Search

It is important to measure retentions and depth profiles of deuterium and tritium on plasma facing materials to understand plasma-wall interactions. The nuclear reaction analysis (NRA) has been established using a deuteron beam with an adjustable energy. By means of the NRA, depth profiles of deuterium and tritium on the surface of JT-60 divertor graphite tiles irradiated with DD burning were obtained. The NRA method showed that the average concentrations of the deuterium retained in the surface of two different were 1.3 × 10 16 and 0.3 × 10 16 cm −2 . α-particles from the T(d, α)n reaction were measured from a dome unit tile of JT-60U W-shaped divertor and the measured activity of tritium was about 18 kBq/cm 2 . The peak energy of the α spectrum obtained with the NRA was about 1.2 MeV. This result implies that energetic tritons were implanted in the divertor graphite tile.

The capability of detachment control in JT-60SC is demonstrated with a divertor code (SOLDOR/NEUT2D). Under the standard operation of the core edge density at 95% of the minor radius, n edge = 3.2 x 10 19 m 3 , the power flow from the core plasma Q c = Q i = 6 MW, and the conductance of inner and outer cryopanels C in = C out = 50 m 3 /s, the inner divertor plasma is partially detached and the outer divertor plasma attached. The inner divertor can be changed from partially detached to attached plasma by inner pump of 150 m 3 /s. The fully detached inner divertor is obtained in high density operation with n sep > 4 x 10 19 m 3 , where n sep is the density at the separatrix outer midplane, The elastic collision in a dense and cold divertor plasma (n c ≥ 5 x 10 20 m 3 . T c < 5 eV) play an important role on detachment.

Damage to a TiC coated Mo divertor plate used in JT-60 for about two years was re-examined from a material viewpoint to obtain data concerning utilization of high- Z plasma facing components in future fusion reactors. Serious damage was observed on the edge of several divertor plates located next to a toroidal gap. Due to very high heat fluxes a apart of the edge was melted and a large crack of about 12 cm wide and 2.5 cm deep was caused near the melted edge along a lamination of the Mo plate. Recrystallization by the heat loading embrittled the subsurface region and resulted in cracking and melting at a lower heat loading. The present work suggests that the maximum temperature of the powder metallurgical Mo should be maintained below the recrystallization temperature to prevent material degradation and plasma radiation losses.

Operational experiences on the performance and material behavior of the JT-60 divertor plates are presented. The outside X-point operation using TiC-coated Mo (TiC/Mo) plates was performed with absorbed power up to 20 MW and duration of 1 s without degradation of the plasma performance by sweeping the separatrix to control the surface temperature below the melting point. The use of graphite divertor plates extended the absorbed power up to 25 MW with duration of 4 s in the lower X-point operation. However, the graphite was eroded and redeposited on the divertor plates and the carbon content in the plasma was increased compared to the outer X-point operation with carbon tiles. A few lower graphite tiles were seriously damaged and radioactivated. Improvements in thermal shock resistance and erosion characteristics of the graphite are necessary. High heat concentration at tile edges and disruption-induced heat loads should be avoided in the future devices.

The behaviour of the heat flux to the JT-60 divertor are investigated in the lower hybrid range of frequency (LHRF) heated and neutral beam heated discharges from the view point of the plasma wall interaction by using an IRTV camera. It is found that the power balance is determined by the distance between the launcher and the separatrix of the plasma with LHRF heating. As the LHRF power increases, the peakedness of the heat flux increases. The parameter dependence of the peaking ratio is investigated for neutral beam heated discharges. It is found that the peaking ratio is a function of the divertor geometry. This behaviour is explained by a simple model that takes into account the response of the divertor plasma to the transport of the heat flux and particle flux.

This paper describes the experimental investigation of the heat load on a lower hybrid (LH) launcher in divertor plasma. It was found that a good coupling of the reflection coefficient, ρ = 15–20%, was obtained by pushing the launcher head a few mm from the first wall even though the distance δ between the separatrix and the launcher head was more than 10 cm. In this condition, the heat load on the launcher head rapidly decreased at δ > 5 cm and was estimated to be about 0.4–0.6 kW/cm 2 at δ = 10 cm, n ¯ e 13 cm −3 during only LH injection. A higher heat load was observed on the ion side of the launcher guard limiter during combined NB + LH heating at n ¯ e > 3 × 10 13 cm −3 when fast ions were observed at the plasma edge due to parametric decay instability. We discuss the implication of these results with respect to LH application on a high power heated plasma.

Outgassing behavior during plasma disruptions in the JT-60 tokamak have been investigated using a newly developed fast pressure gauge. Large amounts of gases are released suddenly in two stages during a plasma disruption. The outgassing peaks coincide with the burst of Hα emission. The first gas release is ascribed to heat flux due to the loss of the plasma thermal energy. The second is caused by the heat flux due to the loss of plasma magnetic energy. Although this outgassing mechanism should be investigated, about three quarters of the released gases then return to the wall.

Remote radiative cooling can significantly reduce the divertor plate heat flux, which may provide a solution to one of the major problems of present and next-step machines. Strong remote radiative cooling of up to 50% of the total input power of 20 MW was observed in JT-60. JT-60 divertor plasmas are dense and cold (nediv ≈ 2.4 × 1020 m−3, Tediv ≈ 26 eV). Remote radiative cooling is a mixture of carbon and hydrogen radiation (+ CX). JT-60 divertor plasmas can be adequately modeled with a simple divertor model with radiative cooling.

In JT-60, MARFEs have been observed frequently in high- I p and high density limited discharges with neutral beam heating after replacement of the first wall from TiC coated molybdenum tiles to graphite ones. The MARFEs occur near the density limit of the JT-60 plasmas. The NB heating is effective to suppress the MARFE. The empirical scaling of the threshold density for MARFE onset taking account of the NB power is obtained, which is useful to predict the MARFE onset. The radiated power from the plasma with MARFE is about 90% of the absorbed power. Both stored energy and central electron temperatures do not change with the MARFE onset in spite of the such intense radiation loss. So the MARFE does not affect the core plasma. A simple model based on the thermal equilibrium in a peripheral plasma can explain the onset condition of the MARFE.

In the last year, ohmic heating experiments were performed in JT-60 for about three months; May–June 1985 and March 1986. The major experimental effort has been concentrated on the divertor configuration in anticipation of H-mode operation of JT-60 in high power supplementary heating experiments starting in August 1986. Stable discharges with a divertor have been obtained at Ip = 1.85 MA and ne = 5.7 × 1019m−3. Radiated power from the main plasma in divertor discharges is reduced to one third of that in material limiter discharges. The electron density has a broad radial profile and the effective safety reaches qeff = 2.5. The neutral gas pressure in the divertor chamber becomes significantly high and the control of particle recycling has been attempted.

Neutral Pressures in the divertor chamber (Pdiv) and around the main plasma (P main ) are measured by ionization gauges which have a delay time constant of 200 to 300 ms. The H-mode shots were observed mainly in the range of n e = 2 to 4 × 10 19 m 3 and P main = 0.8 to 3.5 × 10 −3 Pa in the outside X-point divertor discharges with graphite walls. After the H-mode transition, significant reduction in P div is observed. The compression ratio is around 20 although in the L-mode shots with TiC/Mo walls this value was around 45. Since p div is 1.5 to 13 × 10 −2 Pa during the H-mode phase, the total exhaust rate by divertor pumping system Q ex is 0.1 to 0.8 Pa m 3 /s. This value is 1 30 − 1 4 of the particle fueling rate (~ 3 Pa m 3 /s) by NB injection of 20 MW and 75 keV. Therefore, in the H-mode discharge, improvement of particle exhaust is one of the critical issues.

The first divertor plasma density measurement and the interferometric edge plasma density measurement with boundary condition preserving millimeter waveguides were demonstrated to elucidate the mutual correlation among the divertor plasma, scrape-off layer plasma and the bulk plasma properties in the additionally heated JT-60 plasmas. The electron density in the divertor region exhibited a nonlinear dependence on the bulk plasma density for the joule-heated plasmas. When neutral beam heating is applied on the plasmas with the electron density above 2 × 10 19 /m 3 , however, the bulk plasma density is scraped off from the outer region to lead to density clamping, and the electron density in the divertor region rapidly increases over 1× 10 20 /m 3 , from which we can deduce that the particle flow along the magnetic field is dominant, resulting in the apparent degradation of the particle confinement time. As for the case when neutral beam injection is applied to low-density plasmas, the bulk plasma electron density profile becomes flattened to yield a smaller density increase in the divertor region and no density clamping of the bulk plasma was observed. Simulation analysis which correlates the transport of the divertor plasma and the scrape-off layer plasma was also carried out to find the consistency with the experimental results.

A beam target made of oxygen-free copper with 0.2% silver added was bombarded by a intense neutral hydrogen beam of 75–100 keV, 1.4 MW for 10 s at the prototype injector unit for JT-60 neutral beam injection. Hydrogen gas which was re-emitted from the beam target at the time constant of about 1 s was evacuated by a cryopump with a pumping speed of 960 m 3 /s. We compared the re-emitted hydrogen gas flow rate from the beam target with the number of particles in the hydrogen beam and concluded that the hydrogen gas re-emission coefficient is almost 1.

This paper investigates the current buildup just after the breakdown, and it becomes clear that stable plasma initiation can be obtained by the precise control of particles contained in the graphite wall using clean-up shots with/without gas puff.

JT-60 first walls were originally composed of a number of tiles made of either molybdenum or Inconel 625 with 20 μm thick TiC coating. These first walls, operated after the bakeout at a temperature of 350°C, provided very clean surfaces as well as superior plasma characteristics during Joule heating discharges. Damage of the TiC-coated first walls was modest and localized to specific spots. However, initial results of the additional heating experiments showed degrading effects of locally melted TiC-coated molybdenum and Inconel 625 on the plasma operation. During the venting period from April to May '87 about half of the TiC-coated molybdenum and Inconel 625 was replaced by graphite first walls to prepare for high-power and long-pulse additional heating experiments. After a three-day 300 °c bakeout, 30 h glow discharge cleaning, and 16 h pulsed discharge cleaning at a vessel temperature of around 250°C, a very rapid start-up of the discharge conditioning with plasma current up to 1.0 MA was achieved with 14 shots. Flexibility in plasma operation was increased and the plasma current was increased to 2.9 MA.

Essential divertor functions — density and energy control — have been investigated in the JT-60 divertor with metal walls. Neutral pressure in the divertor chamber increases in proportion to n2e and high recycling state for particle exhaust is realized. At ne ≤ 6 × 1019m−3, wall pumping by the divertor plates is dominant in particle exhaust. However, at ne ≥ 6 × 1019m−3, particle exhaust by active divertor pumping systems becomes essential. Because of the effectiveness of the divertor, radiation loss in the main plasma is reduced to 5–10% of the absorbed power and impurity concentrations are significantly suppressed at very low level (Zeff = 1.5-2.0). In the experiments with graphite walls, short periods of H-mode phases were found in the outside X-point divertor discharges although improvement in energy confinement is limited to within 10%. In limiter discharges, combination of high vessel temperature (210–300 °C) and low current limiter discharges (1.5 MA) without gas puffing is effective as a wall conditioning in obtaining reproducible high density discharges. The maximum parameters, Ip = 3.2 MA, ne = 1.2 × 1020m−3, qeff = 2.2 at Bt = 4.8T, a Murakami parameter of 7.5, and a stored energy of 3.1 MJ have been attained. The maximum density is limited by the occurence of the MARFE. During high power neutral beam heating, enhanced carbon influx was observed possibly due to chemical sputtering. The wall pumping by the inner wall is still effective in decreasing the electron density from the high density region of ne ~ 1 × 1020m−3. An empirical energy confinement scaling of JT-60 has been drawn in terms of an offset linear function as τE = 0.19 I1.9PPabs + 0.062 a1.8p

During the venting period from April to May 1987 about a half of TiC-coated molybdenum and Inconel 625 first walls were removed and graphite first walls were installed. The overall surface area and the weight of the graphite tiles are 200 m2 and 3500 kg, respectively. Two kinds of isotropic graphite (Ibiden ETP-10 and Hitachi-Chemical HCB-18S) were used. Special attention was paid to the contamination of the graphite first walls during the courses of fabrication and installation. Main discharge were tried after 7 d conditioning, which included 3 d bakeout at the temperature of 300°C, 18 h glow discharge cleaning and 36 h Taylor-type discharge cleaning keeping the vessel temperature around 280°C. A 0.5 MA stable plasma was successfully attained at the 6th shot and a very rapid start-up of the discharge conditioning with plasma current up to 1 MA was achieved with 14 shots. An ultimate pressure of 6.4 × 10−7 Pa was obtained with the vessel temperature at RT and the residual gases were water, carbon oxide and carbon hydride.

Itohs' model to explain the H-mode transition is incorporated into the one dimensional transport code. The H-mode transition observed in JT-60 divertor discharge is investigated numerically by using this code. It is shown that the threshold power for the H-mode transition and the threshold density agrees well with the experimental data. The change of the plasma parameter after the transition is also investigated.

JT-60 is a large hydrogen tokamak with a compact poloidal divertor. The maximum input energy into the torus is about 500 MJ, and divertor studies are among the most essential experiments. The divertor in JT-60 has been studied numerically from three different view points, i.e. MHD equilibrium control, divertor plasma control and surface temperature control of the device. These results suggest that a serious heat flux can be avoided and a plasma contacting the divertor plates can be cooled down to 10 eV or less by controlling the magnetic configuration and divertor plasma density. The magnetic configuration with the divertor can be controlled by changing various coil currents, and the divertor plasma density can be controlled by changing the particle recycling condition in the divertor. The particle number in the main plasma can be also controlled by combining the compact divertor and gas influx both in the main chamber and in the divertor chamber.

Impurity control is one of the essential elements in realizing the most favorable operating parameters of a beam-heated tokamak [I]. In JAERI, development of surface coatings of low-Z materials is intensively underway to take advantage of these materials in reducing the undesirable effect of wall-released impurities in JT-60. Work in this area includes development of coating techniques as well as evaluation of candidate coatings. The coating development plan consists of two stages [2,3]. The aim in the first stage is to develop low-Z coatings which are suitable for the attainment of operating parameters for the break-even condition. The goal of the second stage is to develop advanced materials which are applicable to fusion experimental reactors. The coating development includes precoating carried out before wall installation and in-situ coating carried out after the completion of the vacuum vessel.

The Japan Atomic Energy Research Institute has been constructing a large tokamak JT-60, which constitutes a focus of the second phase (1975-□) of the Japanese fusion development program. Its objective is to investigate plasma confinement and heating in reactor-grade hydrogen plasmas and to conduct an integrated test of associated fusion technologies. It has a long-pulse capability of 5–10 s in pulse length and, among other features, has a magnetic limiter (divertor), various methods of supplementary heating, and radiofrequency current drive. Construction of the JT-60 device, which started in April 1978, will be completed in April 1985. Development of heating devices has been successful, and full installation and test of the heating devices of 30 MW in absorbed power in plasmas will be completed by July 1986. The present paper describes the JT-60 project for its objective, device features, status of construction and development, and its experimental program.