Search

The uptake of tritium at 573 K by polycrystalline tungsten limiter material pre-exposed to TEXTOR plasmas was investigated by the imaging plate technique (IP) and found to be mostly non-homogeneous on the plasma facing surface. Particularly high concentrations of tritium were apparent in areas attributed to carbon deposition. The surface density of tritium outside of deposition zones was essentially comparable on both sides of the examined tungsten plate. Under a stream of argon at ambient temperature tritium was predominantly released as tritiated water. While tritium is initially liberated with rates in the (MBq/h) range, after a few days the rate drops to about 100 Bq/h, decreasing even further thereafter. Under atmospheric conditions the concentration of tritium on the surface remained virtually unchanged over a rather extended period of time, i.e. more than 500 d. Tritium in surface zones other than of “deposition” was also firmly trapped at ambient temperature.

Tungsten (W) specimens previously exposed to deuterium (D) plasmas both in the TEXTOR tokamak and high flux linear plasma generator (LPG) were subsequently loaded with tritium at 573 K for 3 h. Ret...

Plasma-facing components (PFCs) in ITER will be castellated by splitting them into small-size blocks to maintain the thermo-mechanical stability. However, there are concerns in particular on retention of codeposited radioactive fuel in the gaps. An R&D program is underway in TEXTOR addressing this acute issue of castellation. Material migration and fuel inventory are investigated using long- and short-term discharge-resolved experiments with castellated structures in TEXTOR. Significant impurity transport to the gaps was detected and results were in part quantitatively reproduced with 3D-GAPS code. Deposits containing up to 70 at.% of tungsten on the gap areas closest to the plasma were detected in recent experiments. Deposition in the gaps accompanied by metal mixing demand for development of effective cleaning techniques. In experiments with ITER-like castellation, the gaps were cleaned from carbonaceous deposits using oxygen plasmas at 350 °C. This contribution contains an overview of experimental and modeling results along with recommendations for PFCs in ITER.

Runaway electrons generated during ITER disruptions are of concern for the integrity of the plasma facing components. It is expected that a power of up to 8 GW is exposed to ITER PFCs. We present ill this article observations from JET and TEXTOR on the generation of runaways and the heat load deposition. Suppression techniques like massive gas injection and resonant magnetic perturbations are discussed. (C) 2009 Elsevier B.V. All rights reserved.

Castellation will be used in divertor and first wall components to provide thermo-mechanical stability of ITER. Radioactive fuel may be stored in the gaps of castellated structures representing a safety issue for ITER. Tungsten castellated structures with different shapes were exposed in TEXTOR to investigate the impact of cell shaping on impurity transport and fuel deposition in the gaps. After exposure a significant intermixing of tungsten was detected in carbon deposits in the gaps reaching 70 at.% of W in the deposited layer. This will provide difficulties in cleaning the gaps in ITER. Poloidal gaps of shaped cells contained a factor or 3 less deuterium than those of rectangular cells, the carbon deposition exhibited only marginal advantages of a new geometry. Poloidal and toroidal gaps contained comparable amount of C and D. Significant deposition at the bottom of gaps was measured which could only partly be reproduced by modeling.

Erosion and deposition behaviour of the TEXTOR main toroidal belt limiter ALT-II has been studied for two generations of limiter tiles with different surface shapes. Two specially prepared tiles with marker holes were subsequently exposed during experimental campaigns with a total plasma duration of > 7000 s and a surface averaged background ion fluence exceeding 2⋅10 25 m -2 . The tile surface relief was measured by optical profilometry and SIMS before and after the exposure. The highest surface erosion of > 10 μm on both tiles was observed in regions with an angle of incidence of the magnetic field α of > ~ 1°, whereas in the regions with α ~ 0.1° both net-erosion and net-deposition were observed depending on the tile shaping. The marker holes were always deposition dominated. Modelling by the ERO code explains qualitatively the surface erosion and deposition behaviour in terms of background flux dilution for shallow angles of incidence.

Tracer experiments have been carried out by injection 13 C marked methane through test limiters exposed to the scrape-off-layer in TEXTOR. The influence of impact energy and flux on depositing 13 C species has been studied. One experiment has been performed with biased test limiter (−300 V) in order to increase energy of positively charged ions and the other one with 10 times reduced 13 CH 4 injection rate compared to previously used injection rate. Biasing of the test limiter increases the resulting 13 C deposition by a factor of ∼6 – post-mortem analysis yields a 13 C deposition efficiency of ∼1.7% compared to ∼0.3% without biasing. Reducing the injection rate increases 13 C deposition efficiency to ∼0.7%, which is more than two times larger compared to experiments with previously used injection rate. ERO modelling shows that enhanced re-erosion of redeposits is still necessary to reproduce measured 13 C deposition efficiencies.

Recent experiments on Ion Cyclotron Wall Conditioning (ICWC) performed in tokamaks TEXTOR and ASDEX Upgrade with standard ICRF antennas operated at fixed frequencies but variable toroidal magnetic field demonstrated rather contrasting parameters of ICWC discharge in scenarios with on-axis fundamental ion cyclotron resonance (ICR) for protons, omega=omega(cH+), and with its high cyclotron harmonics (HCH), omega=10 omega(cH+). HCH scenario: very high antenna coupling to low density RF plasmas (P-pl approximate to 0.9P(RF-G)) and low energy Maxwellian distribution of CX hydrogen atoms with temperature T-H approximate to 350 eV. Fundamental ICR: lower antenna-plasma coupling efficiency (by factor of about 1.5 times) and generation of high energy non-Maxwellian CX hydrogen atoms (with local energy E-perpendicular to H >= 1.0 keV). In the present paper, we analyze the obtained experimental results numerically using (i) newly developed O-D transport code describing the process of plasma production with electron and ion collisional ionization in helium-hydrogen gas mixture and (ii) earlier developed 1-D Dispersion Relation Solver accounting for finite temperature effects and collision absorption mechanisms for all plasma species in addition to conventionally examined Landau/TTPM damping for electrons and cyclotron absorption for ions. The numerical study of plasma production in helium with minor hydrogen content in low and high toroidal magnetic fields is presented. The investigation of the excitation, conversion and absorption of plasma waves as function of B-T-field suggests that only fast waves (FW) may give a crucial impact on antenna coupling and characteristics of the ICWC discharge using standard poloidally polarized ICRF antennas designed to couple RF power mainly to FW. The collisional (non-resonant) absorption by electrons and ions and IC absorption by resonant ions of minor concentration in low T-e plasmas is studied at fundamental ICR and its high harmonics.

The radio-frequency (RF) plasma production technique in the ion cyclotron range of frequency (ICRF) attracts growing attention among fusion experts because of its high potential for solving several basic problems of reactor-oriented superconducting fusion machines, such as ICRF wall conditioning in tokamaks and stellarators (Te = 3−5eV, ne

W nano-structures (fuzz), produced in the linear high plasma device, NAGDIS, were exposed to TEXTOR edge plasmas (ohmic He/D mixed plasma and pure D plasma) to study formation, erosion and C deposition on W fuzz in tokamak plasmas for the first time. Fuzz layers were either completely eroded or covered by C deposit. There was no clear indication of W fuzz growth under the present conditions. There was no significant difference of C deposition between ‘thick’ fuzz (500–600 nm in thickness) and ‘thin’ fuzz (300–400 nm) in the He/D plasma. On the W fuzz surface, C deposition was enhanced probably due to reduction of effective sputtering yield and effective reflection coefficient of carbon ions, similar to roughness effects. Formation and erosion of W fuzz in tokamak devices and role of impurities are discussed.

The three-dimensional (3D) features of plasma edge profiles and wall interaction patterns induced by edge resonant magnetic perturbation fields (RMP) are discussed comparing TEXTOR and DIII-D. We show that the scrape-off layer (SOL) profiles and decay lengths depend the edge safety factor, the RMP base mode as well as on the plasma rotation during RMP application indicating modification of SOL transport by the 3D perturbation fields. This is compatible with channelling of particle and heat efflux along open perturbed field lines in the very edge of the plasma boundary into a completely re-arranged, helically striated 3D divertor footprint. The distribution of the measured divertor heat and particle fluxes at DIII-D match the vacuum modelled magnetic footprint topology in L-mode while in H-mode the striation width exceeds the modelled footprint width by 15–30%. This 3D structure of the measured heat and particle fluxes results in a new situation for the material erosion properties and initial quantification of the net-erosion within the 3D footprint shows in L-mode a 50% decrease of the chemical erosion yield and evidence for a comparably small 15-20% increase in physical sputtering. Extrapolation of these findings to ITER by vacuum modelling of the magnetic footprint for the actual ELM control coils shows a similar vacuum magnetic footprint topology as found at DIII-D during RMP ELM suppression. However, the open field lines escape the CFC covered ITER divertor area potentially transferring net-erosion characteristics from the CFC domain onto the Tungsten including so far unconsidered heat and particle loads on this sensitive material.

This paper focuses on a study of the principal operation aspects of standard ICRF heating antennas in the ion cyclotron wall conditioning (ICWC) mode: (i) ability of the antenna to ignite the cleaning discharge safely and reliably in different gases including those most likely to be used in ITER – He, H 2 , D 2 and their mixtures, (ii) the antenna capacity to couple a large fraction of the RF generator power (>50%) to low density (∼10 16 –10 18 m −3 ) plasmas and (iii) the RF power absorption schemes aimed at improved RF plasma homogeneity and enhanced conditioning effect. The ICWC discharge optimization in terms of RF plasma wave excitation/absorption resulted in successful simulation of the conditioning scenarios for ITER operation at full field (JET) and half-field (TEXTOR, TORE SUPRA, ASDEX Upgrade).

In the present paper, inter-machine studies of Ion Cyclotron Wall Conditioning (ICWC) have been performed in limiter (TEXTOR) and divertor (ASDEX Upgrade, AUG) tokamaks in the presence of toroidal (0.2–2.35 T) and vertical (0–0.04 T) magnetic fields using the conventional ICRF antennas without modifications in hardware. The ICWC effect on both machines was studied by analyzing the removal rate of marker gases which have been loaded to the walls by glow discharge beforehand. Several factors were identified which could have a crucial impact on the conditioning efficiency: (i) RF power coupled to the plasmas; (ii) RF power absorption scheme; (iii) superimposing an additional vertical magnetic field on the toroidal field ( B V B T ). All the observed effects are analyzed in terms of RF plasma wave excitation/absorption and compared with the predictions from 1-D RF and 0-D transport codes. ICWC scenarios for ITER are proposed and analyzed.

A fast Fourier transform (FFT) based wide range millimeter wave diagnostics for spectral characterization of scattered millimeter waves in plasmas has been successfully brought into operation. The scattered millimeter waves are heterodyne downconverted and directly digitized using a fast analog-digital converter and a compact peripheral component interconnect computer. Frequency spectra are obtained by FFT in the time domain of the intermediate frequency signal. The scattered millimeter waves are generated during high power electron cyclotron resonance heating experiments on the TEXTOR tokamak and demonstrate the performance of the diagnostics and, in particular, the usability of direct digitizing and Fourier transformation of millimeter wave signals. The diagnostics is able to acquire 4 GHz wide spectra of signals in the range of 136 -140 GHz. The rate of spectra is tunable and has been tested between 200 000 spectra/s with a frequency resolution of 100 MHz and 120 spectra/s with a frequency resolution of 25 kHz. The respective dynamic ranges are 52 and 88 dB. Major benefits of the new diagnostics are a tunable time and frequency resolution due to postdetection, near-real time processing of the acquired data. This diagnostics has a wider application in astrophysics, earth observation, plasma physics, and molecular spectroscopy for the detection and analysis of millimeter wave radiation, providing high-resolution spectra at high temporal resolution and large dynamic range. (C) 2009 American Institute of Physics. [doi:10.1063/1.3244091]

The Dynamic Ergodic Divertor on TEXTOR imposes three-dimensional stochastic magnetic field lines in the plasma boundary. The resulting heat deposition patterns on the divertor surface depend strongly on a magnetic topology, including radial penetration of magnetic field lines within the first few toroidal revolutions. The influence of the collisionality in the plasma edge is of importance for the resulting deposition patterns. The heat flux formed by the electrons streaming along the field lines with shallow penetration is only seen at low collisionalities, while these field lines reach areas of low electron temperature. The other class of field lines penetrates relatively deep within the first few toroidal turns reaching the plasma region with much higher electron temperature, which makes parallel transport much more effective, even at the higher level of collisionality.

Inter-machine studies of wall conditioning with the ICRF discharges have been performed in the limiter (TEXTOR) and divertor (ASDEX Upgrade (AUG)) tokamaks in the presence of a toroidal magnetic field (⩾2 T) using the conventional ICRF antennas without modifications in hardware. The vessel oxidation treatment by pulsed ICRF discharges in (He + O 2 )-mixture (TEXTOR) is analyzed in terms of ratios of the RF pulse length to the O 2 -puff duration. A successive set of deuterium and helium ICRF discharges was developed for post-oxidation wall cleaning and analyzed in the light of TEXTOR recovery to the normal plasma operation. A new scenario of ICRF wall conditioning in (He + H 2 )-mixture at two frequencies was applied in AUG and compared with the standard glow discharge in terms of outgassing efficiency. Modeling of the absorbed RF power was done to clear up a role of the H 2 concentration in the homogeneity of ICRF plasmas and the generation of high-energy ions.

Penetration processes of rotating helical magnetic perturbation field into tokamak plasmas have been investigated by the dynamic ergodic divertor (DED) in TEXTOR. Experimental observations of the field penetration and field amplification are performed and the data are interpreted by theoretical analyses based on a linearized two-fluid plasma model. It is observed that the growth of the forced magnetic reconnection by the rotating DED-field is accompanied by a change in the plasma fluid rotation. The theoretical model is also applied to the DED experiment in the small tokamak device HYBTOK-II. It is confirmed that the theoretical analyses can explain the observed radial profiles of the DED-field in the plasma by inserting small magnetic pick-up coils in HYBTOK-II.

Experiments have been carried out in the TEXTOR, ASDEX Upgrade (AUG) and Alcator C-Mod (C-Mod) tokamaks to study melt-layer motion, macroscopic W-erosion from the melt as well as the changes of material properties such as grain-size and voids. In addition the effect of multiple exposures is studied to judge the potential amelioration of inflicted melt damage. The parallel heat flux at the radial position of the PFCs in the plasma ranges from around q∥ ∼ 45 MW/m2 at TEXTOR up to q∥ ∼ 500 MW/m2 at C-Mod which covers scenarios close to ITER parameters, allowing samples to be exposed and molten even at shallow divertor angles. Melt-layer motion perpendicular to the magnetic field is observed consistent with a Lorentz-force originating from thermoelectric emission of the hot sample. While melting in the limiter geometry at TEXTOR is rather quiescent causing no severe impact on plasma operation, exposure in the divertors of AUG and C-Mod shows significant impact on operation, leading to subsequent disruptions. The power-handling capabilities are severely degraded by forming exposed hill structures and changing the material structure by re-solidifying and re-crystallizing the original material. Melting of W seems highly unfavorable and needs to be avoided especially in light of uncontrolled transients and misaligned PFCs.

Dust will have severe impact on ITER performance since the accumulation of tritium in dust represents a safety issue, a possible reaction of dust with air and steam imposes an explosion hazard and the penetration of dust in core plasmas may degrade plasma performance by increasing radiative losses. Investigations were performed in TEXTOR where known amounts of pre-characterized carbon, diamond and tungsten dust were mobilized into plasmas using special dust holders. Mobilization of dust changed a balance between plasma–surface interactions processes, significantly increasing net deposition. Immediately after launch dust was dominating both core and edge plasma parameters. Remarkably, in about 100 ms after the launch, the effect of dust on edge and core plasma parameters was vanished: no increase of carbon and tungsten concentrations in the core plasmas was detected suggesting a prompt transport of dust to the nearby plasma-facing components without further residence in the plasma.

A new physical mechanism of formation of runaway electron (RE) beams during plasma disruptions in tokamaks is proposed. The plasma disruption is caused by a strong stochastic magnetic field formed due to nonlinearly excited low-mode number magnetohydrodynamic (MHD) modes. It is conjectured that the runaway electron beam is formed in the central plasma region confined inside the intact magnetic surface located between $q=1$ and the closest low--order rational magnetic surfaces [$q=5/4$ or $q=4/3$, \dots]. It results in that runaway electron beam current has a helical nature with a predominant $m/n=1/1$ component. The thermal quench and current quench times are estimated using the collisional models for electron diffusion and ambipolar particle transport in a stochastic magnetic field, respectively. Possible mechanisms for the decay of the runaway electron current owing to an outward drift electron orbits and resonance interaction of high--energy electrons with the $m/n=1/1$ MHD mode are discussed., Comment: 5 pages, 4 figures, one table. arXiv admin note: text overlap with arXiv:1501.01404

The dynamic ergodic divertor (DED) on the TEXTOR tokamak allows for the reproducible destabilization of the m/n = 2/1 tearing mode which is phase locked to the external static or rotating perturbation field. In combination with its flexible heating systems (co- and counter-neutral beam injection, ion cyclotron resonance heating, electron cyclotron resonance heating (ECRH) with steerable launcher) dedicated experiments to study the mode onset, properties of large islands and mode stabilization can be performed. The dependence of the mode excitation threshold (field penetration) on the plasma rotation shows a resonance character, with minimum threshold when the external perturbation frequency matches the MHD frequency of the 2/1 mode. Mode stabilization by ECRH heating shows that for the TEXTOR plasma heating is more effective than the current drive in O-point. Extrapolation to ITER yields a significant contribution to the mode suppression originating from the temperature increase within the island. Alfven-like modes, which have been previously identified in the vicinity of large islands on FTU (Buratti et al 2005 Nuclear Fusion 45 1446), are found to be created already before island formation above a certain threshold of the externally applied perturbation field.

The error-field penetration process of the DED (Dynamic Ergodic Divertor) into tokamak plasmas has been investigated in terms of numerical simulations based on a linearized reduced set of two-fluid equations (three-field, four-field model) in cylindrical coordinates, in which effects of electron diamagnetic drift are taken into account. Localized radial structures of the perturbations were identified in the vicinity of the rational surface when increasing the plasma viscosity and the density diffusivity. Resonant characteristics of the dependencies of the induced force by the DED on the driving frequency of the DED were also found. It is shown that the force transfers from the DED to the tokamak plasmas so as to bring electron fluid to the rest frame of the DED-field.

(CH4)-C-13 was injected through graphite and tungsten spherical limiters in reproducible TEXTOR discharges. These materials were chosen, as they represent the actual compromise for the plasma facing components in the ITER divertor. C-13 was used in order to distinguish injected and intrinsic carbon in the layer deposited on the limiter surface. Shot-by-shot video recordings show a continuous growth of the deposit near the injection hole. A pronounced difference in the C-13 deposition pattern on the graphite and tungsten limiters was observed. Post-mortem surface analysis showed that the ratios of the locally deposited to the injected amount of carbon are 4% for graphite and 0.3% for tungsten. The margins of the carbon layer deposited on tungsten are significantly steeper in comparison with the graphite limiter case. The large difference in the C-13 deposition efficiency can be explained by direct reflection of carbon from tungsten and the enhanced sputtering of carbon on the tungsten substrate. Nucleation is suggested to be an important mechanism for carbon deposition on tungsten. Monte Carlo impurity transport calculations by the ERO code reproduce reasonably the experimental results for the graphite limiter.

The error-field penetration process of the dynamic ergodic divertor (DED) on the TEXTOR tokamak has been investigated analytically in terms of a single fluid MHD model with a finite plasma resistivity and viscosity in a cylindrical geometry. The linear model produces a localization of the induced current at the resonance surface and predicts a vortex structure of the velocity field near the resonance layer. Moreover, effects of the Alfven resonance for the error-field penetration are identified by two peaks in the radial profiles of the perturbed toroidal current and the perturbed magnetic flux when the relative rotation velocity between the DED and the rotating tokamak plasma is set to large. Fine structures of the vorticity induced by the DED in the vicinity of the rational surface disappear by introducing a finite plasma perpendicular viscosity. In addition, it is shown that the two peaks of the perturbed toroidal current overlap by an anomalous plasma perpendicular viscosity. Likewise, a bifurcation of the penetration process from the suppressed to the excited state is obtained by a quasi-linear approach taking into account modifications of the radial profiles of the equilibrium current and the plasma rotation due to the DED. A comparison with real experimental results of the DED on the TEXTOR tokamak is shown.

The formation of magnetic islands within plasmas confined magnetically within the tori has significant influence upon their confinement and stability. To obtain an experimental insight into the formation and dynamics of such island structures we employed a fast framing camera viewing the plasma tangentially in the toroidal direction. The toroidal viewing direction gives the advantage in that the islands are viewed almost tangentially and this greatly facilitates the reconstruction of the local data from the line integrated ones. We discuss an effective method to do inversion. To study the fluctuations seen in the video images we perform a singular value decomposition, and then we use a truncated least square method to infer their pictures in space.

Intermittent convective transport has been investigated in the edge and the scrape-off layer (SOL) of TEXTOR using Langmuir probe signals. The probability distribution function (PDF) of the density fluctuations and the turbulence-induced flux are all positively skewed, while a Gaussian shape is recorded for the negative fluctuations. The deviation of the signals from Gaussian statistics clearly increases from the plasma edge to the SOL. Conditional averaging reveals that in the SOL region the waveform of intermittent structures is asymmetric in time and the burst events move radially outwards with Eθ × BT/B2 velocities of ~ 450 m s−1. It is found that the large burst fluctuations (≥2.5 × rms) account for nearly 40% of the total transport in the SOL. Statistics of the waiting-time between successive bursts indicate that the PDF of the time interval follows a Poisson-distribution for small-duration events (selected by size ≤2.5 × rms) and changes into a power-law form for larger ones. Moreover, the intermittency density fluctuation data clearly show self-similar characters and long-range time correlations through the presence of (1) sandpile-like frequency spectra and existence of the f−1 region; (2) a long tail in the autocorrelation function and (3) Hurst exponents H > 0.5 from R/S analysis, suggesting a possible role of avalanche-like transport in the turbulence intermittency.

To shed some light on the development of the fast m/n = 1/1 precursor to the sawtooth crash and its influence on plasma transport properties in the vicinity of the q = 1 surface, series of dedicated experiments have been conducted on the Tore Supra and TEXTOR tokamaks. It has been concluded that, before a crash, the hot core gets displaced with respect to the magnetic axis, drifts outwards by as much as 8–10 cm and may change its shape. Observation of the magnetic reconnection process has been made by means of electron cyclotron emission diagnostics. The heat pulse is seen far outside the inversion radius. The colder plasma develops a magnetic island on the former magnetic axis, after the hot core expulsion. Different kinds of behaviour of the m = 1 precursor before the crash, with respect to the displacement of the hot core and the duration of the oscillating phase, have been observed. An ideal kink model alone cannot be used for explanation; therefore, resistive effects play an important role in the mode development. Possible mechanisms that lead an m = 1 mode to such behaviour, and their links to the change in the central q-profile, are discussed. Results have been discussed in the light of various theoretical models of the sawtooth.

Rotating resonant magnetic perturbation (RMP) fields impose a characteristic modulation to the edge electron density n e(r, t) and temperature T e(r, t) fields, which depends on the relative rotation f rel between external RMP field and plasma fluid. The n e(r, t) and T e(r, t) fields measured in the edge (r/a = 0.9–1.05) of TEXTOR L-mode plasmas are in close correlation with the local magnetic vacuum topology for low relative rotation f rel = −0.2 kHz. In comparison with the 3D neutral and plasma transport code EMC3-Eirene, this provides substantial experimental evidence that for low relative rotation level and high resonant field amplitudes (normalized radial field strength ), a stochastic edge with a remnant island chain dominated by diffusive transport exists. Radially outside a helical scrape-off layer, the so-called laminar zone embedded into a stochastic domain is found to exist. In contrast for high relative rotation of f rel = 1.8 kHz, the measured modulation of n e is shifted by π/2 toroidally with respect to the modelled vacuum topology. A pronounced flattening in T e(r) and a reduction in n e(r) is measured at the resonant flux surface and represents a clear signature for a magnetic island, which is phase shifted with respect to the vacuum island position. A correlated shift of the laminar zone radially outwards at the very plasma edge is observed suggesting that the actual near-field structure at the perturbation source is determined by the plasma response as well.

The 1.1–1.5 mm wide gaps between tiles of the main toroidal belt limiter in TEXTOR were utilized to study the long-term impurity deposition and fuel retention in gaps. The tiles were exposed during a full tokamak campaign of 9365 s of plasma to various discharge conditions and wall conditioning, accumulating of up to 30 μm thick layers at the gap entrance. It was found that (i) gaps trap impurities twice as efficient as the top surface, (ii) the deposition in the toroidal gaps is twice as high as in the poloidal, (iii) carbon deposition decays with a fall-off length of about 0.7 mm towards the gap bottom, (iv) deposition on the bottom is significantly higher than on the adjacent side walls of gaps, and (v) the amount of deuterium scales with the amount of carbon with D/C varying from 3% to 30% depending on the surface temperature.

Development of methods to characterise the first wall in ITER and future fusion devices without removal of wall tiles is important to support safety assessments for tritium retention and dust production and to understand plasma wall processes in general. Laser based techniques are presently under investigation to provide these requirements, among which Laser Induced Desorption Spectroscopy (LIDS) is proposed to measure the deuterium and tritium load of the plasma facing surfaces by thermal desorption and spectroscopic detection of the desorbed fuel in the edge of the fusion plasma. The method relies on its capability to desorb the hydrogen isotopes in a laser heated spot. The application of LID on bulk tungsten targets exposed to a wide range of deuterium fluxes, fluences and impact energies under different surface temperatures is investigated in this paper. The results are compared with Thermal Desorption Spectrometry (TDS), Nuclear Reaction Analysis (NRA) and a diffusion model.

At TEXTOR, an O-mode heterodyne reflectometer system is installed and operated for the measurement of plasma density fluctuations and turbulence investigations. With two antenna arrays in the equatorial and top positions having two and three horn antennae, respectively, poloidal correlations are investigated under different plasma scenarios. From the amplitude, cross-phase and coherency spectrum, differences in the ohmic and auxiliary heated discharges are investigated. Furthermore the dynamic behaviour of the turbulence is studied in the SOC-IOC transition and in the precursor phase of a disruption. For the latter an increased integrated power spectral density was observed at the X-point of the mode compared with the O-point. Stationary m = 2 mode activity is observed for the first time at TEXTOR by reflectometry. The fluctuation level is calculated for different conditions and rises significantly increasing heating power which is consistent with the L-mode confinement degradation. Correlation measurements yield the measured phase delays which are used to calculate the poloidal phase velocity perpendicular to the magnetic field. In ohmic plasmas the turbulence rotates like a 'rigid body' with constant angular velocity inside the q = 2 surface. The rigid body rotation is broken up during tangential neutral beam injection. From the deduced poloidal wavenumber of the turbulence, most likely ion temperature gradient modes are the driving mechanism of the turbulence.

It has been observed in the T-10 tokamak that immediately after off-axis electron cyclotron resonance heating (ECRH) switch-off, the core electron temperature stays constant for some time, which can be as long as several tens of milliseconds, i.e. several energy confinement times (tau(E)), before it starts to decrease. Whether or not the effect is observed depends critically on the local magnetic shear in the vicinity of the q = 1 rational surface, which should be close to zero. It is hypothesized that a small shear can induce the formation of an internal transport barrier. Measurements of density fluctuations in the transport barrier with a correlation reflectometer show immediately after the ECRH switch-off a clear reduction in the fluctuation level, corroborating the above results. The delayed temperature decrease has also been observed in similar discharges in the TEXTOR tokamak; however, the delay is restricted to similar to1 x tau(E).

Wall conditioning techniques applicable in the presence of permanent toroidal magnetic field will be required for the operation of ITER, in particular for recovery from disruptions, vent and air leak, isotopic ratio control, recycling control and mitigation of the tritium inventory build-up. Ion Cyclotron Wall Conditioning (ICWC) is one of the most promising options and has been the subject of considerable recent study on current tokamaks. This paper reports on the findings of such studies performed on European tokamaks, covering a range of plasma-facing materials: TORE SUPRA, TEXTOR, ASDEX Upgrade and JET. (C) 2010 Elsevier B.V. All rights reserved.

The behaviour of tungsten (W) plasma-facing components (PFCs) has been investigated in the plasma edge of the TEXTOR tokamak to study melt-layer ejection, macroscopic tungsten erosion from the melt layer as well as the changes of material properties such as grain-size and abundance of voids or bubbles. The parallel heat flux at the radial position of the exposed tungsten tile in the plasma ranges around q ‖ ∼ 45 MW m−2 causing samples to be exposed at an impact angle of 35° to 20–30 MW m−2. Locally the temperature reached up to 6000 K, high levels of evaporation and boiling are causing significant erosion in the form of continuous fine spray or droplet ejection. The amount of fine-spray tungsten emission depends strongly on the material properties: in the case of the tungsten–tantalum alloy the effect of spraying and droplet emission is significantly higher at even low temperatures when compared with regular tungsten or even ultra-high purity tungsten which shows almost no spraying at all. Differences in the material composition, grain structure and size may be related to the different evolution of macroscopic erosion. In addition the re-solidified material is studied and strong differences in terms of re-crystallized grain size and evolution of the grain structure and grain orientation are observed. The build up of large voids has been observed.

The electrostatic potential and density fluctuations have been measured at the edge of the TEXTOR tokamak by two toroidally distant Langmuir probe arrays. The geodesic acoustic mode (GAM) zonal flows (ZFs) are observed in potential fluctuations with a toroidal and poloidal symmetric structure. The GAM frequency, fGAM, changes monotonically with the local temperature and is close to the frequency-dispersion predicted by theories. Bispectral analysis shows clear nonlinear coupling between the GAM and broadband ambient turbulence. The GAM packet has a narrow radial extent with kr 0.5–0.7 cm−1 and exhibits explicitly a radially outward propagation. Furthermore, the radial correlation structure of GAMs and their radial propagation have been investigated in a wide range of parameters by varying plasma density and edge safety factor (5.0 ≤ q(a) ≤ 5.9). It is found that the magnitude of the GAM correlations reduces remarkably with the increase in the plasma density approaching the density limit, while the radial wavelength of GAMs only decreases slightly in higher density and larger q(a) discharges. With increasing plasma density, the radial propagating phase speed of GAMs is strongly reduced along with the drop in the local temperature. The results provide new evidence on the propagation properties of GAM ZFs.

Fuelling efficiency is an important parameter in designing a massive gas injection system for suppression of runaway electrons in ITER. In this work Z-dependence of fuelling efficiency is measured for TEXTOR. The dependence covers the following gases: He, Ne, Ar, Kr, Xe and a 10% Ar–D2 mixture. It is shown that the fuelling efficiency significantly decreases with the gas mass, from above 0.5 for He to below 0.03 for Xe.To explain the variation of the efficiency with the gas mass and pressure a simple model of gas flow from the valve to the plasma edge is developed. The flow model is validated using available laboratory flow measurements of a TEXTOR-like injection system. An unsteady gas flow and a premature plasma disruption are shown to explain the mass dependence of the efficiency.

The parallel and radial transport properties of the plasma edge of TEXTOR are studied using radial electron density and temperature profiles as well as ion temperatures and poloidal velocities in the scrape-off layer (SOL). These quantities are measured by thermal helium beams at the low (LFS) and high field side (HFS) by emission and beam driven charge exchange recombination spectroscopy. We investigate the influence of the safety factor and of a magnetic field reversal on these edge parameters. Especially the field reversal leads to clear effects: a decrease of the density at the LFS, a significant change in the poloidal density distribution, which is identified by comparing LFS and HFS densities, and an increase in the density e-folding length at both poloidal positions. The poloidal ion rotation changes sign in reversed field configuration and thus gives a hint on the role of poloidal drifts in the SOL. For further analysis, we present a simple fluid model including poloidal drift velocities and local ionization sources. With this model we can show that the poloidal E x B drift clearly influences the poloidal density distribution. However, although the model results show the same tendencies as the experimental findings, the impact of the field reversal on the density asymmetry is not that pronounced as in the experiment. The dependence of the density e-folding length on the poloidal and radial drifts as well as on the source distribution is discussed within the frame of analytical estimates.

In high density Ohmically heated discharges in the tokamak TEXTOR a transition from the saturated Ohmic confinement (SOC) to the improved Ohmic confinement (IOC) was observed triggered by a sudden reduction of the external gas flow. The SOC-IOC transition was investigated regarding the influence of the toroidal ITG instability driven by the ion temperature gradient (ITG). The ion temperature profiles were measured with high radial resolution by means of charge-exchange recombination spectroscopy (CXRS) with a high-energetic diagnostic hydrogen beam recently installed at TEXTOR. On the basis of the measured ion temperature distributions the eta(i) parameter (ratio of the density and ion temperature decay lengths) and the growth rate of the toroidal ITG instability were calculated. After the SOC-IOC transition eta(i) drops and lies in a noticeably smaller radial region over the threshold for the toroidal ITG. In consequence of it, the IOC regime is characterized by a clear reduction of the ITG growth rate gamma(ITG) which was calculated including finite Larmor radius effects. The steepening of the plasma density profile after the decrease of the external gas flow is the main reason for the reduction of the ITG growth rate and the subsequent confinement transition to the IOC regime.

The scaling of plasma turbulence, turbulent particle flux and cross-phase with shear in the edge of the TEXTOR-94 tokamak is obtained from measurements from fast scanning probes and compared with various existing analytical theories. It is found that the scaling can be expressed as a second order polynomial and that the cross-phase plays a key role in the suppression of particle flux. The variable rate of shear, kept below the value required to produce a low to high particle confinement transition, was obtained by changing, on a shot to shot basis, the voltage applied to an electrode introduced 4 cm into the plasma.

ITER and future superconducting fusion machines need efficient wall conditioning techniques for routine operation in between shots in the presence of permanent high magnetic field for wall cleaning, surface isotope exchange and to control the in-vessel long term tritium retention. Ion Cyclotron Wall Conditioning (ICWC) based on the ICRF discharge is fully compatible and needs the presence of the magnetic field. The present paper focuses on the principal aspects of the ICWC discharge performance in large-size fusion machines: (i) neutral gas RF breakdown with conventional ICRF heating antennas, (ii) antenna coupling with low density (similar to 10(17) m(-3)) RF plasmas and (iii) ICWC scenarios with improved RF plasma homogeneity in the radial and poloidal directions. All these factors were identified as crucial to achieve an enhanced conditioning effect (e.g. removal rates of selected "marker" masses). All the observed effects are analyzed in terms of RF plasma wave excitation/absorption and compared with the predictions from I-D RF full wave and 0-D RF plasma codes. Numerical modeling and empirical extrapolation from the existing machines give good evidence for the feasibility of using ICWC in ITER with the main ICRF antenna.

The ratio between the heat diffusion coefficients parallel and perpendicular to the magnetic field lines, chi(parallel to)/chi(perpendicular to), influences the flattening of the temperature profile inside magnetic islands and the driving term of neoclassical tearing modes (Fitzpatrick 1995 Phys. Plasmas 2 825). The value of this anisotropy is, however, not easily accessible experimentally. This paper presents a method to determine it from a systematic comparison of temperature measurements at magnetic islands with numerical heat diffusion simulations. The application of the method is demonstrated for a 2/1 magnetic island in the TEXTOR tokamak, where a heat diffusion anisotropy of 108 is observed. This is lower by a factor of 40 than predicted by Spitzer and Harm (Spitzer and Harm 1953 Phys. Rev. 89 997) and a strong indication that the heat flux limit determines the flattening of the electron temperature across magnetic islands.

Neoclassical tearing modes (NTMs) deteriorate high-pressure tokamak plasma confinement and can be suppressed by electron cyclotron current drive (ECCD). In order to obtain efficient suppression, the ECCD power needs to be deposited at the center of an NTM magnetic island. To enhance efficiency, this power also needs to be synchronized in phase with the rotation of the island. The problem is that of real-time detection and precise localization of the island(s) in order to provide the feedback signal required to control the ECCD power deposition area with an accuracy of 1 to 2 cm. Existing schemes based on mode location, equilibrium reconstruction, and plasma profile measurements are limited in positional and temporal accuracy and moreover will become very complex when applied to ITER. To overcome these limitations, it is proposed to provide the feedback signal from electron cyclotron emission (ECE) measurements taken along the identical line of sight as traced by the incident ECCD millimeter-wave beam but in reverse direction. Experiments on TEXTOR have demonstrated a proof of principle. These measurements motivate the further development and the implementation of such an ECCD-aligned ECE system for NTM control in larger fusion machines. Possible implementation of such a system on ASDEX-Upgrade, based on waveguides equipped with a fast directional switch, is presented in this paper. Possible further development for ITER is also discussed.