th t e is actually induced in between the two (BiO) planes and one of these (BiO) plane outermos th s lef i et a t t surface layer. Another important result obtained from s thae Figi surfac a wavth t0 . 1 s y ha efluctuation f atomo s n i s the vertical directions e e Bi-deficienheighth Th t .a t ' bandD t s slightli s y larger than that of the Bi-concentrated B' bands. In this way, the hillocks ' band D e surfac e formee th ar s th n o f deo wit a periodicith m n f abou o y6 . 2 t in the b direction. Such a wavy surface structure is reasonably expected if the modulations in the bulk (BiO) 2 layers are maintained even after the two (BiO) planes are separated. Formations of the wavy surface in the Bi-based superconductors are also reported by Kirk et al. [19] based on the scanning tunneling microscope (STM) observations. 124 3. CONCLUDING REMARKS Because of the formations of the incommensurate modulations from the simple layered structures e analysi e th structure, th f o s se standar th base n o d d diffraction methods are difficult in the present Bi-based superconductors. High-resolution transmission electron microscopy e otheth n r o ,hand n ca , provide us the direct real space information on the characteristic features of e structurath l modulations e HRTETh M. techniqu n alsca e o provide th s u e information on the defects, boundaries and the cleavage surface of the Bi- based superconductors, whic e difficular h y otheb t r diffraction methodsn O . e othe th n generai r s i hand lt i ,quit e difficul o obtait t e accuratth n e data f atoo m position d occupationan s se HRTE baseth Mn o d images e thereforW . e consider that the proper combinations of the direct HRTEM method and the indirect X-ray and/or neutron diffraction methods are required to achieve the finally satisfactory informatio e variouth n o ns aspecte structureth f o s f o s Bi-based and other high~T_ superconducting oxides. References . MaedaH . TanakaY 1, . FukutomM , . AsanoT d an i: Jpn . ApplJ . . Phys, 27 . L209 (1989). . Takayama-MuromachiE 2 . UchidaY , . MatsuiY , . . OnodK KatoM , d aan : Jpn. J. Appl. Phys. 27, L556 (1989). 3 C. Michel, M. Hervieu, M. M. Borel, A. Grandin, F. Deslandes, J. Provost and B. Raveau: Z. Phys. B68, 421 (1987). 4 J. M. Tarascon, Y. Le Page, P. Barboux, B. G. Bagley, L. H. Greene, W. R. McKinnon . . HwangHüllM W . . . GirouM D G , :d , Physan d . Rev. B37, 9382 (1988). . MatsuiY . MaedaH 5 , . Tanak. Y Horiuchi,S d aan : Jpn . J .Appl . Phys, 27 . L361 (1988). 6 Y. Matsui, H. Maeda, Y. Tanaka and S. Horiuchi: Jpn. J. Appl. Phys. 27, L372 (1988). 7 Y. Matsui, S. Takekawa, H. Nozaki, A. Umezono, E. Takayama-Muromachi and S. Horiuchi: Jpn . ApplJ . . Phys , L12427 . 1 (1988). 8 Y. Matsui, S. Takekawa, S. Horiuchi and A. Umezono: Jpn. J. Appl. Phys. 27, L1873 (1988) . MatsuiY . Maeda H ,9 . Tanak. Y Horiuchi,S d an a : Jpn . ApplJ . . Phys, 28 . L946 (1989). . MatsuiY 10 . MaedaH , . TanakaY , . HoriuchiS , . TakekawaS , . TakayamaE , - Muromachi . Umezon A ,. Ibe K d :an o JEOL NEWS 26E 6 (1988),1 . 11 Y. Matsui and S. Horiuchi: Study of Superconductors, Nova Science Pub lishers w YorNe , k e press(1989th n .i ) 125 12 S. Horiuchi, H. Maeda, Y. Tanaka and Y. Matsui: Jpn. J. Appl. Phys. 27, L1172 (1988). 13 A. Yamamoto, M. Onoda, E. Takayama-Muromachi, F. Izumi, T. Ishigaki . AsanoH d an : submitte o Physt d . Rev. 14 S. Ikeda, K. Aota, T. Hatano and K. Ogawa: Jpn. J. Appl. Phys. 27, L2040 (1988). 15 C. H. Chen, D. J. Werder, G. P. Espinosa and A. S. Cooper: Phys. Rev. B39, 4686 (1989). . MatsuiY 16 . . UchinokuraMaedK A , d an a n préparationi : . 17 Y. Matsui, H. Maeda, Y. Tanaka, E. Takayama-Muromachi, S. Takekawa and S. Horiuchi: Jpn. J. Appl. Phys. 27, L827 (1988) 18 S. Horiuchi, K. Shoda, H. Nozaki, Y. Onoda and Y. Matsui: Jpn. J. Appl. Phys. 28, L621 (1989). 19 M. D. Kirk, J. Nogami, A. A. Baski, D. B. Mitzi, A. Kapitulnik, T. H. Geballe and C. F. Quate: Science 242, 1673 (1988). 126 Next page(s) left blank SUPERCONDUCTING MAGNETS AND APPLICATIONS (Session C) MAGNETIZATION STUDIE MULTD1LAMENTARF SO Y STRANDS FOR SUPERCONDUCTING SUPERCOLLIDER (SSC) APPLICATIONS — METHOD CONTROLLINF SO G PROXIMITY-EFFECT COUPLING AND RESIDUAL MAGNETIZATION E.W. COLLINGS*, K.R. MARKEN, Jr., M.D. SUMPTION* Battelle Memorial Institute, Columbus, Ohio, United States of America Abstract Helmholtz coils r modificationo , f theso m (e.g. saddle-coils commonle ar ) y user dfo producing dipolar magneti coile woune th c sf ar i fields t d Bu fro. m superconducting strands, residual magnetization, M™ residen strane th n i td material itsel responsibls i f r efo multipolar distortions of the desired field. It is well known that the height of the M(H) hysteresis loop — AM(H) = (MR+ - MR-), where the signs refer to the trapping (paramagnetic shieldind an ) g (diamagnetic) branches, respectivelys i M(H f - o , ) proportional to the product of filament diameter, d, and critical current density, JC(H). Thus in an attempt to reduce strand magnetization (in the presence of high Jc) and the attendant field distortion strona , g effor bees ha tn produceo undet y commerciaa rwa n o , l scale, multifilamentary strands with smaller and smaller filaments. In order to preserve filament quality (i.e. to prevent thickness undulations, or "sausaging") in small filaments, it bees ha n suggested necessar confino yt ratie eth f filamen oo t spacin filameno t ) g(s t diameter (d) to s/d < 0.15±0.02. The combination of small d with low s/d results in interfilamentary spacings sufficiently close to proximity-effect couple the filaments. For f 0.15o ,example d s/ Cu-matri n a t a , x filaments that have been reduce 5-1/n o di t m 2p diamete beginnine ar r exhibio gt t coupling coupline th d an ;g becomes stil s worsi ld s ea furthe rinterfilamentare reducedth f i t Bu . y matri alloyes xi d with ~0., 5 wt.%Mn couplin barels gi y perceptible even diamete witm (i hl r filaments. Next, having disposef do an excess magnetization due to proximity-effect coupling one is still faced with the inherent magnetization of the NbTi filaments themselves. During the operating cycle (field-increasing magnetC SS e th , thif o ) s magnetizatio diamagnetics ni ; accordingln ca t yi be neutralized by including in the superconducting strand a material with a large positive magnetization suree b i barrier o N , T . , ssucNi havs ha e been incorporated into multifilamentary strands to eliminate proximity-effect interfilamentary coupling, and bulk Ni inserts have been recommended for magnetization compensation in SSC dipoles, but presene th t ide f associatinao i directlN g y wit strane hth r locadfo l magnetization compensation is new. It may turn out to be convenient to add the Ni as an electroplated outside strande layeth th t n shoul f o reo bu ; d interna possible preferree b b y i N l ma e t di eliminato t e couplin compensatd gan residuar efo l shielding magnetizatio singla n ni e operation. INTRODUCTION When a multifilamentary composite superconductor is subjected to a time-varying external magnetic field the "height" of the M(H) hysteresis loop is proportional to the critical current density, Jc, and the diameter, d, of the filaments. Thus in order to minimiz magnetizatioe eth n magnet C residen windingSS e d e th coil e an th ,n i th tf so f so the attendant distortion of its dipolar field, the filaments of the composite strands must be made as small as possible. The maintenance of a high Jc under these conditions requires Also affiliated with Ohio University, Athens, Ohio, United States of America. 129 the preservation of filament quality (the avoidance of "sausaging"), which in turn dictates filament-spacing/filament-diametea f o e us e th r ratio, s/d, preferably withi range nth e 0.13- 0.17. A consequence of this is that filaments whose diameters have been reduced to below about 5 (tm are coupled by proximity effect. This contributes an unwanted excess magnetizatio t leasa o nportiot a M(He th f no ) hysteresis loop, thereby counteractino gt some exten advantage th t e that would otherwise accrue from takin strane gth d through those final reduction stages1'2'3. t wil I showe b l n that interfilamentary couplin suppressee b n gca alloyiny db matrie gth x with a low concentration of Mn4. In the strands prepared for this study, an interfilamentary alloy of Cu-0.5wt.% Mn was used. As a result, filaments as small as d = desig/ime 0.15 f (henco th m 0. t /i 9 thi,d n a i n = s/ 5 s es 2. case 3) were successfully decoupled. But having disposed of the excess magnetization due to proximity-effect couplin stils i le facegon d wit inherene hth t magnetizatio NbTe th f ino filaments themselves. During the operating cycle (field-increasing) of the SSC magnet, this magnetization is diamagnetic; accordingly it can be neutralized by associating the superconducting strand wit materiaha largf o l e positive magnetizatione b o T . , sucNi s ha sure barrieri N , s have been incorporated into multi-filamentary strand eliminato st e proximity-effect interfilamentary coupling buld i insert5an ,k N s have been recommended for magnetization compensation in SSC dipoles6, but the present idea of associating Ni directly with the strand for local magnetization compensation is new. It may turn out to electroplaten a s a i N conveniene e b th doutsid e d strandlayee th ad th t n f ro o e t o bu ; should it be feasible to introduce the Ni between the filaments, it would be possible to eliminate couplin compensatd gan residuar efo l shielding magnetizatio singla n ni e operation. EXPERIMENTAL Magnetization Measurements Magnetizatio measures nwa functios da f NbTo f temperaturc no T i wite th h o t p eu field sweep amplitudes of from a few tens of gauss up to 15 kG. A computerized PAR- EG&G vibrating-sample magnetometer (VSM useds )wa association i , ironG nk wit-7 1 ha core electromagnet powered by a ± 65 A field-controlled bipolar power supply. In completing a full hysteresis loop, including the initial branch from the origin, the instrument records 1,023 data pairs. Thu fiele sth d resolutio experimeny an n ni abous i t t 17200field-sweee th f o p amplitude, which enable l finsal e structure associated with couplinth g magnetizatio fulle b yo nt recorded. Sample Material Samples were prepared commercially from high-homogeneity Nb-46.5wt%Ti rods clad wit thiha n barrier-laye (whosb N f o re presenc ignores date ewa th an di analyses) . Two serie f strandso s were studied Cu-matrixA ) (1 : , 6,000-filament series, z 0.15wit d hs/ , designated RHIC (indicating a class of strand intended for Relativistic Heavy Ion Collider application) 23,000-filamenA ) (2 ; t series, designated CMN, wit interfilamentaryn ha - Cu 0.5wt.%Mn matrix and an s/d of 0.19 3. In each case the strand design called for an annular filamentary bundle surroundind , encasean u C coreu n di C ga . This configuration is illustrated in Figure 1; further details of the strands under study are listed in Table 1. 130 Figure 1 Scanning electron micrograph of CMN-25. Tabl e1 Specification Sample th f so e Materials Strand Fil. Magnetization Test Sample** Sample Diam.,D, Diam.,d*, No. of Length, 2 NbTi Filament Code 10" cm ftm Strands mm Volume, l(T3cm3 RHIC-009 2.54 2.106 80 6.31 10.74 RHIC-013 3.47 2.890 48 6.31 12.13 RHIC-026 6.57 5.490 14 6.44 13.03 Cu-0.5wt.%Mn-Matrix Strands: 22,902 filaments (not heat treated) CMN-5 1.27 0.500 200 5.88 5.28 CMN-10 2.75 1.068 58 5.74 6.83 CMN-15 3.85 I.495 33 5.60 7.43 CMN-21 5.29 2.051 21 6.15 9.77 CMN-25 6.35 2.459 15 5.64 9.20 CMN-115 29.99 II.556 1 5.892 14.15 Obtained by etching-and-weighing using separately measured densit bulf yo k Nb-46.5T = 6.097)( i . Referrin clao gt d samples only. Magnetometer-Sample Preparation samplee Th magnetizatior sfo n measurement consiste cylindricaf do l bundles, about 3 mm in diameter and 6 mm in length, of parallel multifilamentary strands imbedded in epoxy. Depending on the strand diameter, the number of strands in the bundle varied from 1 (CMN-115) to 200 (CMN-5) so as to keep the volume of superconductor roughly constan t aboua t t 0.0 1presene th cm n I 3.t study sample th , e orientatio axis-normas nwa l 131 to the applied field. In general it was customary to prepare two sets of samples: one consistin as-receivee th f go d composite othestrandse th barf ro d ean , NbTi filamentl al - s havinu C e gth been remove etchingy db compariny B . magnetizatione gth clae d th d an f so bare materials influence th , proximitf eo y effect coul clearle db y observed. PROXIMITY EFFECT Isuperconductora f intimatn i s i , s , e contact wit normaha l conductor, n , superelectron pairs will leak throug interfacen si e probabilite hth Th . findinf yo ga superelectron pair in n at a distance x from the interface is P a exp(-V) (1) where, for high-conductivity (so-called "clean") normal conductors, k,,"1, the characteristic decay length gives i , y nb 1 V = h vf/2irkBT (2) Ferme th s i lattee i f whicvelocitv n Th i judgeh/2irs d ri = . cleae n s han h b it n y i o df n t i electronic mean-free-path, mfp, £, is very much greater than a "coherence length", £n, given by en = (h v^otfkßT) (3) Take Cu for example. At 4.2 K the coherence lengths of coppers with residual resistance ratios (RRR) of from 75 to 200 lie between 0.69 and 1.13 /im; the corresponding mfps are 3.1 8.4o 5t 5 /im, respectively above th y eB .criterio n these coppers woul regardee db s da clean with decay lengths as prescribed by Eqn. (2), which for T = 4.2 K, yields k^"1 = 0.45 /im. Thus exampler fo , Cu/NbTn i , i composites stabilizef normao u C ly db purity e th , filaments will be coupled by proximity effect provided the interfilamentary spacing is less than about 0.9 /im; i.e. a filament diameter (at s/d = 0.15) of about 6 /im. Three factors will reduce, and eventually destroy, proximity effect coupling: (1) a reduction of £4 (2) an increas temperaturee th n ei increasn a applie e ) th (3 ;f eo d magnetic field strength. MAGNETIZATION REGIMES - MANIFESTATIONS OF PROXIMITY-EFFECT COUPLING At constant temperature (say 4.2 K), as the applied magnetic field increases, the strand passes through four clearly discernable magnetization regimes, Figur. e2 Filaments in the Meissner State ) Ver(a y Small Applied Field Regime appliee th f I :d fiel vers di y weak t wili , e b l exclude Meissney db r effect fro interfilamentare mth wels a fros a NbTlu e myC th i filaments entire Th . e filamentary bundle acts lik soliea d bloc f superconductoko r witha magnetic volume susceptibility (for field normal to the cylinder axis) of -1/2*. This effect is exhibite initial-magnetizatioe th y db n segmen M(He th f o t) loo Figurn pi . e3 (b) Small Applied Field Regime: As the applied field strength increases beyond what migh termee b t "lowee dth r critical interfilamentare fieldth f "o (whicu yC r hfo CMN-5 at 4.2 K is about 9 gauss) a mixed state is created in which flux is pinned throughout the interior of the filamentary bundle. This state is characterized by excess magnetizations on both the field increasing and the field decreasing branches of what has become an almost four-fold-symmetric M(H) loop. The overall inclination of the loop (cf unclae th tha f o td sample (-l/2ir)Ps )i "flux-exclusio,e wherth s i eP n volume-fractionf "o 132 FILAMENTS IN THE MEISSNER STATE Very Low Field Regime Filament Bundlein APPLIED the Meissner State FIELD INCREASING Low Field Regime Filaments Individually in 8 the Meissner State FILAMENTS IN THE MIXED STATE Intermediate Field Regime Filaments Still Coupled by Proximity Effect High Field Regime B Filaments Uncoupled Figur e2 Field-dependent magnetization regimes. 20- l 10- 0- CMN-5 i -10- -20- 0 50 -1000 0 -500 1000 Magnetic Field Strengthe O , Figure 3 M(H) loops for clad CMN-5 and unclad CMN-5B at 4.2 K illustrating magnetic behavio "vere fieldth w "lod yn lo ri "an w field" regimes. 133 filamentse th filamentr r thicfo Fo t .k filamentbu , s1 whos = sP e radius comparabls i , R , e to the penetration depth, X, P = l-(2/x)I1(x)/I0(x), where I0 and 1} are modified Bessel functions of the 1st kind of order 0 and 1, respectively, and x = R/X. At constant temperature decreaseP , s along d/2 = wit( illustrate s )a hR Figurn di . e4 20 10- 0- I -10 -> CMN-115 -20 0 20 -400 0 -200 400 Magnetic Field Strength, Oe Figure 4 Low-field ( at 4.2 K. cl Filament Mixee th n si d State (a) Filaments Coupled: The filaments themselves enter the mixed state when they are exposed to magnetic fields greater than their Hcls. For thick filaments, this will be the recognize NbTif o t ver l c Bu yd.H fine filaments ente mixee th r d stat t somea e enhanced transverse field give Hy nb cl /vT r examplefo ; , thi r CMN- K gauss s7 fo ,2 fiel60 4. s t di 5.a e mixed-statTh e hysteresis loo NbTr pcurvefo e familias th i l —e sal se labelleo t r d "unclad" in Figure 5. But when the filaments are coupled by proximity effect an additional field-dependent magnetization component arises. The surprising, as-yet unexplained, feature of this component is that it appears only along the field-decreasing (trapping) branches of the M(H) loop. This is particularly evident in the curve labelled "clad Figurn "i e 5(b). (b) Filaments Uncoupled: Eventually the applied field can become sufficiently stron decouplo gt electroe eth n interfilamentar e pairth n si y space, i.e destroo t . e yth proximit thie y se effects e happeninW . "wingse Figure-5(bth e th n gi f "o ) M(H) loop. VISUALIZATIO COUPLINF NO G 3 It is well known that the total height of the magnetization loop, AMV(H) (emu/cm NbTe oth f i componen compositee th f o t relate s i )filament'e th o dt s critical current 7 density, Jc (A/cnr), and diameter, d (cm), by AM^H) = (0.4/3T) Jc d (4) 134 30 20 CMN-) (a 5 unclad I 10 clad 1 -10 -20 -30 0 1 0 -20 0 -1 20 Magnetic Field Strength, kgauss 120 (b)RfflC-009 80 clad 1 40 l 0 a o 'S -40 I -80 -120 -3 -2 -1 Magnetic Field Strength, kgauss Figure 5 Mixed state hysteresis loops at 42 K for clad and unclad samples of (a) CMN-5 and (b) RHIC-009. In the upper figure, the presence of paramagnetia c interfilamentary matri gives xha npositiva e eth tilo t t clad M(H) loop. Particularly eviden lowee th n ri t figure th s ei dominance of coupling magnetization along the trapping branches of H(H) It follows that any magnetization enhancement due to proximity effect coupling can be 8 expresse termn effectivdn i a f so e filament diameter doccurrence y effsa ,Th . da eff feo visualizee b n ca d d through applicatio followineithef e no th f o rmethodso gtw : The Method From Eqn ploa AMy/Jf ) o t. (4 cabsence againsth n i f coupling, eo d t , shoule db linear throug origine hth . Otherwis t smalea l value , positivd f so e departures fro line mth e 1 shoul observee db d indicatin onsee assumee gth b couplingf o tn ca t di s thaf i I .c J t independen oved range f o trth e concerned same th , e shoul ploa true f dr b o te fo versus indicates a , d Figurn di . e6 135 .5 2 - 4 6 8 10 12 Filament Diametern (a , Figur e6 Heigh M(He th f )o t hystersis loop versus nominal filament diameter — derived from Ref.l (Ghos t al)he . The Method insurancs A e agains possible th t e variatio thed J wit avoian o nf t n, o e h d dth c difficult f havinyo measuro gt possiblea y large low-fiel AM^^/AM^e th , c dJ ^ method was devised. Referring agai Eqno nt . afte(4)d an , r associating AMy^ wit couplingha - enhanced obtaide euw , n simply that AMvciad/AMy.bare deff /dlatte, r being constant and close to 1 when coupling is absent. Wit= h decreasine tn g d, departure from the constant value indicate onsee sth f couplingo t seriee . th Figur f Cu-matrir so fo , e7 x RHIC strands indicates that coupling sets in below d = 5.5 pm (i.e. s £ 0.8 /im). COUPLING REDUCTION WITH AN INTERFILAMENTARY Cu-Mn ALLOY Figure 8, for the CMN series of strands with the Cu-0.5 wt.%Mn interfilamentary matrix, shows that d needs to be reduced below 1.1 pm (s ^ 0.2 /im, in this case) before effece w th ef coupling o e begit se o nt . MAGNETIZATION COMPENSATION Introduction and Preliminary Studies bees ha nt I show filamenne th tha n i t t rang f relevanceo SSCe th ,o e t proximit y effect couplin eliminatee b n gca d throug incorporatioe hth f Cu-Mno interfilamentars na y matrix. But as indicated in Figure 5, practically all of the coupling magnetization shows 136 RfflC-Strands kOe 4-- l I &I2- Filament Diameter, /jm Figure7 Relative heights of the M(H) loops at 4.2 K and the fields specified for clad and unclad (bare) samples of RHIC strands. 15 CMN-Strands kOe v 0.05 la- • 0.5 o 1.0 * 10.0 Filament Diameter, pm Figur e8 Relativ fielde e th height M(Hse d specifieth an f so K ) loop2 4. d t sa for claunclad dan d (bare) strandssampleN CM f .so alonp u field-decreasine gth r trappinggo branche f M(H)so , magnewhereaC SS e s i t sth operated along the field-increasing or shielding branches. Thus an issue still to be settled is whether coupling was ever a serious problem for the SSC. In any event, having disposed coupline th f o g problem either operationall r physicallyo stils i le faceyon d wite hth inherent magnetization of the filaments themselves. Since during the operation of the SSC magnet this magnetizatio diamagnetics ni compensatee b n ca t i , d locall associatiny yb e gth strand with a small volume-fraction of ferromagnetic material such as Ni. In our preliminary studies the magnetization of a typical sample was compensated at a field of shora y b t lengtG wirei abouk N 5 f h.o 4- t 137 Conceptual Design of Magnetization-Compensated RHIC and CMN Strands Data neede magneticallo dt y compensat strandN e o RHI t CM field n si 3 d f Can so providee ar G 4 k Tabln di e 2(a)strane Th . d magnetizations referre havo dt e been obtained from the field-increasing branches of their M(H)s; the data for Ni was acquired during a magnetization measurement at 10 K — see Figure 9. Listed in Table 2(b) are: (i) the volume percentages of Ni that need to be associated with the strand for compensation within 3-4 kG; (ii) the number of Ni filaments that would need to be inserted, for compensation at 3.3 kG, should it be decided to replace some of the NbTi filaments with ; (iii thicknese Ni )th platee layea b f so o rt d ont surfac e strande oth th f eo , agair nfo compensation at 3.3 kG, should that route be selected. Table 2 Conceptual Design of Magnetization-Compensated Strands (a) Specific Magnetization of Filamentary NbTi (at 4.2 K) and Ni (at 10 K) Strengt Increasinge th f ho Applied FieldG k , Sample Material 3.0 3.3 4.0 Magnetizatio f NbTino , M<-p, emu/cm RfflC-009 -10.758 -10.176 -9.124 RfflC-013 -13.073 -12.391 -11.185 RHIC-026 -20.881 -19.917 -18.164 CMN-5 - 6.366 - 6.036 - 5.178 CMN-10 -6.223 -5.796 -5.048 CMN-15 - 6.865 - 6.472 - 5.685 Magnetizatio , M,,MNi f no , emu/cm3* Ni + 467.4 + 478.1 + 496.8 ) Volum(b e Percentag Actuad ean l Volum i NeedeN f eo Compensatior dfo n at Various Fields Vol. Pet. Ni, IOORC = IOOA^/ASC*» Thickness Sample No. of Ni of platingft Code 3.0 kG 3.3 kG 4.G 0k Filamentsf t,ftm RHIC-009 2302 2.128 1.837 127 0.6 RHIC-013 2.797 2.592 2.251 154 1.0 RHIC-026 4.468 4.166 3.656 244 3.0 CMN-5 1.362 1.263 1.042 286 0.1 CMN-10 1.331 1.212 1.016 274 0.3 CMN-15 1.469 1.354 1.144 306 0.5 normalization I unio nt t volume densita , 9.0f takens yo 4wa . If M represents a material's specific magnetization and A its cross-sectional area, while subscripts "SC "addd "an " denot compensatine th NbT d an i g addendum (i.e. Ni), then the fractional amounts of addendum material required for compensatio simple = nar c yR t Numbe f Ni-replacero d NbTi filaments totaa r casf 6,10e o lfo ,RHIth f e o 8n i C and 22,902 in the case of CMN, needed for compensation at 3.3 kG. ft Plated layer applied to the outside of the strand (appropriate to 0.33 T operation) computed from the relationship t = (AsC/jrD)Rc. 138 60 •^ 40 i *„ 20 I 'S 0 -20 -40 -60 5 2. -5.0 0 -25 5.0 Magnetic Field Strength, kgauss Figure 9 Specific magnetization of (evidently) unannealed pure Ni wire (0.1 mm diameter actuae )Th measure . l K sampl 0 1 RHIC-00s t da ewa 9 plus Ni wire (see "Preliminary Studies" section) measured above the NbTie th f .o c T Design and Performance of a Ni-Electroplated Strand A length of arbitrarily chosen multifilamentary strand was taken for the electroplating study. Its moment per unit length (Mjg_, emu/cm) at a "design field" of 6 kG was measured. Next desiree th , d plating thicknes calculates swa d accordino gt t = \Mgj*ûa \ (5) where a is the moment per unit volume of Ni at the design field. In practice, a is close to the saturation moment of Ni at liquid-He temperatures. In the present study, with d = 8.226x10-* cm, M£(6kG) = -2.7603xl(r2 emu/cm, and a(6kG, 4.2K) = 515.92 emu/cm3, platinEqn) calla r (5 .sfo g thicknes /«n1 attempn 2. .f A so deposio t t i layeN f thaa to r t thicknes mads swa e usin gcalibratea d Ni-electroplating bath. Estimate weighy db t change, the actual thicknes layee th f r so deposite /*me degre2 2. Th . s whico et d wa h compensation was achieve desige th t da n fiel indicates di Figurn di . e10 139 -9 0 9 18 Magnetic Field Strength, kgauss Figur0 e1 Magnetic hysteresis loop at 4.2 K for a Ni-electroplated multifilamentary strand. Note that moment compensatio increasingH + n( ) lakes plact ea 7.52 kG. Conclusion A few percent of Ni added to a superconducting strand can offset most of its shielding magnetization ove wida r e magnetic field range. Pur i addeN se littlth o et existing magnetic hysteresis oncd an ,t reacheei s saturatio effecs bodilo nit t s i t y shife th t M(H winge directioth M f + directioso ) M e - loo nth e when pth positivi s ni n i nH d ean whenegatives i nH . The addition of Ni to the strand may relieve the SSC magnets's need for fine purposee filamentsth f o whicf e s o als s on , minimizo ht wa e conductor magnetization. Early studies have suggested that magnetization compensation coul achievee db d through the insertion of bulk Ni into the dipole "wedge"6. But in this work we are proposing that the Ni could be associated with the composite strand itself ~ either in the for replacemenf mo t filament coatin a outsid e s th a r n gso o e surface. Furthermore, currently under way is an investigation into the possibility of surrounding the individual filaments with Ni -- a technique reminiscent of diffusion-barrier technology. If it is metallurgically feasibl introduco et needee eth — thiy percenn i w thici sdwa fe N kf o t enough to retain its ferromagnetic properties, and without contaminating the Cu -- a strand could be created that would be both magnetically compensated and immune to proximity effect. ACKNOWLEDGEMENTS The clad and unclad epoxy-potted magnetization samples were prepared by R. D. Smith. The research was sponsored by the U.S. Department of Energy, Division of High- Energy Physics. 140 REFERENCES 1. A. K. Ghosh, W. B. Sampson, E. Gregory, and T. S. Kreilick, "Anomalous low field magnetization in fine filament NbTi conductors", IEEE Trans. Magn. MAG-23, 1724 (1987). 2. A. K. Ghosh, W. B. Sampson, E. Gregory, T. S. Kreilick, and J. Wong, "The effect of magnetic impuritie barrierd magnetizatioe san th n so criticad nan l currenf o t fine filament NbTi composites", Tenth Int. Conf. Magnet Tech., Boston, MA, Sept. 21-25 (1987). 3. E. Gregory, T. S. Kreilick, J. Wong, E. W. Collings, K. R. Marken Jr., R. M. . TaylorE Scanlan . C conducto A " ,d an , r with uncouple diametem /* 5 d2. r filaments designe outee th r dipolrC dfo cabl SS ef e o magnets" , IEEE Trans. Magn. MAG- 25, 1926 (1989). 4. E. W. Collings, "Stabilizer design considerations in ultrafine filamentary Cu/NbTi composites", Sixth NbTi Workshop, Madison , NovWI , . 12-13, 1986 alse ;se o Adv. Cryo. Eng. Materials 34, 867 (1988). . KreilickS . T . GregoryE ,5 . Wong.J d an , , "Geometric consideration desige th n si n and fabricatio multifilamentarf no y superconducting composites", IEEE Trans. Magn. MAG-23. 1344 (1987) alse se ;o Ref. .2 . GreenA . M , "Contro6. highef o l r multipole dipolC SS en s i magnet o t e sdu superconductor magnetization using ferro-magnet materia dipole th n i le wedge", Lawrence Berkeley Laboratory Report LBID-1533, SSC-MAG-661, September 1989. . . WagnerCarJ R . . rW G Jr.d , ,an "Hysteresi . 7 fina n esi filament NbTi composite", Adv. Cryo. Eng. Material (1984)3 92 , .s30 . ShenS . S , "Magneti8. c propertie multifilamentarf so y Nb3Sn composites"n i , Filamentar Superconductors5 yA1 . Clark . F Suenag . M A y , d b Plenu . aan ed , m Press, New York, 1980, pp. 309-320. Next page(s) lef1 t 14 blank EFFEC HEAF TO T CAPACIT MATRID YAN X RESISTIVITY ON STABILIT SUPERCONDUCTORF YO S IN FAST CHANGING FIELDS E.Yu. KLIMENKO, N.N. MARTOVETSKU, S.I. NOVIKOV Kurchatov Institute of Atomic Energy, Moscow, Union of Soviet Socialist Republics Abstract Recently developed theory of stability which takes into account smoothed transition of SC to the normal state allowed to obtain some important results. The theory accounted for stable performanc wireC S f so ewit h high current densit t pooa y r coolin d allowegan o fint d d limitf o s SC wires stable performanc t differena e t conditions. Stability criteria were mostly founded with an approach of "frozen flux" conditions, i.e. unchanged current density distribution during instability evolution. Theory has been confirmed in series of experiments, still as frequency range wireC foS rs spreadi s , some evidence r morfo s e stable wires behavior than theory predicte ar s appeared for thin SC wires in fast changing fields. In this paper a stability of SC wire has been analyzed taking into account heat capacit d matrian y x resistivit d stabilitan y y criterie ar a obtained. It is shown that for thin wires taking into account heat capacity and matrix resistivity permissible rates of field or current changing are tens fold as high as permissible rates for "frozen flux" model. The obtained criteria may remove contradictions between the theory and experiment and allow to estimate stability limits in AC and pulsed magnets. INTRODUCTION agreement between experiment and theory was obtained. However for adequate Modern stabilitC yS theorr fo y description of wire behavior at rapid rates of composites with smoothed transition to the field changing we had to attract heat normal state [1-3] explaine e possibilitth d f o y capacity. At high current charging rates we wire stable performance at poor cooling far saw more stable performance than theory beyond steady state stability conditions [1,2], predicted [1,4]. The numerical simulation of allowed to establish relations between losses wire behavio t a rapir d chargin ] showe[5 g d d stabilitan y [1,3] o t fin, d critical ratef o s that influence of heat capacity appears field changing and current charging; lower significantly even if above mentioned ratio these rate wirC S s stabli e d realizan e s it e between head fielan td diffusion e rateon s i s full current carrying capacity [1,4]. ordeo o tw f rmagnitud o r e rather than some e recentlth Mos f o t y obtained stability units. criteria used assumption that instability e experiencTh f o applicatione C S f o s initially grows under unchanged current or wires in AC windings [6-8] where field magnetic flux distribution conditions (so- changing rates were 100-100 s showeT/ 0 d called "frozen flux" conditions) e grounTh . d that thin wires were still stabl r beyonfa e d for suc n assumptioa h w differentialo s i n l frozen flux model prediction d diffusioan s n resistivity of the superconductor (dE/dj) = rates are becoming comparable. E/j0 at low enough electrical field level. Due o T illustrate comparw t i e e theory thio t s fact diffusion rat f electricaeo l field predictions for known wire [8] and its real turns out to be much slower than that performance e followinth e s wirTh ha .e g of heat inside ( A /c ) or outside (ha/c) the parameters:r0 = 025 mm, jc = 5-10° A/m^ wire. Hence, heat capacity c is getting (whol e coie Th lcros . T s 2 = section B n i ) unimportant because "frozen in" current made of the wire reached 2.1 T and overheats irreversibly the wire with any heat practically fully realized its current carrying capacity as far as such ratio between capacity o T mak. e estimationd ad e w s diffusion rates is valid. Here E - is electrical unreported parameters: temperature field - currenj , t density - currenj , t parameter parameter of electrical field increasing T0 = 0 d coefficienan 0.0K 2 f heao t t transfee th o t r of electrical field increasing which definee th s 2 transition characteristic smoothness, X - helium h=l(P W/m K. thermal conductivity coefficient, c - heat Electrica e takinl th fiel f f whico of gd h e capacitwire th - h hea f , o ty transfer is principle paramete f o stabilitr n frozei y n coefficient, a - dimension of a flux model [1,3] will be: superconductor. 4 Theory predictions were verifien i d Eto = 2hT0/jcr0 = 3.2-10- V/m (1) series of experiments [4] and fair qualitative 143 Recently published pape ] give[9 ra s Boundary condition e slath b n surfaco s e limi f o maximut m field changing ratn i e are: framework of frozen flux model. For -h(T-THc \ dT/dx)= ; dE/dx= -dB/d) (6 t instance, at I = o.9lc one can find that (dB/di) = ma5Ex c/r 0s =6.T/ whic4s i h o simplifT e probleth y assume mw e that approximatel ordero tw y f magnitudo s e lower temperature distribution over conductor cross than really reached e experimenth n i t [8]. section is uniform. It is valid in most The main sources for the discrepancy practical cases and may be expressed as we see in heat capacity and matrix resistivity X/ha»l t h=10A . 3W/m~ Z-S-lO^a d 2Kan t mi omitted in frozen flux model. is enoughXt e b moro e thaw unitfe f no s Actually, even classical stability criterion W/m-K. which gives usually too pessimistic predictions Now we can integrate (4) to obtain: for wire performance still predicts existence of so-called adiabatic limi r thicknesfo t a 2 s slaC S b: oa f dIT f jEdJ ca_= h(T-Tx+ -) _ (7 He=J) i a = (wc(Tc-THc)/)1jc2)l/2 (2) dt 0 Wires with lower thickness should be r stabilitFo y analysi e nee w so kno t d w conditionsy stablan t ea . background electrical field distribution. Fair Here w is a factor of 2.5 - 3 depending approximatio (di/dt)(x-x= r thaE nfo s i t d 0an ) on thermal conductivity coefficient. xo=B/Jc» wher j ecorrespond o t currens t 3 c r exampleFo =2kJ/mc t a , ,Tc =7., K 2 e sladensitth b n o surfacy a e s i [5,10] 0 x , jc=10^ A/m2 we obtain 2a=024 mm, i.e. dept f saturateho d zone. wires with thickness muslesm m s t tha4 02 t After linearization of the set of withstan ratey f an fieldo s d changing. equation by expressing the solution as a sum In contrast, frozen flux model has no of background and perturbed terms E=Eb+ such a limit. )E;T=Th+/1T we obtain: In our paper we present results of study wirC S of e stability under conditions where ratio between rate f o heas t transfed an r magnetic field diffusion may be arbitrary . T value, to find out limits of SC wires stability b EEb/JE"p jT 0 QE4 b in fast changing fields. * _ + ____ _ + __!___*_ _ (8) PROBLEM STATEMENT d/}T . We treat a SC slab stability in external ca__ =JÛEjcdx- (9) magnetic r fielo unded r rapid current dt charging against small virtual perturbations . We assume that current density does not a ter Equatios mha whic) (8 n h regards depend on magnetic field inside the slab. dynamic f backgrouno s d temperatur d thian es Equations which describe a problem are: featur a principl s i e e distinction from earlier used equations for stability analysis [1,2]. E'=0j/e)t)^ (3) We search solution in the form: c(àT/at)=XT"+ JE (4) AE=Y(x)exp(vt/tc)M(dI/dt)(a-x0); Descriptio compositC S r nfo e state: After n a transformation t ge e w s equation: isc-Jc(Tb) E=E exp( c )=JcCst (5) " +(k -v((tt +1)2]Y +v(ftx +l)Wx/(vy+10 )= Jo 1 (10) s i conventionawher c E e l "critical" electrical field, which arises at "critical current density d T=TJan " , (TJ ,s backgrouni d where: temperature). We assume as usual [1,3] that > =f^(dl/dt)(a-x0)/pj0; x=(x-xo)/(a-xQ) j0 and TO parameters are connected as JO/TO =ic(Th,B)/(Tc(B)-T d j0b/jan ) c= <5. We assume tor definiteness that current /= ca(dl/dt)/hj0(a-x0); kx = (dTb/dt)tc/T0; is charged e intslath ob wit a hrat e (dl/dt=const). From the stability standpoint it 1 is equivalent to the case with increasing of W= _jYdx ; k=(l 2 external field with rat ( dB/de t =fAdI/dtj till (U(dI/dt)l /hjc(Tc-THe); full saturation. 00 144 (i - is a ratio between differential At v>kj the stability criterion and resistivity and resistivity of matrix. RESULTS Influenc f Heao e t Capacit WirC S en yo Stability in Fast Changing Field If differential resistivits i y C (E/jS f Qo ) much less than normal resistivity of the neglecn matrixca , thee 1 tw n, « i.e. normal metal shunting effect on stability of SC slab. Solution of (10) has a form: = o (ii) 2 vk where Jn- is a first kind bessel function, OS 10 15 Minimum value f k=ko s m;n correspond e conditioth o t n (dk/dv. 0 = ) FLUX JUMP FIELD, T Eigenvalue v fros m (11 e presente)ar d Fig 2 Field changing rate vs field of in Fig. s dependenca l . k n o e the flux jum ) thi1 ps work (13a)) 2 ; criterion (13 b), 3) frozen flux criterion, o - experiment [4] s e flui th a fielwherp x s f B i jumo d d B e pan a full penetration field. Curve 2 corresponds e criterioth to n obtained earlier froe th m rough model [4]: _ B2=(2-08l_ —— ————3 —+ — )(Tc-THe) (13b) (dB/dt) B As it is seen from comparison, more accurate criterion (13a) has better agreement wit n experimenta h , especiall r higfo y h field changing rates. Curve 3 in Fig.2 corresponds to frozen flux criterion (12) which does not take into account influence of heat capacity. s i seeIt n froe Fig.th m 2 that event a moderate rates of field increasing the At 145 allows to obtain attainable field (or current) frozen flux model and our results is much t infinitela y high changing rate more significant n particularI . e criticath , l charging orden rata s f magnitudi eo r e higher than frozen flux model prediction and the _1.44pc(T= _ c-Tb) (14) differenc s i increasee s chargina d g rats i e grown. Comparing the result with classical one e takinth o gS into account heat capacity (2) one may notice that for high thermal result mucm s h more optimistic permissible conductivity (w=3 for this condition) stable level of field changing than it is expected thickness of the wire, taking into account from frozen flux model e boundarTh . f o y heating up by losses during charging, is less than classical criterion predict a facto m f s o r appreciatio f thio n se writte b mode y s a nma l (3/1.44)1/2 ^ 1.44 So, estimation of the a limit for a slab halfthickness: stable wire thickness presented in the beginnin e pape e th 0.1b f r6o o g t turn t ou s r valuefo m s m chose 4 insteam 02 m n f o d there. Obtained from equation (11)lower this thicknes e muson s t take into consideration heat capacity of the wire Here dependence kmm(<5,^) contains implicitly current I and current charging rate (dl/dt) e defineb y . s 0.0a dma 1 c (foi=l/lty rd can , It is useful to have analytical about 10% accuracy for maximum current). For instance, at i=07 and other approximation for dependence kmjn((f/) for different paramete valuesô f o r : parameter. ic s , definemm 1 d = abov a , e wire r diametethinnefo ; s raboui mm r 2 t wire and for high frequency (some tens of (15) Hz and higher) one can not obtain whic s satisfactorha h y agreement with reasonable estimation e frameworth n i s f o k exact solution. the frozen flux model In Fig 3 the dependences of attainable current on current charging rates are presented for two values of slab thickness 02 Stability criterio r rounfo n d twisted cable t smoothnesa m m 1 sd m parameteman - 0 r with curren changinn i t g held e resTh 001t parameter. s are: h=HPW7m2K, Now we try to spread results obtained j = lo9A/m2, T -T = 3K, =2kJ/m3K In for simple geometr r mosfo y t interesting case c dlc He c from the practical standpoint, namely for e dependenceth Fi g3 s imax( /dt) calculated round twisted multifilamentary wire with from frozen flux criterion (12 e show)ar r fo n transport current in varying magnetic field the comparison It is seen from Fig 3 that for e assumW e that twist pitc s shari h p enough thick wires the last criterion gives satisfactory presentation concerning critical current to neglect the saturation zone due to field charging rate, above this rate attainable changin n comparisogi n wite wirth h e radius. current fall s rata s e increases Howevet a r fast charging rate e frozeth s n flux criterion (dB/dt)L2 becomes too pessimistic The same result was « 1 obtained in paper [5] in numerical simulation e probleoth f m d characteristian c tim f fielo e d changing For thin wires the difference between is much higher than tim f couplino e g currents delay Here \° - is effective resistivity of the 1 2 a twismatrix s i t p pitcL , h Usually these C conditionA r fo s applications are fulfilled. We assume as well, - 0 1 that transport curren s i largt e enougd an h saturated zone where transport current with 08 - 'critical" density flows is comparable with M • wire radius. In the case the stability is mainly «06 - determine y electricab d l field distribution i n the saturated zone where shielding current I > a=lmm a=0.2mm hav same th e e directio s transpore a n Th e on t ~'04- distribution of electrical field is given in paper [9]: 02 - 2 2 E=(dB/dt)sin?[r ro (l-i)]/r(2-i)v (16) 00 2 1 2 ' 10 ' 10 ' 10 0 1 1 * • 0 1 * " 0 1 "10"* where Q< 146 corresponding stability criterion. These Normal matrix influence is essential at parameters are: high electrical fields levels still this influence is much smaller than heat capacity influence. The critical charging rate when current starts 2c(dB/dt)f(i) to fall as rate increases is almost unaffected. (17) This resuln agreemeni s i t t wite obtaineon h d in paper [5]. 2r02jc(dB/dt)f(i) (18) 3*h(Tc-Tb) where f(i)=_ (19) (1-0.5Î) v e th equatioSolutin f o ng (15) simultaneously with set of definitions (17-19) one can find maximal current which can be attained in external varying field. Note, that 10 at 147 s obtainei c j e stabilitth d y proble r themfo m . 5 Klimenko E.Yu., Martovetsky N.N., o thei t o acut s e r e du eb threwoul t eno d StabilitC compositS f o y t a rapie d current order higher heat capacit t a nitrogey n charging and against pulsed heating. IEEE temperature. Trans.on Mag. v.24, N2, 1988, p. 1167. 6. Février A. Latest news about SC A.C. machines, IEEE Trans . Mag.on , , v.24, N2, 1988, p.787. 7. Aksenova E.N., Aksenov P.V., Kruglov V.S, Nikulenkov E.V., Zelensky G.K., Study of serviceability of superconducting REFERENCES z frequencyH 0 wire5 t a s. IEEE Trans. on , Mag. v.25, N3, 1989, p.2116. . 8 Dubot , FévrieP. s , RenarA. r , JC d 1. Klimenko E.Yu., Martovetsky N.N., Tavergnier JP.,, Goye Ky , HoanJ. ra Gi g Novikov S.I., Stability of superconducting NbTi wires with ultra-fine filaments for 50-60 wires with real transition characteristic, z useH : influenc e filamenth f o e t diameter Applied superconductivity in electroenergetics upon losses, IEEE Trans, on Mag., v.21, N2, d eiectrotechnicsan , CMEA Press, Moscow, 1985, p.177. 1986 161. p , . 9. Mints R.G., Rakhmanov A.L., 2. Mints R.G., Rakhmanov, A.L. Current-carrying capacity of twisted Instabilities in superconductors, Nauka, multifilamentary superconducting composites. Moscow, 1984. J.Phys.D. 1988, v20, N5, p.826. . 3 Martovetsky N.N., Some aspectf o s 10. Dorofejev G.L., Klimenko E.Yu., modern theor f applieo y d superconductivity, Soboleva N.M., Preprint IAE-3150, Moscow IEEE Trans, on Mag., v.25, N3, 1989 p.1692. 1979. 4. Klimenko E.Yu., Kozytsyn V.E., 11. E.Yu. Klimenko, N.N. Martovetsky, Martovetsky N.N., Novikov S.I., Experimental S.I. Novikov, StabilitC wireS n i fasf so yt verificatio f o RTC-stabilitn y theoryN DA , changing fields, Superconductivity: Physics, (Sov. Physics Doklady) 1987, v.292, N5, p.1119. Chemistry, Technics, published)e 1989b o ,(t . 148 RATIONAL DESIGN OF HIGH-CURRENT CABLE-IN-CONDUIT SUPERCONDUCTORS* L. DRESNER Oak Ridge National Laboratory, Oak Ridge, Tennessee, United States of America Abstract Cable-in-conduit superconductor e composesar cabla f do f man eo y fine com- posite strands encase stronga n di , protective jacket, with helium coolant filline gth interstice e cableth f so . Becaus hige th hf eo degre f subdivisioeo e composth f no - ite and its consequent large cooled surface, such conductors are capable of stable operatio t quitna e high current densities. The designer of such conductors is frequently given the field at the conductor overalthe land current densit askeyand specifdto remaininthe y g variablethe sof conductor (e.g., the strand diameter, the hydraulic path length, the void fraction e Cu/S e cableth o th fd Can , ratio). This paper outline rationaa s l procedurr efo determinin e mosgth t problematic variables compositioo tw e th , n variables that determin proportione eth copperf so , superconductor heliud cable ,an th men i space. All other variables of the conductor are assumed known. Two thermodynamic stateheliue th f so m coolant, supercritical He- 1-atd Ian m He-II (superfluid helium), are considered. For these states of helium, hydrodynamic phenomena exist that add to the stability of the superconductor. The allowed com- positions with heliu supercriticae th mn i l stat limitee ear threy db e constraints) (1 : that the stability margin be single valued, (2) that the quench pressure not exceed some preset value, and (3) that there be sufficient superconductor to carry the trans- port current. These three constraints defin allowen a e d regiocompositioe th f no n plane (variables: fraction of strands in the cable space and the fraction of copper in strands)e th . Wit heliue superfluie hth th mn i d state stabilitye th , margi singls ni e valued constraina t bu , composition o t n involving stability arises fro Kapitze mth a interfacial resistance. Thus thin ,i s case, too, ther three ear e constraints eacr Fo .h stat heliumf eo computea , r program plot allowee th s d regiocompositioe th f no n plan drawd ean s contour stabilite th f so y pressuree margith d nan compositio A . n desiree oth f d conducto then chosee rca n b n rationally. Two examples, an 8-T fusion magnet and a 15-T detector magnet, are given and are discussed in detail. * Research sponsored by the Office of Fusion Energy, United States Department of Energy, under contract DE-AC05-84OR21400 with Martin Marietta Energy Systems, Inc. 149 INTRODUCTION Cable-in-conduit superconductor e composesar cabla f do f man eo y fine com- posite strands encased in a strong, protective jacket; the interstices of the cable are filled with helium. Because of the high degree of subdivision of the composite and its consequent large cooled surface, such conductors are capable of stable operation at quite high current densities. e designeTh sucf o r h conductor frequentls si y giveconductoe th fiele nt th a d r overale th d lan current densit asked yan specifo dt remainine th y g variablee th f so conductor (e.g., the wire diameter, the hydraulic path length, the void fraction of the Cu/Se cableth d C an ,variablese ratio)th f O .mose th , t problematio tw e th e car composition variables that determin proportione eth f copperso , superconductor, and helium in the cable space. In this paper, all the variables of the conductor are assumed known except the two composition variables, and a rational procedure is outline their dfo r determination. Two thermodynamic states of the helium coolant are considered in this paper: supercritical He-I and 1-atm He-II (superfluid helium). For both of these states of helium, hydrodynamic phenomena exist that contribut hige th h o et stabilit e th f yo superconductor. When the helium is in the supercritical state, it undergoes strong heating-induced flow transients during recovery that augment heat transfer from metal to helium and greatly improve stability. When the helium is in the superfluid state e higth , h heat transfer cause Gorter-Melliny db k counterflow lead vero t s y high stability. A quantitative measure of the stability of a cable-in-conduit conductor is the so-called stability margin, i.e., the largest, sudden, uniform energy deposition in the strands still allowing recover superconductine th f yo g state. Whe e heliunth ms i in the supercritical state, the stability margin is sometimes multivalued. That is, as the energy deposition is progressively increased in a series of experiments, the conductor sometimes exhibits the double sequence of responses recover, quench, recover, quench rather than the single sequence recover, quench. Which response occurs depends strongl currene th n yo t density semiempiricaA . l correlation existo st calculate the limiting current density that avoids multiple stability and its attendant reductio stabilitn ni y margin. Another limitatio conductoe th n no r desig tha s ni maximue th t m quench pres- sure followin e worsgth te sudde casth f eo n normalizatio n entira f no e hydraulic path shoul t exceedno d some preset limit formulA . a computatioe existth r sfo f no this quench pressure. A third, obvious limitation is that there should be enough superconducto carro rt entire yth e operating current. These three limitations define an allowed region of the composition plane (vari- ables: fractio f strand o ne cabl th e fraction i s th spac d f coppee o nan e th n i r strands) computeA . r program plots this regio d drawnan e s th contourf o t i n i s 150 stability margin and the maximum quench pressure. The program also identifies e pointth f maximuo s m stabilit minimud yan m quench pressure. With this plot before him, the designer is able to choose rationally the composition of the desired conductor. When the helium is in the superfluid state, the stability margin is single valued. A theory computatioe existth r sfo stabilite th f no y margin; this theory indicates that ther constraina s ei stabilitn o t y arising fro e Kapitzmth a interfacial resis- tance that limits the possible conductor compositions. Thus in this case, too, there are three constraints. A computer program again plots the allowed region of the composition plane and plots contours of stability margin and pressure. Two examples are given of conductor design carried out with the procedure outlined above. One is a 4.0-K NbTi conductor destined for use in a fusion magnet at a maximum field of 8 T. The second is a 1.8-K NbsSn conductor intended for use in a detector magnet at 15 T. PHENOMENOLOG CABLE-IN-CONDUIF YO T SUPERCONDUCTORS As mentioned earlier, a cable-in-conduit conductor is one in which the super- conductor is contained in a braided cable that is enclosed in a protective jacket. Helium, usually supercritical but also possibly superfluid, fills the interstices of the cable strande Th . s composin cable gmadth e matriea ar f eo x material (usually cop- per but sometimes aluminum) in which are embedded many fine filaments of pure superconductor. Such cabled conductors contrast sharply in form with pool-cooled conductors in which the matrix and superconducting filaments form a solid bar that is immersed in a pool of boiling helium. mose Th t important propert superconductorf yo theis si r abilit recoveo yt e th r superconducting state afte suddea r n heat input. Such inputs typically arise from conductor motion causeLorente th y db z force t othebu , r cause possible sar e (e.g., plasma disruption tokamaks)n si heae th tf I inpu. largs i t e enough causy ma e t i , the superconductor to become resistive. Then, because the conductor is carrying current t producei , s mor well-enougs ei heatt i f I . h heliume cooley th y ma db t i , recover the superconducting state. If not, the supply of current to the conductor mus interruptede b t conductoe th destroyede r b o , y rma . Whe conductoe nth r fails cooo t l dow recoved nan e superconductinth r g state sais i "quench.o t dt i , e Th " recovery of pool-cooled conductors can be guaranteed by supplying enough copper time0 2 amoune (ofteo sth t 0 1 f nsuperconductor) o t . Such conductor callee ar s d "cryostable." Becaus e largth f eeo amoun f coppeo t r they contain, their average current density is low. For example, the pool-cooled conductors of the International Energy Agency's Large Coil Task had current densities over the winding pack of about 2.6 kA/cm2 [1]. 151 Cable-in-conduit conductors contain less copper than cryostable pool-cooled conductor accordingld an s y operat t higheea r current densities than pool-cooled conductors. For example, the Westinghouse coil of the Large Coil Task, which was wound with a cable-in-conduit conductor, had a current density of 4.3 kA/cm2 over the conductor area. The advantage of higher current density is offset by the disadvantage that cable-in-conduit conductor t cryostableno e sar reaso e Th . n they thiss i t : no recover e ar y fro msuddea n heat input takes only ten f millisecondsso , whereas the helium often resides in the conductor for minutes. Therefore, the helium inventory available for promoting recovery is limited. A large-enough heat inpu n thuca ts quenc e conductorhth n cryostablI . e pool-cooled conductorsn i , contrast, the helium inventory available for recovery is effectively infinite because helium vaporize contacy db t wit e conductohth immediatels i r y replace coly db d liquid helium fro poole mth . (However windine th f i , g heliupace th tighs ki d man t passages are small, vapor locking of these passages may occur, and the conductor fai y recover.o t l ma ) Small heat inputs do not quench cable-in-conduit conductors; large ones do. The largest, sudden, uniform heat input to the strands after which recovery of the superconducting state is still possible is called the "stability margin." As noted earlier, the stability margin is a quantitative measure of the stability of the su- perconductor. Cable-in-conduit conductor usable applicatioy sar an n ei whicn ni h the stability margin is larger than any heat inputs to the conductor. The stability margin is limited by the enthalpy difference of the helium between ambient temper- current-sharine th atur d ean g threshold temperatur superconductore th f eo . (The current-sharing threshold temperatur temperature th s ei t whicea supercone hth - ductor is no longer able to carry all the current and some current spills over into the matrix.) Not all of this enthalpy is available, because the Joule heat produced during recovery mus subtractee b t d fro. mit The stability margin of some cable-in-conduit conductors cooled with supercrit- ical heliu bees mha n measure experimentn di whicn si conductoe hth exposes rwa d to a succession of heat inputs, each of which was larger than its predecessor. In experiments done at the Oak Ridge National Laboratory (ORNL) [2], we expected n eventuaa l switce outcom th e experiment n th hi f o e s from recover quencho t y . This was usually what happened, but in some experiments a double switch was observed sequence ; thath , is t outcomef eo recoverys swa , quench, recovery, quench. date th an I from ORNL show example r Fign ni fo , 1 . , whe transpore nth t current , heaA t0 inpute metai38 s th o lt s strands less tha mJ/cm0 n5 3 lea recovo dt - ery; input mJ/cm0 s 9 betwee d an 3 lea0 nquenching5 o dt ; inputs betweed an 0 n9 3 300 mJ/cm 3 again lead to recovery; and inputs above 300 mJ/cm again lead to quenching e resultTh .f mano s y experiments don t ORNa e e summariseLar n di three-dimensionae th l sketc Fign hi . Plotte 2 vertica.e th n do l stabilite scalth s ei y 152 i SINGLT b N E TRIPLEX ijiw = 1.0 mm, Pats =. 5.0 atm Th =16.7ms,vHe = 0 Lsampla = 3-8 m, B = 6.0 T 20 I I 340 360 380 400 420 440 U, A Fig. 1. Experimental data from the Oak Ridge National Laboratory [2] showing the double sequence of outcome» recovery, quench, recovery, quench. CURRENT, I Fig. 2. A three-dimensional sketch summarizing the results of many experiments done at the Oak Ridge National Laboratory. The fold in the surface is connected with the double sequence of outcomes recovery, quench, recovery, quench. 153 margin plotte d horizontae ,an th n do imposele currene scaleth th d e sar an td helium flow velocity. (The imposed flow is created by the refrigerator or by pumps.) The surface that represents the stability margin is folded, and the fold is connected with the double switch in outcomes. Inside the shaded region under the fold, the sequence of outcomes as the heat input is raised is recovery, quench, recovery, quench. Outside this region, the se- quenc recoverys ei , quench. This unusual situatio causes nheating-inducee i th y db d flow transients that occur during recovery [2]. It turns out that, on the upper sheet of the folded surface, the stability margin is approximately equal to the available enthalpy of the helium, whereas, on the lower sheet, it is much smaller. Accordingly, worts i t i h whil desigo et n cable-in-conduit conductor operato st t currentea s less than the current at point B in Fig. 2. In this way, we can fully exploit the available enthalpy of the helium. currene Th t poina t whic, B t calles hi "limitine dth g current, determinee b n "ca d for any conductor from the results of the experiments already carried out and a scaling rule given in Ref. [3]. The validity of the scaling rule itself has been tested experimentall rule th e d corroboratedan ] y[4 . The information summarized previously enables us to determine by calculation stabilite th y margi proposef no d cable-in-conduit conductors coole supercriticay db l heliudesigo t d mnan conductor t likelsno thaquencho ye t ar t . But thef ,i y should quench muse w ,sure b t e that they wilt suffeno l r damage. When cable-in-conduit conductors quench exposee b , they higa o yma dht internal pressure, which they mus designee b t withstando dt worse th n tI .entirn casa f eo e hydraulic path going normal all at once, the maximum quench pressure can be calculated by using a simple formula given in Ref. [5]. This formula has been tested for quench pressures between 4 and 200 atm [5, 6] and found to be accurate. stabilite Th y margi cable-in-conduif no t superconductors cooled with superfluid He-II is denned slightly differently from that for superconductors cooled with su- percritical helium. It is defined as the largest, sudden, uniform energy deposition in the strands still allowing recovery of the superconducting state without the he- lium temperature being raised abov He-II-He-e eth I transition temperaturee Th . restrictio temperature th n no heliue th f emo necessars i alloo yt w computatiof no the stability margin [7, 8]. Conductors cooled with He-II may actually be stable against larger perturbations. When define thin di s way stabilitye th , margi singls ni e valued; thus limie th , - tatioallowee th f no dcompositioe areth n ai n plane havin wit o limitine d hth o g t g current no longer applies. However, a new limitation, having to do with the Kapitza interfacial resistance, take s placenormal-state sit th f I . e Joule heat flux froe mth conducto heliue th o rmt largs i e enough temperature ,th e difference acros phase sth e boundary induced by the Kapitza resistance will be large enough to keep the metal 154 temperature above the current-sharing threshold. Recovery is then impossible. So again ther three ear e limitations that defin allowen a e d regiocompositioe th f no n plane that can be plotted together with contours of the stability margin. stabilite Th f cable-in-conduiyo t superconductors coole superfluiy db d helium does not depend on heating-induced flow. In fact, the coefficient of thermal ex- pansion of superfluid helium is much smaller than that of normal helium [9]; thus, e verb theryy littlema e induce de superfluid th flo n wi . Moreover, ther s gooei d evidence that the heat transfer coefficient to superfluid helium is independent of the cross-flow velocity [10]. Instead, stability in He-II depends on heat transfer highle byth y efficient Gorter-Mellink counterflow. Therefore, tight confinemenf o t heliue th m bathin r e strandsupercritical t gnecessarth fo no s i s si t i s ya helium, where it promotes axial heating-induced flow over the strands. For example, when coolante He-Ith s i Iconductoe jackee th th , f o t r migh perforatee b t allowed dan d communicato t e wit outsidn ha e space, thereby providing pressure relief duringa quench. Thus, quench pressurcriticaa t no s lei limitatio r cable-in-conduinfo - su t perconductors cooled with superfluid He-II. The cable-in-conduit superconductor is still advantageous because of the fine subdivision of the strands which provides larga e cooled surface jackete Th . , whether closed, perforateda n i r ope o ,s n(a U-shaped channel), provides mechanical protectio strandse th r nfo . Furthermore, t servei s co-wouna s d structure that prevent buildue sth e Lorent th f po z forcn o e the cable. EXAMPL NbTi/CT 8- N A u: E 1 CONDUCTO R RFO A FUSION MAGNET desige Th nfirse goath t n examplli achievo t s ei ecurrena t density ove cable rth e NbTi/Ca n i T 8 ud cable-in-conduian spackA/cmK 0 2 0 f e4. o t a 2 t conductore Th . critical curren take134s e b wa 0o tnt densitT A/mm 8 d purn yan ,i 2 K e 0 NbT4. t a i following the data reported by Larbalestier et al. [11]. According to these authors, this current density was typical of the best available industrial material at the time (1986) they wrote their article residuae Th . l resistivity coppee rati th takes f oo rwa n to be 100, and the strand diameter was taken to be 0.7 mm, which is the strand diamete Westinghouse th f ro Large e th coi f eo l Coil Task ambiene Th . t pressurf eo the heliu atm5 lengte take s e hydraulia Th b .m f wa ho o nt c pattakes e b hwa o nt quence Th . 2h0m pressur limites moro ewa n o edt thaatm0 n50 . Figure 3 shows the allowed area of the composition plane under the constraints. Three sets of contours cross this area, one set for the stability margin, one set for ratie quencth e transporf r oth o fo t h se pressure criticao e t t on d l currentan , e Th . largest stability margin attainable is 123 mJ/cm3; the smallest quench pressure attainabl atm8 25 s .ei Note that these extremes refe differeno t r t conductors. 155 1.0 0.9 0.8 258 atm 31 mj/cm3 f 0.7 '62 mJ/cm3 0.6 93mJ/cm3 0.5 0.4 0.7 0.8 0.9 1.0 allowee Th Fig . d3 compositioe . areth f ao n e NbTi/Cpianth r efo u cable-in- condui. T) t 8 conducto d an kA/cmK 0 0 r(2 4. t a 2 We deem a stability margin of 123 mJ/cm3 to be adequate mainly on the ex- perimental evidence of Lue and Miller [12], who operated an experimental NbTi cable-in-conduit coil at 7.7 T and 4.2 K and at 8.1 T and 3.9 K. The measured stability margin of the conductor was less than 50 mJ/cm3, and the magnet never quenched spontaneously Lue-Millee Th . r 36-strand rathea cabl s ewa r loose eon (void fractio 43%f no thud )an s presumabl greatea d yha r potentia stranr fo l- dmo tion than the tighter cables shown in Fig. 3 (void fractions of 10-20%). It would therefore appear tha stabilitya t margin aroun mJ/cm0 d10 3 shoul sufficiene db o t t ensure stable operation. The quench pressures in the allowed area are rather high. For example, the conducto r Cu/S a voidwit % d h20 san C rati 1.8f o o whicr , fo stabilite hth y margin is about 95 mJ/cm, 3 has a maximum quench pressure of roughly 350 atm. This figure corresponds to a 20-m hydraulic path length, which is rather small for a fusion magnet and means added complexity in the plumbing. So it appears that, for the rather high current density we have chosen to aim at, the limiting factor is quench pressure t stabilityno , . A variation of the ordinary cable-in-conduit conductor that may allow reduction e quencoth f h pressur s showei e inne FigTh n ni r . condui.46 s perforatei t d an d allows the cable space to communicate with the four open helium spaces in the corners. These open space intendee ar s relievo dt e quenceth h pressure. They hav e disadvantageth e tha innee tth therd conduiyan t tak spacp eu e thath n i t 156 conducto devotes f Figi o r o 4 . carryino dt g current conductore th f I . f Figsso o 4 . and 46 are to carry the same total current, the current density over the cable space in the conductor of Fig. 46 must be higher than that of the conductor of Fig. 4a. quantifo T y these considerations I too, physicae kth l dimension plaie th f nso cable-in-conduit conducto Westinghouse those b th f o f Figeo t o r a 4 . e conductor (i.e., 20.8- by 20.8-mm exterior dimensions with a jacket thickness of 1.75 mm). The thickness of the inner jacket I took to be 0.5 mm and assumed it to have a circular shape just tangen interioe square th th o t tf ro e jacket helo T .p reduce eth Joule heat production I assume, innee dth r jackesame made b th f o eet o t copper strande uses th a n di s (RR R100)= orden I . carro t r same yth e currene th s a t conducto currene Figf th o r , .4a t density ove cable th r e conductoe spacth n ei f o r Fig. 46 must be 28.7 kA/cm2. plaie Th n cable-in-conduit conducto length-to-diametea s Figf ha ro a .4 r (L/D) ratio of about 5 x 104 (the hydraulic diameter is roughly 0.2 mm for 20% voids, and the half-length of a flow path is 10 m). For such a large L/D ratio, the inertial terms in the momentum equation can be neglected; the pressure gradient is expended overcominn i g wall friction acceleratinn i t no , fluide gth . This simplifications a , explained in Ref. [5], leads to a simple formula for the maximum quench pressure that has been used to calculate the quench pressure for the conductor of Fig. 4a. experimente Th s that corroborated this formula were carrie t witratioD dou hL/ s of 1.23 x 10s and 6.16 x 104 [5]. The hydraulic diameter of the corner spaces in the conductor of Fig. 46 is onlw ratiD yno L/ s 4810oi e th s thismal o I .e o to s us allo o , t l e wth 2.0mm 8 frictionae oth f l theory previously described conditioe Th ?validite e th r th nf fo y o frictional theor thas yi tFannine , fL/Dwherth 1 s i > ; ge/ friction e factoth f o r flow path. For typical Reynolds numbers ranging from 10s to 108, the smooth-tube friction factor ranges from 2.5 to 4.0 x 10~; 3 thus, L/D should exceed the 250 to 400 range in order for the frictional theory to be valid. Hence, even for the conductor frictionae ofth Fig.46e us e l w ,theor calculato yt quence eth h pressure. PLAIN CABLE-IN-CONDUIT CABLE-IN-DOUBLE CONDUIT TUBE IN CABLE-IN-CONDUIT Fig. 4. A sketch of a plain cable-in-conduit conductor and two variations that may allow reduction of the quench pressure. 157 Figure 5 shows the allowed area of the composition plane together with the contours of stability margin, quench pressure, and fraction of critical current. The cable e evear s n tighter than before, having voi e r lessdo Th fraction.% 15 f o s stability-optimized conducto stabilita s rha y margi 56.f no 6 mJ/cm maximua d an s m quenc appeahe pressurw o atmhav4 o S rt .14 ef eachieveo appreciabln da e reduction in quench pressure at the expense of some reduction in stability. 0.8 104 atm 0.7 - 0.75 0.6 0.5 56.6 mJ/cm3 f 0.4 0.3 0.2 0.1 0.8 0.9 1.0 1.1 alloweThe Figd5. compositio .arethe aof n cablplana doublthe ein efor e conduit (Fig. 46). An even more convenient dispositio e opeth f nno helium spac s showi e n ni . WheFig4c e diamete. n th e perforate th f o r d inner copper tub s reducedei e th , volum t eoccupiei made b n e sca smaller . Therefore currene th , t densit cable th en yi space need not be as high as that for the conductor of Fig. 46. For example, if the outethickness it t coppee r th bu diamete f sm o r m tub5 reduces keps 0. ri ei t a t o dt 7 mm, the current density in the cable space must be 22.5 kA/cm 2 to match the current in the conductor of Fig. 4a. The L/D ratio is 1667, which is probably enough to continue usin e limitingth g frictional theor computo yt e quenceth h pressure. Figure 6 shows the allowed area of the composition plane together with the contours of stability margin, quench pressure, and fraction of critical current. The void 158 1.0 0.9 0.8 72.3 atm 0.75 0.7 0.50 0.6 f 0.5 97.2 mj/cm3 179 atm 0.4 0.3 0.2 0.1 0.7 0.8 0.9 1.0 1.1 fco allowee Th Fig . d6 .compositio e areth f ao n tube-in-cable planth r efo e config- uration (Fig. 4c). fraction is 20% or less; the stability-optimized conductor has a stability margin of 97.2 mJ/cm3 and a quench pressure of 109 atm. So with this arrangement we have increased the stability margin and reduced the quench pressure modestly. According to Lue et al. [13], the current density over the entire winding pack coul 35-65e db tha%f o t ove cable rth e space precise ,th e value dependin size th e n go and shape of the cable-in-conduit conductor and the amount of external insulation wrappe basie d th arounthi f sn o sO estimate. dit targea , t kA/cmvalu0 1 f eo 2 over entire th e winding pack seem reasonablsa e 8-Ttargen a r ,fo t4.0- K NbTi cable-in- conduit conductor wisI . emphasizo ht e agai t thina s point that suc hconductoa r should operate stably in spite of its high current density. EXAMPLE 2: A 15-T Nb3Sn/Cu CONDUCTOR FOR A HIGH-FIELD DIPOLE MAGNET desige Th nsecone goath n i ld exampl achievo t s ei largs ea ecurrena t density possibls a e ove cable rth e space consistent wit stabilitha y margimJ/cm0 10 f no t 3a 1.8 K and 15 T in a Nb3Sn cable-in-conduit conductor. The critical current density 159 take370s e b wa 0o nt A/mmT 5 1 d , 2 followinpuran n i K e 8 Nbdate 1. g Sth t naa 3 reported by Foner et al. [14] for (Nb-4 at. % Ta)3Sn. The residual resistivity ratio coppee o th f agais strane th rwa n d dtake100e an b diamete , o nt agais rwa n taken to be 0.7 mm. The ambient pressure of the helium, which is now in the superfluid state, was taken to be 1 atm. As mentioned earlier stabilite th , cablea f yo d superconductor cooled wit- hsu perfluid helium depends heating-inducen o t no , dheliue floth f wo m ove surfaces rit , hean o t t transfe bu highle th y yb r efficient Gorter-Mellink counterflow. Thus, tight confinement of the helium is not necessary. If we start by assuming the conduit to be perforated and to communicate hydraulically with a large plenum allowing pressure relief, we can ignore the constraint of quench pressure. Figure 7 shows the allowed compositioe areth f ao n plane along with contour stabilitf so y margi fractiod nan n of critical current for a current density over the cable space of 45 kA/cm2. The stability-optimized conducto stabilita has r y margi 98.nof 6 mJ/cm basithe s 3On . of this estimate, target value f 15-3so 0 kA/cm2 ove entire th r e winding pack seem a reasonable target for a 15-T, 1.8-K Nb3Sn cabled conductor. Again I emphasize that such a conductor should operate stably in spite of its high current density. 1.0 l T l T I 0.9 0.8 98.6 mj/cmji 0.7 0.6 f 0.5 0.25 - 0.4 0.3 0.2 0.1 0.5 0.6 0.7 0.8 0.9 1.0 1.1 U e alloweTh . 7 Figde compositio are.th f ao n e (Nb—planth . r at e4fo Ta)jSn/Cu cable-in-condui. T) 5 t 1 conducto d an kA/cmK 5 8 (4 r 1. t a 2 conductorIfthe s surveye Figdin .wer7 e confine imperviouan in d s conduit wit hhydraulia c quence patth , h m hlengt 0 pressur2 f ho e they would e suffeth n i r event of the simultaneous quench of a whole hydraulic path would be in excess of 800 atm. (Thi vere b s y t numbe accuratno y ma r e extensioowine th e o gt th f no theor outsid ] f Refyo [5 . e rangeth r whics beeefo ha nt hi checked experimentally, 160 but there is no question that the quench pressure would be very high.) Therefore, for a plain cable-in-conduit conductor, a current density over the cable space of kA/cm5 4 largeo to s i 2. However reduce w currenf e i , eth t density ove cable th r e space to 20 kA/cm2, we see (from Fig. 8) that conductors with acceptable quench pressures existallowee Th . d regioe compositioth f no n plan rathes ei r larget bu , only the lenticular region above and to the left of the 275-mJ/cm3 contour is really interesting. This is the region of conductors with quench pressures around 200 atm and stability margins around 300 mJ/cm. 3 These conductors have Cu/SC ratios in the neighborhood of 3 and void fractions of roughly 60%. If such cables are believed looseo to e ,b movino t voio gt d fraction woul% 30 df increasso quence eth h pressure to roughly 300 atm and reduce the stability margin to roughly 150 mJ/cm. 3 Some improvemen s doubtlesi t s possibl usiny e b e configurationgth , 4c d f Figsan o s 6 4 . but the best solution to the quench pressure problem in the superfluid case is to abando idee nth f tigh ao t confinemen heliume th f o t . This will, however, vitiate the advantage cable-in-conduit conductors have of being completely surrounded by insulatio thud nbeinf an s o g abl withstano et largda e voltage. 1.0 l l l l l l l 0.9 0.8 0.7 0.6 f 0.5 0.4 0.3 0.2 0.1 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0 U e alloweTh Fig. d8 compositioe .areth f ao n (Nb-e planth % r . 4efo at Ta)sSn/Cu cable-in-conduit conducto lowea t a r r current density than thaf o t . T) 5 1 d an kA/cmK 0 Fig8 (2 1. 7 . t a 2 CONCLUDING REMARKS have Asw e seen, cable-in-conduit superconductors shoul capable db f stableo e operatio hignat h current densities, which means smalle thereforand r e cheaper magnets. Furthermore, smaller magnets may mean smaller machines overall, so that the savings won by the high current density may be magnified beyond just the savingmagnete th n so s themselves e capabilitTh . r withstandinyfo g high voltage 161 (>10 kV) due to their unbroken exterior insulation is another advantage of these conductors. In addition, because the jacket acts as co-wound structure and because the conductors may be potted in epoxy, the winding pack can be quite rigid. Finally, the helium inventory is lower than that in pool-cooled magnets. Taken all together, these advantages make cable-in-conduit superconductors an attractive alternative for superconducting magnets. The preceding rational design procedure, which is fruie basif th o t c research carrie t ove lase dou rth t decade, eliminate guesswore sth k fro desige mth thesf no e conductors. REFERENCES . "Th1 LargA eIE e Coil Task, . BeardS . "D . Shimamoto S , . VecseG d yan , (Eds.), Fusion Engineering and Design 7 (1,2): 1-230 (1988). 2. "Stabilit Cable-in-Conduif yo t Superconductors, . Miller. LueR W . d J ,. "J an , L. Dresner . ApplJ , . Phys (1)1 5 . : 772-783 (1980). . "Parametri3 c Stude Stabilitth f yo y Margi f Cable-in-Conduino t Superconduc- tors: Theory," Lawrence Dresner, IEEE Trans. Magn. MAG-17 (1): 753-756 (1981). . "Parametri4 c Stude Stabilitth f yo y Margi f Cable-in-Conduino t Superconduc- . MillerR tors . J :,d IEEExperiment,an e E Lu Trans . W . ."J Magn. MAG-17 (1): 757-760 (1981). . "Pressur5 e Rise Durin e Quencgth Superconductina f ho g Magnet Using Inter- nally Cooled Conductors," J. R. Miller, L. Dresner, J. W. Lue, S. S. Shen, and H. T. Yeh, Proceedings of the Eighth International Cryogenic Engineering Con- ference (ICEC-8), Genoa, Italy, June 3-6, 1980, IPC Science and Technology Press, Guildford, Surrey, UK, C. Rizzuto (Ed.), pp. 321-329. 6. "Simulation of the Quenching of an Internally Cooled Superconducting Mag- net," J. W. Lue, J. R. Miller, L. Dresner, and S. S. Shen, Proceedings of the Ninth International Cryogenic Engineering Conference (ICEC-9), Kobe, Japan, May 11-14, 1982, Butterworths, Guildford, . YasukochSurreyK , UK , i (Ed.), pp. 814-818. RapidA " . 7 , Semiempirical Metho f Calculatindo e Stabilitgth y Margin- Su f so perconductors Cooled with Sub-Cooled He-II," Lawrence Dresner, IEEE Trans. Magn. MAG-23 (2): 918-921 (1987). 8. "The Stability Margi f Superconductoro n s Cooled with Subcooled He-II," Lawrence Dresner, Proceedings of the Japan-US Seminar on Basic Mechanisms of Helium Heat Transfe Related an r d Influenc Stabilitn eo Superconductinf yo g Magnets, Fukuoka, Japan, August 29-Septembe , 19882 r , paper 4A-2e b o t , published in Cryogenics. 9. "The Propertie f Liquiso Solid dan d Helium, . Wilks"J , Clarendon Press- Ox , ford, 1967, Chapte. 2 r 162 10. "Heat Transfer to Forced Flow Helium II," H. P. Kramer, Proceedings of the Twelfth International Cryogenic Engineering Conference (ICEC-12), Southamp- ton, UK, July 12-15,1988, Butterworths, Guildford, Surrey, UK, R. G. Scurlock and C. A. Bailey (Eds.), pp. 299-304. 11. "New Developments in Niobium-Titanium Superconductors," D. C. Lar- balestier, P. J. Lee, Li Chengren, and W. H. Warnes, Proceedings of the ICFA Worksho n Superconductino p g Magnet d Cryogenicsan s y 12-17Ma , , 1986, Brookhaven National Laboratory. 12. J. R. Miller, J. W. Lue, R. L. Brown, and W. J. Kenney, IEEE Trans. Magn. MAG-17: 2250 (1981). 13. "High-Field, High-Current-Density, Stable Superconducting Magnetics for Fu- sion Machines," J. W. Lue, L. Dresner, and M. S. Lubell, Proceedings of the Seventh International Workshop on Stellerators, Oak Ridge, TN, April 10-14, 1989. 14. "High Field Properties of Multifilamentary (Nb-4 at. % Ta)3Sn," S. Foner, E. J. McNiff . Ozeryanski, . Jr.M Schwall E . . G , R d , IEEan , E Trans. Magn. MAG-23 (2): 984-987 (1987). Next page(s) left3 blan16 k DEVELOPMEN PULSEF TO D SUPERCONDUCTING MAGNETS AND THEIR OPERATING CHARACTERISTICS T. ONISHI Electrotechnical Laboratory, Tsukuba City, Ibaraki, Japan Abstract The large scale pulsed magnets named PSM-1 and PSM-2, respectively which were both wound with NbTi wire having almost the same characteris- tics were developed. Those stored energies are 3 MJ and 4 MJ at rating curren f 550o t, respectively 0A presene th n I t .paper , repetitive pulse operations of the PSM-1 will be described firstly. It showed stable per- formances against 1000 time pulsf so e operation lop/It a s % ce 0.7 th 5n o loa 6 T/secd 3. lind .an e The pool-cooled Nb3Sn pulsed magnets named HPSH-1 and HPSM-2, resepec- tively were also developed by a wind-and-react approach. Their designed , respectivelT field0 1 d san wertheit T a y 8 e r rating currents HPSe Th M. -1 was successfully operated up to 8 T in a background field of 4.9 T at 3.4 T/sec. The turn to turn transmission of electromagnetic force was analysed, and the ac loss was also measured. The electromagnetic and mechanical properties are compared with those of HPSM-2. INTRODUCTION Many pulsed superconducting magnets have beenr developefo r fa o s d application to fusion reactors, pulsed energy storage equipments, and so on1'3. It is essentially important to improve the reliability withstan- ding repetitive cycle f pulso s e operation n ordei s o makt rr efo thet fi m practical use. However, repetitive operation f magnetso s havt beeno en studied yet from a viewpoint of fatigue. e recenTh t desig f Tokamao n k fusion reactors requires high field ohmic heating coils producing more than 10 T. Although Nb3Sn wire will be indispensabl r sucfo e h magnets s difficuli t i , o pulseo applt t t i yd magnets because of the brittle characteristics. e presenIth n t paper e dynamith , c propertie e pulseth f do s magnet wound with NbTi wire will be described firstly, and then the electromagnetic and mechanical properties of the two wind-and-react Nb3Sn pulsed magnets will be discussed. NbTi PULSED MAGNET (1) Structure and parameters of PSM-1 e conductoTh ra multi-stag uses wa d e twisted cabl s showa e n Fig.li n , e criticaanth d l curren s mor Ä Normewa t e tha. xT ntap6 800t ea 0A finae thickwouns m th m widm lwa n m o 5 d cabl) e ,7 0. ( e spiralln a t a y interval of 8 mm. The insulated cable was co-wound in a double pancake fashion together with the insulated stainless steel strip having approxi- mately the same width as the cable. 165 CuNi-ctad strand basic strand first-level able vsecond-level able final cable Fig.l Schemati f cablo c e cross section. e magneTh s fabricatewa t 2 doubl 1 usiny b de eth g pancake coilse Th . spacers betwee e coil th e 8-mn sar m thick GFRP. Liquid-bubble separating plates with the slope of 3° from horizontal were inserted in the gap as shown in Fig.2. After the 12 double pancake coils were stacked , it was pre-compresse y aboub d0 tons 40 tthed ,an n fastene e rodsti .6 d1 wite th h The main parameters of the magnet are listed in Table 1. Three strain gauges were attached separating each other by 120° on the outer support band f eaco s h pancake coils. Thos e schematicallar e y shown in Fig.3. ( A-A ) Double Pancake Bubble-liquid c Bubble-liquid separating plate Spacer separating plate Pig.2 Illustration of liquid-bubble separating plate. 166 Table 1. Parameters of PSM-1 ST-1-3 AA_ ST-1-A1 Coil 12 double pancakes i 1 T r v Winding i.d. 400 •>- Winding o.d. 860 •• Winding height 625 •• Total turns 792 Inductance Henr2 0. y I 5500 A op 30.3 A/M2 J op, coi•l! B 6.6 T HH 1* 300m 400mm 922mm Fig.3 Positio f straio n n gauges. (2) Tests results and discussions e magneTh s firstlwa t y w ratoperate lo f chango ea t f magnetia o ed c fiel d generatean d e winding th a maximud n o t 550T a s m0A 6 fiel6. f o d without any quenches. Then, in pulse mode operations, about 6.5 T and 6 T were generated at 2.2 T/sec and 4 T/sec, respectively. These values were limited by the capacity of the power supply. The load line and operating currents are shown in Fig.4. In orde o investigatt r e fatiguth e e properties s thewa nt i ,operate d repititively in bipolar field sweep iodes by using the energy storage and transfer system4 which was composed of PSM-1 and PSM-2. A typical current waveform is shown in Fig.5. These operations were repeated and the total number of operations amounted to 1000 times in all by ragarding one operation as an operation from 0 to a peak current and then to 0. Repetitive Pulse Operation 0 I 2345678 6 tlme(s) Fig.4 Load line and operating Fig.5 Typical current waveform of current of PSM-1. a bypolar field sweep. 167 During the repetitive pulse operations, the strains of the windings and e totath l countE signa f o sl were measured. Those result e showar sn i n Figs.6 and 7. Any changes of strain did not occur up to 1000 times of pulse operations and that the ÂE counts gradually decreased5. The reduc- tion of the total ÂE counts means that the dominant part of the AE in the present magnet came froe wirth me motions. Therefore e rigidite th , th f o y windings is supposed to be improved during the repetitive pulse operations o 100ut p 0 time f operationso t leasta s maximuA . m strai f approximateo n - s obtainewa m pp d 0 whicl 20 ys wel wa hl consistent with calculation. ST&-1 200- E 3 O LU ..«-..*—-».< «-SB-11 -200 0 1 10 102 103 104 -300 0 10090 0 080 0 70 0 60 0 50 Pulse operation Times Number of Pu]se Operations Fig.7 Total AE counts vs pulse Fig.6 Strain s pulsv s e operations. operation times. Nb3Sn PULSED MAGNETS ) HPSM-(1 1 1-1 Cable structure The cabl composes wa e5 subcable 1 f o d n whici s Nb4 h 3Sn basic strands and 3 copper wires were twisted together. The Nb3Sn filament was about 2 n diameteri jwm e percentagTh . e strandth f coppeo n ei s abou.% wa r 6 3 t A critical current was about 4200 A at 8 T. The illustration of the final cabl s showi e n Fig.8i n orden I .o reduc t r coupline th e g current loss betwee e basith n c strands e stabilizinth , g coppee centeth rn i r wert pu e island of the strand and the surface sheath was designed to be turned into bronze after heat treatment. The stainless steel strips (0.25 ram in thickness, 8 mm in width) for internal reinforcement of the cable were coated with AlgO o reduct s e coupling curren m thicm t1 k 0. losswove e .Th n glass-fiber tape with high silica content was wrapped by three layers on 168 Strand (0.5mm) 0 Cu(0.5mm0) SUS Strip Coating) l XJUOUOC -1m 1m Fig.8 Illustratio f finao n l cable. the final cable at first, but it was found to become rauch brittle after heat treatment of around 700 °C under some compressional load, although it showed elasticity provided that it was heat treated under no load. There- fore e insulatoth , s replacewa r d with Normex tape (0.3 m thick8m , 11.m 2m o layerstw widd an e) whee windingth n s reassembledwa s . 1-2 Magnet construction The six double pancake coils were stacked axially with GFRP spacers (4 mm in thickness) interleaved between each double pancake coil. The number p percentagga e th d ofean spacer betwee0 3 e ar sn spacerf o s abou% i s 4 3 t e overalth l windin e feature ge th magne surfaceth f o f o tse On desig. s i n that the windings are not epoxyimpregnated and hence liquid helium will infiltrate well inte cableth o , leadin o bettet g r cooling. Aftee th r magnet was precompresed axially with about 36 ton, it was fastened with 8 stainless steel tie rods (10 mm in diameter) in the inner radius side of e windinge oute th 6 oneth 1 n rd si radiuan s s side A Young'. s moduluf o s the coil Wingdings before heat treatment against axial compression was a Stressleveobtaine n i e mor b a eo MP t d tha 0 l 32 ncorrespondin n a o t g electromagnetic forca ratin t a e g curren f themagneto t A photograp. f o h the completed magnet being set into a background field magnet is shown in e maiFig.9th nd parameter,an e listedar s n Tabli 6 . 2 e 1-3 Transmission of electromagnetic force Turn to turn transmission of electromagnetic force was analysed in a pancake coil in the center of the magnet for the case of an operating curren a backgroun f 251e calculaten o ti Th 0A . T d9 dfiel4. strainf o d s e outeoth f r suppor ts mad i ban e showt eI ar dclean Fig i n. r10 . froe th m result that the electromagnetic force is well transmitted from the inner turn to the outer one of the windings if a compressive Young's modulus is assumed to be a magnitude on the order of 200 GPa and the corresponding 169 Table 2. Parameters of HPSM-1 Winding l.d, 10m m 1 Winding o.d, 286 mm Winding Height 180 mm Nof Turo , n 252 2510 A 2 Jc,coil 38,1 A/mm Bm (BalQST 8 : 4.9T) Coil PancakW 6 e Coils Pig.9 Photograp f HPSM-1o h . 300 8T(4.9Bias) Llop:2510 A 200 c I t/5 100 2 0 1 6 2 0 1 6 10 2 2 0 1 E Compressive Young% Modulu) a MP s( Pig.10 Calculated strains of outer support ban f HPSM-1o d . straine supporth f o st band «ill amoun o about t0 ppm30 t . Howevern i , a magnitudcas 0 MPaf e orde o e10 e strain th ,f th o rn o e expecte ar s o t d be quite small, that is, the electromagnetic force will not be well transmitted from turn to turn. In the experiments, about 30 ppm of strain of the outer support band was obtained at 8 T (2510 A) in a background . T Bot 9 hfiel 4. experimenta f o d d analyticaan l l results indicate thaa t Young's modulus perpendicula e flatteneth o t r d surface cablth ef o e will be less than 100 MPa. 1-4 Test results The magnet was successfully charged at a rate of 3.4 T/sec up to the rating current of 2510 A in a background field of 4.9 T. The correspon- ding field was 8 T in the magnet bore. The load curve and short sample B- I characteristics are shown in Fig. 11. Although the windings had a fairly low rigidity, no remarkable wire motion was observed while it was charged and discharged at high dB/dT rate. The electromagnetic force was not necessarily transmitted sufficiently from the inner turn to the outer accordingld onan e y measured strains were considerably smaller than expected. 170 lop «c =I / 0.62 (Vertical) Bm 3 0123456789 10 B (T) Fig.1 loaA 1 d lin f o HPSM-1e . Ac losses were also measure n electricaa y b d l method e resulTh . s i t n comparisoshowi 2 n Fig1 i n. n with calculation, w fielIlo n d regione th , measured losses were roughl e one th e sams s th ya e calculated assumine th g basic strands to be independent electrically each other. However, they increase y approximatelb d e ordeon y f magnitudo r e compared with calcula- tio t highea n r field. Thi s supposei s d mainlo t y o reductiot e bdu e f o n contact resistance between strands by compression of the electromagnetic force. u i i i i i — — — —— | ii 1 1 ( i iJ — — — —— | 11 U1 1 i i i ir— — — : — : Exp. : o B, . : 0 3 bias 10 r + B,_ . : 4 . 9T | 1 : bias g ]] : "o 10 * : 9 CJ \ o ° ® ' 13 o w o Total '•. /; ° ^ 10 l r • ° çy/'^ H : 10° i o °y/ "" i : ' / / o/' 10 "' , ,^,,,,i/, . ,,,„., 10 ' 10 ' 10 S K I„ (A) Fig.1 c losA f 2HPSM-1so . 171 (2) HPSM-2 It consists of 10 double pancake coils which were fabricated by wind- and-react approach usin e flatteneth g d Nb3Sn cable eon similae th o t r shown in Fig.8. For the case of the present cable, a stainless steel wire subcable centes placeth th wa e f n o ri dn whic i e h four Nb3Sn wires and two copper wires were twisted around the stainless steel wire. The diameter of those wires are 0.5 ram. The filament diameter and twist pitch respectively, mm 0 1 d e percentagan Th m s abou. i p 3 t f coppeo e s aboui r t 34.8% in the strand cross-section. The final cable size is 10.9 mm by 3.0 mm and its critical current is more than 3000 A at 10 T. It was co-wound with the insulator and stainless steel strip (0.25 mm in thickness) coated with A1 e coil20Th 3. s were stacked with GFR Pn thicknessi spacer m m 2 ( s ) between each coil and fastened with tie rods under precompression of about maie 6 ton3 Th n. parameter listee ar s Tabln i d . e3 Table 3. Parameters of HPSM-2 Winding i.d. m m 0 10 Winding o.d. 302 mm Winding Height 256 mm No. of Turn 540 I op 2350 A J 49.1 A/mm2 c,coil B (Design) 10 T (Bias: 5.IT) Coil PancakW 0 1 e Coils 5 e T/seHPSM- 2. Th a ratd produs charge f o ct an 235wa o et a 2 0-A p u d ced 4.9 T by itself. A maximum field attained was 7.1 T in a background field of 2.95 T at present6. A maximum measured strains of the outer support band at 7.1 T was about 75 ppm. A maximum calculated strain amount o about s tt t wil10T a I 200 e allowabl.m b l 0pp e from viewpoinf o t NbsSn wire characteristics. Therefore, it is expected to be able to produce 10 T just from viewpoint of mechanical strength of the cable in the present magnet. As to ac loss, the measured one was fairly small and about 0.1 % of a total stored e magneenerg a cyclth n f i to ye o 235frot 0 A m(4.9T d )an 5 T/sec2. t .a 0 bac o t k CONCLUDING REMARKS o largTw e scale pulsed magnets wound with NbTi wire were developedd an , e dynamith c propertief theo me on (PSM-1 f o s s investigated)wa s wa t I . clarifie e stablb o t de against 1000 time f pulso s e operations, althougt i h was the pool-cooled pulsed magnets. Two wind-and-react Nb3Sn pulsed magnets which were not epoxy-impregated were fabricate d operatean d pulsa n i de e HPSM-modeTh s chargewa .1 d up to 8 T at 3.4 T/sec in a background field of 4.9 T. Although ac losses 172 were fairly large, the magnet was stably operated because of sufficient cooling. In the HPSM-2, a ralatively high winding rigidity and low ac loss properties were attained. As to the latter, a stainless steel wire was placed in the center of the subcable and accordingly the coupling loss was considerably decreased. ACKNOWLEDGEMENTS The authors would like to express their gratitudes to Dr. S. Toraiyaraa for his kind support and encouragement, and Drs. M. Nagata and T. Oku of Sumitomo Electric Industries . SatoT ,f Mitubish. o hLtdDr d an . i Electric Company for constructions and experiments of the magnets. REFERENCES 1) S.H. Kirn, et al.I Proc. 9th Syrap. Engineering Problems of Fusion Research (Chicago,Oct. 26-29, 1981) 2) T. Onishi, et al.I IEEE Trans. Mag. Vol. HAG-21, No.2, 799 (1985) 3) T. Shintorai and H. Masudal Proc. of the US-Japan Workshop on Supercon- ductive Energy Storag Madison( e , Oct. 19-23, 1981) P.442 Tateishi. H ) 4 t al.,e : Advance Cryogenin i s c Engineering, Vol.319 15 , (1986) 5) H. Nomura, et al.: Proc. of 36th Japanese Cryogenic Engineering Conf. (Kanagawa, Nov. 19-21, 1986) Ä2-5 6) T. Onishi et al. : llth Intern. Conf. on Magnet Technology (Tsukuba, Aug. 28-Sept.l, 1989) KH-01 Next page(s) left blank 173 ITER MAGNET DEVELOPMENT PROGRAE TH T MA LAWRENCE LIVERMORE NATIONAL LABORATORY, USA* S.S. SHEN Lawrence Livermore National Laboratory, Universit f Californiayo , Livermore, California, United States of America Abstract The International Thermonuclear Experimental Reactor (ITER) magnet development program at Lawrence Livermore National Laboratory mainly involves developing high-performance, radiation-tolerant magnet concepts. Model coils, full-scale conductors, and insulation systems will be fabricated and tested. This paper briefly describe prograe sth m plan. Also include descriptioe th s di e th f no Fusion Engineering International Experimental Magnet Facility (FENIX) whics hi currently under constructio testinr nfo g full-scale ITERd conductoran A k 0 4 o st 14 T. Its design parameters and schedule are reviewed. INTRODUCTION Internationae Th l Thermonuclear Experimental Reactor (ITER) projects ha 1 just completed the definition phase and is proceeding to a three-year design effort. Meanwhile extensivn a , e superconducting magne prograD alsR& ts mha o been initiated prograe Th . mdeemes i d necessar meeo yt t need botr sfo h ignitiod nan steady burn of the plasma in ITER. Toroidal field coils of 12 T and of a large size require high-current, niobium- tin superconductor operato st e under condition higf so h stress, cyclic loadingc a , losses, and nuclear heating. As for the poloidal coil system, the conductors must operate near 40 k A and at more than 12 T in the central solenoid to provide sufficient flux swin plasmr gfo a startu burningd pan . Furthermore magnetF T , s need to be made radiation tolerant. Employing advanced insulators, forced-flow magnets should possess high neutron heat-removal capability and good stability. MAGNET DEVELOPMENT PROGRAM majoe Th r objectiv ITEe th Rf eo magnet development progra develoo t ms i p high-performance, radiation-tolerant magnet concepts using advanced conductor designs (generally applicable to TF as well as PF systems) and nonconventional bonding/insulation techniques. Model coils, full-scale conductors, and insulation * This work was performed under the auspices of the U.S. Department of Energy by Lawrence th e Livermore National Laboratory under Contract W-7405-Eng-48. 175 systems will be fabricated and tested and related to the engineering, component database detailey sb d analyses. Table 1 outlines major tasks for the ITER R&D program at LLNL. Tablet ITER R & D PROGRAMME A. Radiation Tolerant Magnets 1. NbaSn Conductor Development 2. Structural Component Development and Characterization 3. Conductor System Analysis 4. Model Coil Development and Test B. Insulation Development and Characterization______ Descriptions of each prject are given below: A-l. NbsSn Conductor Development Objective: Verify superconductor performance on a scale appropriate to ITER magnet system demonstratd san predictio s basie it er th sfo n froe mth component database. Major tasks include: Performing characterization test strann so subscald dan e CICC. Operatio FENIf no X facilit full-scalr yfo e conductor testing. A-2. Structural Component Developmen Characterizatiod an t n Objective: Select and characterize materials suitable for the structural component high-currenf so t cable-in-conduit conductor systemf so ITER TF and PF coil systems. Major tasks include: 1) Characterizing candidate sheath materials for tensile, fracture toughness, and fatigue properties at 4.2 K. 2) Characterizing mode windinf lo g pac orthotropir kfo c elastic properties, failure modes, tensile properties, and fatigue at 4.2 K. 3) Producing appropriate structural components by industrial processes. 176 A-3. Conductor Systems Analysis Objective: Provide analysis, supporte diverse th y db e databases presentle th n yi literature and by the databases produced by the other tasks, that will accurately describe the performance expected of conductors in the complex ITER environment of nuclear radiation, ac heating, and multiaxial loading. Major tasks include: 1) Providing predictive relation for Jc (B,T, e, fluence) for NbaSn. 2) Developing code predicting quench behavior, thermal margins stabilitd an y performanc f conductoreo ITEe th Rn si environment. A-4. Model Coils Developmen Tesd an tt Objective: Develo demonstratd pan reasonabla n eo e scal conductoe eth coid lran technologies required for the design and construction of ITER magnet systems demonstratand , e reliable performance with mechanical stresse distributed san d heat loads comparabl thoso et e expecten di ITER. Major tasks include: 1) Developing techniques appropriat insulate-wind-reacte th r efo - impregnate method of coil construction using CICCs, such as joint insulationsd an s . 2) Fabricating and testing model coils with winding current density of 40 A/mm2 at 15 T. B. Insulation Developmen Characterizatiod an t n Objective: Develop high-strength, radiation-tolerant insulations, and characteriz tolerancr efo complexe eth , multiaxial loading conditions radiatiod an n environmen ITEe th componenF Rf T o t t system. The FENIX Magnet Facility The FENIX Magnet Facility, under construction at Lawrence Livermore National Laboratory (LLNL), is a significant step forward in meeting the testing requirements necessary for the development of superconductor for large-scale, superconducting magnets. Proposed conductors for ITER measure ~ 35 mm on one 177 testine Th . f go T sid wild 4 e1 t fieldan l a ~ operat A f sk o 0 4 t currenteo a t p u f so conductors and associated components, such as joints, will require large-bore, high- field magnet facilities. A 14-T, transverse field over a test volume of 150 x 60 x 150 mm in length will be capable of testing conductors of ITER size. FENIX is being constructed using the existing A20 and A2i magnets from the idle MFTF (Mirror Fusion Test Facility easwese d ) pair 2 (FigTh an t tA . s .1) wil e lb mounted together to form a split-pair solenoid. The pairs of magnets will be installe 4.0-a n di m cryostat vessel locate HFTe th n dFi (High-Field Test Facility) buildin LLNLt ga . Each magne cryostatn encloses i t ow s existine it th ,n di g 4.0-m vessel serving only as a vacuum chamber. Figure 2 illustrates the facility arrangements. Conductor test weü (wifh tail) Liquid helium reservoir I2o magnets... Crossover busses A,», magnets Magnet supports FENIX Magnet Facility FIG. 1. • 4 METER \CRYOSTAT HIGH FIELD AREA «» SEE PLOT FIQ. 1 METER FIG . FENI2 . X test facility. 178 With iron cores inserte high-fiele boree th th f sn do i d magnetse ,th maximu . Forced-flomT 4 fiel1 s di w heliu mt adjustabla e temperatures, pressures, and flows wil providede lb . severaSampleo t fro e e b mn on l sca meter lengthn si , with about 0.1 uniformbeine 5m th gn i , high-field region. Power suppliee sar available to provide currents up to 40 kA. Anticipated completio FENIe th f nXo abous i t April 1990. System testind gan shakedown should take another three to six months. The system should then be available for conductor testing. REFERENCES 1. C. D. Henning and J. R. Miller, "Magnet Systems For the International Thermonuclear Experimental Reactor." IEEE Trans MAG, 2 . . VolNo , .25 pp 1469-1472, Mar. 1989. 2. D. S. Slack, et. al., "The FENIX Test Facility," UCID-21750, Lawrence Livermore National Laboratory, Aug. 1989. Next page(s) left blank 179 RESEARCH AND DEVELOPMENT PROJECT OF SUPERCONDUCTING HELICAL COIL AT THE NATIONAL INSTITUTE FOR FUSION SCIENCE, JAPAN . YAMAMOTJ O National Institute for Fusion Science, Nagoya, Japan Abstract National Institut r Fusiofo e n Sciencs wa e established on May 29, 1989. The main plasma device of the institute is Superconducting Helical Device, whic mads i hf o e superconducting helical coils and poloidal coils. However superconducting technolog s growha yp enough u n e th , this i s first helical winding coil for fusion device. Several research and development program f superconductino s g coid an l conductor are planned now. Major paramete f largo r e helical device Table 1 shows the major parameters of the Large Helical Device (LHD). The major radius is 4 m, coil minor radius is plasm, m 0.9 fiel, 6 m a 6 minod 0. perio - r m 5 radiu1/ d0. s i s 2/10, magnetic fielplasm, T 4 d a duratio heatin, s 0 1 n g powes i r 20 MW. The magnetic energy stored in the coil is 2 GJ. Three TABLE 1. LHD SPECIFICATIONS . HELICA1 L COIL a 0 4. MAJOR RADIUS MINOR RADIUS (COIL) 0.96 • MINOR RADIU6 S0. - (PLASMA 5 0. ) 1/m 2/10 MAXIMUM FIELD (PLASMA) 4 7 MAXIMUM FIELD T (COIL 8 ) MAGNETOMOTIVT MA E8 FORCE CURREN 0 A/mm4 T DENSIT2 Y STORED ENERGY 2 GJ 2. POLOIDAL COIL 0V R=5.63m, Z= 1.55m, -5.02MAT IS R=3.10m, Z= 2.03m, -3.22MAT IV R=2.13m s 1.25mZ , , 6.11MAT . TIM3 E SCHEDULE YEAR 89 90 91 92 93 94 95 96 R&D POLÖIDAL HELICAL EXPERIMENT 181 sets of poloidal coil is also made of superconducting coil. The construction wile completeb l n 1995i de mosTh . t important e improvementh s i D f plasmLH poino t f o ta confinement timf o e currentless plasma. Historical Background of Superconducting Magnet Technology for Fusion Science at Universities Monbusho (Ministr f Educationo y , Scienc d Culturean e s ha ) funded superconducting magnet technology at universities these ten year y grant-in-aib s r scientififo d c researc n o fusioh n science. It also founded three superconducting research labora- tories r materialikfo e l researc t a Tohokh u Universityr fo , cryogenics and power application at Osaka University, and for magnet technology at Kyushu University. Therefore the universities accumulated enough knowledge for superconducting technolog r fusiofo y n science device. Befor e decisioth e f o n the LHD proposal, precise design study on superconducting helical coil was performed with researchers from universities. From 1987, Plasma Physics Laborator f o yKyot o University has successfully developed a small helical winding coil named KYOTO-SC e firswhicth ts i hful l torus superconducting coin i l the world. Superconducting split coir conductofo l r tess i t also prepared there. ProjecD & R NIFt a t S e constructioth r Fo f LHDe instituto n th ,o t develo s ha e p the high current conductor, winding technology, large stress supporting apparatus d coolinan , g technolog e coilth .r fo y National Institute for Fusion Science (NIFS) will prepare cryogenic d superconductinan s g test w facilitieTokne e i th t a s site in middle of 1990 in a new cryogenics building which area is 2600 . m s NIFstarte ha D coilSR& 5 ds named TOKI series includin windina g g test coil, medium size full torus coilsa , poloidal helicaa coil d an , l module coi s showa l n Tabli n . 2 e Experimental result f theso s e coils wil e obtaineb l d e froth m mid 1990. COILTABLD R& S . E2 NAHE TYPE COOLING COIL SPEC. MANUFACT. KYOTO-SC HELICAL BATH-COOLED a=0.3n r=0.063m •»16 1-2 FULL TORUS B=2.0T I=775A MITSUBISHI TOKI-WT HELICAL SECTIONAL O. B RM r=0.96m »=10 1=2 MODED LH L Cu conductor HITACHI TOKI-HB HELICAL BATH-COOLED R'O.Sn r=0.2m •>3 1»! FULL TORUS B»3.0T I«8930A HITACHI TOKI-TF HELICAL FORCE-COOLED R*0.9n r =0.25« 4 1-=m* 1 FULL TORUS B=2.77T I=8080A TOSHIBA TOKI-PP POLOIDAL FORCE-COOLED Ri*0.6m Ro=0.82B DOUBL2 E CAKE B«2.76T I=25000A TOSHIBA TOKI-MC S-SHAPE BATH-COOLED RJ.aO.8n Ro*1.4a LHD MODULE B=7.5T I=20000A MITSUBISHI 182 Selectio f reao n l size conducto d coolinan r g- methoan s i d other important points. NIFS asks several d coiconductoan l r manufactures to prepare 10 kA class conductor. Critical current measurement will star thin i t s Octobe Kyott a r o University with the split coil and- 30kA power supply. Superconducting coil and conductor test facility The first building at Toki site is a cryogenics building, which will be constructed mid 1990. In the building, the super- conducting coid conductoan l r test facilities wil e lb prepare d as show in Fig. 1. SAS WARMER 30KA POWER SUPPLT MECHANICAL TESTING MACHINE l 1000 TON PRESS U MECHANICAL TESTING CLEAN BOOM ! i CRYOSTAT l [ SPLIT COIL > "*C7^fo£I:.TPD tM-m »TOKI-P» •PURS WATER- 75KA POWER SUPPLY SUPPLY 2 TANN [i±ÎI. KIL l 3*4 LAYOUT OP CRYOGENICS BUILDING S TANOA KX H Fig.l. Layout of superconducting coil and conductor test facility 1. Helium Liquéfier and supercritical helium generator 0 liter/hou20 f liquefie o rf o refrigeratio W 0 e supplH d50 r o y n power at 4.5 K, also 50 g/s of supercritical helium generation. 10000 liters liquid helium storage dewar. distributior fo x bo n R&D coils. Storage gas tank of 400m0 in volume and 20bars in pressure. C 2CurrenD . t Supply 75 kA, 21 V current supply for superconducting coils and conduc- V curren21 , tkA 6 suppl d an a coilytor , . 3. Mechanical Testing Machine A press machine with capacit f lOOOtoo y n itself, cryostat capac- ity is 500 ton for a coil pack test at liquid helium tempera- a conducto r fo n rto ture tes0 t liqui15 ,a t d helium temperature up to 30 kA of current with 9T of magnetic field. Also a 20ton Instron mechanical testing machine with a temperature control- lable cryostat. 183 R&D coil KYOTO-SC Kyoto Universit NIFd an yS have develope a smald l full torus superconducting coil which aims realization of superconducting helical coil as future plasma device. The coil made of two helical coil produco t s e magnetic field surface. Tabl 3 eshow s the main coid conductoan l r parameter 2 d Figshowe an s. th s structure of the coil. Recent experimental results show the coil is fairly rigid and the critical current of the coil is almost same as the short sample test results as shown Fig. 3. Superconducting Helical Coil TABLE 3. KYOTO-SC PARAMETERS Coil Support Coil Bobbin Helical Coil Major radius 300.0mm Minor radius 62.7mm Number of coiIs 2 Poloiclal pilch number 16 Winding method Geodesic Winding per coi 1 turris>:21 ] 2 layers Coil cross section 36.lmmx45.3mra Total inductance 255mH Norl_____Maximuma n ni . Current 775A 1530A Central field 2. IT 4. IT Maximum field 2.9T 5.7T ,x- 8 a CoilC ross se Fig.3. Operating currenf o t KTOTO-SC under different 0 2 4 B 8 10 quench number Magnet ic F ie ld[T] 184 D coiR& l TOKI-WT TOKI-W sectionaa s i T l 110° winding test coif o whicl h geometry is the same as LED as shown in Fig. 4. Copper conductor having the .same mechanical properties as superconductor is used there. Fig.4. TOKI-WT D coiR& l TOKI-HB TOKI-HB is a pool boiling type helical winding coil which aims magnetic field accurac f helicao y l winding coil, stresd an s helicae th strai f lo ncoi l under full current excitatiod an n cooling process, and stability test. The size of TOKI-HB is almost 1/5 of LED. m and 1 number are 3 and 1 respectively. Load rati o t shoro te sams th sampla LEDes i e e .Th specificatio e structurth d e showan n ar ed FigTabln i nan . .5 4 e TABL TOKI-H. E4 B PARAMETERS Item Parameter Major radius 0.8m Minor radius 0.2m Winding rule 3 m= 1 1= Magnetmotive force 1MAT Operating current 8930A Numbe f turno r s 112turns Inductance 0.048H Stored energy 1.9MJ Maximum field 3T Cool ing method Poo I-coo I ing Size of conductor 16mm widthX8mm height Critical current 17580T 3 t Aa Fig1.5. Helical can and support of StabiIity parameter 0.67 suppor f TOKI-Ho t B 185 R&D coil TOKI-TF TOKI-TF is a forced flow type helical winding coil which aims demonstration of forced flow type helical coil, mechanical properties of the coil, and support system of the coil. The d an numbe1 4 d coie e LHDf almoss an th o sizar ri l m f 5 .o e 1/ t 1 respectively. Major and minor radii of the coil is 900 mm and 250 mm, respectively. Nominal current is 8.08 kA at 1.7 T. Stored energy will be 13 MJ at nominal current. The conductor is a NbTi cable-in-conduit type, of which size is 13 mm x 13 mm including insulation. Numbe f stran o rcabla n s i d324i e f o , which copper ratio is 4.8. The operation temperature, pressure, , d K 10.an flo 2 w00 4. g/s1. rats bare i d e.Th an , specification is tabulated in Table 5. TABLE 5. TOKI-TF PARAMETERS Major radius mm 900 Minor r ad i ua mm 250 Magnetomotive fore« MAT 0. 8 Operating current kA 8. 08 Average current density Urn' 47. 8 Max. field T 2. 77 Inductance mH 41. 1 Stored energy MJ 1. 34 Superconductor: Typ f conductoo e r Cable— in-condult. Forcedw o 1 —f Material of conductor NbTi Critical current kA T 4 t 1a 6 Condu 1 t n o i a n e m 1 d mm' 11X11 th Ickness mm 1. 0 Vo id f ract 1 on 0. 4 Strand diameter mm 0. 428 Numbe f strando r s 324 NbTI:Cu ratio 8 . 1:4 Stability margin ILl/CC 699 Coo 1 an t : Inlet temperature K 4. 2 bar 10. 0 Pressure drop bar/I 0. 0113 R&D coll TOKI-PF TOKI-PF is a forced flow type poloidal test coil, which e aimdemonstratioth s forcede th f o n flow long conductod an r cooling technology e coi composeTh s .i l doubl2 f o d e pancakes, and its inner and outer diameters of the coil are 1200 mm and , respectivelymm 1640 .e fielth Nomina d dan , l kA curren 5 2 s i t is 2.76 T. The conductor is a NbTi cable-in-conduit type, which outer size is 22 mm x 27.5 mm including 0.5 mm thick insulation e Th cabln i condui. e s i madt e 6 fro48 m superconducting strands, of which operation temperature, pressure d , 10.floK an , w 2 0 0 4. barg/s rat4. e ar ed ,,an respectively e specificatioTh . s tabulatei n Tabln i d . 6 e 186 TABL TOKI-P. E6 F PARAMETERS Coil parameters: mm 600 Ou ter rad lus mm 820 Magnetomotive force MAT 1 Operating current kA 25 Urn' 41. 3 Max. field T 2. 76 I nduc tance mH 3. 35 Stored energy MJ 1. 05 S dt f conductoo e p y T r Cable-in-conduit, w o 1 Forf — eed Materia f conductoo l r NbTl Critical current kA 50 at 7 T Condu 1 t d ime ns 1 on mm1 17X22. 5 thickness mm 1. O Void fraction 0. 4 Strand diameter mm 0. 67 Numbe f strando r s 480 NbTi :Cu :Cu-Ni ratio 1:1. 6:0. 5 Stab y margt i l1 in tl/cc 449 Coo 1 an t : Inlet temperature K 4. 2 Inlet pressure bar 10. 0 Pressure drop bar/i 0. 0078 D coiR& l TOKI-MC TOKI-MC is a module coil for simulation of helical coil, and aims to get engineering data of large electromagnetic force, simulating LHD. The shape of the TOKI-MC is S type as shown in Fig. 6. Inner and outer diameters are 800 mm and 1400 mm, re- spectively. The operating current is 20 kA to get 7.5 T. Pure aluminum stabilized bath cooled type conducto s adoptedi r . 0 »0 ,.li* L l NNER D l AMETER 800mm OUTER DIAMETER 400m1 ~ m AVERAGE CURRENT DENSITY 40A/mm' D EL I F M MU I X MA 7.5T OPERATING CURRENT 20kA CONDUCTOR SIZE 2Ommx20mm (PURE ALUMINUM STA ZED)I L Bl MAXIMUM TORSIONAL ANGLE IS' MAXIMUM CHANGING RATE F TORSIONAO L ANGLE 4O*/m Fig.6. Structure and design parameters of TOKI-MC 187 Conductor Test Program For the purpose of development and decision of the superconductor and cooling method of the LHD, we are planning to A clask tes0 1 s t superconductor using Kyoto University's split type superconducting coil and 30 kA power supply. Conductors are manufacturing now at several wire company. Besides the critical current, mechanical propertie d an windins g characteristics like coil elasticity and electrical insulation are important decision points. 188 MODEL COIL TESTING TO COMPLETE THE DATABASE FUSIOT NE FOE N RDEVICTH E J.V. MINERVINI . MITCHELN , L NET Team, Max-Planck-Institut für Plasmaphysik, Garching, Federal Republi f Germanco y Abstract e magnetiTh c a fieldfuturT (Nex f o NE se t European Torus) fusion device woule b d generated by two main sets of superconducting coils : the system of the toroidal field n numbei coils, 6 d abou1 an r t 7xl2 e e systepoloida th th n bore m i f d mo an ,l field coils with diameter e maximuTh s . ranginm m 0 f field2 gcoilo e n o eac frot t o ar s se hm4 about 12.5 T, requiring the use of A15 force flow cooled superconductors. Coils with these specification t availablno e ar se a considerabltoday t bu , e experienc n i large e superconducting magnet technology has been gained by laboratories and industry from fusion related R and D since early 1977. There is confidence that industry will be able to produce the coils in due time as a result of a vigorous mission oriented development being carrie t unde ou e dMagne th r t Technology Programme. Ther e threar e e main e conductostageth n i s r development programm: e . 1 Developmen f subsizo t e conductor n loni s gT inser 2 length1 t d testea coian ss la d e SULTA th n Switzerlan i n i I NPS t facilita d] [1 (underway)y , 2. Development of short lengths of full size conductors [3] tested in either SULTAN e LLNe completeb th o r Lo (t FENIS n 1990)i dU e Xth , facilitn i ] [2 y 3. Industrial fabrication of long lengths of full size conductors and manufacturing of modeTOSKK Kf ] le facilite completee Atesteb [7 coilb th o o n t i (t dsy n 1993)i d . e technologn stagI th 1 e f Nb3$o y n conductor s i developes d experiencan d s i e gained with fabrication and performance on a laboratory scale using conductors and coils of moderate size. e developmenTh f o full-sizt e conductor s i performes n i stago dmakin tw e g extensiv f o industria e us e l facilities y n industriathiI wa s. l feasibilit s i provey d an n the firms are better prepared to respond to orders for long lengths of full-size conductors require n stagi d. 3 e In stage 3 four circular model coils will be manufactured by using the conductors which have satisfie e qualifyinth d g straight sample e testsmodeTh . l coile sth wile b l smallest size thas e i fulcompatiblth t l f o siz e eus conductoe wite th th hd an r achievement of the full operating field levels. A circular stack is chosen as the most efficient shap r achievinfo e g this. The intention is that the coil manufacturing process approach to the maximum possible extent that envisaged for the NET TF and PF coils. It must be representative not only for the manufacture of the NET coil winding but also for the terminations 189 d jointsan , current leads, feed throughs e heliuth , m supply e pipeworth d an k instrumentation. The manufacturing process will in particular demonstrate that the conductor e robusar s t enoug o t withstanh e industriath d l handling during coil manufactur e successfuth d an e l operation will serv s finaa e l proof e thadecisioth t n to proceed with construction of the NET coils can be taken in confidence. n thiI s pape e definw re testh et requirement e coith l d configuratioan s e th f o n model solenoid test facility envisaged to fulfill them. GENERAL TEST OBJECTIVES e modeTh l coil test programme must achieve certain well-defined goal n ordei s r to prove the feasibility of the conductor and coil concepts chosen for the NET machine. These general objectives are summarized below : developmen e entirth f eo t coil manufacturing process including coil windinf o g double pancakes, conductor termination and joints, coil insulation, assembly, vacuum impregnation, current leads, feed throughs and instrumentation, verificatio e e succesindustriath th f f o o ns l manufacturing proces y testinb s e th g coils under operation as close as possible to NET conditions, validation of design codes for stress analysis, AC losses and quench behaviour and validation of predictions of performance made on the basis of sub-size component tests, performance of tests that can only be made in a large coil test and that may point up synergistic effects, i.e. combined stress degradation of superconductor performance in the coil compared to single wire or cable short sample test, selection among the conductor options based upon test performance, manufacturing considerations, and cost . e modeTh l coils should have geometrical dimensions whic e representativar h r fo e T coil n NE i certais n limits. They should reach d fielstresan d s values whice ar h similar to those of the NET coils and demonstrate that the manufacturing techniques of conductor and coil are ruled by engineering standards and are reproducible. Althoug T relevanNE h t stress levels wile achieve b e lmode th n i ld coils, this test will not be sufficient for evaluating the mechanical fatigue life of the conductor jacket or coil structure. Therefore this mus e assesseb n tadditionaa y b d l materials testing programme. Sinc e coild th conductorean s e prototypar s e components t i mus , e possiblb t o t e complet a substantiae e testinlth t paravailablef no o o e gt coit s on i evee l f du i n, delays or failure. 190 CONDUCTORS Several type f conductoro s s have beeT magnets f nNE theo designel e mal th , r fo d using an A15 compound as superconductor and the forced flow of helium as coolant with a helium tight, reinforcing stainless steel jacket. These are being developed in winA d reack an dtw0 t4 o a mai reacA F coild d wink T an an nt6 ,e d 1 typeth typ a r ; fo se e centrath typr fo le solenoi F coildT 3 versions(CS( sd an ) ) [3,4]. AdditionallA k 0 4 a y conducto n NbTi r s i beini g developee oute th F coils P r r fo d. Figur 1 showe e th s conductor types. These conductors should meee specificationth t s liste n Tabli d . 1 e The most promising conductor designs from the tests of short lengths will be manufactured wit a m hlengteac K f ho abou 2 h 1- usint g manufacturing process whic s ha possibl approacr fa T e econductorsfinas th a NE thosh r l fo e , including connections and current lead feed throughs. SCALi h E I = 16 U »s a = «oo SA U « = I LUI 9-- 4600 M h- M « = J AB8 = •(809 0W FIG. Mode1 . l coil configuratio conductorsd nan . 191 Table 1 SUMMARY OF CONDUCTOR SPECIFICATION Toroidal Field B'ma, x 12 T Bmax s (disruptionT/ 0 4 ) 0.3 T/s (operation) 2 T (disruption) $B2dt 50 T2/s (disruption) T1op 4.5 K Jcable space at operation 50 MA/m2 Jcrit f°r cable space at operating B, T 2 and eexternal = 0 80 MA/m External strain : longitudinal % 3 0. < transverse > 0.2 % Nuclear heating : mW/cm1 < ^ (max) < 1 kW/TF coil (av) Unit conductor length <400m Central Solenoid max 12.5 T "max 15 T/s (disruption) 2 T/s (operation) (ßdt s (disruptionT/ l ) 10 T2/s (disruption) AI 6 kA (disruption) Jcable space at operation 50 MA/m2 Jr cablcrfo jt e spac t operatina e T , B g and eexternal = 0 80 MA/m2 External strain : longitudinal % 3 <0. transverse -0.> % 2 Unit conductor length : <400m COIL SYSTEM DESCRIPTION General e modeTh l coil geometr s beeha y n chose s circulara n , wite e inneradiuth th h f ro s bore determined by the minimum bending radius of the conductor (typically 1 m, limite y winA keystonind b reack d an d0 t4 A reac d bendink r an fo g6 t 1 d strai an r g fo n wind). The circular shape enables a minimum volume of superconductor to be used to generate the required fields, [5], and gives a predictable stress/strain level in the coil. 192 e mechanicaTh l problem f o vacuus m impregnated coilr tokamakfo s e welar s l understood from copper coil s t i importanrooa s t I m . temperaturo select K t 7 a 7 t d an e simple coil geometry where the force/stress situation can be easily understood, to allow interpretatio e superconductoth f o n r measurements, rather tha a ncoi l where winding problems dominate the behaviour. The complete magnet will be composed of four subcoils. This allows different conductors to be manufactured and tested in representative lengths, reduces the risk of o t inpropefailur e du e r heat treatmen r insulatioo t n impregnation d followan ,e th s modula T centraNE r e l th desigsolenoi f o n d thus providing experience making intercoil connections. e fouTh r subcoils wile assembleb l a coaxia o t d l solenoid e assembl Th e fouth . rf o y coils wits supportinit h e gaccomodateb structuro t e existinth s ha n ei d g TOSK vacuum tank at the KfK Karlsruhe (Fig. 2), with the overall dimensions being restricte e availablth y b d e tank dimensions e supportinTh . e gbuilb structuro tt s ha e in such a way that the NET standard operating conditions can be simulated and the limit f operatioo se assessedb e n assemblca Th n . y should also forese e possibilitth e f o y testing additional pancakes. The coils will be assembled in such a way that after the tests they can form a high field test facility by using the internal bore of the solenoid assembly. FIG. 2. Possible configuration of the model coil in the TOSKA facility [7J. 193 Proposed Layout The layout of the facility is shown in figure 1. The solenoid consists of a stack of four separate coils, each nominally about 400 mm long with inner and outer radii of 1.0 and about 1.5 m. Optimisation procedures [5] have shown that for the field and stress limits of interest, this geometry uses the minimum volume of superconductor. Each sub-coil provide a nominas A k 0 A (i.el4 0 turnM 12 capabilitf . o 8 s 4. f o y superconductor, wit a totah l lengt f abouo h 1 Km/coil)t e coil e wounTh ar s. s doubla d e pancakes with cooling inlet on the inner bore and outlets, current leads and pancake connection e outside th e singl n Th o s .e pancake cooling length wit a A typicahk 0 4 l e pancakTh . e m conducto connection5 e 12 outsidth s i n r o es should occupe th y minimu f radiamo lT centra spacNE s requirea e le th solenoi y b d d stack d shoulan t , ac d as a test of the connection concepts proposed for the solenoid. Pancake connection concepte toroidath r fo sl field coils have fewer space limitations. Overall structural suppord flangeen s i provideo t stw linkey b dy bolts b d , which will provide precompression to keep the coils in contact and support any outward o t opposinforce e du s g coil currents. Thin steel sheet n separatca s e individuath e l coils but intermediate structural support or cases will not be used to allow representative vertical stresses to be achieved. At present four Nb3Sn superconductor e proposee ar coilssth r , fo deac h having slightly different dimension o desigt e r curreno ndu s t capability. e solenoiTh n als NbTA e useca k db o o test di0 4 tconductor s propose e outeth r r fo d PF coils by making provisions for a double pancake of NbTi to be placed on the end of e stackth . o alloT w fully instrumented pancake e b teste o t ds e withouth e risf th o tk instrumentation itself leadin o t coig l failure ,n extra spac r a fo edoubl e pancake th n i e e staccentrth f ko e wile providedb l . All coils are considered to have a steel fraction of about 0.27 corresponding to a jacket wal conductoA k l thicknes0 . 4 f Thia abouo mm r n s3 o st allow s maximum hoop stressee jacketh o n t approaci ts e allowablth h e limir statid fo fatigut an c e stresses (500-600 MPa). In the tokamak coils, either a thicker jacket can be provided or extra stee e b cowounl n strica pd e witconductoth h o t providr e reinforcementh e t necessary. Operating Conditions e solenoiTh d steady operating conditions which defin e mechanicath e l limite ar s summarize r normafo n tabli d 2 le operating conditions wit l foual h r coils charged an d in table 3 for operation with three coils charged. In a first rough approximation the stresses are evaluated using the conventional expression r hoofo s p tensio n i thicn k cylinders under internal pressure. Fieldd an s forces are calculated numerically. It is clear that at least 12 T can be achieved even 194 with 3 coils, with some overcurrent in the conductor, but without exceeding the stress limits. The tensile hoop strain at about 0.25% is close to the operating levels of the coils . e averagTh e modulu e e windintransversth th f n o i sg a epac s GP i abouk 0 5 t directio o transverss n e strai s i aboun t -0.1%. Thi s i slightls y lower than that proposed (tablT fo. NE r1) e The electrical conditions for the coils are determined by the test procedure. The insulatio a nomina o t s i builn p u l t voltag e betweeV k toleranc 0 2 n f o eterminal o t d an s e fulth groundl n coilsT i siz s NE ea ., Table 2 4 subunits energised 120 turns/subunit Conductor Maxim. Field Minim. Field Aver, vertical Peak Tensile Current on Conductor in bore on stres t centra s e stres steen i s l Conductor jacket T T MPa MPa 40 10.7 8.0 43.2 439 44.92 12.0 9.0 54.5 553 46.78 12.5 9.4 59.1 600 Table 3 3 subunits energised 120 turns/subunit Conductor Maxim. Field Minim. Field Aver, vertical Peak Tensile Current on Conductor in bore on stress at centre stres steen i s l Conductor jacket T T MPa MPa 40 9.53 7.4 32.2 369 50.36 12.0 9.3 51.1 584 POWER SUPPLIES The testing capability of the solenoid is strongly determined by the available power supplies e totaTh l . stored magneticf I e energsolenoith . f o MJ ys i abou d0 35 t chargine coilth f sa timescal o g (0-1 n o ) 2T e comparable with that expectea n i d tokama s i requiredk , ) s (abouthe 0 n 4 t eithe a power r supply capabl f o deliverine g about 20 MVA or an energy storage system with a capacity of about 0.5 GJ is necessary. 195 This coul e takeb d n directly froe grir th mo frod a mflywhee l generator e onlth , y restriction being financial. e testh t f I programm e completeb n ca e d withou a fult l charging capability thea n substantial reduction in the energy supply capacity is possible, at the expense of more extensive switching systems and discharge resistors. Detailed calculations [6] show that a 0.5 MVA supply (i.e. 50 kA, 10 V) is sufficient to charge the coils in 30 min. and that switching of resistors and transfering energy between the subcoils can meet e necessaralmosth l al t y test requirements e exceptioon a e swin s Th i n g. through zero ) whicT 2 h-1 unavoidablo fielt dT (i.e2 +1 y. require n externaa s l energy source. Chargin e subcoilth f o n gi opposit s e directions allow a limites d swing through zero current at low (2 T) fields. TESTING PROGRAMME Tests for both the TF and OH coils can generally be divided into two categories, namely tests at standard NET operating conditions and tests to determine the limits of operation. Given the different objectives for the two coil types, the specific test requirement r eacfo s h type wile differentb l e testTh s . should simulat s closa e s a e possible the actual NET operating conditions including magnetic field, current, strain and transient operation. This means, for the TF coils, DC current operation with steady statd transienan e t heat loadd transienan s t external F magnetiP e th r c fo fields d an , coils, AC or fast ramp current operation with simultaneous transient magnetic fields. The coils will be designed and instrumented to extract as much information as possible. However, the experimental nature of these coils and the imposed instrumentation shoul t interferno d e wite d safreliablth an he e operatio e coilth f so n or otherwise compromise the chances for a successful test. To achieve this it will be required to make one double pancake, fully instrumented with embedded heaters, voltage taps, etc. This will allow for full data acquisition and complete characterisation of the performance without jeopardizing the operation of the bulk of e modeth l coil. All test requirement e modeth r lfo s coile basear s d upo e assumptioth n n thaa t significant basis of knowledge of conductor performance and parameters has been determined by exhaustive tests on sub-size components and full-size conductors in shor e t extenth lengths o tT tha. t exhaustive characterizatioe b t nno testn ca s complete n sub-sizo d e component d coilsan t si become, s imperative that thee b y e modeth n carriei l t coiou dl r exampletestFo . t i migh, e t difficulb sufficient ge o t t t data on quench behaviour from a sub-size test facility thus requiring more extensive e modetestinth n i lge identificatiocoilTh . f differenceo n s between component tests and the model coil tests, and their explanation, will form an important conclusion of 196 e testh t programme e modea Th substitut t .l no r coi extensivs fo i el e component testing. e actuaTh l operatin ge modeth rang f o le coils depends significantl e poweth n ro y e coil e th connecte ar sw ho suppld thi d an ds an ythe n limit e actuath s l teste b s n thaca t carried out. Howeve t i woulr e desirablb d e followin th o carr t et ou y g test: s Normal operation Nominal peak current Nominal peak magnetic fields Nominal helium coolant conditions Nominal global winding pack stresses which influenc e conductoth e r (hoop tension, axial compression) Simulate pulsed operation scenario (PF conductor) Simulate external pulsed field (TF conductor) Simulate nuclear heat load (TF conductor) Plasma disruption From nominal operating conditions simulate a plasma disruption with field components transverse to the conductor only with external field change only (TF conductor) and current change (PF conductor) e followinTh g additional testd measurementan s s shoul e performeb d : d Measure the energy deposition due to AC loss and mechanical hysteresis during a pulsed cycle Measure coil strai r verificatio fo ncalculationM FE f o nd globa an s l winding pack parameters Determine the DC limits of operation by measuring the critical current as a function of magnetic field and temperature The limits to stable AC operation should be determined by increasing the rates of field change until the coil quenches. Alternatively, energy perturbations can be superimpose e normath n o dl opening cycl o t determine e stabilitth e y limitsA . parameter should be the time during the cycle when the quench is initiated Protection system verification test under normal operation conditions The tests described above should be performed over a wide range of helium inlet temperatures, pressures and flow rates Tes f o coit l insulation syste y statib m r o pulsc e voltage tesf o coit l ground insulation and by development of NET relevant internal voltage distributions. 197 TIMESCALE A general timescale for the production of the conductors and model coils is given by the events listed in Table 4. Table 4 TIMESCALE OF THE MODEL COIL TEST Event Date Testing of full-size, short length conductors and pre-selection Mid-1990 Modification o TOSKt s A Facility 1 - Jan9 9 . Jan8 . Detailed Design of Model Coils Sept. 90- June92 Conductor Manufacture Jan. 91 - Oct. 92 Coil Manufactur Assembl& e y - Jun 3 2 9 Jan9 e . Assembly in TOSKA June 93 - Dec. 93 Testing Begins Jan4 9 . CONCLUSIONS A model coil test facility has been described that will provide the required manufacturing, test, and operational database to proceed with the construction of the NET magnet system with confidence. The various conductors proposed for the NET coils wile produceb l n i significand t length d testean s sub-coila de facilitth n o i t sy e maximuth m possible extent. 198 REFERENCES [1] The SULTAN III Project, A. délia Corte, G. Pasotti et al., 15th SOFT, Utrecht, Sept. 1988. e FENI[2Th ] X Magnet Facility, D.S. Slac t e al.k , Livermore National Laboratory, Internal Report No. UCID 2 1750, Aug. 1989. [3] J.V. Minervini . R Poehlchen, t e al., , "Conductor Desig r Superconductinfo n g Poloidal Field Coils of NET", Proceedings of the 15th SOFT, Utrecht Sept. 1988 . [4] J.V. Minervini, N. Mitchell, et al., "Progress in the Development of a 40 kA SuperconductoT CentraNE e l th Solenoid" r fo r , presente e lltth ht a dMagne t Technology Conference (MT11), Tsukuba, Japan (1989). [5] N. Mitchell, "Preliminary Calculations for a Solenoid Test Facility", NET Internal Document, N/R/0220/3/A, 11.5.89. [6] J.V. Minervini, N. Mitchell, R. Poehlchen, Requirement Definition Document for F ModeP d lan e Manufactur th CoilsT F ReporT NE ,T d tTestinNE an e f o g N/P/0220/2/A, 1989. [7] NET Modell Coil Test Possibilities/ Study Report KfK 4355 KernforschungsZentrum Karlsruhe, November 87. Next page(s) lef9 t19 blank DEVELOPMENT OF SUPERCONDUCTING MAGNETS FOR A TOKAMAK FUSION MACHIN JAERIT EA , JAPAN (Abstract) . ANDOT . SHIMAMOTS , O Naka Fusion Research Establishment, Japan Atomic Energy Research Institute, Naka-machi, Naka-gun, Ibaraki-ken, Japan As development of superconducting magnets for the Fusion Experimental Reactor (FER) which is the next generation fusion machine afte JT-6e th r 0 machine programo ,tw s (the Proto Toroidal Coil DemPrograe th o d Poloidaman l Coil Program progresn i e )ar s t JAERIa e purposProte Th .th of eo Toroida l Coil Prograo t s i m verif e availabilitth y e conducto windine th th f o yd gan r structure, designed for the toroidal coil of the FER. The coil Prote systeth of o mToroida l Coil Program consist protoe th f -so toroidal coil, two LCT coils and two LCT backup coils. The proto toroidal racetraca coi s i l k form wit minimua h m winding diameter of 1.5 m. The operating current of 30 kA at 12 T. For the proto toroidal coil, there are three candidate conductors (TMC-FF type, Performed Armor CICC typd Advancean e d Disk type). Trial manufacture of them began from 1988. Three conductors are tested and evaluated in 1989. The best conductor will be chosen for the proto toroidal coil. The purpose of the Demo Poloidal Coil Program is to develop a technique requireconstructioe th r fo d f superconductino n g poloidal coils. This coil system consists of three coils (DPC- d DPC-EX)an 2 e windinU Th , . Ul g inner diamete f theso r e coils i s 1 m and the total energy is around 40 MJ. The DPC-U1, U2 coils which are made of a NbTi conductor, have an operating current of changina d an T g 7 fiel t a d3A 0k T/s7 rat DPC-Ee f o e.Th X whic NbgSa mads i hf o en conductor n operatina s ha , g currenf o t changina d an T g 0 fiel1 t a d T/s0 A 1 ratk Marcn f 0 .I o e1 f o h 1989, the DPC-U1, U2 has been completed and the preliminary test has been carried out with a pulsed operation of 7 T/s. In the end of 1989, the DPC-U1, U2 with the DPC-EX is tested. After that, the US-DPC in Japan-US collaboration program will be tested n thiI s wit. e DPC-UpaperU2 th h e statu d th 1an , f boto se th h Proto Toroidal Coil Program and the Demo Poloidal Coil Program are described. Next page(s) left blank 201 THE SULTAN-m PROJECT A. della CORTE, G. PASOTTI, M. RICCI, N. SACCHETTI . SPADONM , I Centre Ricerche Energia Frascati, Associazione EURATOM-ENEA sulla Fusione, Frascati, Rome, Italy l MUTGDa . . SPIGOG , . VEARDG , O Anslado Componenti, Genoa, Italy J.A. ROETERDINK, J.D. ELEN, A.C. GLTZE, W.M.P. FRANKEN Netherlands Energy Research Foundation (ECN), Petten, Netherlands . AEBLIE . HORVATHI , . JAKOBB , . MARINUCCIC , , P. MING, G. PASTOR, V. VÉCSEY, P. WEYMUTH Paul-Scherrer-Institute, Villigen, Switzerland Abstract e SULTATh N e majotesth f t o r e Europea facilite toolth on n i ss i y n Communit e developmenth r yfo f higo t h current superconductors Presentl calleo s e dyth SULTAN phas superconductina ( I eI g magnet system witn i hfrea m 8 e5 bor0 f eo diameteusablT 2 testfiela r 1 ef fo d reactef do so an r d conductor loop r coilsso closs )i completiono et relatively B . y small change SULTAf so flexibilite th NI facilitconsiderable e I b th f n yo yca e enhance facn I dsplia t t coil configuratioe b n nca realize substitutiny db presene gth sectioT 8 t n witintermediato htw e A-1 splittin y 5parb T coil 6 d t e switan s gth hwa previously realized with a pancake structure In this way the facility will be used for NET conductor prototypes and series acceptance tests as well as for measurements on subsize NET pancakes In this paper we will present the SULTAN splitting project (SULTAN III) which is a collaborative effort among ECN, ENEA and PSI, giving detail mechanicaw s aboune e th t magne T le splitte6 structur th e n th i t dd configurationan e o tw e th , intermediate A 15 coils, the new equipment necessary for high measuring repetition rate including a superconducting transformer and the cryogenic circuitry layout being influenced by successive performance tests of the 1. INTRODUCTION first set of coils. For such an undertaking prior to final maie Th naree tas magneth f ao n ki t technologo t a yl design, risks reduceneee b o dt d significantl impley yb - develo ptoroidaa l magne twit T systehNE superr mfo - mentin a comprehensiv f go e development programme conducting coils for about 8m times llm bore for gen- closing the gap LOT-NET in several areas as: crating maximum fiel 10-12Tf do , highly reliable th n ei Tokamak environmen storea t a t d energy leve tenf o l s « design studies, of Gigajoulcs. A significant present step in this task is • high field conductor development including also the International Large Coil Task Project (LOT), ini- conductors for the more advanced objective of tiate 197dm joina d s an 7 a t A experimenUS e th y b t pulsed superconductin coilsH gO , its partners EURATOM, Japa Switzerlandd nan e Th . • coil technology verification tests, experience gained in the LOT programme has to be ex- • full size conductor tests, tended in coil size by factor 3x and from 8 to 12 Tesla • full size and/or subsize pancake tests. in max field strength. Ultimately, integration of coils ECN, ENE d PSIAan , presently engage collaba n di - t thaa t size int oTokamaa k configuration will require oration on the development and testing of high field the most extensive development effort. forced flow superconductors for fusion devices, are now A single full size TF coil for NET represents an in preparing conductor inserd san t coil orden si demono rt - vestment of the order of 40 MUG. Therefore, it is very strate the feasibility of s.c. technology for high current, probable that the full aize coil fabrication and tests will high field applications based on a react and wind tech only be possible as a first phase of the final commis- nique. Their programme include modificatioe th s f o n sionin f NETgo , wit e sequenchth f coieo l fabrication the existing SULTAN facility1. 203 2. SPLIT PAIR CONFIGURATION Z direction, it has only a limited influence on the peak e maiTh n componen superconductina s i t g magnet fields showed in Figure 2. system wit hfrea e bor testinr efo conductof go r loopr so 11 coils in a background field of 12 Tesla at 0.58 m diam Brnod (T) eter after realization of the SULTAN Phase II stage 12 Upgrading of SULTAN-II by split coils replacing the Tcsl8 a section will enhanc e flexibilite th e testh t f o y facility considerably This upgrading represent relasa - tively cheap solution to fulfill the special needs required by the NET programme In order to test fullsize NET MIDDLE conductors a 12 Tesla magnet system with at least 4 m i r - l - 1 - inner bore or a smaller magnet system with split coil ENEA I configuration is needed It is proposed to construct a "IT split coll facility using as many of the existing mag- nets of SULTAN-II as possible. In fact only one of the present NbTi coils has to be replaced by two coils, due to the fact of having in the new facility, called SULTAN-III a splitted magnet system The "new" in- termediate coils must have a higher current density and Z (cm) as a consequence to withstand a higher magnetic Meld FIGURE2 than the "old" NbTi coil. The new magnet will have Field distribution in the gap and max field at the innei therefor buile b A-1 y o b et 5 conductors splittede Th . - SULTAN-IIe radith f o i I coils III facility will provid epossibilita teso yt t short sam- presencn i ) energizea (a f eo d test pancake ples from industrial pilot production fulf so l size NET- (b) magnet system alone conductors which coul extendee db testino dt largef go r coiled samples (pancakes of 2.0m diameter). factn I , withou tese th tt pancak peae eth e k th fiel n do A15 insert lowers fro whilT m 12 12.2 eo t therT pracs ei - tically no influence on the rest of the coils. Forces: The forces acting on each coil of the mag- netpresencn i , tesa f eto pancake, have been calculated using the local value of field and current. Each coil has been divided into 256 elements, the field being calcu- 1290 e middllateth n di e poin eacf o t h elemen e totaTh lt d forceF an z actin x F eacsn go h coischematicalle ar l y reporte verticae axith X e s i se Figurn di on l Th . e3 There are no net forces in the Y direction because of the symmetry. It is worth noting the huge centering force acting betwee tese nth t pancakmagnee th d ean t Aare Berg FIGURE 1 Arrangemen f splio t t coil r SULTAN-IIfo s I (dim in mm) . ANALYSI3 MAGNEF SO T CONFIGURATION e magnetiField:Th c field distributio s beenha n calculated using the code GFUN. The iron parts of the faculty have been divided intelements0 o35 maxe th , - imum allowecodee th y .d b Figur showe2 radiae sth l distributio e fielth df no modulus e centee th th t f a ,o r gap, produce magnee th y db t system pres n aloni d - ean enctesa f eto pancake energized Figurn I .fiele th de2 distributio e inneth t rna coile radi th presencn si f o i e otesa f t pancak s reportedei e fiels beeTh ha d. n cal- culate eacr dfo h coil alon generatrie gth whicn xo e hth field modulus reache s maximuit s m value. Since th e FIGURE 3 self-field of the lest pancake decays rapidly along the Forces acting on the different coils 204 asymmetrie th syste o t e mdu c positio tese th t f panno - the hot spot quench temperatures for ail the magnets cak shows ea Figurn ni . Thie3 s force necessitatee sth at about the same level. A faster discharge means on application of a rugged central flange and the applica- the other hand tha larga t e fractio magnetif no c energy tion of a lOrnrn stainless steel liner inside the ENEA wiltransferree b l e otheth o rdt coils. 6T coil. The liner has to prevent possible radial move- Figure 4 gives an impression of the current decay in ments in the pancake wound 6T solenoid. the 3 subsystems. It is shown that the current in the In absence of a lest pancake, the axial forces re- 6T section rises initially from 4.5 kA up to 5.4 kA. The ported in Figure 3 have practically the same values t spoho t temperature coile respectivele th sar n si K y34 but, in the vertical direction, the six coils of the mag- (12T), 42K (9T) and 43K (6T). net are subjected only to a total downward force of abou tons3 t . This forc generates ei asymmetn a y db - ric distributio e iroth n f nmasseo s with respece th o t t 10 middle magnete planth f eo . Sirènes: e SULTAN-11e mesth Th f o h 1 facility with the imposed boundary conditions consists of 448 I I ï isoparametric elements wit node0 h2 d reduce san - din T w3 2 tegration interfac6 (C3D20R13 f o ed element)an s (IN- ) place8 R d coilo TE aroun ordetw n i s e sepo t drth - 0.01 0.10 1.00 10.00 100.0 arate thes componento tw e « e fro e th flang mth d an e « (sectu ) stainless steel cylinder. The mesh represents only one SULTAe halth f fo N facility becaus magnetie eth c loads FIGURE4 and the geometry are symmetric with respect to the xy Fast discharge of SULTAN-III coil system; plane. The displacements of the flat sid« of the flange current decay ofthe three subsystems fulle ar y contained coilo cornere tw s th e nea n th I .f r so the corner between flang d cylinderan e , separatt ebu MAGNET 48 . T coincident nodes have been introduced in order to have The present ENEA 6T coil is a pancake wound sole- distinct normals of coatact. The coincident nodes have noid, wher doubl0 e4 e pancake stackee sar d together11. also been constrained to have the same displacements In orde reaco t r splie hth t pair configurationT 6 e th , and rotations by mean» of the multi-point constraints relations (MFCS). Three different materials have been coil will be divided into two halves. The new mechani- considered in the model: stainless steel and two differ- cal structure has been designed in a way to use as much ent equivalent materials for the two coils. The stainless s possibla e existinth f eo g components maie - Th .n ad steel and the ENEA coil have been supposed Isotropie ditional new element is a system of two stainless steel coilT whil9 e s haveth e been assumed orthotropice Th . central flange spacersa d flangee san Th .necessare sar y loads due to the Lorentz force have been applied to the to obtain two separate semi-coils. The spacer, 10cm mode bods a l y force. thick, will be inserted between the flanges to withstand axial compress!ve forces. A centrai hole in the spacer As a result the peak value of the von Mises stress will allow radial introductio sample th f no e holder fo r are: conductor tests. This assembl rathef yo r thick stainless « on the ENEA coil 80 MPa steel elements necessitate its own cooling circuit. • on the liner 200 MPa « on the 9T coil 180 MPa a MP flange 0 th 30 n e o • These values are well below the allowable stresses for d ENEan AT 9 e th s a thr flangee efa th lines d A .an r coils are concerned, the calculated stresses are rather high but still within the capabilities of these windings. Energy Discharge: The magnet system is divided int osubsystem3 r powerinfo s d energgan y discharge sectio T current)A 12 k sectio T vize 6 9 n( th .e 2 th ,n(1 kA current) and the 6T section (4.5 kA current). In case of detection of a normal zone in one of the coil powee sth r supplie switchec sar frof mage dof m th - FIGURE5 net system and dump resistors are switched into the e splitteTh d magnet wit centrae hth l flange. circuit. The hole to insert test cables is shown coilT 9 sw wilne e l operatTh currena t ea t density being about twice as high as in the other coils. As the dissipated energy inside a conductor during quench is For this purpos epattera f canano l serie- s ob wil e b l about proportional to / j* dt, the energy of the new taine machininy db space e t flangee th ge gth d o t r an s dischargee b o t coil T s 9 ha s d faste coilw r ne sfro e mth some rectangular quarries close stainlesy db s stee- he l than fro existine mth gdetectioe coilth t sa normaa f no l lium tight welded strips. Once the two halves of the zone magnetsth e f insido e .eon This enable keeo t s p coil wil e reassembledb l internae th , l surface will have 205 to be machined in order to fit for coupling with the othee linerth n r O han. d surface layers wit reduce. ha d roughnes requiree sar loweo dt frictioe rth n heat gener- ated between the stainless steel liner and the EN E A coil by movements produced by the compression Lorentz forces. The use of special antifriction materials is also being considered. The new general view of the magnet showing the hole where the superconducting cable will be inserted for tests is given in Figure 5. COILT O . DESIG5 E S TH P NO Starting point for the 9T coils is the use of the 12T niobium-tin composite conductors as developed for SULTAN stage II1-3. Use of these pre-reacted conduc- tors in the new 9T coils saves costs for development testin d productiogan e conductorsth f no . These conductors have been designe a trans r fo d - z krnrsutlnt port current of 6 kA at 12.2 T and 4.5K.AI 9T, being the maximum field on the winding package of the new middle magnets, the conductors can carry a transport FIGURE 6 takinA k curren 2 g1 f into t o accoun e samth t e mar- Global sketch of the ECN coil criticae ginth o st l current e criticaTh . l current wile b l (9TabouA k , 8 4.5K)1 t . Lorentz-forces acting on the conductors are roughly a facto coilsT large2 12 f t wilI o r.e e rb l th tha r fo n clear that additional structural material is necessary I ConductorPS I Concept: PS concepe é Th th r fo t o support e conductorsth t , therefor stainles2 e s steel conductor is based on a flat cable of NbjSn strands strip e soldere ar se conductors th o t d . Both coils will cooled by forced flow of supercritical helium. The con- e layeredb layer0 turn4 1 2 : f o s each constitute th e ductor, designed to carry 12 kA at 9T and 4.5K is winding pack of 240 turns. The outer stainless cylin- shown in Figure 7. It consists of a pre-reacted flat der will be used to support the coils. cable sandwiched between two copper strips and po- Both coils wil e cooleb l y supercriticab d l helium. sitioned between two cooling tubes. To support the coiT 9 l eacN EC h e layeInth r will hav s individuaeit l hoop stress of about 190 MPa acting on the conductor coolant inlet, locate outee th t rda radial side-planf eo e coilreinforcino inth tw , g stainless steel 8 strip1. f so the coil, coolant outlets are on the transition from one eac m assemblyaddee e m har th o dt sevel Al . n conduc- layer to another so that 2 layers have one common out- tor components are joined together in one operation by coiI lPS onlheliue e th yon let n I m. inleoutled an te ar t soldering. The superconducting cable design is based foresee outee th t rna radia l plane. o na Nb diamete3m Snm stran6 0. rf e do madth y eb In the ECN coil 5 individual conductor lengths (about interna r bronzo n ti l e technique o reac T e speci. th h - longm 0 , 18 being sufficien layers2 r c usedfo t ar ) n I . fied critical current, a cable desigrt was made using 112 the PSI coil only one conductor length of about 900m strandcablino tw n si g stages. is used r botFo h. coil e electricath s l connectione ar s locateoutee th t r da radia l side-plane. Coil character- ECN Conductor Concept: The conductor is based istics are given in Table I; a global sketch of the ECN NbjSn o n materia powdeN l produceEC e r th pro y d-b coi s givei l Figurn ni . e6 cess. Stran diametern di materiam usee 1m b s d.i o t l Thirty-six strand cablee sar flaa o td t Rutherfor- dca ble, having a thickness of 1.84 mm and a width of 18.1 mm. The cable is enclosed in copper necessary for addi- tional stabilization and current by-pass in case of nor- TABLE I l zones coppeo ma Tw . r cooling channel presente ar s . Coil Characteristics equivalene Th t diamete coolane th f ro t channel 2.7s si 5 mm. Helium massflo conductoe th wg/sn 5 i - 2. .s Av r i ECN PSI erage pressure drop ove conducto1 r r kPalaye0 6 s .i r Operation»! current (9T.4.5K) kA 12 12 Two stainless steel strips (thickness 1.5 rnm) are placed Critical current (9T.4.5K) kA 18 18 Numbe f windingo r » 240 240 bottod conductoe an th p f omo nlo reinforcements a r . Number of layer» - 10 10 The cable is heat treated for 48h at700°C in argon at- Numbe windingf o r r laye»pe r 24 24 mosphere. After heat treatment the cable is soldered to Ampere- turns MA-tumi 2.88 2.88 Inner diameter winding package m 1.088 1.064 the other components of the composite conductor. The Outer diameter winding package m 1.270 1.276 component7 conductoe th f o s solderee ar r d together Lengt f windinho g package m 0.6T5 0.676 in onle solderinon y g run o havT . a gooe d bonding Average length of 1 winding m 3.71 3.68 Total conductor length m 891 886 between copper and stainless steel the stainless steel OTerall current dentity (wind.package) MA/mJ 46.8 40.3 strip pretinnede sar . 206 TABLE II n .300 A current leads Characteristic dimensions composite Nb3Sn conductor 4K helium Inlet pumping lube BOK helium Inlet ECN PSI 90K helium outlet Cross section mm2 % mm1 % Nb3Sn + Powder+ Nb 11.31 4.7 8.0 2.87 4K helium outlet Copper 109.84 45.5 97.0 34.88 Bronze + Ta 7.8 2.8 vncuuiti tulle Stainless steel 78.00 32.3 94.6 34.0 radiation shield Insulation 21.12 8.8 25.6 9.2 cryostat Helium 15.12 6.3 25.1 9.02 secondary winding Solder 5.81 2.4 15.3 5.50 J-T valve .primary winding Voids - - 4.7 1.69 NVV 500 4ieat exchanger Tabl I giveeI dimensione composito th s tw e th f eso helium supply conductors and the distribution of the different mate- rials. . MEASUREMEN6 T OPTIONS current connection Short Sample Teils:orden I reaco t r h high rep- etition rates for sample measurements a special sam- ple insert unit allowing fast sample insertion without warming up the whole magnet system is foreseen. This unit - similar to a telescope - will be mounted together FIGURE 7 with the sample holder on top of the facility. The insert Top of sample holder with cryostat overaln a diametea uni s d ha tl an lengt 0.5mf m ro 6 f ho . and superconducting transformer A superconducting transformer wil usee b l s curda - rent supplshore th r t yfo samples . Suc hsolutioa n elim- inates large and power consuming current leads. The transformer is specified for currents up to 50 kA. It sample th wil f mountee - o b elin p holdee to th n do n i r sert uni avoio t t d larg complicated ean d current feed through» (see Fig. .7) Pancake facilite Teil:Th y wil modifiee b l sucn di h thay twa alsa o small pancake NET-conductorf so n sca REFERENCES be tested in the midplane gap of the magnet system. A free gap region of 100 mm is reserved for the windings e pancake o desige th fe spli th th r f t no coiFo . l system 1. J.D.EIen, N.Sacchetti, G.Vecsey, 12 Tesla Split Coils "standarda " double pancak s takeei n into considera- SULTAe foth r N Test Facility, Project Proposal, tion. This standard 8 doubl1 = 9 e pancakx 2 s ha e January (1988) turns; the inner diameter of the pancake is 1.6 m and 2. J.A.Roeterdink, M.W.Brieko, A.C.Gijze, H.V. Mer- the outer diameter is 2.1 m. The width of the double tens, IEEE Trans.Mag. Vol.24, (1988),1429 pancake is 76 mm. The conductor current taken into account for the design is 16 kA. The position of the . G.Pasztor3 , B.Jakob, I.Horvath, P.Ming, G.Vecsey standard pancake is given in Figure 3. The maximum P.Weymulh, IEEE Trans.Mag. Vol.24 (1988),1080 field in the gap region is 11.7 T; the self field of the 4. G.Pasottiet. al., IEEE Trans.Mag. MAG-17, (1981), pancake (0.7T includes )i thin di s figure. 2007 Next page(s) left blank 207 CONSTRUCTIO HIGH-FIELW NE A F NO D LABORATORE TH T YA NATIONAL RESEARCH INSTITUT METALSR EFO , JAPAN H. WADA, K. INOUE, T. KIYOSHI, T. ASANO, K. ITOH, H. AOKI National Research Institut r Metalsefo , Tsukuba City, Ibaraki, Japan Abstract The National Research Institut Metalsr fo e ( bees NRIha n) M developin seriega high-fielf o s d magnets sinc year2 eparticipana s sa t organizatio Science th Technologd n i nan e y s Multi-CorAgency)' A ST ( e Research Project on Superconductivity which focuses on the research and developmen high-temperaturf o t e oxide superconductors. These magnets should be used to evaluate the properties of superconductors developed in the project. They will eventuall gatheree yb Tsukubt a d w ford ne an ama high-field laboratory, a center for various kinds of experiments using high magnetic fields. The magnet development program involves an clasT 0 8 s long-pulse magnet systemclasT 0 4 s ,a hybri d magnet systema , 20 T class superconducting large-bore magnet system, and three NMR quality superconducting magnet systems. INTRODUCTION Stimulated by the recent discovery of high temperature oxide superconductors Science ,th Technologd ean y AgencyGovernmen, ) A (ST f to Japan, initiated an integrated research project in 1987 which aimed at supporting research activities on oxide superconductors. The project was named "the Multi-Core Research Projec Superconductivityn o t s i d an , " collaboratioe baseth n o d n among research organizations belongine th o gt " Multi-Core e ider STATh fo a . sucs i " h that each participant organ- ization offers its relevant researchers and facilities as (a) research core(s) for (a) research area(s) where it has high research potential, d interactan s with research organizations both insid outsidd an e STAe th e . Research cores and responsible institutes are shown in Table 1. Thus Nationae ,th l Research Institut Metalsr fo e (bees NRIha n) M contributin Multi-Core th o gt e projec l threal n eti areas majos It .r contributio projece o develoth t o s t ni t seriea p f largo s e high-field magnets useful for the evaluation of the properties of superconductors 209 develope projecte th n e developmeni dTh . t progra n facI m t Involvee th s development of an 80 T class long-pulse magnet system, a 40 T class hybrid magnet system, a 20 T class superconducting large-bore magnet system thred qualitR ,an eNM y 12~1 6clasT s superconducting magnet systems. These magnets wil installee lb NRIM'a n laboratorw i d ne s o t y be constructe Tsukubat a d . Table 1: Organization for Multi-Core Project. Areas Research Cores Institutes Fundamental Theory NRIM13 Data Base NRIM Synthesis Materials Exploration NIRIM2' & Materials Purification NRIM Fabrication Thin Films NRIM Single Crystals NIRIM Microfabrications IPCR33 Composite Conductors NRIM Space Environment Application NSDAJ"" Evaluation High-Field Magnets NRIM & TEM NIRIM Characterization Neutron Irradiation JAERI5> Application Patent Development RDCJS) 1)NRIM (National Research Institute for Metals) 21NIRIM (National Institute for Research in Inorganic Materials) 3)IPCR (Institut Physicaf eo Chemicad lan l Research) NSDAJ (National Space Development Agenc f Japanyo ) JAERI (Japan Atomic Energy Research Institute) RDCJ (Research Development Corporation of Japan) MAIN SPECIFICATION MAGNEE TH F SO T SYSTEMS (A) High field magnet systems clasT 0 8 s lon® g pulse magnet system Main specifications are shown in Table 2. e mosTh t critical proble systee develoo t th s o i mt m p conductor materials which have high electrical conductivit d largan y e mechanical strength. Strengthenin s accomplishegi alloyiny db g which usually deteriorates conductivity o overcomT . e this dilemma, intensivD & eR work on Cu-based alloys is going on, and candidate materials at present are CuNb and CuCr. Coil design is also being studied by carrying out a 210 serie destructivf o s e coil testcapacitoJ M s6 usin1. e r gth ban k newly installed. Attainable fields so far are in the range of 50 T. This system wil completee lb 1992n i d . Tabl Specification: e2 ClasT 0 8 s f o Lons g Pulse Magnet System. magnet maximum field; 80 T, operation temperature; 77 K conductor materials; CuNb, CuCr rise time to 80 T; 4 msec, bore; 10 mm © 40 T hybrid magnet system Main specification e show ar sTabln i n . e3 successfuo t y ke e lTh constructio thif o n s syste agais i me nth development of conductor materials with good conductivity and large strength. R & D work similar to that for the pulse magnet system is being carried out d promisinan , g material e CuNw b d see sno o an b t m Cu-Al203. Coil desig d fabricatioan n n studie spolihelie baseth n o d x concep proceedine ar t g which include meltin larga f o ge volume ingod an t spark cuttin a poliheli f go f xsuco coiht ingotlou monohelie Th . x coil concept is being examined under a collaboration program among NRIM, MIT and Toshiba. R & D studies on superconducting coil materials are also underway. Different fabrication methods, such as the Nb-tube method, are being applie Ti-dopeo t d d Nb3Sn conductors. Attention pais i de th als o t o fabricatio NbsAf o n l conductors. Coil design involves gradina n i g pancake consisting of 2 outer NbTi and 2 inner Nb3Sn conductor sections. All conductors are designed as cryostabilized with Cu, while stabiliza- tion using Al is also under consideration. e completioTh f thio n s system n 1994wili e lb . 211 Table 3: Specifications of 40 T Hybrid Magnet System. normal conducting magnet (water-cooled) maximum field; 25 T, operation temperature;~ room temperature conductor materials; CuNb, CuAlsOs coil geometry; polihelix bore diameters;3, T 0 4 r bore; 400 mm paid for Ti-doped Nb3Sn conductors fabricated by the Nb-tube method. In the coil design NbTd Nban i3S n coil e centricallsar y placed, formin- 3 g 4 layers; NbTi/NbaSn/NbsSn/NbaSn. 2 outer coils will be cryostable. Al stabilize undes i r r consideration. 1.8 K operation is assumed for the magnet system design, where saturated superfluid helium is used as the coolant. This system be completed in 1993; Tabl : Specificatione4 SuperconductinT 0 2 f o s g Large-Bore Magnet System. maximum field; 20., 5T stored energy; 47 MJ, operation temperatureK 8 1. ; conductor materials; NbTi, (Nb,Ti)3Sn, , K 8 1. t r 20.5a bores c m fo 6m d Tan 0 4 ; , K 8 1. t a d an T 1S r çfo m ôm 0 16 310 mmçô for 15 T and 1.8 K. 212 (B) High-resolution magnet systems, high-resolutio3 n magnet undew systemno re constructionsar n I . each system design parameters have carefully been selected, enabling the specifications, such as sample space geometries, operating temperatures, field decay time, etc.wele b lo ,t compatible witrequiremente th h r sfo measurements of very high quality. Thus, they will have the highest possible performance attained using the current cryo-magnet technology. (D Broad-band solid state NMR study system: Main specifications are show Tabln i n . e5 Table 5: Specifications of Broad-Band Solid State NMR Study System. operating fields; 0-15.5 T, field homogeneity; 10~5/10 mmdsv(in driven mode), field decay; 10~c/hr(in driven mode), bore; 70 mm 0 at 4.2K, sample 0 spacem m 0 5 ; operating temperatures; 0.35-300 K(He/ He), 4 equipment; high power puls R spectrometer(4KWeNM 3 ) D Hig( h resolution solid stat R studeNM y system: Main specifications are shown in Table 6. Tabl : Specificatione6 High-Resolutiof o s n Solid State NMR Study System. operating field; 11.7, T 5 bore; 89 mm 0 at room temperature, field homogeneity; 2.6xlO~6/20mm dsv(superconducting shim), field decay; 8xlO~B/hr, probes; cross-polarization/magic angle spinning probes. 213 (D Quantum oscillation study system: Main specifications are shown in Tabldilutioa ( e7 n refrigerato equipped)s i r . Tabl : Specificatione7 Quantuf o s m Oscillation Study System. operating fields; 0-1, 6T homogeneity; 10~VlOnim dsv, t 4.2K Various equipment probed an s s wil developee b l d offerean d r fo d supportin varietga measurementf o y s using these high-resolution magnet systems. These systems, together with the measurement equipments and probes, will be completed in 1990. SUMMARY researce Baseth n do h potential establishe varioun di s research programs involving application f superconductivityso , suc - s thosha nu n o e clear fusion, NRIM has been developing a series of high-field magnets, each havin world'e gth qualityp to s . These magnets wil e installeb l n di the near future, in a laboratory to be newly constructed, and offered for collaborative research programs propose eithey b d r NRIM's research staff members or outside researchers. Emphasis will then be placed on international collaboration. Already, NRIM has been establishing collaborative relationship with the Francis Bitter National Magnet Laborator MITf yo , Cambridge High-Fiele th d an , d Laborator f Centryo e Nationa Researcha l e ld e Scientifique, Grenoble e examplOn . f suceo h collaboratio internationae e founth b n y i d nma l worksho " Develop n o p - ment and Applications of High-Magnetic Fields " to be held in October 198 Tokyon i 9 . This wil e hostelb NRIMy b d , with participants froe th m other 2 institutes. 214 LIST OF PARTICIPANTS Ando, T. Japan Atomic Energy Research Institute Tokai Research Establishment Tokai-mura Ibaraki-ken, 319-11, Japan Aral, K. Electrotechnical Labour 1-1-4 Umezono, Tsukuba Ibaraki 305, Japan Beard, D. US Department Energy ER-531 Washington, D.C. 20545, United State Americf so a Collings, E. Battelle Advanced Materials Group 505 King Avenue, Columbus Ohio 43201-2693, United States of America Cornells, J. SCK/CEN Boeretang 200 B-2400 Mol, Belgium Dresner, L. Ridgk Oa e National Lab. P.O2009x RidgBo k . ,Oa e 37831-8054N T , United State Americf o s a Elen, J.D. Stichting Energieonderzoek Centrum Netherland Netherlands Energy Research Foundation ECN Postbus l 175 PettenG Z 5 , Netherlands Higuchi, N. Electrotechnical Labour 1-1-4 Umezono, Tsukuba Ibaraki 305, Japan Ivanov, D.P. Kurchatov's Atomic Energy Institute Kurchatov's Sq. Moscow, 123182, Union of Soviet Socialist Republics Klimenko, E.Y. Kurchatov's Atomic Energy Institute Kurchatov's Sq. Moscow, 123182, Unio Sovief no t Socialist Republics Machalek. ,M DepartmenS U t Energy ER-531 Washington, D.C. 20545, United State Americf o s a Matsui, Y. National Institute Research Inorganic Materials 1-1 Namiki, Tsukuba Ibaraki 305, Japan Matsushita, T. Departmen Electronicsf to , Kyushu University 6-10-1 Hakozaki, Higashi-ku Fukuoka 812, Japan 215 Mavrin, A.S. USSR State Committee on the Utilizatio Atomif no c Energy Staromonetny Per6 2 . Moscow, 109180, Union of Soviet Socialist Republics Miervini. ,J CEC TeaT NE m Max Planck Institu Plasmaphysir fü t k Boltzmannstrasse 2 D-8046 Garching, Germany Mitchell. ,C CEC NET Team Max Planck Institu Plasmaphysir fü t k Boltzmannstrasse2 D-8046 Garching, Germany Murase. ,S Toshib D CenteaR& r 4-1 Ukishima-cho, Kawasaki-ku Kawasaki 210, Japan Nagata. ,A Mining Colleage, Akita University Electric Ind. Ltd. 1-1 Tegata-gakuen-machi Akita 010, Japan Nikulin, A.D. Bochvar' Uniol sAl n Inorganic Materials Institute Rogoff's Str5 . Moscow, 123060, Unio Sovief no t Socialist Republics Noto, K. Facult Engineerinf yo g Iwate University 4-3-5 Ueda, Morioka Iwate 020, Japan Ohmatsu, K. Osaka Research Lab. Sumitomo Electric Ind. Ltd. 1-1-3 Shimaya Konohana-ku, Osaka 554, Japan Ohnishi. ,T Electrotechnical Labour 1-1-4 Umezono, Tsukuba Ibaraki 305, Japan Okada, T. Institute of Science Ind. Research Osaka University 8-1 Mihogaoka, Ibaraki Osaka 567, Japan Sacchetti, N. ENEA FrascatE ,CR i Via E. Fermi-27-C.P. 65 00044 Frascati (Rome), Italy Sato, K. Osaka Research Lab. Sumitomo Electric Ind. Ltd. 1-1-3 Shimaya Konohana-ku, Osaka 554, Japan 216 Sekine. ,H National Research Institute for Metals Sengen 1-2-1, Tsukuba Ibaraki 305, Japan Shen, S. Lawrence Livermore National Lab. L-643, P.O 551x .Bo 1 Livermore, CA 94550, United States of America Suenaga. ,M Brookhaven National Labour Bldg. 480 Upton 11973Y ,N , United State Americf so a Tachikawa, K. Faculty of Engineering Tokai University Kitakaname 1117, Hiratasuka Kanagawa 259-12, Japan Tada. ,N Hitachi Research Lab. 4026 Kuji-cho, Hitachi Ibaraki 319-12, Japan Takeuchi. ,T National Research Institut Metalr fo e s Sengen 1-2-1, Tsukuba Ibaraki 305, Japan Tanaka. ,Y Superconductin D DeptgR& . Yokohama R&D Lab. Furukawa Electric Co., Ltd. 2-4-3 Okano, Nishi-ku Yokohama 220, Japan Van de Klundert, L.J.M. University of Twenty P.O. Box 217, Enschede NL-7500-AE, Netherlands Wada, H. National Research Institute for Metals Sengen 1-2-1, Tsukuba Ibaraki 305, Japan Yamamoto. ,J National Institute for Fusion Science Furo-cho, Chikusa-ku Nagoya 464-01, Japan Yoshida. ,H Research Reactor Institute, Kyoto University Kumatori-cho, Sennan-gun Osaka 590-04, Japan Kupitz, J. (Scientific Secretary) IAEA 0 10 P.Ox Bo . A-1400 Vienna, Austria 217 HOW TO ORDER IAEA PUBLICATIONS An exclusive sales agent for IAEA publications, to whom all orders inquiried an s shoul addressede db bees ha ,n appointed followine inth g country: UNITED STATES OF AMERICA UNIPUB, 4611-F Assembly Drive, Lanham, MD 20706-4391 in the following countries IAEA publications may be purchased from the sales agent r booksellerso s liste r througdo h major local booksellers. Payment can be made in local currenc witr yo h UNESCO coupons. ARGENTINA Comisiôn Nacional de Energfa Atômica, Avenida del Libertador 8250, RA-1429 Buenos Aires AUSTRALIA Hunter Publications, 58 AGipps Street, Collingwood, Victoria 3066 BELGIUM Service Courrier UNESCO, 202, Avenue du Roi,B-1060 Brussels CHILE Comisiön Chilene de Energfa Nuclear, Venta de Publicaciones, Amunategui 95, Casilla 188-D, Santiago CHINA IAEA Publications in Chinese: China Nuclear Energy Industry Corporation,Translation Section, P.O. Box 2103, Beijing IAEA Publications other than in Chinese: China National Publications Import & Export Corporation, Deutsche Abteilung , Beijin88 x gBo , P.O. CZECHOSLOVAKIA S.N.T.L.,Mikulandska4,CS-11686 Prague 1 Alfa, Publishers, Hurbanovo nemestie 3, CS-81589 Bratislava FRANCE Office International de Documentation et Librairie,48, rue Gay-Lussac, F-75240 Paris Cedex 05 HUNGARY Kultura, Hungarian Foreign Trading Company, P.O. Box 149, H-1389 Budapest 62 INDIA Oxford Boo Stationerd kan Par, y17 kCo. Street, , Calcutta-706 001 Oxford Boo Stationerd kan y Co.,Scindia House Delhi-11000w Ne , 1 ISRAEL Heiliger & Co. Ltd. Kere3 2 n Hayesod Street, Jerusalem 94188 ITALY Libreria Scientifica, Dott. Lucio de Biasio "aeiou", Via Meravigl , 1-201216 i 3 Milan JAPAN Maruzen Company, Ltd5050x Bo , P.O, . 100-31 Tokyo International PAKISTAN Mirza Book Agency, 65, Shahrah Quaid-e-Azam, P.O. Box 729, Lahore 3 POLAND Ars Polona-Ruch, Centrale Handlu Zagranicznego, Krakowskie Przedmiescie 7, PL-00-068 Warsaw ROMANIA llexim, P O. Box 136-137, Bucharest SOUTH AFRICA Van Schaik Bookstore (Pty) Ltd 724x , ,Bo P.OPretori. a 0001 SPAIN Di'a Santose zd , Lagasc , E-2800a95 6 Madrid Diaz de Santos, Balmes 417, E-08022 Barcelona SWEDEN AB Fritzes Kungl. Hovbokhandel, Fredsgatan 2, P.O. Box 16356, S-103 27 Stockholm UNITED KINGDOM HMSO, Publications Centre, Agency Section, 51 Nine Elms Lane, London SW8 5DR USSR Mezhdunarodnaya Kniga,Smolenskaya-Sennaya 32-34, Moscow G-200 YUGOSLAVIA Jugoslovenska Knjiga,Terazij , YU-110036 x Bo 1, P.O . Belgrade27 e Orders from countries where sales agents have not yet been appointed and request r informatiosfo n shoul addressee db d directl: yto Divisio f Publicationno s International Atomic Energy Agency Wagramerstrasse 5, P.O. Box 100, A-1400 Vienna, Austria