Supported by US DoE OFE DE-FC20-93ER54186

Abstract--A magnetic Pulse Test Facility (PTF), in which samples of CICC electrical joints from each ITER home team will be tested, has been fabricated at the MIT Plasma Fusion Center under an ITER task agreement. Construction of this facility has recently been completed, and an initial test phase on the first CICC joint sample has begun. PTF includes capabilities for sample currents up to 50 kA from a superconducting transformer developed by the University of Twente, magnetic fields up to 6.6 T with ramp rates to +1.5 T/s and -20 T/s, and a cryogenic interface, supplying supercritical helium with flow rates to 20 g/s through each CICC leg at controlled temperatures to 10 K and pressures to 10 atmospheres. A sophisticated, multiple-channel data acquisition system is provided to processed, digitally recorded sensor signals from both the sample and the facility. The facility is totally remote-controlled from a control room through a fiber optic link, and qualified users worldwide are afforded secured access to test data on a 24-hour basis via the Internet. The facility has successfully exercised the first joint sample over the ITER test spectrum with positive results.

I. INTRODUCTION

The Pulse Test Facility (PTF) is being used to test full scale ITER joint samples [1] in a pulsed magnetic background field with sample currents up to 50 kA. Specifications on the test sample lead spacing (100 mm), sample length (3.5 m) and sample sensor electrical connectors assure that samples designed for testing in the PTF are also be compatible for DC field testing in SULTAN [2], as required in the ITER R&D program. The PTF design was described in a previous paper [3], prior to actual construction. Since construction of the facility was completed in May 1996, this paper provides an updated description based on as-built conditions and improvements. In addition, the first ITER joint test sample [4] has undergone its initial phase of parallel-field test activity and those results will be discussed briefly.

II. PTF INTERFACES TO THE TEST SAMPLE

PTF is sited in a high-bay experimental hall equipped with a 10-ton overhead crane, and with access to nearby services, including dc converters, high and low pressure, distilled, deionized water, compressed air and 208/480 V, 3-phase ac power. For safety purposes, the facility control room is located in an adjacent room, and is equipped with two video cameras for sample cryostat and pulse magnet viewing, an audio monitor and two fiber optic links for communications between the test cell and the control room.

A. Pulse magnets

Capabilities of the pulse magnet system are summarized in Table 1. The pulse magnet system for ITER comprises a split pair, water-cooled, copper solenoid set that is mounted in a structural frame permitting manual magnet rotation about an axis normal to the solenoid axis. The coils are powered up to 25 kA by 6-dc converters connected in a 12-pulse configuration. The test sample is inserted vertically into the high field region, either in the bore of the solenoids, for parallel field testing, or in the gap between solenoid coils for transverse field testing. The present configuration employs a removable 0.22 m G-10 spacer between solenoid coils, and a set of iron pole inserts which are removed for parallel field testing. Short pulses, up to about a 7-s equivalent square wave at maximum field, are possible with magnet cooling from only the low pressure, deionized water loop, while pulses up to 60 s are available with high pressure water cooling.

B. Sample current

Current to the sample is provided from a +/-18 V, +/-280 A (four quadrant) power supply through a superconducting current transformer with an integrated superconducting current metering circuit in the secondary output lead [5]. The metering circuit (together with a controller) enables operation that may be switched from the control room to either voltage or current control mode. The transformer secondary is connected to the sample through a pair of high-current, copper feedthroughs which penetrate the bottom flange of the liquid helium can housing the transformer. Each feedthrough employs a helium-to-vacuum seal using a stainless steel bellows insulated from the conducting feed-through. The feed-through copper is shunted at the top with NbTi cable that is identical in cross-section to the transformer secondary cable, and at the bottom with an Nb3Sn monolith [6] , both of which are soldered into blind holes in the copper. The transformer has driven the test sample current to 50 kA at rates up to about 5 kA/s.

Vertical field (parallel to sample axis) 
     No gap between coils 
Maximum field (B//,max), T                        6.6 
          Length of high field region1, m         0.29 
     With 0.22 m gap between coils 
          Maximum field (B//,max), T              5.0 
          Length of high field region¡, m         0.66

Horizontal field (transverse to sample axis)                                     
     With 0.22 m gap between coils and with iron poles                                            
          Maximum field (BÝ,max), T               5.2               
          Length of high field region¡, m         0.26

Maximum positive ramp rate for                    1.5         
dB = Bmax, T/s                                             
Maximum negative ramp rate for                    20          
dB = -Bmax, T/s                
_________________________
¡Field within 90% of Bmax                        
                                             

C. Cryogenic services

An overall cryogenic schematic is shown in Fig. 1. Pressure in the closed, supercritical helium loop in the cryostat is controlled at room temperature with a regulator in the line from standard high pressure gas bottles and by circuit vent valves. High pressure gas flows through a liquid nitrogen-cooled purifier and enters the circuit at a temperature close to 80 K. Supercritical helium is pumped through the sample up to 10 atm and 20 g/s/leg with a differential pressure of up to 1 atm using a helium pump manufactured by Barber-Nichols. Flow enters the sample at the bottom and exits at the top, just below the sample's clamped connection to the feedthroughs. During cooling and experimental runs, heat is removed from the helium circuit and sample using two independent heat exchangers, one upstream and one downstream from the sample. The entire facility, including the test sample, can be cooled from room temperature to 80 K in about 48 hours when the heat exchanger and current transformer containers are filled with liquid nitrogen. The circuit is cooled at 8-9 atm pressure using the liquid helium pump as a compressor, first through the downstream heat exchanger (at the pump inlet), with the sample bypassed. After the pump inlet temperature drops to about 80 K, part of the flow is sent through the upstream heat exchanger to cool the sample. When temperature throughout the helium circuit is close to 80 K, liquid nitrogen is blown out from both helium reservoirs; they are purged with helium gas, and liquid helium cooling is initiated from two 500 l dewars. The facility and sample cooling time from 80 K to 4.5 K is about another 24 hours, including time to fill the current transformer container with liquid helium.

Electrical heaters at the sample inlet, with a power up to 200 W on each helium leg line, and the valving of the individual heat exchangers enable temperature control of the inlet helium over the range of 4.4 to 10 K. Valves at the sample inlet and outlet serve to balance (or imbalance, if desired) the flow in each of the two sample legs. Volumetric helium flow is measured by a set of two magnetically shielded turbine flow meters (Hoffer Flow Controls) installed upstream of the heaters in each leg. Total helium mass flow rate in each sample leg is calculated from volumetric flow measurements, combined with nearby measurements of the helium pressure and temperature. Helium circuit temperature sensors are Cernox RTD's (Lake Shore Cryotronics) epoxied with Stycast to the tube outer surface. Temperature is monitored at each sample leg inlet and outlet and at the pump inlet. All pressure transducers are thermally anchored to the bottom of the heat exchanger container which has a constant temperature close to 4.2 K. The differential pressure is measured across each sample leg.

Flow through the sample is controlled by the pump speed (1000-10,000 rpm) and the settings of the twelve proportional valves. Depending on the helium test temperatures and the number of sample quenches, a typical test campaign lasting about 1 week, including cooldown will consume about 3000 l of liquid helium and helium gas from 6-8 standard (150-170 atm, 40 normal l) gas cylinders.

Fig. 1. PTF Cryogenic schematic. Schematic shows arrangement of (1) test sample, (2) superconducting transformer helium container, (3) helium pump/motor, (4) Upstream, and (5) Downstream heat exchangers, (6) liquid nitrogen heat exchanger and He purifier and (7) inlet-line helium heaters. Facility flow meters (F), and pressure (P), temperature (T) and differential pressure (dP) sensor locations are also shown.

D. Control, monitoring, and data acquisition system

The facility control and monitoring and the experimental data acquisition functions of PTF were designed to be separate, but codependent systems. Major facility functions for monitoring and control include the background magnetic field, sample current and sample coolant. Each of these has supporting subsystems which are also monitored and controlled.

1) Hardware overview

The facility control and monitoring system utilizes a PC-based industrial process control software product, FIX-32reg. (Intellution, Inc., Norwood, MA), and an Allen-Bradley Programmable Logic Controller (PLC). The experimental data acquisition system comprises a bank of sixty dual-channel, fixed-gain isolation amplifiers and seven 16-channel programmable transient recorders (analog/digital converters) in a standard CAMAC chassis, along with a programmable 4-channel clock and a programmable 4-channel waveform generator. One fiber optic link runs from the CAMAC crate to a SCSI serial highway driver in the control room, connected in turn to the data acquisition computer, a DEC AlphaStationreg. running VMS. The second fiber optic link runs from the PLC processor to an ethernet transducer in the control room. Two operator interface computers communicate with the PLC processor over the ethernet.

2) Facility monitoring and control

The PLC, programmed through a separate dedicated PC, serves multiple roles: A) it is the source of control signals for facility components such as electro-pnuematically-actuated cryogenic valves, the helium pump, and various power converters; B) it monitors all the signals from facility sensors, such as the helium temperature, pressure, and flow transducers, pump drive frequency, dc converters, superconducting transformer, compressed air pressure, cooling water temperature, pressure, and resistivity; C) it monitors all safety interlocks; D) it monitors operator commands that are input from the on-screen mouse/keyboard-driven user interface. The PLC continuously processes these inputs according to the user-programmed logic, thus controlling operation of all facility `services' - the background field, sample current, and sample cooling, and in the event of a fault or an out-of-range parameter, shutting down or preventing start-up of any service.

The primary operator interface computer is equipped with a four-port video adapter, along with a virtual screen driver, so that four monitors, arranged in a 2x2 array, appear to the operating system to be a single large screen. During operation, the FIX-32 application can be set up to display four windows in this large virtual screen, one on each monitor, and thus, four separate configurations of control sliders, push-buttons, data-entry boxes, and real-time data plots and digital displays are available concurrently for use by the operator. From this computer, the operator can set up the initial sample coolant parameters, monitor and reset fault conditions, and trigger the waveform generator and data acquisition systems.

The second operator interface computer is also equipped with a multiport video adapter/driver and at present is running two monitors dedicated to real-time strip-chart-like displays of selected `slow' facility data trends of such parameters as coolant flow rates, temperatures and pressures, cryostat radiation shield temperatures, as well as some sample temperatures.

3) Experimental Data Acquisition and Display

The PTF data acquisition system is configured in accordance with the ITER test requirements, with the inventory of data channels as follows: 10 voltage tap channels, 25 Hall probe channels, 38 pickup coil channels, and 12 temperature channels [7]. In addition to these data sources from the ITER test sample, the data acquisition system also stores facility data, vital to the interpretation of the experimental data: supercritical helium volumetric flow rates; inlet temperatures and pressures, outlet temperature and differential pressures for each sample leg; background field magnet current; superconducting transformer primary and secondary currents and voltages, and quench detector signals.

Some signal sources, such as the flow meters and the superconducting transformer, have integrated signal processing, but most of the signals are first input to a fixed-gain isolation amplifier, and thence to the CAMAC transient recorders. These A/D converters are fully software programmable from the control room, as is the external clock that drives their sampling rate. All 112 digitizer channels are currently driven at 100 samples per second, but are capable of 1 Hz to 40 MHz operation. With 1 MB of on-board memory per module, the current configuration allows a maximum sampling time per `shot' of 655 seconds. During the recent test of the first ITER sample, the typical stored data duration has been 240 seconds, generating about 2.56 x 106 samples per shot.

The reference voltages that drive the magnet and superconducting transformer power converters are output from a CAMAC waveform generator. This D/A converter module is driven by one of the programmable clock channels at 10 Hz. Prior to each `shot', the operator generates the desired background field and sample current waveforms at the primary data acquisition computer screen (AlphaStation), either with the mouse or by keyboard data entry. These waveforms are downloaded into the D/A converter's memory during CAMAC initialization. The waveform `commands' are stored with the rest of the experimental data, along with the actual field and current values that are generated during the shot.

The data acquisition and display software, MDS-Plus(c) [8] was co-authored at PFC and has been in use for a number of years at the PFC Alcator fusion experiment and other sites around the world because of its advanced user interface.

An MDS data acquisition database is based on a multi-level `tree' file structure, in which the user enters all the relevant hardware parameters for the experiment, and in which all the digitized data is stored. Additionally, a user can store signal processing codes and conversion routines that are executed on demand by the MDS data display tool, SCOPE. That is, only the hardware descriptions, raw data and the conversion and computational codes are stored in data files for each shot -- signal conversion to engineering units, and other complex computational processing is carried out only on demand during plotting.

There are both x-window and command-line interfaces to MDS. The package, which runs on VMS (porting to UNIX and Windows NTreg./95reg. platforms is underway), has been tightly integrated with a commercially available data processing and display product, IDLreg. (Interactive Data Language, Research Systems, Inc., Boulder CO). The combination of IDL and MDS-Plus affords a powerful set of software tools for accessing, processing, analyzing and displaying PTF test data. Access to the computer cluster on which the MDS database is stored is password protected, and read privileges to the database itself are separately granted to ITER collaborators.

III. FIRST US JOINT SAMPLE TEST--EARLY RESULTS

Parallel field testing of the US pre-prototype joint sample has commenced in accordance with ITER test requirements [7]. The PTF facility was specifically arranged for these tests to allow electrometric and calorimetric determination of joint stability, joint losses, and helium flow properties in the presence of a pulsed, parallel background magnetic field. Data collected from Hall probe and pick-up coil sensors mounted on the joint sample during the tests will facilitate the determination of current density distributions within both the conductor and joint regions.

The joint sample was tested at 50 kA transport current in parallel background field at ramp rates to 1T/s, fields to 4 T, and helium flow rates from 3.5~15 g/s/leg and found to be completely stable at helium inlet temperatures below 8.5 K. A trapezoidal background magnetic field pulse with 7-s flat-top was used during these tests to facilitate collection of joint resistance vs. field data during the field flattop.

  Background     Resistance     
   field [T]     [n-ohm]   
       0         4.6 - 4.8      
       1         4.7 - 5.2      
       2         4.8 - 5.3      
       4         5.4 - 5.9    

Table 2 summarizes the joint resistances observed during these stability tests. Variation in sample temperature, pressure and flow rate had no effect on the joint resistance within the range of uncertainty of the measurements. The increased uncertainty for the resistance measurements in the presence of background field results because the joint's current distribution and resistance had not completely stabilized before the conclusion of the 7-s field flattop. The joint's measured resistance at 50 kA was roughly 4.6~4.8 n-ohm in zero background field, and increased to about 5.4~5.9 n-ohm at 4 T.

Joint loss data has been collected for the US sample at 4.4 K and zero transport current by applying trapezoidal waveform background field pulses to the joint and monitoring the change in helium flow enthalpy through the sample. Ramp rates during these tests ranged from 0.25-2.0 T/s with peak fields amplitudes from 1-4 T. Preliminary calibrations using 5-10 s duration 500-2000 J heat pulses applied to the helium circuit heaters indicate that this calorimetric method is accurate to about 20%.

The bulk of the data collected during initial testing of the US pre-prototype joint sample is presently under analysis and will be presented at a later date, following completion of transverse field testing.

IV. CONCLUSION

A newly operational facility for testing CICC conductor joints in a pulsed magnetic background field has been described . The magnets may be oriented for either parallel or transverse field relative to the test sample axis, with maximum fields up to 6.6 T, depending on orientation, and field ramp rates of +1.5 T/s and -20 T/s. The facility provides full cryogenic services to the test sample using 4.5-10 K supercritical helium at 4-10 atm and up to 20 g/s, depending on the pressure and temperature. Computer-based instrumentation, control and data acquisition systems, with interfaces to a large number of test sensors, significant storage and remote access capability round out a comprehensive testing environment. Early, successful parallel-field testing of the first ITER joint has shown that the facility can provide valuable research data in support of the ITER program, and offers promising capabilities to other large-scale conductors for the future.

ACKNOWLEDGMENTS

The authors wish to express their gratitude to Marvin Cohen and Jeff Hoy at the Office of Fusion Energy, US Department of Energy, whose cooperation and guidance made this project not only possible but a success. Thanks are also given to all the others, too numerous to cite by name.

REFERENCES

[1] P. Bruzzone, et al., "Design and R&D Results of the Joints for the ITER Conductor", Presentation at this conference.

[2] B. Blau, et al, "First performance tests of the 12 T split coil test facility SULTAN III," IEEE Trans. Applied Superconductivity, Vol. 3, No. 1, pp. 361-364, March 1993.

[3] B. Smith, et al. "A Pulsed Magnetic Field Test Facility for Conductors and Joints," IEEE Trans. Magnetics, vol. 32, pp. 2292-2295, July 1996.

[4] C.Y. Gung, et al., "Design and Manufacturing of US-ITER Pre Prototype Joint Sample," Presentation at this conference.

[5] H.G. Knoopers, et al., "Fast Ramp 50 kA Superconducting Transformer for Testing Full-Size ITER Cable Joints", Presented at ICEC16/ICMC, Kitakyushu, Japan 20-24 May 1996

[6] R.M. Scanlan, J.P. Zbasnik, Y. Furuto, M. Ikeda and S. Meguro, "Fabrication of a Cryostable Nb3Sn Superconductor for the Mirror Fusion Test Facility (MFTF-B)," IEEE Trans. Magnetics, vol. 21, pp. 1087-1090, March 1985.

[7] S. Shen, et. al., "ITER Full-Scale Conductor Test Program," ITER/US/96/EV-MAG/S. Shen/6.06/-1, June 6, 1996.

[8] G. Manduchi, G. Fregonese, C. Taliercio, T.W. Fredian, J.A. Stillerman, "The design of the user interface for a large physics experiment," Information and Software Technology, 1994 36 (12) 733-742.