A tungsten (W) deposition layer is formed while a sample is exposed to a helium (He) plasma at the surface temperature of 473 or 573 K. The formed He-W co-deposition layer was composed of fine grains, the size of which increased with the surface temperature. The samples with the co-deposition layer was exposed to a deuterium (D₂) plasma, and the D retention was investigated using thermal desorption spectroscopy. It was found that the co-deposition layer has quite different D retention characteristics compared with bulk W.

The magnetic field dependence of a linear argon plasma is examined with a tomography system in PANTA. It is found that a plateau region exists around a particular region of magnetic field (∼600 G), below and above which the plasma changes the properties of emission and its fluctuations. A model is proposed to explain the observed dependence, and the comparison demonstrates that the dependence should be ascribed to the change in the Lamor motion inside the plasma production source using helicon wave and the plasma transport after the production.

A compact hard X-ray camera has been developed to study the behavior of fast electrons accelerated by lower hybrid waves in the TST-2 spherical tokamak. The camera consists of a pinhole made of tungsten alloy, a thin resin window, a thin LYSO scintillator disk, 5 fiber optic light guides and 5 photomultiplier tubes and shields. The camera is installed on a tangential port, and two-dimensional images are obtained by rotating the system around the axis of the port on a shot by shot basis. The typical energy range is 10 - 250 keV, and the energy resolution is about 100% (FWHM) at 100 keV. The measured images of a lower hybrid wave driven plasma show bright region at the inboard side (i.e., high field side).

The realization of advanced fusion reactors rests upon improvements of plasma-facing materials for divertors. The toughness of tungsten is a critical issue for these improvements. As a solution, we developed dispersion-strengthened (DS) W-Ti alloys using a new process which is based on a combination of mechanical alloying (MA) and hot isostatic pressing. In this paper, the microstructure of new DS-W-Ti alloys and the formation mechanism of nano-strengthening particles are investigated. The produced DS-W-Ti alloys are dispersed with TiO₂. The chemical analysis shows that new precipitation processes occurred during the MA based on redox reactions between carbon and oxygen.

A data-driven approach called sparsity-promoting dynamic mode decomposition (SP-DMD) is applied to the plasma turbulence signals obtained with an azimuthal probe array. The spatiotemporal turbulence can be reasonably decomposed into seven modes which capture the azimuthal bunching of the turbulence. A superiority of the DMD analysis to the conventional stationary analysis is demonstrated.

The feasibility of operation with high plasma density assumed in DEMO conceptual designs is one of the major concerns. Although the operation density would be reduced if peaked density profiles are realized, accumulation of impurity ions injected for divertor heat load reduction may be a problem. We analyzed the argon (Ar) ion transport in the core plasma with various density profile shapes, under the condition of the given fusion power maintained by the feedback control of injection frequency of deuterium-tritium pellets, using the TOTAL code. The ratio of the Ar density to the electron density was fixed at 0.5%, the expected value in JA DEMO, on the plasma surface. For the peaked density profiles, Ar was accumulated in the central region, which caused larger increment in the electron density, but the increment was smaller than the reduction of operation density by peaking the fuel density profile in moderately peaked cases. As a result, the line-averaged electron density and the pedestal electron density were in the feasible ranges reported in the previous experimental study, for moderately peaked density profiles. It was revealed that making the peaked density profiles can improve the feasibility of the DEMO operation density.

The stability of contact discontinuities was studied by means of one-dimensional electrostatic (ES) hybrid- and full-Vlasov simulations with the initial parameter based on observational data. The ES hybrid-Vlasov simulations show that the sharp gradient of the ion number density is generated at the early stage and is subsequently maintained for a long term. On the other hand, the sharp gradient is absent in the ES full-Vlasov simulation. The generalized Ohm's law shows that the electron pressure gradient accounts for the electric field on the ion time scale in both ES hybrid- and full-Vlasov simulations. It is shown that there is a difference in the time evolution of the electron pressure between the ES hybrid- and full-Vlasov simulations, which is mainly caused by the electron heat flux.

A 2.45 GHz microwave power injected through a quartz glass window produced a stable sheet-shaped plasma with the 11 cm wide and 4 cm high rectangular cross section in a static linear magnetic field. A thin slot gas conduit opened near the surface of the window supplied a discharge gas forming a stream of neutral atoms that intersected perpendicularly the plasma flowing toward the window along the magnetic field. The gas stream caused the change of the luminous intensity distribution of the plasma around the microwave window indicating reduction of the local plasma heat load. The window did not show any damage for a continuous operation of plasma up to 2 kW microwave input power.

An innovative concept for power and particle removal from the divertor is proposed. This scheme takes full advantage of both liquid metal convection and conduction to remove heat from the divertor, which is the most difficult issue for fusion reactor design. We propose that a liquid metal (LM) should replace the solid divertor plates on the bottom of the vacuum vessel. The LM is continuously supplied from openings located at the inner separatrix strike point on the floor of the LM container on the bottom of the vacuum vessel, and exhausted from openings located at the outer separatrix strike point on the floor of the LM container. The LM flow is guided along the field line to reduce MHD drag. In the event of a disruption, the current induced in the LM during the current quench is in the same direction of the plasma current. The induced LM current would either attract the plasma toward the LM divertor (leading to a benign Vertical Displacement Event), or force the LM toward the core plasma, providing automatic disruption mitigation, not requiring a learning process. The use of liquid tin instead of liquid lithium would provide greater stability against Rayleigh-Taylor and Kelvin-Helmholtz instabilities in quiescent plasmas.

We are developing a Lagrange (LG)-Monte-Carlo (MC) scheme for three-dimensional (3D) SOL/Divertor plasma fluid modeling. By using test particles, the scheme is suitable for handling 3D complex geometries. The semi-implicit treatment of the pressure gradient term enables us to improve the robustness of the coupling of the continuity and the momentum equations. Detailed numerical checks of the integrated scheme of LG-MC have been done for a simple 1D geometry. Benchmark tests between the new LG-MC and a conventional Finite-Volume scheme were carried out and good agreement was obtained. A first test calculation for a 3D cylindrical geometry has been also successfully done.

Proliferation of dusty layers inside the Tokamak prompts deleterious effect in Tokamak operation. The stability of the fusion plasma is adversely affected by building-up of dusty layers inside the tokamak which actuate disruptions in tokamak's performance. The current study is based on probing the mechanisms of dust accumulation inside the Tokamak device by simulating the process with the support of pulsed plasma accelerator. Experimentally it was observed that upon interacting with the pulsed power plasma stream, the materials used for designing the in-vessel elements of Tokamak such as graphite, lost its crystalline structure which in turn affects its inherent properties like ductility, malleability, rust resistance etc., thus emanation of dust paricles take place as a result of collisions between plasma species such as electrons, ions and neutrals with molecules of graphite as the tokamak material, ergo causing erosion in the tokamak material. The concatenated dusty layers formed due to accumulation of dust particles inside the Tokamak can cause probable instability in fusion plasma that would mar the function of the Tokamak device, besides the generated dust would become chemically reactive, radioactive or toxic. The dynamics of dust grains can be realised by Coulomb force, ion drag force, frictional forces on ions, ion pre-sheath drifts etc. and can also be interpreted from the kinetic theory of dusty plasma, to account this the numerical studies have been carried out with DUSTT and UEDGE codes to understand the dust dynamics and transport mechanisms.

The cryogenic coolant is important as an electrical insulation in an immersion cooling. One of the dielectric breakdown mechanisms of the cryogenic coolant is via bubbles, caused by partial discharge (PD). The insulation performance has been discussed with PD parameters and breakdown voltage. However, the essential PD mechanism has not been clarified in detail. In this paper, PDs in gaseous and liquid nitrogen were discriminated based on characteristics of the original PD waveform. It was shown that the difference in charge behavior appears significantly in the fall time of the original PD waveform.

The hybrid method based on the collocation element-free Galerkin method and the boundary element method has been applied to the 2D steady-state scattering problem of the electromagnetic wave. In addition, the performance of the proposed method has been investigated numerically. In this study, the numerical solution of the proposed method has been obtained by using the GMRES(m) method for the complex linear system. The results of computations show that the relatively smooth distribution of electric field is obtained regardless of the boundary shape. Therefore, it is found that the proposed method can be used as one of the tools for solving the 2D steady-state scattering problems of the electromagnetic wave.

The use of cesium leads complicates ion source operation and requires regular maintenance for continuous operation. The development of a negative-ion source without Cs seeding is desired in neutral beam injectors. A magnetized-sheet plasma producing negative ions using volume production without Cs seeding was designed. The experiment was performed by a TPDsheet-U and tested using steady-state hydrogen plasma. Two different types of grid structures were used in the experiment to extract the negative-ion beam: single- and multi-aperture grids. The multi-aperture grids have been developed to achieve more beam current. The negative hydrogen ions were successfully extracted from the sheet plasma using both single- and multi-aperture grids. The current densities of the ion beams increased with increasing discharge current and extraction voltage. At an extraction voltage of 9.5 kV and a discharge current of 80 A, the approximate current density of the negative hydrogen ion beam was 8.4 mA/cm² for the case of single-aperture grids. At an extraction voltage of 9.5 kV and a discharge current of 50 A, the approximate current density of the negative hydrogen ion beam was 23 mA for the case of multi-aperture grids.

Parametric decay instabilities (PDIs) in lower hybrid (LH) frequency range were observed in TST-2 in a various parameter range. LH wave measurement has been performed using four RF magnetic probes. We have been studied experiments aiming at efficient non-inductive current drive with the outboard and the top antennas installed on the outboard side and the top side of the vacuum vessel, respectively. Since PDIs are believed to deteriorate the current drive efficiency, we investigated conditions to avoid them. Operations with various combination of antennas (outboard, top), toroidal magnetic field strength and direction (CW, CCW B_T), working gas (hydrogen, deuterium) and density (normal, high) were investigated. In the outboard launch case, PDI sideband frequency δf associated with ion cyclotron quasi-modes was observed with large degree in any B_T strength in both hydrogen and deuterium plasma. On the other hand, no clear such sideband δf peak was observed only in deuterium plasma with the top launch case. Among various scenarios, the top launch with CW and CCW toroidal field in deuterium plasma seems to be desirable in terms of avoiding PDIs.

A new Langmuir probe has been designed and installed in TST-2 for measurements of Scrape-Off Layer plasmas (SOL). Non-inductive current drive is considered essential for spherical tokamak reactors. It has previously been shown that a large amount of injected Lower Hybrid Wave (LHW) power is lost in the SOL [1]. A full density profile of SOL conditions is necessary to accurately simulate the propagation of LHW in TST-2. A new probe was designed for durability, larger signal and Mach probe measurements. The new probe has been installed in TST-2 and results have been obtained. Temperature measurements show T_e = 30 - 50 eV during RF injection and <10 eV otherwise. Density measurements show n_e = 2.0 × 10¹⁵ m⁻³ and 1.5 × 10¹⁶ m⁻³ during flat-top RF power injection from Outboard- and Top-launch antennas, respectively. This is above the cut-off density for the 200 MHz LHW (5 × 10¹⁴ m⁻³) in TST-2, thus LHW can propagate through SOL plasma.

Removal of the central solenoid is considered essential to realize a spherical tokamak fusion reactor, but non-inductive plasma start-up is a challenge. Start-up using lower-hybrid (LH) waves has been studied on the TST-2 spherical tokamak at the University of Tokyo. The equilibrium poloidal field is believed to be generated mostly by the wave driven fast electrons, which are highly non-thermal and have large orbit excursions from the flux surfaces due to low plasma current. Such an equilibrium can be qualitatively different from the Grad-Shafranov equilibrium routinely used for internal magnetic field reconstruction in a tokamak. In this work, the effect of fast electrons on the MHD equilibrium was investigated by considering the equilibrium solution of the hybrid-MHD model [Y. Todo and A. Bierwage, Plasma Fusion Res. 9, 3403068]. The fast electron distribution function was estimated using a LH current drive simulation based on ray-tracing and an orbit-averaged Fokker-Planck solver. The equilibrium solution of the hybrid-MHD model was successfully fitted to the magnetic and kinetic measurements. The resulting poloidal flux function was more skewed towards the outboard side when fast electrons were introduced, which was more consistent with the density profile measured by the Thomson scattering diagnostic.

Arc trails found in heliotron/stellarator devices Large Helical Device (LHD) and Wendelstein 7-X (W7-X) were inspected; arcing occurred on different locations at various situations. In LHD, a helium-plasma-induced tungsten nanostructure sample was installed and exposed to a LHD plasma. Many arc trails were formed only on the sample with nanostructures, suggesting an easy initiation of arcing compared to pristine tungsten. After the completion of annual campaign 2011 in LHD, arc trails appeared on a graphite divertor tile which was taken out for inspection. Because the arc trails had a linear shape, the arcing was likely caused by main discharge operation. In W7-X, some Langmuir-probes installed within the limiter tiles suffered severe damage with arc trails during the operational phase 1.1. After the operational phase 1.2b, in-vessel inspection was performed for the first time for all the plasma facing components focused on arcing . No arc trails appeared on graphite components including test divertor units. In turn, considerable number of the trails appeared on non-plasma exposed region. Most of the arc trails had no clear linearity, indicating that arcing initiated during glow discharge cleaning phase. Only a few trails seemed to be affected by the existence of an external magnetic field. A vast majority (80%) of arcs were initiated from the edge of surfaces, while half of arc trails starting on the interior region of surfaces were accompanied by deposition spots. Broad arc trails appeared on the surfaces of diagnostic ports and mirrors.

Fast Ejecting System of Targeted sAmple called FESTA has been developed to carry out the measurement of dynamic hydrogen retention by a test sample. A sample can be exposed and extracted from the targeted plasma at any time using FESTA, however, when exposing the sample, the test chamber wall gets coated by some hydrogen as it is open to the QUEST vacuum vessel. We refer to this as the plasma-induced background. To measure the amount of hydrogen retained by the sample itself, the contribution from the plasma induced background must be subtracted from the measurements. To measure the accurate dynamic retention from plasma-exposed sample, a background subtraction model has been developed and tested. The initial testing shows that the FESTA system and model can estimate the dynamic hydrogen retention by a target test sample.

Ion Cyclotron Range of Frequencies (ICRF) heating is one of the plasma heating methods in the Large Helical Device (LHD). The wave injected from the ICRF antenna can propagate in the plasma even if the plasma density is extremely high. It was shown that the high power absorption occurs in the plasma core with the second harmonic heating in the case of high-density deuterium plasma in LHD by the calculation with a simple model of the ICRF heating. Wave number perpendicular to the static magnetic field increases with the plasma density, and it enhances the finite Larmor radius effect in the second harmonic heating. Enhanced finite Larmor radius effect and a large amount of resonant ions enable the intense power absorption. By increasing the frequency, the third harmonic heating will also be possible. Though the intensity of the power absorption will decrease, more localized heating on the magnetic axis will be realized because the finite Larmor radius effect works better in the third harmonic heating.

Neutronics aspects of a compact D-D neutron generator as a neutron source for the neutron calibration in magnetic confinement fusion devices are assessed by the MCNP calculation. The neutron emission distribution of the compact D-D neutron generator has a large anisotropy not only due to the scattering with the neutron generator body but also due to the intrinsic anisotropy of the differential cross-section of the d(d,n)³He reaction. The angular neutron distribution at the target of the compact D-D neutron generator is calculated with the PHITS code where the slowing down on the accelerated deuterons in the target material is considered. The calibration experiments are simulated by using the MCNP-6 code for the ITER neutron flux monitor (NFM) to be installed in the equatorial port. The detection efficiency of NFM is calculated for a D-D plasma neutron source, an idealistic D-D neutron source, a ²⁵²Cf neutron source and the compact neutron generator. It is found that the detection efficiency of NFM for the compact neutron generator is approximately 50% larger than that for the idealistic D-D neutron source. The discrepancy is improved to be 25% by the intention of the target 20cm from the body of the compact neutron generator.

Previously, the collisional merging formation of FRCs has been developed to generate high performance FRCs. A drastic increase in the excluded flux and total temperature were observed in the FRCs generated with this technique; furthermore, these phenomena strongly depend on the translation velocity of individual plasmoids. However, the dependence of the merged-FRC performance on the translation velocity has not been studied in detail. To study the dependence, the effect of installed auxiliary coils on the formation region on the translation velocity of FRCs in the FAT-CM device has been evaluated by the simulation with a two-dimensional resistive magnetohydrodynamics code, which is called Lamy Ridge. The control of translation velocity by using the auxiliary coils has beenwas demonstrated on the order of several tens of km/s in the simulation. The trapped poloidal flux and the total temperature of simulated FRCs have increased depending on the translation velocity.

A fusion burning plasma is sustained by deuterium-tritium (DT)-born energetic alpha particles. Therefore, energetic alpha particles must be well confined. In a deuterium experiment, 1 MeV tritons are created by D(d,p)T reactions. 1 MeV triton is regarded as simulated DT-born alpha particles because their kinetic parameters are almost same. A study of 1 MeV triton confinement has been widely and intensively performed in fusion devices in order to understand alpha particle confinement. To understand 1 MeV triton confinement/loss in EAST plasmas, 1 MeV triton orbit analysis is performed in various plasma current (I_p) cases using LORBIT codes. It is shown that the number of lost tritons decreases with an increase in I_p. The number of lost tritons rapidly increases at 10⁻⁶ s; then, it is almost saturated at 10⁻⁵ s regardless of I_p. The pitch angle distribution of confined 1 MeV triton shows that tritons that exist in a wider pitch angle range can be confined in the higher I_p case compared with the lower I_p case.

We study fuel layering for the Fast Ignition Realization EXperiment (FIREX) cryogenic target with a foam shell. A void free solid fuel layer within a porous foam material must be formed ideally. We have demonstrated the residual void fraction of ∼1% in a foam wedge with temperature controlled solidification. ANSYS simulations have shown that the residual void reduction technique will be applicable to the FIREX target. We examined each step in the simulated solidification process using a dummy foam shell target. In several attempts, solid fuel formation with a reduced void fraction in the foam shell succeeded.

To investigate the behavior of hydrogen isotopes in a large fusion test device, a new technique using infrared absorption spectroscopy is applied for the monitoring of hydrocarbons in exhaust gas. Fourier transform infrared (FT-IR) spectroscopy combined with a gas cell, which has an optical path length of 16 m, was installed at the plasma exhaust line. In this configuration, the detection limit of CH₄, C₂H₆, C₂H₄, and CO is at the level of sub-ppm. The exhaust gas observations were conducted during hydrogen glow discharge and during the regeneration operation of cryosorption pumps during the hydrogen plasma phase. As a result, hydrocarbons and carbon monoxide were detected in the exhaust gas, and the exhaust behavior and the ratio of gas components were determined. It is also demonstrated that light and the heavy hydrocarbons can be discriminated by the FT-IR system. Our infrared absorption spectroscopy study provides positive prospects for its application to exhaust gas analyses.

Joint winding of high-temperature superconducting (HTS) helical coil with conductor segments that are connected using bridge-type mechanical lap joints is considered as a promising method of fabricating magnet for a FFHR heliotron-type fusion reactor. Although methods for joining large-scale HTS conductor has been developed using the “simple-stack” and “joint-piece” procedures, the difference between these procedures is unclear. In this study, the two-row-four-layer joint samples were fabricated via the two joining procedures and compared in terms of contact resistivity. Joint thickness and joint resistance were measured; the contact area at the contact interface was evaluated using an X-ray computer tomography scan, to obtain the precise contact resistivity of the joint. The contact resistivity of the sample fabricated via the “simple-stack procedure” ranged from 2.41 - 5.15 pΩm², whereas that of the sample fabricated using the “joint-piece procedure” ranged from 1.98 - 6.07 pΩm². There was no significant difference between the procedures in terms of contact resistivity range. Considering the characteristics of joint thickness and location distribution of each lap joint, the inhomogeneous joining pressure was the primary factor affecting contact resistivity. As the similarity of electrical performances of the two procedures was clarified, future studies should focus on the manufacturability of large-scale joints.

Fuel retention and hydrogen permeation in the first wall of future fusion devices are crucial factors. Due to safety issues and in order to guarantee an economical reactor operation, tritium accumulation into reactor walls and permeation through walls have to be estimated and prevented. Therefore, studies of permeation in the fusion materials are performed and the need for tritium permeation barriers (TPB) is verified. The development of TPB layers is explained. A reliable way of comparing different TPB layers and the estimation of the permeation reduction effect of a TPB layer on different bulk materials is enabled by calculation of the layer permeability.

This paper presents an evaluation of a rapid pretreatment method for tritium analysis, that uses ion exchange resins (the batch method). The standard water samples were prepared with river water collected in Okinawa, Japan. Powdex resin was used in this study, which is an ion exchange resins in the form of powder beads. First, 100 mL of the standard sample was stirred with the cation exchange (PCH) and anion exchange (PAO) resins, as well as with activated carbon. To determine the optimal stirring time, we tested stirring times of 5, 10, 20 and 30 min. Additionally, to determine a suitable amount of ion exchange resin for environmental water samples, variations in the amount of resin and activated carbon were tested. Under the conditions of Experiment-3 (0.3 g, 0.3 g and 0.6 g of PCH, PAO and activated carbon, respectively) and with a stirring time of 5 min, electrical conductivity values of less than 1 mSm⁻¹ were achieved. These results demonstrate the effectiveness of the removal of impurities in the sample water by ion exchange resins and confirm that such an effect can be achieved during a short period of time (i.e., in 5 min).

For a multi-stacked (MS) pancake coil wound with Nb₃Sn CIC conductors, quench detection methods using voltage measurements were investigated. In this study, we utilized two methods: the balanced voltage between two coil windings and the balanced voltage between one coil winding and a disk-shaped pickup coil. During the energization of the MS pancake coil, voltage measurements of the coil were conducted to compare the measurement results of the two quench detection methods. The measurement results indicate that the quench detection based on the balanced voltage between two coil windings is more reliable than the other. In the measurements, spike voltages were observed when there was a rapid change of a coil current. Magnetization of the MS pancake coil is considered one cause of the spikes. When determining criteria of quench detection, spike voltages should be taken into account to prevent malfunction of a quench detection system.

In order to understand the decomposition process of hydrocarbons in a hydrogen (H) plasma, a Monte Carlo simulation of collisional transport of a methane (CH₄) molecule was developed. The model simulates collision reactions with plasma ions and electrons (including dissociation, excitation, ionization, and charge exchange) and elastic collisions with residual H₂ gas. The interaction with a surrounding wall was also considered (reflection from the wall, deposition on the wall, and reemission of carbon (C) and hydrocarbons (CH_x) by physical and chemical sputtering). In a low-temperature plasma, because the decomposition process was mainly dominated by charge exchange with plasma ions followed by dissociative recombination with electrons, many neutral C and CH_x species were obtained. At high temperature, the ionized C^y+ and CH_x⁺ species were the dominant ones because of the dissociative ionization and excitation by electrons. Comparable to our previous experiment, the calculated decomposition rate of CH₄ into neutral and ionized C atoms was ∼50% for a temperature of 15 eV and a density of 3.5 × 10¹⁷ m⁻³. Nevertheless, the calculated distribution of C and CH_x deposits on the vessel wall were localized in the upstream of the plasma, which was different from the experimental setup.