Search

Disruption mitigation is of high priority for future tokamaks like ITER and DEMO. Massive gas injection (MGI) has proven to be an effective method in medium size machines and will likely be part of future disruption mitigation systems. For further research, the large superconducting tokamak JT-60SA will be equipped with a MGI system as an experimental equipment. This system will consist of two in-vessel MGI valves, which are mounted in opposite segments of the machine, vacuum feed throughs, a gas preparation system and an industrial PLC for control. The MGI valves are a scaled version of the spring-driven valve used in ADSEX Upgrade with an internal gas reservoir of 815 cm³, a maximum mitigation gas pressure of 6.5 MPa, a closing pressure of about 2 MPa, a nozzle diameter of 28 mm and an opening time below 2 ms. CFD simulations with common gas mixtures indicate a peak flow rate of 3.8 kg/s after 1.6 ms. The valve has a size of 140 mm x 110 mm x 292 mm. The gas preparation system allows easy and reproducible mixing of two gases by using an electronic pressure controller.

Developments of a two directional beam steering electron cyclotron heating/current drive launcher for JT-60SA have been progressed based on a linear motion antenna concept. A full-scale (˜7 m) mock-up of the mirror steering structure with two kinds of bellows structures and an in-vessel support structure has been fabricated and tested in vacuum successfully. The support structure consists of a pair of angular ball bearings made of Si3N4 for rotational motion and a linear guide assembled by a rail and a housing made of non-magnetic stainless steel and Si3N4 balls. Solid lubricant has been used for the linear guide and it results in the improvement of the life time by a factor greater than 10 with respect to the case without lubricant. Considering that the mock-up test results satisfied the requirements for the bellows and support structures (reliable 40 cm linear movement, 105 cycles and angular rotation 30°, 104 cycles), the final design of the launcher is underway targeting the start of its fabrication.

Disruption is a critical phenomenon in a tokamak reactor. Although disruption causes serious damage to the reactor, its physical mechanism remains unclear. To realize a tokamak reactor, it is necessary to understand and control the disruption phenomenon. The present research constructs a disruption predictor using experimental high-beta plasma data in the JT-60U tokamak. The predictor was constructed using a support vector machine as a linear discriminant, and we focus on a variable selection problem for the binary classification by sparse modeling, specifically, exhaustively searching the best combinations of variables which maximize the predictor performance. By the sparse modeling, we found that the six input parameters as the best combinations. The selected parameters were the n = 1 mode amplitude |Brn=1| and its time derivative d|Brn=1|/dt, the plasma density (relative to the Greenwald density limit) and its time derivative, and the time derivatives of the plasma internal inductance and plasma elongation. In particular, it was identified that the parameter d|Brn=1|/dt, plays a key role on plasma disruption. We should notice that the combination with other plasma parameters is indispensable and remarkably make it possible to improve the performance of disruption prediction.

Disruption is a critical phenomenon in a tokamak reactor. Although disruption causes serious damage to the reactor, its physical mechanism remains unclear. To realize a tokamak reactor, it is necessary to understand and control the disruption phenomenon. The present research constructs a disruption predictor using experimental high-beta plasma data in the JT-60U tokamak. The predictor was constructed using a support vector machine as a linear discriminant, and we focus on a variable selection problem for the binary classification by sparse modeling, specifically, exhaustively searching the best combinations of variables which maximize the predictor performance. By the sparse modeling, we found that the six input parameters as the best combinations. The selected parameters were the n = 1 mode amplitude | B r n = 1 | and its time derivative d | B r n = 1 | / d t , the plasma density (relative to the Greenwald density limit) and its time derivative, and the time derivatives of the plasma internal inductance and plasma elongation. In particular, it was identified that the parameter d | B r n = 1 | / d t , plays a key role on plasma disruption. We should notice that the combination with other plasma parameters is indispensable and remarkably make it possible to improve the performance of disruption prediction.

JT-60SAでは、プラズマの加熱・電流駆動のほか、プラズマ着火やプラズマ対向壁洗浄などにも電子サイクロトロン(EC)波が用いられる。本論文では、プラズマ着火やプラズマ対向壁洗浄など、プラズマへのEC波吸収パワーが低い場合におけるプラズマ対向壁への熱負荷を評価した。今回の解析では、EC波周波数や入射角度を変えた場合において、プラズマ対向壁で1回および2回反射した際のパワー密度を評価した。また、JT-60SAのCADモデルを用いて、非吸収EC波パワー測定装置が設置可能な場所を提示した。さらに、スイスプラズマセンターのTCV装置で行われたプラズマ対向壁洗浄実験の際の非吸収EC波パワーを解析し、プラズマの有無により非吸収EC波検出装置の信号値が大きく変わり、同実験における非吸収EC波パワーの評価法として有効であることを示した。

Eighteen error field correction coils (EFCCs) consisting of copper conductor and epoxy resin are installed in the vacuum vessel of JT-60SA in order to correct magnetic fields caused by manufacturing and installation errors of superconducting coils. The EFCCs are supported by pedestals made of stainless steel 316 L, and the pedestal has spring bars made of NCF625. The pedestals need to withstand electromagnetic force induced by EFCC current and magnetic field during plasma operations. They also need to allow relative displacement between the EFCC and the vacuum vessel during baking operations caused by the temperature difference between the vacuum vessel of 200 °C and EFCC of 40 °C. Diameter of the spring bars and connection structure for fixing the spring bars were designed through FEM analysis.

Eighteen error field correction coils (EFCCs) consisting of copper conductor and epoxy resin are installed in the vacuum vessel of JT-60SA in order to correct magnetic fields caused by manufacturing and installation errors of superconducting coils. The EFCCs are supported by pedestals made of stainless steel 316 L, and the pedestal has spring bars made of NCF625. The pedestals need to withstand electromagnetic force induced by EFCC current and magnetic field during plasma operations. They also need to allow relative displacement between the EFCC and the vacuum vessel during baking operations caused by the temperature difference between the vacuum vessel of 200 ◦C and EFCC of 40 ◦C. Diameter of the spring bars and connection structure for fixing the spring bars were designed through FEM analysis.

The stabilizing plate (SP) of JT-60SA has been designed based on an electromagnetic and a structural analysis. The SP plays a role of both a passive stabilizer of magnetohydrodynamics (MHD) instability and a first wall at low field side in combination with a graphite tile. The SP has a double skin structure with 10 mm thickness each in order to have simultaneously high resistivity in the toroidal direction and high strength against plasma disruption as well as a seismic event. A finite element method for the calculation of the electromagnetic force induced by disruption and the structural analysis has been applied. The most serious event which is fast major disruption, is mainly considered. The eddy current reaches up to 100 MA/m2, which induces electromagnetic force

Minimum sets of in-vessel components were installed for initial operation of JT-60SA in order to product and control the plasma and to know the plasma basic information, including inboard first wall, upper divertor, protection limiter, glow electrodes and magnetic sensors. All of magnetic sensors were installed within the required accuracy with in-situ measurement with laser tracker. The inboard first wall and the upper divertor were installed for the limiter configuration at the plasma initiation and the divertor configuration, respectively. They have carbon tiles as the first wall, which must be installed with accuracy of ± 1 mm in order to avoid the heat concentration. However, the vacuum-vessel wall had some deformation. Therefore, we measured in advance with laser tracker the position of the vacuum vessel where the bases of the inboard first wall and upper divertor would be welded. By using these position data and 3D CAD, we customized all bases. Owing to these procedures, we could install the first wall with required accuracy. Also, all other components were installed with required accuracy.

First divertor cassette covering a 10° sector in the toroidal direction has been manufactured for JT-60SA. Modularized plasma facing components such as inner and outer targets, baffles, and domes are assembled into the divertor cassette. All plasma facing components are actively cooled. Carbon armor tiles bolted to water-cooled copper alloy heat sinks are used to remove heat loads of 0.3–1MW/m2 for 100 s and 10 MW/m2 for less than 5 s. The divertor cassette has two pipe joints between it and the inlet and outlet pipes on the vacuum vessel side. After removing the divertor cassette, the inner target can be replaced by cutting and welding the other two pipe joints with the divertor cassette. A remote pipe welding tool has been developed both to install the divertor cassette into the vacuum vessel and to connect the inner target with the divertor cassette. The space around the cooling pipes is so limited that they are cut and welded remotely from the inside for replacement. The laser welding method was used, and the focusing mirror inside the pipe was rotated to perform circumferential welding. As a countermeasure for welding groove misalignment, the aiming accuracy of the laser was improved by simultaneously controlling the rotation and lifting of the tool. Thereby, reliable welding is achieved even with large gaps of up to 0.5 mm and grooves with a maximum angular deviation of 0.5 ° between connecting pipes. The outer diameter of the cooling pipe is 59.8 mm, the wall thickness is 2.8 mm, and the material is SUS316L. After connecting the inner target, the other plasma facing components have been assembled into the first lower divertor cassette. When installing the divertor cassette towards the vacuum vessel, the laser welding tool described above is used to connect the cooling pipes between divertor cassette and vacuum vessel in JT-60SA, because the outer diameter and wall thickness of the cooling pipe are the same as the inner target.

The stabilizing plate (SP) of JT-60SA has been designed based on an electromagnetic and a structural analysis. The SP plays a role of both a passive stabilizer of magnetohydrodynamics (MHD) instability and a first wall at low field side in combination with a graphite tile. The SP has a double skin structure with 10 mm thickness each in order to have simultaneously high resistivity in the toroidal direction and high strength against plasma disruption as well as a seismic event. A finite element method for the calculation of the electromagnetic force induced by disruption and the structural analysis has been applied. The most serious event which is fast major disruption, is mainly considered. The eddy current reaches up to 100 MA/m2, which induces electromagnetic force

Mechanical and quasi-optical design of an electron cyclotron heating/current drive launcher for JT-60SA is in progress. A full-scale mock-up of the steering structure, which enables linear and rotation motions of the first mirror of the launcher, has been fabricated for cyclic test of the bellows part. Moreover, an improved design enables easy replacement of the bellows for rotation for maintenance. Quasi-optical characteristics of the antenna mirrors have been studied to evaluate its transmission efficiency and beam focusing property. In calculation, it was found that the antenna is applicable to two frequency operation at 110 GHz and 138 GHz. It was quantitatively shown that the transmission efficiency of ∼99% (not including Ohmic loss) is obtained even with the higher order mode (LP11even) fraction of 10% by optimizations of the shape of the first mirror. These results contribute to optimization/finalization of the launcher design toward fabrication of the launcher for JT-60SA.

Development of an antenna that features the linear motion (LM) concept for long-pulse electron-cyclotron range of frequency heating and current drive for the JT-60SA is in progress. Combining a linearly movable first mirror and a fixed curved second mirror allows the injection-beam angle to be controlled. Cooling water is fed through the drive shaft for the first mirror and through the fixed support for the second mirror. The shaft support structure uses a metallic sliding bearing with a solid lubricant. The sliding bearing supports combined linear and rotational motion, whereas a conventional ball bearing supports either linear or rotational motion. Therefore, the sliding bearing offers the advantage of reducing the support-structure volume, which is important in the design of the relatively narrow port duct of the JT-60SA. Recently, the sliding bearing has been installed into the mockup. Results of a vacuum test with a mass analyzer indicate the presence of hydrocarbons during shaft motion. The injection-beam profile obtained from a toroidal beam scan is checked against low-power measurements taken on the mockup. Finally, for typical JT-60SA experimental scenarios, heating- and current-drive characteristics of the LM antenna are investigated theoretically.

The ITER Plasma Control System (PCS) requires an extensive set of about 50 diagnostic systems to measure the plasma response and about 20 actuators to act on the plasma to carry out its control functions. The specifications and real limitations of the actuators and diagnostics are being assessed as part of the ongoing conceptual design of the PCS to understand the potential impact on plasma control. The actuators include magnetic coils (central solenoid (CS), poloidal field (PF), vertical stability (VS), edge localized mode (ELM), correction coils (CC)), heating and current drive (electron cyclotron (EC), ion cyclotron (IC), neutral beam injection (NBI), and possibly lower hybrid (LH)), glow discharge cleaning, fueling and impurity gas and pellet injection, vacuum pumping, and disruption mitigation systems. Diagnostic systems are prioritized according to their role in machine protection (MP), basic control (BC), advanced control (AC), and physics studies (PS). At the conceptual design phase, detailed control algorithms do not yet need to be specified, but conceptual solutions must be chosen that satisfy the PCS requirements for control within the real constraints of the diagnostics and actuators. The feasibility of the chosen solutions must be proven either through established control schemes on existing machines or through an R&D program to develop them before they will be required on ITER. The diagnostic and actuator requirements of the PCS will evolve from first plasma through the high performance DT phase. A comparison is made of the expected requirements to control vertical stability, sawteeth, neoclassical tearing modes (NTMs), edge localized modes (ELMs), error fields, resistive wall modes (RWMs), Alfven eigenmodes, and disruptions with the ITER baseline actuator and diagnostic specifications. (c) 2012 ITER Organization. Published by Elsevier B.V. All rights reserved.

Structural, mechanical and optical design work on antennas/launchers for the electron cyclotron range of frequency heating and current drive system in JT-60 Super Advanced (JT-60SA) have been advanced based on a linear motion antenna concept. A CAD model of the launcher was built with realistic component sizes. A mock-up of the steering structure consisting of two different bellows sections for poloidal and toroidal beam scanning was fabricated to test movement of the bellows. The poloidal (−40° to +20°) and toroidal (−15° to +15°) injection angle ranges required in JT-60SA were shown to be realized by this steering structure and mirrors.

To contribute to high performance long pulse (∼65 s) experiments in JT-60U, the target of the electron cyclotron (EC) operation in long pulse is 0.6 MW for 30 s with four gyrotrons, though 10 MJ (2.8 MW and 3.6 s) was achieved in high power operation before 2003. One of the critical issues for the long pulse operation is detuning due to decay in beam current of the gyrotron. This decay comes from the cathode cooling by continuous electron emission. As a countermeasure for this issue, active adjustments for the heater current and anode voltage during the pulse have successfully extended the duration of a good oscillation condition for the gyrotron. As a result, 0.4 MW for 16 s with one gyrotron to the dummy load and for 8.7 s to the plasma have been achieved up to now.

In order to investigate the effect of polarization change on measurements of electron temperature profile from electron cyclotron emission (ECE) in a tokamak plasma, we have calculated the spatial evolution of the electric field of ECE propagating in the plasma. In the case of almost on-axis propagation, the Boyd's prediction is a good approximation, that is, the electric vector rotates with magnetic field. On the other hand, along horizontal sight line far from the axis the Boyd's prediction has a discrepancy of 3% from the numerical results. The effect of the polarization change is important for the measurement of ECE imaging in large tokamaks.

Numerical analysis of the island width of the neoclassical tearing mode, which is induced by the helical bootstrap current along the island, have been carried out to clarify effects of profiles on stability and to obtain conditions for stabilization by electron cyclotron current drive (ECCD). When the neoclassical tearing mode emerges, the saturated island grows up to some 10% of the minor radius at the beta value well below the ideal kink limit. The saturated island width of m/n mode can be small when the magnetic shear is large and the pressure gradient is low at the corresponding rational surface. Therefore, in the broad pressure and high li (internal inductance) plasma, the neoclassical tearing mode is more stable when qmin is slightly above one or two. In the broad pressure and low li plasma, the dependence of the island width on q0 is weak. A localized additional current at the O-point of the island by the ECCD significantly stabilizes the mode. The localization of the external current is effective, however, the stabilization also works well when the island width is smaller the radial width of the local current. The required current for the stabilization does not depend on the profiles very much.

Efficient transmission and polarization in 110 GHz transmission line for JT-60U was studied. A 40-m run transmission test line (inner diameter of 31.75 mm) with eight miter bends, which included a pair of polarizers was designed and constructed. The transmission at 1 MW was performed and the transmission efficiency of 89% (without mirror optical unit (MOU)) was obtained. In basis of the good result, a 1 MW, 110 GHz electron cyclotron range of frequency (ECRF) system was constructed in JT-60U. Transmissions of 1 MW-2.0 s and 0.32 MW-5.0 s were performed in JT-60U ECRF system. The polarization of the transmitted wave in the transmission line was measured and the dependence of plasma electron heating on polarization was investigated. The optimum elliptical polarization of the incidence wave for the ECRF experiments in JT-60U was calculated using the cold dispersion relation with the measured polarization. The optimum polarization was applied to the electron heating experiment of JT-60U.

An electron cyclotron emission (ECE) measurement system has been developed to obtain electron temperature with high time and spatial resolution for JT-60U plasmas. The system consists of three different diagnostic systems. These systems have different advantages and have been used complementarily. Several kinds of calibration are periodically carried out in the systems to secure reliable measurements. In order to reduce the transmission loss, a corrugated waveguide transmission line was installed and the reduction of the loss of ca. 3 dB was attained. As preliminary applications toward the operation of future fusion reactors, the feedback control of electron temperature profile and the automatic determination of the sawtooth inversion radius with an adaptive neural network were successfully demonstrated.

The 110 GHz 1 MW electron cyclotron range of frequency (ECRF) system was designed and constructed on JT-60U to locally heat and control the plasmas. The gyrotron has a diamond window to transmit RF power with Gaussian mode, which is easily transformed to HE11 mode for the transmission line of the corrugated waveguide. The second diamond window is installed at the inlet of the antenna for a vacuum seal between the transmission line and the JT-60U tokamak. The total length of the transmission line from the gyrotron to the antenna is about 60 m including nine meter bends, The antenna has a focusing mirror and a flat steerable one to focus and to control the RF beam angle mainly in the poloidal direction. In the initial operation, the power of PEC∼0.75 MW for 2 s was successfully injected into plasma when the gyrotron generated the power up to 1 MW. The total transmission efficiency from the gyrotron to the plasma was about 75%. A controllability of local electron heating with the deposition width of =15 cm was well demonstrated by using the steerable mirror. A large downshift in the deposition position was observed at the high Te plasma. Strong central electron heating was obtained from 2.2 to 6.6 keV for PEC∼0.75 MW, 0.3 s at the optimized polarization. An effective electron heating was also obtained up to ∼10 keV during EC injection for ∼1.6 s in the high βp H-mode plasma produced by NBI.