It has been observed in the superconducting helical coils of the Large Helical Device (LHD) that the balance voltage signals measured between the corresponding pairs of the coil blocks contain a number of spike signals during ramp-up and ramp-down processes of excitation. The spike signals might be generated by rapid changes of the self-inductances of the coil windings due to mechanical disturbances caused by large electromagnetic forces. Pulse height analysis (PHA) has been applied to analyze these signals in order to investigate the changes of mechanical properties of the coil windings as the excitation and cooling cycles proceed. In addition, acoustic emission (AE) sensors attached to the helical coil cans are also used to detect mechanical disturbances.

During the past six experimental cycles, the mechanical properties of the helical coils (HC) have been investigated in the Large Helical Device (LHD) using the balance voltage signals measured between the corresponding pairs of the HC blocks. During excitation processes, a number of spike signals are observed on the balance voltages. They might be generated by rapid changes of the self-inductances of the coil windings due to mechanical displacements (conductor motions) caused by large electromagnetic forces. The spike signals can be quantitatively analyzed using the technique of pulse height analysis (PHA). One of the most important results of this analysis is that the total intensity of the spike signals significantly reduces from the second excitation with the same operation condition [I, 2]. This observation is very similar to the ones that have been universally seen by acoustic emiSSIOn (AE) sensors for many other superconducting magnets [3]. In this connection, in the seventh cycle in 2003, four AE sensors were attached to the HC-cans and we started direct comparisons between the results obtained with balance voltage signals and the ones by AE sensors. Figure I shows a photograph of one of the AE sensors attached to the surface of the HC-can. The signals from the AE sensors are taken out of the LHD cryostat through feed-through connectors and fed into AE analyzers through preamplifiers. The envelopes of the actual AE signals are provided by the analyzers and they are digitized with a sampling frequency of 10kHz. The AE data can be observed and stored by computers in the control room via a LAN with optical fibers. After the AE sensors were installed, the stainless-steel surface of the HC-cans and the supporting structure were hit by a hammer before closing the cryostat in order to check the sensors by applying artificial mechanical disturbances to them. The obtained results showed that the AE sensors clearly detected the mechanical disturbance signals that propagated in the structures with the sound velocity. Figure 2 shows a typical example of the measured AE signals during an excitation process up to the central toroidal magnetic field of 2.7 T with the # 1-0 operation mode. The balance voltage ofH-1 (the innermost blocks of HC) is simultaneously obtained with two different

Performance of the superconducting helical coils of the Large Helical Device (LHD) during the past 12 cooling cycles is reviewed. The pair of helical coils are pool cooled by liquid helium and wound with aluminum-stabilized NbTi/Cu composite-type superconductors. Intensive efforts have been made to reliably carry out excitations, as more than 20 temporary normal transitions were observed. It was found that the minimum propagation current was about 10% lower than the nominal operation current. To improve the cryogenic stability, subcooled liquid helium has been supplied since 2006 using cold compressors, and the inlet temperature is lowered to be 3.2 K. The toroidal magnetic field has been raised by 5% and the plasma parameters are being enhanced. Pulse-height analysis is successfully applied on the balance voltage and acoustic emission signals to investigate the mechanical properties of the windings and their changes in years of operation. Short-duration normal transitions are automatically detected using a sophisticated monitoring system and careful operations are continued.

Multilayer insulator (MLI) is a strong, tool for use as a radiation heat shield, though the use of MLI has disadvantages in construction and evacuation for a long superconducting power cable. We have proposed the "MLI-free" radiation heat shielding for DC superconducting power cable and have measured the radiation heat transfer for thermally-isolated double-pipes with different surfaces. Here, Zn coating, MLI, and Al-foil sheet were tested. Consequently, from the radiation heat of 9.7 W/m for bare stainless-steel pipe, Zn-coated stainless-steel surface reduced to 2.6 W/m, whereas the use of NLI reduced to 0.2 W/m. It is expected that the simultaneous use of Zn coating and MLI can reduce the number of total MLI sheets to reduce the evacuation time.

Different from low temperature superconductors, Bi-2223/Ag HTS tapes have lower ldquonrdquo value. This characteristic brings rather ldquofuzzyrdquo notification of the critical current. It depends on the cooling condition. This means that the critical current of Bi-2223/Ag tape can not be decided absolutely. Rather we should focus on the temperature rise characteristics of the coil against over-current exceeding the critical current. If the temperature of the coil keeps a finite value without thermal runaway, the coil can be used continuously. We call the uppermost current without thermal runaway as ldquotolerance currentrdquo. Considering the situation mentioned above, the authors have studied the temperature rise characteristics of HTS coil for both DC and AC currents. Two cooling conditions are adopted. One is a cooling by immersion in liquid nitrogen bath, and the other is a conduction cooling by GM refrigerator. From the experimental result, the ldquotolerance currentrdquo of the Bi-2223/Ag coil for various conditions are obtained.

As well known, an inrush current of transformer reaches about several times larger than the rated current. When such a large inrush current flows into a superconducting transformer, it will induce a quench of superconducting windings. In this paper, we fabricated a small experimental superconducting transformer, and investigated the behavior of superconducting winding against the inrush current. The experimental result shows that a superconducting winding quenches for such a large inrush current, but it returns quickly to superconducting state in few cycles spontaneously.

Peltier current leads (PCL) were applied to a cryogen-free superconducting magnet (CFM) of 5 T/90 A. We measured the temperature profiles on the current lead of the CFM. Compared with the conventional copper current lead, the PCL reduced the temperature and the temperature gradient on the first stage of the cryocooler of the CFM at zero current. The current operation of the CFM improved the reductions due to the heat pump of the Peltier element. These results are the direct proof of the heat leakage reduction by the PCL

We have installed Peltier current leads (PCLs) in a commercial, cryogen free magnet to reduce heat flow to the GM cooler. Experimental results show that the first stage temperatures of the GM cooler are lower with the application of the PCLs than with conventional current leads (CCLs). At each operating current, from 0 to 75 A, the temperature gradient across the copper portion of the PCL is lower than that across the CCL. Heat flow into the cryogen-free magnet has been reduced from 16 to 30% by the application of the PCLs. Heat loss reduction was achieved by the installation of PCLs.

The authors have measured AE signals from the helical coils of the LHD (Large Helical Device) system at NIFS (National Institute of Fusion Science, Gifu, Japan) in order to monitor the state of the superconducting coils. Four AE sensors are attached on the surface of the vessel containing the helical coils, and preamplifiers are installed close to the sensors in the area with a leakage magnetic field of about 0.02 T. The measuring system is remotely controlled by PC's via a LAN system. The AE signals are recorded and analyzed. Their relation with the balance voltage signal of the coil has been investigated. The obtained experimental result shows that (1) the AE signals have been successfully recorded despite the leakage magnetic field on preamplifier, (2) the AE signal shows a good correlation with the balance voltages, (3) the AE signal is observed only when the excitation current is changing, (4) the pattern of the observed AE signal is stable for each excitation pattern.

Recovery time of Bi-2223/Ag HTS tapes after quench by short-time AC over-current has been observed. The experiment is carried out using an electromagnetic switching circuit. In order to confirm the state of the tape, a small DC current is fed continuously to the tape. The HTS tape is immersed in liquid nitrogen bath. The duration of the over-current is about 100-200 ms (5-10 cycles in 50 Hz), and the magnitude of the transient AC current is 150 A which is about 3 times larger than the DC critical current of the tape. The experimental result shows that the normal voltage generated by the over-current disappears about 2 ms after the end of the over-current. In addition to the experiment on HTS tapes, an experiment on a HTS coil wound by Bi-2223/Ag HTS tape is carried out. Also from this experiment good superconducting-to-normal-conducting recovery characteristics are confirmed. These experimental results will encourage the possibility of many applications of HTS tapes to power apparatus.

We made a co-axial copper tube measurement device in order to measure AC losses of HTS tapes. This device can minimize the external disturbance for high accuracy measurement. Also it is easy to fix the tape inside the copper tube. We used a linear amplifier in order to measure the voltage in high accuracy because the transport current was large. The voltage taps are located at the center and the edge of the tapes to measure the voltage drop on the samples. Our experimental data show that the inductive voltages are proportional to the frequency, and therefore this means that the main loss depends on hysteresis loss. In this study, the AC loss characteristics of twisted Bi-2223 tapes and untwisted tapes carrying AC transport current in self field is investigated. The magnetic flux generated by the transport current flowing in the HTS tape can be measured by using the same experimental setup. If the superconducting filament is twisted to reduce the AC losses, a relatively higher magnetic flux is induced at low frequencies, this is measured by winding the measuring wire around the tape - winding pick up method.

The Central Solenoid Model Coil (CSMC) was designed and built from 1993 to 1999 by an ITER collaboration between the U.S. and Japan, with contributions from the European Union and the Russian Federation. The main goal of the project was to establish the superconducting magnet technology necessary for a large-scale fusion experimental reactor. Three heavily instrumented insert coils were built to cover a wide operational space for testing. The CS Insert, built by Japan, was tested in April-August of 2000. The TF Insert, built by Russian Federation, will be tested in the fall of 2001. The NbAl Insert, built by Japan, will be tested in 2002. The testing takes place in the CSMC Test Facility at the Japan Atomic Energy Research Institute, Naka, Japan. The CSMC was charged successfully without training to its design current of 46 kA to produce 13 T in the magnet bore. The stored energy at 46 kA was 640 MJ. This paper presents the main results of the CSMC and the CS Insert testing-magnet critical parameters, ac losses, joint performance, quench characteristics and some results of the post-test analysis.

In 2000, Japan Atomic Energy Research Institute (JAERI) and its collaboration team accomplished many kinds of experiments under the magnetic field of 13 T for the ITER Project. The target coils are the central solenoid (CS) model coil and the CS insert coil. In 2001, the test using both the CS model coil and the toroidal field (TF) insert coil was carried out and successfully finished. During the experiment, we have measured the change in the amount of mechanical disturbances inside the coil using acoustic emission (AE) technology. In this paper, we report the general trend of AE characteristics obtained in the experiments for two years. That is to say, as for the CS model coil, we investigated the training characteristics of the CS model coil that experienced one cooling cycle from 4.2 K to room temperature. As a result, we confirmed the training effect of the CS model coil wound by forced flow CIC conductors. On the other hand, as for the insert coil, some peculiar AE signals were observed during the CS insert coil cyclic test. On this matter, a re-examination was carried out.

This paper presents acoustic emission (AE) signals induced from the Central Solenoid (CS) model coil of the International Thermonuclear Experimental Reactor (ITER) program. Envelopes of the AE signals were quasi-continuously measured with high-time resolution of 100 μs using six multi-channel AE sensors in order to detect the disturbances in the CS model coil during the process of the series of direct current (DC) operations. The AE signals were considered to be originally induced by motion of superconducting cables in cable-in-conduit (CIC) conductors and local motion of the conductors, judging from the close correlation between the AE signals and voltage spikes in the coil, especially during the virgin current operation. The multi-channel measurements provide us with information about the distribution of disturbances that could be detected acoustically by the AE sensors installed at each point of the CS model coil. The observation of AE envelopes with high-time resolution showed that the disturbances at each location of the CS model coil decreased by repeatedly charging-up the coil, judging from instantaneous AE levels, AE energy and the number of AE events. Direct measurements of the number of AE events that were carried out at another point on the coil confirmed that the disturbance dependence on the number of operations was similar to that mentioned above. The transfer function methods using one pair of AE sensors enabled us to analyze changes that might occur in either the coil structure or in the disturbance in frequency region during the repeating of the charging-up processes of the coil.