Search

Toroidal current penetration is studied in current ramp experiments on the JIPP T-IIU tokamak. The poloidal magnetic field profile in the peripheral region of a plasma (0.5 ≤ r/a ≤ 1.0) has been measured directly with a newly developed fast response Zeeman polarimeter. The experimental results indicate that a skin effect of the toroidal current density is clearly observed during both the current ramp-up (CRU) and the current ramp-down (CRD) experiments. The experimentally obtained toroidal current density profiles are well described by the profiles calculated on the assumption of neoclassical electrical resistivity, although the difference between the neoclassical electrical resistivity and the Spitzer electrical resistivity is not clearly distinguished because of the relatively collisional peripheral plasma region. Quasi-linear Δ' analysis of tearing modes for the measured current density profile is consistent with the time behaviour of the coherent magnetohydrodynamic (MHD) modes such as m = 4/n = 1 or m = 3/n = 1 (m is the poloidal mode number and n is the toroidal mode number) often observed during the CRU phase. An obvious enhancement of the current penetration by the MHD modes is not detected in this ramp-up experiment, where the relative amplitude of the poloidal field fluctuations is less than 0.8%.

Laser scattering signals that indicate the presence of small dust particles (diameter ≤ 2 μm) have been occasionally observed in the JIPPT-IIU tokamak chamber. This phenomenon was reproduced deliberately by spreading carbon dust from the top of the vacuum chamber. No noticeable effect on the plasma was observed for dust falls of up to at least 106 particles (10 μg) in 20 ms during discharges. Dust falling just before plasma startup seemed to be confined but it was soon ejected (in less than 30 ms)

Multiple and small sample volume measurements of the density turbulence and potential profile measurements in tokamak plasmas were conducted using a heavy ion beam probe. The obtained wavenumber-frequency spectrum S(k, omega ) shows that the cross-sections of neutral beam injection (NBI) heated plasmas are divided into three regions of different turbulence characteristics outside the layer where the direction of the density turbulence propagation reverses, a low frequency and low wavenumber mode travelling in the ion diamagnetic drift direction dominates. The region enclosed by this reversal layer is divided into two parts during nearly perpendicular NBI heating: a region where the propagation velocity is near the Etau /Bt poloidal rotation velocity and a bad curvature region of very small wavenumber and high propagation velocity. The region of high propagation velocity, found in NBI plasmas, disappears in the ohmic plasmas. In addition, a small component that propagates in the ion diamagnetic drift direction is observed in NBI plasmas

A pellet injection system with controllability of the injection angle has been developed and installed in the JIPP T-IIU tokamak. A long helical ablation cloud (‘tail’) has been observed when an ice pellet is injected into the tokamak plasma using the controllable injection angle system. The ‘tail’ does not flow along the direction of the total magnetic field lines of the torus. In order to understand the details of the ice pellet ablation structure in the ‘tail mode’, the local parameters of electron density and temperature within the ablation cloud are obtained by a spectroscopic method. The electron density determined from Stark broadening of the Balmer β line is in the range 1.0 × 10 16 −1.0 × 10 17 cm −3 while the temperature obtained from the line-to-continuum intensity ratios is in the range 1–4 eV. It has also been observed by a heavy ion beam probe (HIBP) measurement that the potential at the plasma centre becomes more negative and the potential profile becomes deeper negatively after pellet injection.

Fast changes of electric potential with different polarities are observed during sawtooth oscil- lations in the core region of a tokamak plasma using a heavy ion beam probe. Near the inversion radius the polarity of the observed change of the potential is found to be dependent on the swift movement of the hot core at the crash and is consistent with the prediction of one- fluid MHD theory. Near the magnetic axis the change of the potentials is positive and outside the inversion radius the change is negative. This is in contradiction with the MHD prediction. A periodic crash of the central electron tempera- ture called a sawtooth oscillation is a common fea- ture of tokamak plasmas (l-61. After the crash, a flattening of the pressure throughout the q = 1 sur- face is observed. Kadomtsev proposed a theoretical model of the resistive reconnection (7). The theory predicted the flattening of the q profile as well as the pressure profile inside the q = 1 surface. The predic- tion was, however, found to be in contradiction with recent measurements of the current density in toka- mak plasmas (8). In addition, experiments in large machines such as JET and TFTR showed much faster crashes than those predicted by resistive reconnection (2-61. More refined theories, which take account of the kinetic effect and ergodicity on the reconnections, have now been proposed (9-121. The present Letter reports the first measurement of potential changes during sawtooth oscillations and discusses the impli- cations of the results. The experiment is performed with the JIPP T-IIU tokamak (13). Its major radius is 93 cm and the maximum toroidal field is 3 T. The experiment is conducted at a relatively low electron density of about 2 x 1013 cm3, so that good beam penetra- tion into the centre of the plasma could be obtained. A 450 keV singly ionized thallium beam is injected

Thin lithium layers are deposited on metal or graphite walls by evaporation in vacuum. Clean lithium surfaces have wide chemical activities on such gases as O 2 , CO and CH 4 . In particular, a strong gettering effect on oxygen is observed; the minimum number of 0 atoms gettered by the lithium layer is about one half the total number of Li atoms deposited on the wall, thus suggesting the formation of Li 2 O. On the other hand, H 2 gas hardly reacts with the lithium surface which, however, displays a large pumping effect in a hydrogen glow discharge. The maximum number of H atoms pumped by the lithium layer below 180°C is approximately equal to the number of Li atoms in the vessel, probably due to the formation of LiH. Helium glow conditioning of the used lithium layer allows the partial recovery of lithium from losses to hydrogen pumping or oxygen gettering. This lithium evaporation method has been applied to JIPP T-IIU. 100-250 mg of lithium was deposited onto a limited vessel area of ∼ 1 m 2 . The lithium coating leads to 20-50% reduction in oxygen and carbon impurities with less hydrogen recycling in ohmic and NBI discharges

By the use of simultaneous fast toroidal and slow poloidal sweeps of a heavy ion beam, plasma potential profiles in the JIPP T-IIU tokamak are measured at the rate of 120 spatial profiles per second. A new method to eliminate the error due to the change of out-of-plane entrance angle caused by large plasma current is successfully applied. One of the key factors is the very homogeneous characteristics of the energy analyser along the wide slit length. A shaped electrode system, instead of guard rings with a resistor chain, successfully increases the homogeneity of the analyser. Error due to the significant change of in-plane entrance angle during a poloidal sweep is carefully minimized and calibrated by a secondary beam ionized by neutral gas introduced in the vacuum vessel. The depth of the measured potential at the plasma centre reaches more than 1.5 keV in ohmic plasmas with an ion temperature of 600 eV. In some cases, a rather wide region of positive potential and a sharp decrease of the potential in the centre of the plasma are observed.

An experiment in a new regime of ion Bernstein wave (IBW) heating was carried out using 130 MHz high power transmitters in the JIPP T-IIU tokamak. The heating regime utilized the IBW branch between the 3rd and 4th harmonics of the hydrogen ion cyclotron frequencies. This harmonic number is the highest one used in IBW experiments conducted previously. The net radiofrequency (RF) power injected into the plasma is around 400 kW and is limited by the transmitter output power. Core heating of ions and electrons was confirmed in the experiment and density profile peaking was found to be a special feature of IBW heating. Peaking of the density profile was also found when IBWs were injected into neutral beam heated discharges. An analysis, using a transport code with these experimental data, indicates that particle and energy confinement should be improved in the plasma core region upon application of IBW heating. It is also found that the ion energy distribution function observed during IBW heating has a smaller high energy tail than those observed in conventional fast magnetosonic wave ICRF heating regimes. The ion energy distribution function obtained during IBW heating is in reasonable agreement with that calculated using the quasi-linear RF diffusion/Fokker-Planck model

A series of experiments has been conducted on the JIPP TII-U tokamak since 1989, using the newly constructed 130 MHz radiofrequency system. It has been predicted theoretically that the fast wave in this frequency range interacts weakly with particles. Two mechanisms of wave absorption have been identified in the experiment: electron Landau damping/transit time damping and 3rd harmonic ion cyclotron heating. The former mechanism is intimately connected with fast wave current drive and the latter can provide a new regime of plasma heating or a possible method of controlling the transport of alpha particles. It is found that the efficiency of the 3rd harmonic ion cyclotron heating is improved by using it in combination with neutral beam injection and ion cyclotron range of frequency heating. The heating efficiency obtained is as high as that of conventional heating. The experimental results are also analysed on the basis of a global wave theory which takes into account wave-particle interactions. These mechanisms of interaction are competing with each other; this will also be the case under more realistic reactor conditions.

Ice pellet injection experiments were carried out in the JIPP T-IIU tokamak in order to study thermal (cooling) transport just after injection. The cut-off problem of ECE signals due to the rise in density has been resolved by careful measurements of the temperature profile at a high time resolution during its decay phase. The phenomenon of ultra-fast cooling (so-called precooling) has been identified using the two different methods of ECE and soft X-ray measurements. In the outer region of the plasma the cooling propagation velocity is comparable to or slightly greater than the pellet velocity, while in the central region the propagation velocity is significantly greater than the pellet velocity. Ice pellets were injected into various kinds of JIPP T-IIU plasmas, the current and sawtooth phase of which had different values, including a no-sawtooth plasma. The existence of the q=1 surface and arrival of a pellet near the q=1 surface have turned out to be necessary conditions for precooling, and even just after the sawtooth crash the precooling starts around the q=1 surface, not at the plasma center. Simultaneous measurements of electron temperature and density profiles indicate that the central temperature always decreases before the central density increases. Some anomalous transport might be induced by pellet injection at the central region.

Toroidal plasma rotation driven not by the momentum input but by the electric field induced by loss of ions has been observed for plasmas heated by perpendicular neutral beam injection. The rotation is found to increase in the direction opposite to the plasma current when the plasma ions are heated by ion cyclotron resonance frequency waves. The rotation driven by the electric field is quantitatively consistent with that inferred from the ambipolarity of particle fluxes.

Fast wave absorption and heating at a relatively high ion cyclotron harmonic (ω ΩD) were studied on JIPP T-IIU. In spite of the low absorption efficiency predicted by a simple theory for present experimental parameters, appreciable electron and ion heating was observed. Heating is most effective at relatively low densities (e 1 × 1019 m−3). An ion tail on the majority deuterium energy spectrum was observed, but not an ion tail on the minority hydrogen spectrum, which suggests the possibility of direct absorption of mode converted ion Bernstein wave power by deuterium ions.

The H-mode, an improved confinement regime, was attained recently in JIPP T-IIU high power heating experiments, in a limiter configuration without any shaping of the plasma cross-section. This H-mode is unusual because it was obtained with heating in the ion cyclotron range of frequencies (ICRF). It was also attained with combined ICRF and neutral beam injection (NBI) heating. The threshold power level obtained with ICRF alone is similar to or less than that obtained in the combined heating case. The dependence of the power threshold of the H-mode on various plasma parameters has been studied. It increases with the plasma current and is insensitive to the plasma density, and there is an optimum value of the toroidal field intensity. The power deposition profile for ICRF heating has been analysed with a ray tracing code and used to explain the observed dependence on the toroidal magnetic field. The paper also discusses a class of discharges with improved confinement observed in the same series of experiments. These discharges had a power level close to the H-mode threshold power and exhibited a marked improvement of confinement. They were, however, different from H-mode discharges in the time evolution of the profiles.

Several improvements in the high‐voltage heavy‐ion‐beam probe (HIBP) are discussed. (1) It is clearly found that the precision slide mount of the detector plates 30° parallel to the base electrode is very effective for the determination of the in‐plane entrance angle of the beam in the analyzer to estimate the error in the potential measurement. (2) A two‐staged optical trap in the HIBP greatly reduced the effect of the UV radiation in the analyzer. (3) A multiple‐plate detector up to 13 measurement points clearly showed the direction of the propagation of the turbulence and path‐integral effects.

A high frequency 130 MHz ion Bernstein wave heating (IBWH) experiment was carried out in JIPP T‐IIU tokamak. Three‐elements center‐fed antenna permitted the launching of a relatively well defined nni spectrum. The 3ΩH heating represents the highest hydrogen harmonic experiment thus far performed. Core plasma heating of both ions and electrons by IBWH alone and also in conjunction with NBI heating was observed. The radiated antenna power reached to 400 kW. An interesting density profile peaking was observed during IBWH, implying that the particle diffusivity is greatly reduced in the half‐radius region of the plasma. This IBWH confinement improvement was retained even with additional high power NBI heating. Some plasma transport analyses have been performed for those discharge.

In a circular cross-section plasma bounded by a limiter, the {ital H}-mode transition is triggered by a rapid rampdown in plasma current during auxiliary heating, even in the case when the edge electron temperature gradually decreases prior to the transition. This result suggests that the transition is governed by the enhancement of the magnetic shear near the plasma edge, associated with the radial modification of the edge current-density profile.

Practicability of cleaning Thomson scattering windows by a laser blow-off method is demonstrated and discussed in terms of sweep speed and window damage probability. Optimum sweep speeds of ~6 cm^2/s are obtainable with a 30-Hz, 0.4-J/pulse Nd:YAG laser.

By using the JIPP T-IIU device the wall surface carbonization experiment was carried out with ECR discharge (2.45 GHz, 1.6 kW) in a gas mixture (CH4/H2 = 5050, 1.5 × 10−2 Pa). The film is free from oxygen and metal. The mean deposition rate (on the first wall except special plane like limiter) can be predicted by simple measurement of CH4 partial pressure. Wall heating (up to 300°C or more) and ion bombardment (E > 100 eV) are effective to decrease the hydrogen concentration in the film.

The paper discusses the results of a heating experiment in the ion cyclotron range of frequencies (ICRF) conducted on the JIPP T-IIU device with a five-element antenna array. This antenna system provides precisely shaped k∥, spectra which vary according to the relative phasing angle of the neighbouring antennas, Δ. In the experiment, Δ is scanned in a close pitch and it is found that the electron heating efficiency is drastically improved with Δ = π and the lowest efficiency is obtained with Δ = 0. This observation is analysed by calculating the k∥ spectra and then the power deposition profiles using a ray tracing code. It is also found that the effect of the phasing on the impurity problem is unexpectedly small. The reasons for the discrepancies between these data and theoretical expectations are also discussed.

Data of JIPP T-II ICRF heating experiments are presented. The experiment covers three typical cases: the low-concentration hydrogen minority case, the high-concentration hydrogen minority case, and the 3He-minority case. The best heating efficiency is obtained for the 3He-minority case. It is shown through power balance analysis that the two H-minority cases are different in the wave energy deposition profile. The difference is explained by the presence of a local cavity mode in the high-concentration minority case. The ion temperature stops rising at a power density level of 0.65 W-cm−3 for the hydrogen minority experiments. No such deterioration is found in the case of the 3He minority experiment up to a power density level of 0.45 kW-cm−3. This is the maximum possible to attain within the maximum power injection (180 kW) up to which the experiment is conducted. An analytic solution of the Fokker-Planck equation is derived in order to interpret the deterioration of the heating efficiency.

In situ carbon coating was carried out with a glow discharge in methane-hydrogen mixtures on the wall of the JIPP T-IIU tokamak at room temperature. The concentration of hydrogen and deuterium in C-films was analyzed by thermal desorption spectroscopy (TDS) and elastic recoil detection analysis (ERDA). Good agreement was obtained between results deduced from these two methods. Hydrogen concentration is 40–55 at. % in the C-films. Replacement of H in the C-layer was tried using a pure D2 glow discharge. In order to achieve complete replacement, a fluence of 2 × 1017 cm−2 was needed at room temperature. The replacement process was monitored by a quadrupole mass analyzer and results are compared with Hα/Dα observation of tokamak discharges and the results of the surface analysis.

The effectiveness of the in-situ carbon coating (carbonization) has been demonstrated to reduce the radiation loss by iron impurities during ICRF heating in the JIPP T-IIU tokamak. As a result of carbonization, the total radiation loss decreased down to one fifth of the RF power, which resulted in an increase in electrons and total stored energy compared with these conditions before carbonization. The thickness of the carbon layer was 300–900 A, and its toroidal uniformity was within a factor of 3, although only one anode and one gas-inlet were used. A thin carbide layer is formed between the C-film and the stainless steel substrate with carbonization at room temperature. The hydrogen concentration is 40–50 at.% in the carbon layer. Deposition of carbon was observed on window materials. The deposition rate was relatively less on electrical insulators compared to the deposition rate on metals.

A plasma current is initiated and raised to a quasistationary level of about 20 kA by injection of the lower hybrid wave into a cold and low-density plasma produced by electron cyclotron resonance. The plasma current rises more slowly than the experimentally obtained L/sub p//R/sub p/ magnetic diffusion time of the bulk plasma. The current rise time is inversely proportional to the bulk electron density, and agrees well with the collision time of the current-carrying high-energy electrons with the bulk plasma.

Suprathermal electrons which carry a seed current are generated by non-resonant parametric decay instability during the initial phase of lower-hybrid current startup in the JIPP T-IIU tokamak. From the numerical analysis it is found that parametrically excited lower-hybrid waves at the lower side-band can bridge the spectral gap between the thermal-velocity region and the low-velocity end of the pump power spectrum.

In the JIPP T-IIU tokamak, a high power ICRF heating experiment has been conducted, up to an extremely high power density (~2 MW·m−3), with a total RF power of PRF = 2 MW. Great attention has initially been paid to the problem of impurities, and it has been found that (a) the adoption of low Z materials for the limiter, (b) in situ carbon coating (i.e. carbonization) and (c) adequate gas puffing synchronized to the RF pulse are very effective in suppressing radiation loss. With these methods, a remarkable reduction in metal impurities (especially in iron impurity) has been achieved; the total radiation loss has been reduced to less than 30-40% of the input power. In these reduced radiation loss plasmas, the characteristics of ICRF heated plasmas have been studied intensively. With an increase in the ICRF heating power, a deterioration of the energy confinement time has been observed, indicating quantitative agreement with the Kaye-Goldston L-mode scaling. It is shown that the so-called profile consistency, which is the leading feature in neutral beam heated plasmas, also holds in ICRF heated plasma. It has been observed that the electron temperature profile only responds to the safety factor q(a) and does not change when the deposition profile is controlled by tailoring the k1 spectrum.

Replacing stainless steel limiters by graphite limiters, we found that radiations from iron and titanium ions were significantly reduced. Total radiation and loop voltage also decreased. This indicates that the limiters are the major impurity sources both in the ohmic and RF heating phases. Although titanium radiations increased with RF power injected by an antenna with a titanium Faraday shield, the maximum intensity was much smaller than in the experiment where the titanium-flashed stainless steel limiters were used. Thus it has been found that the Faraday shield is less important as an impurity source than limiters. Toroidal asymmetry observed for O II radiation suggest that the energetic charge-exchange neutrals play a role in releasing oxygen from the wall and that those energetic particles are relatively abundant in the toroidal sections near the antenna. The Hα + Dα radiation decreases during the RF pulse around the limiter, which may be due to the change in hydrogen/ deuterium recycling at the limiter. The reduction of Hα + Dα is greater with graphite limiters than with stainless steel limiters. The relation between recycling and impurity release is briefly discussed.

Surfaces of two sets of limiters were analyzed by PIXE and RBS. One is made of stainless steel SUS 304 and the other made of graphite. The two limiters are the same shape and size and have been exposed to similar numbers of tokamak discharges. Molybdenum and titanium were observed on both limiters. Although the deposited metals were larger on the graphite limiters than the amount on the steel limiters, the plasma radiation intensity from Mo or Ti was smaller with the graphite limiter than with the steel limiter during tokamak discharges. The sputtering rate of the metals on the graphite limiter was investiaged with an argon-ion beam. The apparent sputtering yields of metals are small on carbon limiters, and this results in the lower metal contamination of tokamak plasmas with carbon limiters than with metal limiters.

MHD activity and turbulence near the plasma edge are studied on the JIPP T-IIU tokamak using a new high-speed visible-light image-converter video-camera system. Different from conventional cinefilm and photo-diode array systems, this system is convenient for the instantaneous display of the high-speed optical plasma images after plasma discharges. The effectiveness of this instrument for research on the plasma-wall interaction is demonstrated in this experiment. The observed characteristics of the edge-plasma behavior are as follows: (1) The helical mode structure of the luminous plasma boundary suggesting plasma-surface interaction is identified in the case of OH or ICRF-heated discharge. (2) In the LH-current drive case, no clear large-scale coherent modes are identified, however, on the initial stage a medium-scale turbulence ( λ ~ a few cm, f ~10 kHz ) is found. (3) Before current disruptions, an m = 2 or m = 3 helical mode is found and up-down asymmetric light emissions are often observed during disruptions.

After several hours of discharge cleaning, it has been observed that the decay time of the water‐vapor partial pressure changes from one well defined value to another. A long decay time observed in the later phase can be interpreted as a result of a slow deoxidization rate of chromium oxide, which may dominate the cleaning process in this phase. Optimization of the electron density for the cleaning is discussed, comparing the experimental results on density dependence of the water‐vapor partial pressure with a result based on a zero‐dimensional calculation for particle balance. One of the essential points for effective cleaning is found to be raising the electron density of the plasma high enough that the dissociation loss rate of H2O is as large as the sticking loss rate. A density as high as 1011 cm−3 is required for a clean surface condition where sticking probability is presumed to be around 0.5.

Ion Bernstein Wave (IBW) heating has been examined on the JIPP T-IIU tokamak under two different conditions referred to as Mode-I and Mode-II. In the Mode-I regime, a wave is launched on an IBW branch between the third and fourth cyclotron harmonics of deuterium ions. In the Mode-II regime, a wave is launched on a branch between the second and third cyclotron harmonics. These two modes show quite different heating characteristics. The causes of this difference are analysed by using a simple model to determine the k|| spectrum of the excited wave and by applying a ray tracing code. In connection with the Mode-I experiment as discussed in a previous report (1985), two important new experimental results are obtained. It is shown that an IBW heats the core of the plasma rather than causing plasma-edge interaction, as anticipated. It is also shown that the energy tail of the hydrogen ions is higher than that of the deuterium ions, which indicates that the responsible heating mechanisms are different.

The energy confinement characteristics of ICRF-heated Tokamak plasmas are studied at high input power density approximately 2 MW m-3 volume averaged, on the JIPP T-IIU device (R=0.91 m, a=0.23 m). High electron and ion temperatures (Te approximately 2.5 keV, Ti approximately 2.0 keV, at each maximum) have been achieved by the operation at a plasma current IP of 280 kA, plasma line-averaged electron density ne of 7*1013 cm-3 and input power of 2 MW, with a suppression of total radiation loss (30-40% of the total input power) by a carbon coating on the vacuum vessel. The fraction of ICRF power absorbed by the plasma, alpha , is determined experimentally from the decay of the stored plasma energy just after the ICRF pulse is terminated. The value of alpha increases slightly with increasing electron density and decreases from 90 to 70% as the ICRF power is increased from 1 to 2 MW m-3 volume averaged. The global energy confinement time tau epsilon , defined by WP/(POH+ alpha Pr.f), decreases by a factor of 2 approximately 3 from that in Ohmic plasmas as the heating power increases up to 2 MW. It is found that the energy confinement time has a strong line-averaged electron density dependence as tau E varies as ne0.6, which is obtained by the use of the measured absorbed power, while the Kaye-Goldston scaling predicts tau E varies as ne0.26.

Current drive by fast magnetosonic waves is successfully performed in the JIPP T-IIU tokamak by means of a four-element dipole antenna array with a Faraday shield. A plasma current of about 50 kA is sustained by an RF power of about 80 kW in a low-density plasma (3 x 10/sup 12/ cm/sup -3/) with an efficiency comparable to that of slow wave current drive. A density limit for fast magnetosonic current drive is observed contrary to theoretical expectations based on linear wave propagation.

aNational Institute for Fusion Science, Toki 509-5292, Japan bResearch Institute for Applied Mechanics, Kyushu University, Fukuoka 816-8580, Japan cEuratom0UKAEA Fusion Association, Culham Science Centre, Abingdon OX14 3EA, United Kingdom dArt Science and Technology Center for Cooperative Research, Kyushu University, Kasuga 816-8560, Japan ePlasma Research Center, University of Tsukuba, Tsukuba 305-8577, Japan fGraduate School of Frontier Sciences, The University of Tokyo, Kashiwa 277-8561, Japan gInstitute of Plasma Physics, National Science Center, Kharkov Institute of Physics and Technology 1, Akademicheskaya St., Kharkov, 61108, Ukraine hFaculty of Engineering Science, Kansai University, Osaka 564-8680, Japan iTokyo Institute of Technology, Tokyo 152-8550, Japan jGraduate University for Advanced Studies, Toki 509-5292, Japan

In ICRF heating experiments the deposition profile of the rf power directly coupled with electrons has been derived experimentally, and the electron thermal diffusivity has been evaluated. The rf power delivered to electrons is estimated to be \hatPe∼60% of the radiated power from antennas. It is found that the deposition profile is relatively broad (Prf,e(r)=Prf,e(0)(1-(r/a)2)kwithk=1.8), and this is interpreted in terms of characteristics of the mode-converted ion Bernstein wave. With the deposition profile experimentally obtained, the electron thermal diffusivity χehas been calculated, resulting in χe∼0.65 m2/s at n̄e=9×1019m-3.

The effect of the C-coating has been demonstrated for tokamak plasmas with high power heating. In situ carbon coating has been made with a glow discharge in a methane/hydrogen mixture gas. Without C-coating, iron contamination has been so severe in megawatts of ICRF heating that the total radiation loss exceeds the rf-heating power, and the electron temperature, after once heating up, then decreases continuously during the latter half of 55 ms rf pulse. With C-coating, the radiation loss is reduced to be one fifth of the rf power and the electron temperature T e becomes stationary at the end of the rf pulse. As a result of this change in T e , the total stored energy increases more than the one without C-coating.