JT-60U Divertor Research Articles

Investigation of carbon dust accumulation in the JT-60U tokamak vacuum vessel

Abstract Dust generated by plasma–wall interaction is a potential source of tritium retention in a fusion reactor. Evaluation of the dust accumulation in the entire vacuum vessel is required to estimate the total amount of tritium retention, but it was particularly difficult to measure for plasma-unexposed areas behind the PFC structures, i.e. “shadow areas”. Dust samples were collected at 3, 5 and 2–4 different toroidal locations on the first wall, divertor surface and the exhaust route under the divertor in JT-60U, respectively. On the tile surface, large mass area density was found at the inner divertor and baffle, in particular, upper tiles compared to the lower target tile where the thick deposition layers were produced. Mass area density was significantly increased at the shadow areas, i.e. under the divertor structure such as the divertor and baffle tiles and the divertor dome. It was found that the poloidal distribution is relatively symmetrical in the toroidal direction within a factor of three. In comparison with the previous collection just before major change of the plasma operations, dust accumulation was increased both at the exposed and shadow areas due to change in the operating conditions.

Spatial characterization of a divertor in the presence of an X-point MARFE by collisional-radiative modeling and line profile calculations

High-resolution C IV n=6–7 (λ=722.6nm) line spectra measured in the divertor of JT-60U in presence of an X-point MARFE are analyzed using line profiles accounting for Stark and Doppler broadenings but excluding Zeeman effect which was suppressed experimentally. Spectra are fitted assuming either a single or two plasma layers depending on the crossed part of the MARFE. The ion temperature is taken equal to that of the electrons, the latter being deduced from C IV line intensity ratios. A density of 1020m−3 has been determined after fitting peripheral spectra with respect to the MARFE. For spectra related to central parts of the MARFE, two homogeneous plasma layers were assumed: a low-density layer (Ne(l)=1020m-3) and a high-density one with Te(h)=1-3eV. For few values of Te(h), a fitting procedure was used in which Te(l) (low-density layer temperature) and Ne(h) (high-density layer density) were searched for in the domains 1–25eV and 1020–1021m−3.

Chemical Binding States of Carbon Atoms Migrated in Tungsten Coating Layer Exposed to JT-60U Divertor Plasmas

Carbon migration in the tungsten coating layer exposed to JT-60U divertor plasmas has been investigated by analysis of chemical binding states of the carbon atoms. More than 1% of carbon atoms were accumulated as graphitic carbon, amorphous carbon and/or carbon-deuterium bonds. This concentration was more than five orders of magnitude higher than the solubility of carbon atoms in tungsten lattice. Up to 20% of ditungsten carbide (W2C) was also formed in the tungsten coating layer. These findings suggested the following carbon migration mechanism in the tungsten coating layer. The incident carbon migrates along grain boundaries and defects such as pores over the depth which is evaluated by the carbon diffusion coefficient in tungsten lattice. The carbon atoms trapped on grain surface penetrate and diffuse in the grains. The carbon atoms exceeded the solubility of carbon atoms in tungsten lattice chemically bind to tungsten atoms and form W2C. c

Effects of carbon impurity on deuterium retention in VPS-tungsten coatings exposed to JT-60U divertor plasmas

Carbon eroded from carbon armor tiles during plasma discharge was implanted into and accumulated in tungsten coating exposed to JT-60U divertor plasmas. The D/C ratio of 0.06±0.02 evaluated in the tungsten coating was half to one-quarter that in carbon codeposits formed at similar temperature of the tungsten coating. These results suggest that simultaneous use of carbon and tungsten coating would enhance tritium retention in the tungsten coating in future deuterium–tritium fusion devices. To investigate the carbon diffusion mechanism in the tungsten coating exposed to JT-60U divertor plasmas, the carbon diffusion coefficient in tungsten coating was measured by tracer methods. Using the apparent carbon diffusion coefficient obtained in this study (∼8×10−19m2/s), the carbon diffusion length in the tungsten coating exposed to JT-60U divertor plasmas was evaluated to ∼100nm. This diffusion length was quite shorter than that observed in the tungsten coating exposed to JT-60U divertor plasmas. Therefore, it remains possible that diffusion of implanted carbon in tungsten coating would be enhanced by other diffusion mechanisms which did not arise in the diffusion experiments or heat loads to the tungsten coating during transient events and plasma discharges with a strike point positioned on the tungsten-coated tiles.

Diagnostics of JT-60U divertor plasmas by Stark–Doppler broadening of carbon spectral lines

Line broadening by Stark and Doppler effects of C iv line spectra is suggested for the diagnostic of detached JT-60U divertor plasmas. Owing to the competition between Stark and Doppler broadening mechanisms, the C IV n=6–7 (λ∼7726Å) line profile depends on both the plasma electron density and temperature and the C3+ ion temperature. C IV n=6–7 line spectra, measured during an X-point MARFE phase in the JT-60U divertor using a high-resolution visible spectrometer, are fitted assuming the same temperature for both electrons and ions in order to determine the average electron density along the line of sight. These average electron densities deduced from the fit of C IV n=6–7 spectral line shapes corroborate those obtained using C IV line intensities and collisional-radiative modeling. Line broadening of C IV spectra is proposed to be used for the spatial characterization of the JT-60U divertor.

Temporal evolutions of electron temperature and density with edge localized mode in the JT-60U divertor plasma

From the intensity ratios of the three He I lines measured at 20 kHz, the temporal evolutions of the electron temperature and density during and after the power and the particle flow into the divertor plasma caused by edge localized modes are determined. The electron temperature increases from 70 eV to 80 eV with increasing Dα intensity. Then, at the peak of Dα intensity, the electron temperature starts decreasing down to 60 eV. The electron density increases from 0.1 × 1019 m−3 to 0.3 × 1019 m−3 with increasing Dα intensity, and then starts to decrease more gradually compared with the electron temperature after the peak of Dα intensity. It is interpreted that the increase of the electron temperature is ascribed to the power and the particle flow into the divertor plasma, and that the decrease of the electron temperature and the increase of the electron density are ascribed to the ionization of the recycled neutrals, which consumes the electron energy and produces electrons.

Analysis on EAST LHCD Operation Space by Using Simple Core-SOL-Divertor Model

A simple core-SOL-divertor model (CSD model) was developed to investigate qualitatively the overall features of the operational space for the integrated core and edge plasma. In the CSD model, the core plasma model of ITER physics guidelines and the two-point SOL-divertor model are applied. This CSD model is validated by the two dimensional divertor transport code (B2-EIRINE) and by the JT-60U divertor recycling database, and this model is applicable to the low- and high-recycling state of the divertor plasma. The CSD model is applied to the study of the EAST operational space with lower hybrid current drive under various kinds of trade-off for the basic plasma parameters, and the relationship between the operational space and the plasma discharge duration is also discussed.

Tritium distribution measurement of the tile gap of JT-60U

Tritium retention in the gaps or side surfaces of JT-60U divertor tiles was measured by an imaging plate (IP) technique and a combustion method. On the toroidal side surfaces, the retained amount decreased exponentially from the front (plasma facing) end to the rear end with an e-folding length of 3 mm for all tiles. On the other hand, the tritium retention characteristics on the poloidal sides varied depending on the location of the gap. Retention of tritium in redeposited carbon layers was observed in the gaps between inner divertor target tiles and the bottom side of dome outer wing tiles that faced the outer pumping slot. From the extrapolation of these IP and combustion results to the full toroidal direction, the total tritium retention in the divertor gaps was estimated to be around 0.07% of the total tritium produced by D–D reactions or 0.13% of thermalized tritium.

Two-Dimensional Spectroscopic Measurement of Hydrogen Emission in JT-60U Divertor Plasmas

In JT-60U divertor plasmas, deuterium Balmer-series line emission has been measured with a wide-spectral-band spectrometer, which has 92 viewing chords with a ˜1cm spatial resolution. Two-dimensional spatial distribution of the Balmer line intensities has been reconstructed using a computer tomography technique (maximum entropy method). In an inner-detached and outer-attached divertor plasma, the intensity of Dβ and Dγ lines were stronger above the strike point in the inner divertor and near the strike point in the outer divertor. The ratio of the Dγ line intensity to the Dβ line intensity was 0.3 - 0.5 above the strike point in the inner divertor and 0 - 0.2 near the strike point in the outer divertor. It suggested that the line emission were attributed to the plasma recombination above the strike point in the inner divertor and the plasma ionization near the strike point in the outer divertor.

Observation of divertor and core radiation in JT-60U by means of bolometric imaging

An imaging bolometer has been realized for the first time on a tokamak (JT-60U). The imaging bolometer utilizes a 7cm×9cm×2.5μm gold foil and an Omega/Micron infrared (IR) camera by Indigo/FLIR (128×164 pixels, 70mK, 30fps). The imaging bolometer has an array of 12 (toroidal)×16 (poloidal) (192 in total) channels, a noise equivalent power density of greater than 350μW/cm2 and a frame rate of 30fps. The field of view of the imaging bolometer is semi-tangential covering the divertor for more than 100° toroidally in addition to the entire poloidal cross-section. The data clearly shows examples of the radiation changing from strong divertor radiation to core dominated radiation in agreement with the resistive bolometer data.

Modeling of asymmetric redeposition distribution between inner and outer regions of the W-shaped divertor in JT-60U

Poloidal distributions of erosion depth and deposition thickness on components in the inner and outer regions of the W-shaped divertor in JT-60U are modeled using EDDY. Hydrocarbons released from the outer divertor plate are immediately ionized when entering the plasma and are redeposited near the release position. But they are subjected to re-erosion by the successive bombardment of plasma ions, resulting in small effective sticking. On the whole, most area of the outer divetor plate are eroded agreeing with observed distributions of erosion depth. In contrast, the inner divertor plate is dominated by deposition and the observed poloidal distribution of the thickness of the redeposited carbon layers agrees with incoming carbon flux from the plasma without re-erosion owing to lower-temperature. Due to much lower-temperature (∼1eV) of the private plasma in the outer region, the neutral carbon/hydrocarbon species are locally redeposited at the bottom edge of the outer dome wing adjacent to the bottom of the outer divertor plate.

Long-term erosion and re-deposition of carbon in the divertor region of JT-60U

Erosion and redeposition profiles of carbon tiles used in the W-shaped divertor of JT-60U with all carbon plasma facing wall are studied. The inner divertor is mostly covered by carbon redeposited layers, while the outer divertor mostly eroded. In the dome region, the erosion dominates on the inner dome-wing, while redeposited on the outer dome-wing. The redeposited layers on the outer dome-wing show very clear columnar structures indicating local carbon transport form the outer divertor to the outer dome-wing. The weight gain by the redeposition extrapolated to the whole divertor area is 0.55 kg. Since the extrapolated total erosion is about 0.33 kg, the remaining 0.22 kg must be originated from the main chamber erosion. Significant amount of the redeposition is caused locally by multiple processes of erosion, ionization and prompt redeposition toward inboard direction owing to gyration along magnetic filed line. This inboard transport is one of the reasons for small redeposition on the plasma shadowed area of the W shaped divertor of JT-60U with pumping slots placed at the bottoms side.

Thermal properties of redeposition layers in the JT-60U divertor region

Thermal properties of the redeposition layer on the inner plate of the W-shaped divertor of JT-60U have been measured with laser flash method so as to estimate transient heat loads onto the divertor. Morphology analysis of the redeposition layer was conducted with a scanning electron microscope. Measurement of a redeposition layer sample of more than 200 μm thick, which had been produced near the most frequent striking point, showed following results: (1) the bulk density of the redeposition layer is about half of that of carbon fiber composite material; (2) the specific heat of the layer is roughly equal to that of the isotropic graphite; (3) the thermal conductivity of the redeposition layer is two orders of magnitude smaller than that of the carbon fiber composite. This low thermal conductivity of the redeposition layer is considered to be caused by a low graphitization degree of the redeposition layer. The difference between the divertor heat loads and the loss of the plasma stored energy becomes smaller taking account of thermal properties of the redeposition layer on the inner divertor, whereas estimated heat loads due to the ELMs is still larger than the loss. This is probably caused by the poloidal distribution of the thermal properties.

Comparison of tritium retention and carbon deposition in JET and JT-60U

Profiles of T retention in the Mark IIA divertor of JET and H/D and T retention in the W-shaped divertor of JT-60U are compared. Hydrogen (H, D and T) is retained in carbon deposited layers with nearly constant concentration throughout the layers, except high energy triton directly impinging into more than 1μm in depth. However, carbon deposition profiles and hydrogen retention are strongly influenced by geometrical structure of the divertor and tile alignment as well as by magnetic field lines. Carbon deposition on the divertor base tile in JET shows stripes parallel to magnetic filed lines, suggesting a direct plasma deposition process. In JT-60U, the temperature of the co-deposited layers during operation plays critically important role on the hydrogen retention. It seems possible to reduce tritium inventory significantly by increasing the surface temperature of the plasma facing components.

Parallel and radial transport of ELM plasma in the SOL and divertor of JT-60U

Parallel and radial transport of ELMs in the Scrape-Off-Layer and divertor were investigated in JT-60U. The time lag from the start of magnetic fluctuations to the peak of the ion saturation current (js) at low-field-side (LFS) divertor was consistent with convective transport of the ELM plasma along field lines. However, the time lag to the start of the js enhancement was shorter than the convection time between the LFS midplane and divertor. In addition, js started to increase both at high-field-side and LFS divertors simultaneously, which is not explained by the convection model. Large multi-peaks appeared in the time traces of js at the midplane and near the X-point. The radial distribution of the first peak show that the ELM particle flux is transported to outer SOL flux surfaces up to 15cm from the separatrix. The radial expansion speed at the midplane was in the range of 1–2.5km/s.

Depth profile and retention of hydrogen isotopes in graphite tiles used in the W-shaped divertor of JT-60U

Depth profiles of deuterium and hydrogen retained in graphite tiles placed in the divertor region of JT-60U were investigated by secondary ion mass spectroscopy. The deuterium in the near surface of all graphite tiles was replaced by hydrogen due to exposure to hydrogen plasmas at the final stage operations. The depth profiles of the H/ 12C and D/ 12C signal intensity ratios varied not only with the location of the tile but also with the existence of redeposited layers. The integrated intensity of (H + D)/ 12C within the depth of 1.7 μm for the deposition dominated tiles was much smaller than that for the erosion dominated tiles. This suggests that thermal contact of the redeposited layer might be too poor to remove the heat loads from plasma to the substrate resulting in the surface temperature of the redeposited layer becoming much higher than that of the substrate.

In–out asymmetry of low- Z impurity deposition on the JT-60U divertor tiles

Low- Z impurity ( 7Be) on the JT-60U divertor tiles was analyzed to study the impurity behavior in the divertor region. The amount of the 7Be increased approximately one hundred times after B 4C-tile installation in the outer divertor. The 7Be was probably produced by 10B(p,α) 7Be nuclear reaction on the divertor tiles in the hydrogen experiment with ion cyclotron range of frequency heating. The 7Be was distributed asymmetrically in the poloidal and the toroidal direction. The highest 7Be concentration was found at the inner divertor whose boron content (B/(B+C) ∼ 20%) was lower than the B 4C tiles (B/(B+C) ∼ 80%) of the outer divertor. This result may imply impurity transport from the outer divertor to the inner divertor.

Analyses of Hydrogen Isotope Distributions in the Outer Target Tile used in the W-shaped Divertor of JT-60U

Hydrogen isotope distributions of the outer divertor target CFC tile, in the W-shaped divertor of JT-60U where erosion dominated, were analyzed by secondary ion mass spectroscopy (SIMS) to study the correlation between the concentrations of hydrogen isotopes and erosion/deposition rate. The chemical states of the CFC tile surfaces were analyzed by x-ray photoelectron spectroscopy (XPS). The existence of re-deposition layers was observed by scanning electron microscope (SEM). The erosion/deposition depth was also evaluated by a surface profilometer, namely dial gauge. The hydrogen and deuterium retention profiles were compared with the tritium profile obtained by the tritium imaging plate technique (TIPT).It was found that almost all the deuterium once retained in the near surface region of the erosion dominated area was replaced by hydrogen during the final HH discharge and/or exposure to the atmosphere. The hydrogen retention profile was controlled by surface temperature and plasma flux, and the total hydrogen retention was much less than that in the re-deposition layers observed at the plasma shadowed area in this particular tile.

Impurity Transport and Its Application to Ion Temperature Measurement in JT-60U Divertor Plasmas

Impurity Transport and Its Application to Ion Temperature Measurement in JT-60U Divertor Plasmas

Analyses of erosion and re-deposition layers on graphite tiles used in the W-shaped divertor region of JT-60U

Erosion and re-deposition profiles were studied on graphite tiles used in the W-shaped divertor of JT-60U in June 1997–October 1998 periods, operated with all-carbon walls with boronizations and inner-private flux pumping. Continuous re-deposition layers were found neither on the dome top nor on the outer wing, while re-deposition layers of around 20 μm thickness were found on the inner wing, in the region close to the dome top. On the outer divertor target, erosion was found to be dominant: maximum erosion depth of around 20 μm was measured, while on the inner target, re-deposition was dominant: columnar structure layers of maximum thickness at around 30 μm on the inner zone while laminar/columnar-layered structures of maximum thickness around 60 μm were found on the outer zone. Poloidal distributions of the erosion depth/re-deposition layer thickness were well correlated with the frequency histograms of strike point position, which were weighted with total power of neutral beam injection, on both the outer and inner targets. Through X-ray photoelectron spectroscopy, composition of the re-deposition layers at a mid zone on the inner target were 3–4 at.% B and <0.6 at.% O, Fe, Cr, and Ni with remaining C. Boron atoms are mostly bound to C atoms but some may precipitated as boron.