Executive summary

Fusion Engineering and Design North-Holland, Amsterdam

EXECUTIVE

7 (IOgg)

3-12

3

SUMMARY

A multinational program of cooperative research, development, demonstrations, and exchanges of information on superconducting magnets for fusion was initiated in 1977 under an IEA agreement. The first major step in the development of TF magnets was called the Large Coil Task. Participants in LCT were the U.S. DOE, EURATOM, JAERI, and the Wpartement Federal de I’InteriCur of Switzerland. The goals of LCT were to obtain experimental data, to demonstrate reliable operation of large superconducting coils, and to prove design principles and fabrication techniques being considered for the toroidal magnets of thermonuclear reactors. These goals were to be accomplished through coordinated but largely independent design, development, and construction of six test coils, followed by collaborative testing in a compact toroidal test array at fields of 8 T and higher. Under the terms of the IEA Agreement, the United States built and operated the test facility at Oak Ridge and provided three test coils. The other participants provided one coil each. information on design and manufacturing and all test data were shared by all. The LCT team of each participant included a government laboratory and industrial partners or contractors, ss shown in fig. 1. The last coil was completed in 1985, and the test assembly was completed in October of that year (see fig. 2). Over the next 23 months, the G-coil array was cooled down and extensive testing was performed. Results were gratifying, ss tests achieved designpoint performance and well beyond. (Each coil reached a peak field of 9 T.) Experiments elucidated coil behavior, delineated limits of operability, and demonstrated coil safety. This special issue of fision Engineering and Design makes available to all potential users the LCT results and experience, which are described in detail sufficient for useful guidance of further work on toroidal superconducting magnets.

Specifications Coil specifications were written to ensure compatibility with the test array and relevance to anticipated tokamak reactors while providing significant freedom in making important design choices. Performance requirements and critical dimensions were precisely defined,

but the design of conductor, winding, and structure were left to each design team. The size and shape of the coils were specified to leave only a reasonable extrapolation of fabrication methods and test results to produce full-size reactor coils. A D-shaped bore, about 2.5 by 3.5 m, was specified. Conductor current had to be in the range of 10

JAPAN

PARTICIPANT

LABORATORY

TEST COIL MANUFACTURER TEST

FACILITY

DESIGN CONSTRUCT INSTALL

COILS

PROVIDE

SPECIAL

OPERATE

FACILITY

TEST

EGUIPMENT

AND ANALYZE

Fig. 1. Responsibilities

0920-3796/88/%03.50 (North-Holland

Physics

@ Elsevier Publishing

Science Publishers Division)

B.V.

in the IEA Large

Coil Task.

SWITZERLAND

4

Ezecutive

Fig. 2. Test stand

with

to 18 kA, ensuring that manufacturing processes would be directly applicable. The peak field requirement was set at 8 T in the horizontal midplane, which was considered to be the practical operating limit for the most highly developed superconductor (NbTi) at the temperature of boiling helium (4.2 K). Coils had to be designed to enable cooling with helium gas to superconducting temperature in 120 h if facility capabilities were not limiting. During operation, the cooling mode could be either a PB bath of LHe at atmospheric pressure or FF of helium at high pressure. With tokamak requirements in view, coils had to be designed for tolerable ac losses in specified poloidal field pulses and for cryostability to recover spontaneously to the fully ( i.e., the ability superconducting state from normal zones produced by credible internal events). Coil and facility specifications also provided for extraordinarily thorough testing. Structures had to be strong enough to allow operation to the limits of superconducting capability. Diagnostic instrumentation sufficient for thorough analyses of behavior was required. By agreement among participants, not all paragraphs of the ORNL coil specifications were mandatory for nonU.S. coils. However, as a minimum, all coils had to be

Summary

six LCT

coils-October

1985.

capable of operating reliably as a background coil in all prescribed tests of other coils. The test program goals were first to obtain conclusive data from experiments that determined the ahilil! of each coil to meet all specified design-point reqliirrments and then to determine actual margins by exploring the limits of operability. The planned sequence was: startup tests, including cooldown; preliminary tests to provide calibration data and to prove the capability for operation at full current with other coils deenergized; design-point tests, including internal heating and stability experiments with the coil at 8 T; extended-condition tests to highest currents and fields and extreme out-ofplane loads; warmup; and inspection. Coil safety tests on quench initiation and propagation and interruption of coolant flow during high-current operation were later added to the program. Participants jointly worked out and formally defined a management structure headed by the Executive Committee, which was established by the IEA Agreement. Leadership in technical planning and review was by a Project Officer for each participant. On-site representatives at the test facility actively participated in coil testing and had a voice in day-to-day decisions within

Ezecutive

the constraints by the Executive

Coil

of the Test Committee.

Program

document

.%mmary

issued

The principal distinctive features of the six X”P coils are compared in table 1. Numerous other significant differences in details reflected differences in both design philosophies and perceptions of needs of future fusion magnets. Figure 3 illustrates the great differences among the conductor configurations; the left column shows PBcooled designs, and the right column shows FF designs. As befitted their purpose, all coils were heavily equipped with diagnostic sensors, more than 199 on each coil. These included voltage taps for detection of resistive zones, strain gauges on conductor and structure, sensors to measure temperatures and pressures, acoustic emission sensors, and transducers to *measure displacement of the winding relative to the coil case. Five coils had heaters embedded in the winding that were used to simulate nuclear heating or to drive a length of conductor normal to determine the stability of the coil. LCT coil teams settled on a variety of manufacturing and assembly methods. The winding packs of the EU and CH coils, made of fiberglass-wrapped, internally cooled conductors, were impregnated en bloc with epoxy resin. The EU winding was potted in a mold, after which it was inserted in the coil case, and bladders were injected with epoxy to fill the clearance space, The CH winding was placed in its case, and epoxy was then injected to impregnate the winding and fill the space between winding and case. The conductor for

characteristics

By the time the LCT teams began their work, magnet technology offered several promising options for meeting the special requirements of tokamak coils. Favored candidates were either NbTi or NbsSn superconductors in various composite conductor forms, cooled either by immersion in a helium bath or by forced circulation of helium through passages in the conductor, wound in pancakes or layers, and supported by an external coil case or by grooved plates. Each concept had advocates, but there was not enough directly applicable experience to ensure selection of the best one. In this situation, not surprisingly, there were significant differences among the choices of the six LCT design teams. Each of the teams proposed to use NbTi rather than NbsSn, which was potentially more capable but definitely more difficult. However, the WH proposal emphasized that with their conductor concept either material could be used. Therefore, DOE directed them to go ahead with the development necessary for NbsSn. On the choice of cooling mode, the teams were evenly divided, three choosing PB cooling and three FF. The two teams designing FF NbTi coils requested the lowest inlet temperature reasonably attainable in the Oak Ridge facility. This turned out to be 3.8 K.

Table 1 Distinctive

features

of the LCT

5

test coils

Participant

United

EURATOM

Japan

Switzerland

Direction

ORNL

States

ORNL

United

States

United ORNL

States

KfK

JAERI

SIN

coil

GD/C

GE/OR

WH

Siemens

Hitachi

BBC

Conductor

IGC/GD

IGC

Airco

Vacuumschmelre

Hitachi Cable

SASM/BBC

Superconductor

NbTi

NbTi

NbsSn

NbTi

Cooling

PB

PB

FF

FF

Design

modea current

NbTi

NbTi

PB

FF

10,200

10,500

17,760

11,400

10,220

13,000

Current density, winding (MA/ml)

27.4

24.7

20.15

24.1

26.6

30.7

Current density, coil (MA/m*)*

15.3

15.7

17.6

16.3

16.0

17.9

Design

6.40

‘6.53

7.53

6.70

6.73

5.95

Winding configuration

Edge wound, 14 layers

Flat wound, 6 double pancakes

In grooves in 24 plates

Flat wound, 7 double pancakes, impregnated

Edge wound, 20 double pancakes

Wound in 11 double pancakes, impregnated

Structure

Type 304L SS, Type 316LN SS, welded case bolted and welded csse

Al alloy 2219-T87, bolted plates

Type 316LN SS, bolted, sealed csse

Type 304LN SS, bolted and welded case

Type 316L/316LN SS, bolted csse

MA turns

(A)

-

Ltd.

;PB: boibng at 0.1 MPa (4.2 K); FF: 1.2 MPa (3.6 K). ‘Average over total cross section of winding and structure %cludes plate structure inside the outermost conductor

in nose region. boundary.

6

Facility

lGC/GE

CHANNELS VACUUMSCHMELZE

IGClGD

Fig. 3. Configurations

AIRCOAVH

of conductors

in the LCT

coils.

the WH coil was heat-treated to form NbsSn and then bent to the contours of the coil, wrapped with insulation, and fitted into grooves in the plates, which were bolted together to make the coil. The JA coil winding, with passages among the conductors for PB cooling, was compressed in the final stages of assembly. The inner and outer rings were shrink-fitted; then, the side plates were forced into place with jacks and bolts. The GD winding was placed in the case, and polyurethane was injected between the case and a membrane around the winding to fill the clearance. For the GE/OR coil, insulating shims were inserted between winding and coil case.

description

The test facility, whose official name was the International Fusion Superconducting Magnet Test Farili ty? consisted of the test stand, holding six coils in a toroit1a.l array, and ancillary systems. The 420-t test stand was enclosed in an 11-m vacuum tank. Each coil was fastened to a central bucking post by wedges inserted through upper and lower collars into slots in the coil’s structure. Torque rings supported the outer corners of each coil against put-of-plane forces. An assembly including a pair of 1.3-m-diam copper coils was mounted on a track that threaded the bores of the test coils. A system of ratchets and wedges actuated by helium-gas-driven pistons moved and locked the assembly. In operation, the coils produced a pulsed poloidal field of 0.2 T at the selected test coil windings. LN cooling was provided through flexible hoses. The entire test stand rested on a structure in which LN intercepted heat conducted from the outside. The vacuum tank was lined with panels cooled with LN and blanketed with superinsulation. (See fig. 2.) The helium cooling system was designed to circulate up to 300 g/s at 1.5 MPa and 3.8 K through the three FF coils while supplying liquid at 0.1 MPa and 4.2 K to the other coils and the bucking post. In this mode, the primary liquefier/refrigerator provided simultaneously 0.2 kW of refrigeration at 4.2 K, 1.5 kW at 3.6 K, and liquefaction of 360 L/h from warm gas. This refrigerator was intentionally not sized to be capable of quick cooldown of the test array or of keeping up with helium boiloff under some test conditions. The fa.rility also included two smaller helium liquefiers, which provided up to 100 L/h when required. Typically, during high-current tests, a reserve of liquid in a 19,000-L dewar was steadily depleted and then was recuperated a.t night and on weekends. Each coil had a separate 12-V power supply with a control system that enabled coordinated ramping of currents. A protection system detected any normal zone in a coil and automatically discharged the coils through individual dump resistors outside the vacuum tank. The transitions from external, room-temperature conductors were through 12 vapor-cooled leads in dewars arrayed around the tank (visible in fig. 2). More than 1600 sensor leads from the coils and test stand came out of the tank to patch panels where those pertinent to the test in progress were connected to the data acquisition system. This system recorded and analyzed data in a cluster of computers linked to the test facility. Chronology

and.

operating

experience

In 1981-82, the facility operations group was formed and began training and acceptance testing. The first coil (from Japan) was delivered inNovemberl982, after testing at Naka. In 1983, the helium refrigerator was commissioned and the JA and GD coils wcrc installed. An attempted cooldown was thwarted by Ira.ks at seal welds of tubing on the GD coil. While repairs

Ezecutive

were being made, the CH coil arrived and was put into the test stand; it was connected structurally and cryogenically but had no high-current leads. This led to operation of the facility for partial-array coil tests in June-September 1984. The EU coil was delivered in November 1984, after having been tested at Karlsruhe. The GE/CR coil was delivered immediately thereafter, leaving only the WH NbsSn coil to be completed. After delays caused by various technical difficulties, that coil was delivered in August 1985. It was promptly installed along with the pulse coil system, thus completing the test stand assembly in October 1985. The six-coil test program was accomplished over the next 23 months. The chronology of operations and tests is briefly outlined in table 2. (The coil tests are more fully described later.)

Summary

7

and alternating periods of preparations for testing). For the 22-month period from October 1985 until September 1987, the fraction was 0.57. Table 3 lists principal causes of downtime or necessary, unplanned activities.

Table 3 Facility problems

with

coil testing 1987

between

Delays

system

Data acquisition Planned services of operation

delayed

Air leakage into helium Abnormal heat leaks Leak in cooldown heat exchanger Leak in oil cooler Helium compressors Other helium system components Test coil current systems Pulsed-field

Table 2 Chronology

that

November 1985 and September Problem area

system (air, water,

and electricity)

full test array

Locate and repair leaks and evacuate tank Start cooldown and replace leaking heat exchanger Cool down and begin superconducting Single-coil checks and tests to full current Test controls in six-coil operation Six-coil tests at rated current and 8 T Tests with pulsed poloidal field Single-coil high-temperature and high-current tests Six-coil higher-field tests Five-coil tests with extreme out-of-plane loads Coil safety tests Highest-symmetric-field tests Warmup

(d)

124 14

29 . 17 31 3 5 7

8 22 260

Nov.-Dec. Dec.-Jan.

1985 1986

Jan.-Feb. Mar.-June June-July July-Nov. Dec.-Mar. Mar.-June

1986 1986 1986 1986 1987 1987

July-Aug. 1987 Aug. 1987 Aug. Sept. Sept.

1987 1987 1987

The lengthy period of sustained operation produced useful information on performance of supporting systems and overall availability of the facility for use in coil tests. The helium refrigeration system was of particular interest. Six-coil operation confirmed that the capacity was adequate, although cooldown and recovery after some tests were undesirably slow. The performance of vacuum and LN systems met all requirements. The coil protection system also worked quite well. Coil instrumentation and data acquisition systems proved capable of acquiring abundant, reliable data for the penetrating analysis of coil performance that was the goal of the program. An exception to the generally satisfactory performance was the system for moving and locking the pulse coils, which became inoperable after tests of the first three coils. Repairs would have required warmup; so, rather than extend the program, plans for testing the other coils in pulsed fields were dropped. Facility availability was defined as the fraction of time spent in previously planned activities (coil testing

After the final test, the assembly was gradualI\ warmed up to room temperature. On October 9, 1987, the vacuum tank lid was removed after having lIeen sealed for more than 22 months. After inspections of the test stand, partial disassembly began. The JA coil was removed in November, the EU coil in December, and the CH coil in January. Each was further inspected, packaged, and turned over to the owner for shipment,

Single-coil

tests

The first tests of each coil were essentially single-coil tests. After electrical checks and leak tests at room temperature, temperature distributions in each coil were monitored during cooldown, and cold leak tests were performed. Subsequently, each coil was independently energized for various tests. For each coil, in turn, the current was gradually raised, with pauses to analyze diagnostic data at each new level, until the rated current was reached. No evidence of training and no impediment to operation were encountered with any coil in this series of tests, which: l

l

l l

l

on

measured characteristic parameters such as inductance and dump time constant; verified performance of power supplies, quench detectors, and dump systems; measured heat leakage into each coil; recorded strains, acoustic emissions, and voltage spikes during charging and discharging; and demonstrated ability to operate stably at rated current. The nature of the the coil design and

stability test varied, on special provisions

depending for pulsed

8

Executive

heating. Each PB-cooled coil had been designed to recover spontaneously from large normal zones. Tests of each coil demonstrated recovery after 5 to 11 m (half or full turn) of conductor was driven normal by attached heaters. Each FF coil had been designed to tolerate some amount of heat without formation of any normal zone. (The chosen design goal varied greatly among the three coils.) The designed stability margin of the CH coil was verified by tests in which a heater embedded in the winding was pulsed to produce up to 10 J over a 0.6-m length without creation of a detectable normal zone. The EU coil was tested by pulsed heating of helium supplied to one coolant channel, which was increased until current-sharing was observed. Tests of the WH coil used resistive heaters on the conductor sheath and inductive heating of the conductor strands. Heat pulses were stepped up to observe the onset of a normalcy and the largest normal zone from which the coil could recover. Information relevant to capabilities for handling nuclear heating was obtained from single-coil tests at high current in which heat inputs were more widely spread and more prolonged but at lower levels than in the stability tests. The capability of the NbsSn conductor for operation at higher temperature was demonstrated by operating the WH coil at full current (17.8 kA) with its coolant inlet temperature at 8 K, more than twice its usual value. Both the EU and the CH coils proved capable of stable operation at full current with drastically reduced coolant flow rates. Other extended-condition, single-coil tests consisted in carefully observing coil behavior while the current was gradually raised until either the test objective was reached or the coil quenched. The EU coil reached its goal of 140% current without quenching. The GE/OR coil began to show current-sharing at 139% but did not quench. The JA coil quenched at 138% without warning. The WH and GD coils quenched at 131% and 120% of rated current, respectively. The CH coil reached its goal of 120% current without quenching after having quenched at 112% and 119% current. The effects of a pulsed poloidal field were measured for the EU, CH, and JA coils, which were each subjected to 300 to 400 pulse field cycles, in single-coil and toroidal configuration. (The pulse coil movement and locking system became inoperative before the other coils could be tested.) These tests were done with the test coil energized at several levels up to rated current. Heating effects of the field pulses were in reasonably good agreement with design calculations and were satisfactorily low: under prescribed LCT test conditions, from 4 to 14 W in the winding and about 5 to 10 W in the coil case. Effects of torque on the test coil winding due to interaction with the pulsed vertical field (up to 7 MN.m) produced effects in some tests that were clearly observed by displacement transducers and strain gauges. The effects of pulses on the quench detection systems required some adjustments in the facility sys-

Summary

tern but not compensation

Full-array

in

the for

EU system, with inductive voltages.

its

more

effective

tests

The most prominent objective of the LCT was the testing and demonstration of the operability of each coil at its design point (i.e., rated current and a magnetic field of 8 T at the specified location in the winding). Results of the series of design-point tests were gratifying, as each of the coils reached the 8-T design point without quenching or otherwise encountering a limiting condition within the test coil. Following the successful achievement of this objective, extended-condition tests were performed to determine the highest field/current that could be attained and to observe the response of each coil to extraordinarily large out-of-plane loads. In each of the 8-T design-point tests, when the test coil was at rated current, the other coils were typically operating at 70% to 90% of their rated currents to provide the desired background field. Preparations for each test included reconnection of signal processing channels to provide maximum information on the coil being subjected to the highest field and forces, followed by thorough checkout of diagnostic and protection systems. Typically, in the first approach to full field, currents in all six coils were stepped up to the specified values over a period of 2 to 4 h, while extensive data were collected and examined. Currents were then ramped down over a period of an hour or so. Among the interesting observations during the initial 8-T design-point tests were strains in the conductor and structure, acoustic emissions from the coils (indicative of slight movements under the changing loads), and displacement of the winding pack relative to the coil case in those coils equipped for these measurements. Various experiments were performed while the coils were at high field and current. In the EU test, for example, the coil continued to operate stably while the flow rate of helium coolant was reduced to one-fifth of the value originally required by the owner. Stability tests on the three PB-cooled coils (GD, GE, and JA) proved their abilities to recover from large normal zones (produced by heaters embedded in their windings) while operating at rated field and current. The ability of each coil to cope with distributed nuclear heating was investigated by various tests. Over the course of the tests, the WH coil began to develop detectable resistance when its current reached about half of the rated 18 kA. This was unexpected, as it indicated that the NbsSn superconductor was not performing nearly as well as had been predicted on the basis of short-sample tests. Otherwise, the tests of all coils at design field indicated quite satisfactory performance. Some of the high-current tests (design-point and extended-condition) entailed coil energy dumps, initiated either manually or by the coil protection system upon detection of a quench in any coil. Some quenches

9

were intentionally produced by heating, some occurred when the test coil was deliberately driven to its limit, and a few quenches of the CH coil occurred during tests of other coils. Analysis of coil data during and after dumps afforded useful information on heating of the structure and conductor by transient field effects. The quenches at high current also tested the ability of the helium supply and pressure control systems to cope with effects of resistive heating that sometimes boiled off thousands of liters of helium in a few seconds and violently expelled helium from the FF coils. Substantial amounts of helium were sometimes released (up to 300 kg in early dumps), but no equipment was ever damaged. Information on the response of the winding packs and structures to lateral loads much greater than encountered in the design-point tests was obtained by a series of experiments. In each experiment, five coils were run up to high currents while one remained deenergized. Both PB and FF coils performed satisfactorily, sustaining lateral forces as high as 27 MN without excessive strains in conductor or structure. The coil test program was concluded with a demonstration and experiment in which the ratios of coil currents were chosen to produce equal fields in the windings of all six coils. In the first run, all coils operated stably at peak fields of approximately 8.5 T. In the next run, currents were raised gradually until a spontaneous quench occurred, initiated by resistive heating in the WH coil. At that point, the peak fields in the six coils ranged from 9.0 to 9.2 T, and the stored energy of the array was 944 MJ. These and other high-field tests revealed that each of the six coils possessed substantial performance margins beyond the 8-T requirement of the coil specifications. Safety

tests

Several significant questions facing designers of coil protection systems were investigated in experiments with three LCT coils. One set of questions involved temperatures (and consequent thermal stresses) that would be reached in windings if a normal zone should occur and become quite large before the coil current could be dumped. In experiments with the GE/OR and WH coils, the a111omatic dump was delayed, allowing propagating normal zones initiated by internal heaters to grow. With thr EU coil, quenches were initiated at the helium inlet, and hot-spot temperatures were inferred from the temperature of expelled helium. Results showed that the adiabatic approximations in general use for choosing thresholds and delay times in coil protection systems were quite conservative for the three different LCT coils tested. Other experiments investigated the behavior of FF coils in occurrences involving drastic reduction or complete stoppage of coolant flow with the coils at high current. (Interruption of LHe supply to the PB coils

for periods up to an hour had already been shown to be inconsequential.) Both the EU coil and the CH coil proved capable of continuing high-current operation indefinitely with flow rates much lower than were previously requested by their owners to ensure stable operation. In a series of tests with the EU coil, flow was completely stopped for varying periods of time before being restarted. Finally, flow was stopped and kept stopped while the coil current was held at its rated value for 10 min and then was slowly run down to zero over the next 50 min. No evidence of a normal zone or other undesirable effect was observed. These tests showed that interruptions of refrigeration system operation could be accommodated without requiring rapid controLaction to prevent damage to the superconducting magnet coils. International

management

The LCT was not only a major technical undertaking but also a significant step forward in international collaboration in fusion technology development. It is therefore worthwhile to examine the important elements of this collaboration. In the mid-1970s, the Soviet Union decided to build tokamak experiments (T-7 and T-15) with superconducting coils to be developed specifically for each. France soon did the same with Tore Supra. In other fusion programs, decisions were made to use copper coils in tokamak experiments (TFTR, JT-60, and JET). Simultaneously, the United States, Japan, and the European Community were each considering how best to accomplish the expensive task of scaled-up development of the superconducting coils that would be needed in the next generation of tokamaks. The idea of international collaboration originated spontaneously within the several programs. Organization within the esisl.irlg framework of the IEA was soon recognized and acrrptcd as a logical course. At the same time, it was realized that the envisioned collaboration was, in somr important respects, unprecedented and, for it to br effective and equitable, it would have to be carefully organized. The approach was an IEA Agreement that established a program of research and development on magnets for fusion, comprising individually defined tasks, each with a specific group of participants. The first (and thus far, the only) task to be defined was the LCT. Control of the LCT was vested in the Executive Committee, which consisted of one high-level scientist/administrator from each of the four LCT participants. Each Committee member in turn designated a Project Officer to represent the participant in technical planning and guidance of the LCT. The Committee, with assistance from the Project Officers, prepared and issued a document that formally established goals and the broad outline of the test program. A more-detailed plan for implementation of the test program wa.s developed by the Project Officers through extensive intrrartion among scientists and engineers of all participants in decisions on mandatory coil specifications, definition

Ezecutive

10

of facility equipment and special equipment to be provided by each participant, and detailed planning of experiments. It had been foreseen that the purposes of LCT would be furthered by exchanging information on practical aspects of manufacturing conductors and coils. The Annex therefore required participants to arrange visits to coil fabrication facilities. This provision was fully implemented, with every production line and assembly operation being visited not only by government laboratory personnel but also by employees of industrial firms engaged in similar work. Further information dissemination was accomplished by exchange of monthly reports and final design reports among the LCT participants. The Executive Committee and the Project Officers met semiannually, most often at Oak Ridge. A vital element of the LCT was the active participation of long-term, on-site representatives at Oak Ridge in the day-to-day implementation of the test program. Assignments began in 1979 and continued into 1988. The number of non-U.S. personnel stationed at the test facility averaged six at any time, with length of assignments typically ranging from one month to t,wo yea.rs. Each participant was represented on a local commit,tcc that met at least weekly to review progress a.nd drcidr near-term plans. There was complete integration of all participants in the coil testing and analysis activities but not in the facility operation, which was performed by specially trained crews of ORNL personnel. At the conclusion of the test program, coils belonging to Japan, EURATOM, and Switzerland were returned to the country of origin. The final act of the collaboration in LCT was the evaluation of results, agreement on conclusions most relevant to future magnets for fusion programs, and joint preparation of a summary report.

Summary

l

l

l

l

l

l

s

l

s

s

s

Conclusions All 11 critical objectives identified at the outset of the LCT were accomplished. s The goals of coil and facility designs were achieved. l Limits of coil operability were explored and found to be greater than design points by substantial margins. l Data from the coil tests were of high quality and of unprecedented depth, enabling conclusions as to feasible and optimal designs of toroidal magnets for tokamaks. l The mature state of development of NbTi superconductors was demonstrated again in LCT, as the performance of several kilometers wound into some of the coils was practically equal to that of short samples. l The current-carrying ability of the NbsSn conductor fell short of expectations because of imperfections scattered along the 4-km length in the coil, l

l

l

s

s

the result apparently of problems in conductor production. More R&D is needed. Some LCT tests demonstrated the expected capability of NbsSn conductors for operation at temperatures much higher than would be tolerable for NbTi conductors. The practicality of both PB and FF cooling for TF coils of this size was demonstrated. Heat-removal capabilities commensurate with nuclear heating in the TF magnet of a tokamak reactor are practicable with either PB-cooled or FF-cooled coils. LCT tests demonstrated outstanding stabilit! agamst thermal drsturbances in the PB-cooled coils of this size. LCT tests demonstrated that thermal disturbances in FF-cooled coils can be minimized by the construction techniques possible with internally cooled conductors. Satisfactory stability of much larger tokamak magnets should be achievable through use of the design procedures tested in LCT. High-voltage insulation of conductors and the consequent feasibility of rapid discharge of FF-cooled coils in the event of a quench were demonstrated in LCT. Loss of flow in an FF-cooled coil need not demand extremely rapid action to prevent damage. Energy that is magnetically stored in TF coils of a tokamak can be harmlessly dissipated by practical means. Evaluation of LCT results, in conjunction with other magnet R&D findings, suggest that 1°F cooling is preferable for TF coils much larger than those in LCT. Further R&D is needed, however, on the stability of FF-cooled coils with much longer cooling channels. Solutions to the problem of detecting a normal zone in the LCT coils were demonstrated, but further development, including alternate methods, is desird able. There were enough similarities between operation of the LCT test stand and the operation of magnet systems in a tokamak to make the LCT experience valuable. LCT facility operation demonstrated good availability and revealed components and systems nceding improvement. The most important problem in LCT facility operation was air leakage into the helium, emphasizing the importance of this failure mode in other helium refrigeration systems that have sections at subatmospheric pressure. Effective collaboration in the LCT, involving as it did the integration of large-scale, advancedtechnology components that were cooperatively designed and produced in several countries, presages success in larger such ventures.

Ezecetive

Acknowledgements This report was compiled by P. N. Haubenreich,” P. Komarek,b S. Shimamoto,’ G. Vecsey,* L. Dresner,” W. A. Fietq” T. Kate,’ M. S. Lubell,a J. W. Lue,O J. N. Luton,’ W. Maurer,b K. Okuno,c S. W. Schwenterly,” Y. Takahashi,’ A. R. Ulbricht,b and J. A. Zichy.* The persons named in the preceding paragraph are only those who were heavily involved in the compilation of this report. In the LCT, the magnitude and complexity of the work and the numbers of people involved were so great that it is impractical to give proper credit here to all who deserve recognition because of their contributions. Nevertheless, special mention should be made

ORNL/Martin R. R. J. J. J. S. R. P. R. C. J. P. W. C. R. G. R. T. J. W.

E. Bohanan L. Brown F. Ellis R. May R. Miller S. Shen E. Stamps L. Walstrom K. Kibbe M. Amonett E. Batey B. Burn D. Cain E. Childress G. Clark M. Goodwin 0. Hussung L. Mann A. Mayhall D. Shipley

Westinghouse J. P. P. D. T. C. T. E. A. S.

L. H. C. T. E. J. D. A. J. K.

Young Eckels Gaberson Hackworth Hampton Heyne Hordubay Ibrahim Jarabak Singh Airco

P. A. Sanger G. Grabinsky W. Marancik

0. D. J. R. C. D. W. J. J. T. W. L. J. C. B. C. L. S. K.

Marietta

D. Ballinger T. Fehling M. Finger L. Hendren R. Howell L. James J. Kenney 0. Kiggans N. Money C. Patrick J. Redmond Riester P. Rudd R. Schaich C. Smith A. Wallace S. Scruggs F. Vaughan L. Zell

General D. G. R. C. E. D. D.

Summary

11

of two who were prominent in the founding of the LCT but who did not live to enjoy the satisfying fruition of their vision and their efforts. They are Prof. Ko Yasukochi and Prof. Werner Heinz. The following people also made notable contributions to this ll-year, international effort.

“Oak Ridge United States.

National

Laboratory,

bKernforschungseentrum Germany. ‘Japan Atomic Energy ken, Japan. *Paul

Energy L. C. B. W. J. R. L. J. D. C. Il. R. J. W. S. R. D. J. J. .J.

Scherrer

Systems,

Ksrlsruhe, Research

Institute,

Villigen,

Ridge,

Federal

Tennessee, Republic

Institute,

Nska,

Ibsraki-

Switaerland.

Inc.

R. Baylor Caruthers C. Duggins M. Fletcher S. Goddard L. Hansard R. Layman L. Murphy H. Thompson T. Wilson E. Wintenberg J. Wood V. Byrd R. Carver L. Guffy L. Jenkins Medovich R. Rider R. Taylor R. McGuffey _-___-

P. J. T. L. H. J. L. D. R. C. K. J. W. D. P. C. T. J. B. C.

B. Thompson K. Jones L. Ryan V. Wilson E. Miller R. Monday W. Nelms J. Taylor D. Benson Bridgman K. Chipley A. Clinard H. Gray R. Hatfield S. Litherland J. Long J. McManamy W. Moore E. Nelson C. Queen

General

Dynamics/Convair

S. Hackley S. Kruse E. Bailey Baldi H. Christensen Lieurance L. Walker

Oak

J. C. J. R. C. R.

-

Electric

J. Ferrante H. Gerwig P. Heinrich F. Koenig L. Linkinhoker Quay

Intermagnetics Corporation

General

B. A. Zeitlin C. H. Rosner --

of

12

Kernforschungszentrum H. S. G. S. W.

Bayer Foerster Friesinger Gauss Heep Siemens

C. Albrecht W. Else1 P. Henniger W. Laumann H. Marsing

W. H. H. W. A. G.

Hem Katheder Krauth Lehmann Naschwitz Noether

G. A. K. G. H.

Matthaeus Maurer Maier Ries Salzburger

(Erlangen)

Japan

F. Iida R. Saito R. Takahashi

R. A. H. H. J. E. H.

Atomic

Energy

J. M. F. G. H. (Hanau)

Toshiba H. Mukai Y. Sanada

Paul

Spiegel Suesser Wuechner Zahn Zehlein

Krupp

(Essen)

H. Hurnicki H. Janz U. Kolberg

Bezouska Brill Hillmann Hoeflich Rudolf Wagner Weber Research

Institute-Naka

K. Koizumi H. Nakajima M. Nishi

E. Tada H. Tsuji K. Yoshida

Kawasaki

Mitsubishi

Hitachi

S. Kamiya

Y. Hattori

Y. Ishigami

Scherrer

I. Horvath B. Jakob

Institute

(formerly

K. Kwasnitza C. Marinucci G. Pas&or Brown

H. Benz Th. Hilpert A. Koch

Nyilas Padligur Rietzschel Schmidt Siewerdt Spath

Vacuumschmelze

T. Ando T. Hiyama K. Kawano Hitachi

Karlsruhe A. U. K. C. L. F.

Boveri F. R. H. G.

Koenig K. Maix P. Marti Meier

Although in the final analysis one can conclude that the LCT was a great success and its benefits amply justified its costs, there were times when such an outcome was by no means certain. In the face of technical difficulties and costs greater than had been anticipated, strong commitment of the sponsoring agencies was vital for the ultimate achievement of the LCT goals. The U.S. government, through the Department of Energy, provided funds for design and construction of the test facility, for the work of three coil teams, installation and removal of coils, operation of the facility throughout the test program, and a large share of the coil testing and analysis activities. The European Atomic Energy Community, in association with the Federal Republic of Germany, supported the ELJ coil activities. The Gov-

Cable

SIN) P. Weymuth J. Zellweger

Company J. Rauch Th. Roman A. Segesseman

ernment of Japan, through the Japan Atomic Energy Research Institute, funded the JA coil activities. Swiss activities were funded by the Department of the Interior and the Brown Boveri Company. Technical leadership and management functions were provided by ORNL, KfK, JAERI, SIN, and the operating contractor for DOE facilities at Oak Ridge (originally Union, Carbide Corporation, later Martin Marietta Energy Systems, Inc.). Many industrial subcontractors committed highly capable personnel to the design, manufacturing, and testing of LCT coils, as indicated in the preceding list. Some also made substantial contributions above and beyond the limits of available government funding.