The Japan Atomic Energy Agency (JAEA) is procuring conductors for the ITER central solenoid (CS). To evaluate the superconducting performance of the CS conductor, a test coil “CS insert” was fabricated and tested at the CS model coil test facility in JAEA Naka. The CS insert is a single-layer, nine-turn solenoid using a piece of the actual CS conductor. The test was performed in 2015, and its duration was five months including initial cool-down and final warm-up. At the beginning, the nominal conditions of 40 kA - 13 T and 45.1 kA - 12.5 T were achieved. The change in current-sharing temperature (Tcs) was then measured during 16,000 cycles of electromagnetic force and three cycles of warm-up to room temperature and cool-down. There was no apparent degradation of Tcs observed. The variation in Tcs resulting from conductor strain in the longitudinal direction was also measured using direct and reverse charges. Quenching behavior was measured using an inductive heater installed at the middle point of the CS insert. From the tests, a database to estimate the performance of the CS conductor for the actual ITER CS was established.
The performance of the ITER central solenoid insert (CSI) conductor was tested in 2015. The current-sharing temperatures (Tcs) were measured over 16,000 electromagnetic cycles, including three thermal cycles between 4.2 K and room temperature. Tcs under the initial magnetization (IM) condition (13 T, 40 kA) of the CSI conductor not only increased, but also decreased between 6.71 and 6.84 K against cycling; then Tcs became almost constant at 6.74 K. Thus Tcs under the IM condition, was approximately 1.5 K higher than the specification of 5.2 K throughout the test. The slope of the hoop strain (εhoop) on the CSI conductor against the electromagnetic force was 1.55×10-4 % m/kN (in εhoop > 0) and 1.39×10-4 % m/kN (in εhoop < 0). Taking the effect of εhoop into account, the Tcs of the CSI under the SULTAN simulated condition (11.5 T, 45.1 kA) was equivalent to that of the SULTAN test after around 10,000 cycles. Before around 10,000 cycles, especially at the initial charge, the Tcs of the SULTAN test was lower than that of the CSI test. It is assumed that the hoop strain in the CSI test accelerated a strain relaxation, which increased the Tcs from the initial charge. When the strain fully relaxed and Tcs stopped increasing after around 10,000 cycles, the Tcs of the SULTAN test became equivalent to that of the CSI test. Given this perspective, the CSI test and SULTAN test were consistent. In εhoop > 0, the absolute value of the effective strain (εeff) of the CSI test decreased (i.e., Tcs increased) against the electromagnetic force (Fr) because the effect of the positive εhoop on the increase in Tcs exceeded the effect of the Fr on the decrease in Tcs. The line of εeff −εhoop of the CSI test against Fr was nearly symmetric about the y-axis (Fr=0). Comparing the εeff −εhoop of the CSI test and the εeff of the SULTAN test, the slopes of the strain against Fr were almost the same between the CSI test and SULTAN test before cycling. The εeff of the SULTAN test became close to the εeff−εhoop of the CSI test after cycling. This CSI test demonstrated that mass-produced CS conductors are highly capable of being used in the ITER.
For the stable operation of the Central Solenoid (CS) coil of the ITER without quenching, it is important to know the threshold of allowable external heating energy the CS conductor can be subjected to during operation. To evaluate the minimum quench energy of the CS conductor for the ITER, an inductive heating test was performed during the CS Insert Coil (CSIC) test campaign. A 59-turn inductive heater installed on the central turn of the CSIC was used to apply the heat energy. The heating energy from the inductive heater was calibrated by calorimetry using short conductor samples with inductive heater windings and a resistive heater. A series of inductive heating tests was performed while applying a 45.1 kA current and 12.5 T backup field on the CSIC. The alternating current (AC) applied for the inductive heater was 1,000 Hz in 40 ms, and the amplitude of the AC was varied until a quench occurred. As the result, it was obtained that the minimum quench energy for the CSIC heated by eddy current was 0.23 J/cm3 without including the joule heating energy of the heater itself.
The Central Solenoid (CS) insert coil consists of a 42-m-long CS conductor, of which the specifications are the same as that of the ITER CS. In order to investigate normal zone propagation and hot spot temperature, a quench test was carried out on the CS insert under End-of-Burn condition at 12.5 T and 45.1 kA of after 16,000 cycles. External heat was applied at nearly the center of the CS insert using an inductive heater, and quench was induced. A current of 45.1 kA was dumped 9.5 s (7 s) after voltage generation (Quench detection, QD). The Normal zone propagation length reached 23.4 m, and the maxim propagation velocity was 3.1 m/s just before dumping. Considering the distribution of temperature, which is calculated by GANDALF, hot spot temperature was expected to reach 227 K. As the result, it was found that the hot spot temperature exceeded the criteria of 150 K which is designed on ITER. However, heating the CS insert to 227 K did not influence conductor performance, because the current sharing temperature was maintained after the quench test. Therefore, the quench detection has a margin of approximately 9.5 s (7 s) after voltage generation (QD) in view of the conductor performance under the conditions applied in this quench test. If the hot spot temperature is kept to less than 150 K, the current should be dumped 7.5 s (5 s) before voltage generation (QD). These results are very useful for designing quench protection of the ITER CS.