Search

In order to reduce the risks for ITER Plasma Facing Components (PFCs), it is proposed to equip Tore Supra with a full tungsten divertor, benefitting from the unique long pulse capabilities, the high installed RF power and the long experience with actively cooled high heat flux components of the Tore Supra platform. The transformation from the current circular limiter geometry to the required X-point configuration will be achieved by installing a set of copper poloidal coils inside the vacuum vessel. The new configuration will allow for H-mode access, providing relevant plasma conditions for PFC technology validation. Furthermore, attractive steady-state regimes are expected to be achievable. The lower divertor target design will be closely based on that currently envisaged for ITER (W monoblocks), while the upper divertor region will be used to qualify the main first wall heat sink technology adopted for the ITER blanket modules (CuCrZr copper/stainless steel) with a tungsten coating (in place of the Be tiles which ITER will use). Extended plasma exposure will provide access to ITER critical issues such as PFC lifetime (melting, cracking, etc.), tokamak operation on damaged metallic surfaces, real time heat flux control through PFC monitoring, fuel retention and dust production.

Fusion experiments with wall cladding of different chemical elements face the problem of material mixing by the plasma–wall interaction. Understanding of experiments that investigate this long term material re-distribution require an iterative approach that takes the local erosion, the global transport of material by the scrape-off layer (SOL) plasma and the local re-deposition into account. We present a model that combines parameterized DIVIMP calculations and a simplified sputtering model to simulate the time evolution of first wall material migration in a fusion experiment. The simulations are benchmarked by Be evaporation experiments performed previously at JET. The time evolution of the spectroscopically measured Be erosion fluxes for 800 s total plasma time at the main wall and inner divertor can be reproduced by the model using an extended calculation grid for DIVIMP and an increased cross-field diffusivity for impurity ions in the SOL. Discrepancies were found in the outer divertor region.

Micro and nanocrystalline diamond layers have been deposited on molybdenum substrates by hot-filament CVD, and tested in the Mega Amp Spherical Tokamak as a protective coating of fusion plasma-facing materials. The modification of surface properties induced by high density plasma was investigated by SEM, X-ray photoelectron spectroscopy and Raman spectroscopy. Although some modifications of the coating, amorphization and some traces of arcing, diamond proved to be a viable protection as most of the samples were still coated after the plasma exposure. Fuel retention measurements, evaluated by nuclear reaction analysis, showed that a small amount of deuterium was trapped in the surface of the coatings. The chemical erosion of the layers was 50% lower than graphite, as evaluated by dedicated experiments in Pilot-PSI, a linear plasma simulator device.

A new version of the B2SOLPS5.0 transport code, which is free from numerical problems in the barrier region, has been used to simulate H-mode shots from ASDEX-Upgrade and MAST. The radial electric field inside the edge transport barrier and in the pedestal region is close to the neoclassical prediction. The shear of poloidal E → × B → drift at the inner side of the barrier is close to the value before the transition, while inside the barrier it is significantly larger. It is demonstrated that to match the experimental density and temperature radial profiles the drop in the diffusion coefficient within the barrier should be significantly larger than the drop in the electron heat conductivity.

Spectroscopy of high-n Balmer line transitions provides a means of measuring ne and Te in recombining plasmas [J. L. Terry et al., Phys. Plasmas 5, 1579 (1998)]. The relative intensities of Rydberg series lines near the ionization limit are a sensitive diagnostic of Te, for Te < 1.5 eV. Stark broadening of these same lines provides a measure of local ne and, with less accuracy, of Te. The accuracy of different theoretical models for Stark broadening [H.R. Griem, Spectral Line Broadening by Plasmas (Academic, New York, 1974); E. Oks, Stark Broadening of Hydrogen and Hydrogenlike Spectral Lines in Plasmas: The Physical Insight (Alpha Science International, Oxford, UK, 2006)] is evaluated by comparing values of ne and Te measured on DIII-D by divertor Thomson scattering (DTS) with those deduced from spectral profile analysis of Balmer series deuterium lines. In particular, the detailed dependence of line width on principal quantum number provides a sensitive metric for distinguishing which model best accords with experiment.

Recent research in scrape-off layer (SOL) and divertor physics is reviewed; new and existing data from a variety of experiments have been used to make cross-experiment comparisons with implications for further research and ITER. Studies of the region near the separatrix have addressed the relationship of profiles to turbulence as well as the scaling of the parallel power flow. Enhanced low-field side radial transport is implicated as driving parallel flows to the inboard side. The medium-n nature of edge localized modes (ELMs) has been elucidated and new measurements have determined that they carry ~10?20% of the ELM energy to the far SOL with implications for ITER limiters and the upper divertor. The predicted divertor power loads for ITER disruptions are reduced while those to main chamber plasma facing components (PFCs) increase. Disruption mitigation through massive gas puffing is successful at reducing PFC heat loads. New estimates of ITER tritium retention have shown tile sides to play a significant role; tritium cleanup may be necessary every few days to weeks. ITER's use of mixed materials gives rise to a reduction of surface melting temperatures and chemical sputtering. Advances in modelling of the ITER divertor and flows have enhanced the capability to match experimental data and predict ITER performance.

The high neutral densities and short neutral mean-free-paths in theAlcator C-Mod divertor have provided a unique testing ground for our understanding of the role of neutrals in a tokamak. The high neutral pressures found in the C-Mod divertor can be reproduced in models only by including such processes as ion-neutral and neutral-neutral collisions and neutral viscosity, as well as taking into account the plasma in the private flux region. After detachment, when the divertor plate ion flux has dropped by more than an order of magnitude, the divertor pressure still remains high. High neutral collisionality and the plasma in the private flux region again help keep neutrals in the divertor along with the large source of neutrals due to recombination. Likewise, diffusive neutrals are the explanation for the divertor neutral pressure's insensitivity to strike point position. Closure of neutral leakage pathways did not lead to a decrease in neutral pressures in the region outside the divertor-the main chamber. This observation prompted further research, which showed that ion fluxes to main chamber surfaces rival those reaching the divertor plates; the main chamber pressure can be primarily determined by the level of ion transport perpendicular to the magnetic field. This finding has spawned a host of studies (active and passive) both at C-Mod and other tokamaks to understand how radial transport can be so large.

The Mega Ampère Spherical Tokamak (MAST) programme is strongly focused on addressing key physics issues in preparation for operation of ITER as well as providing solutions for DEMO design choices. In this regard, MAST has provided key results in understanding and optimizing H-mode confinement, operating with smaller edge localized modes (ELMs), predicting and handling plasma exhaust and tailoring auxiliary current drive. In all cases, the high-resolution diagnostic capability on MAST is complemented by sophisticated numerical modelling to facilitate a deeper understanding. Mitigation of ELMs with resonant magnetic perturbations (RMPs) with toroidal mode number n RMP = 2, 3, 4, 6 has been demonstrated: at high and low collisionality; for the first ELM following the transition to high confinement operation; during the current ramp-up; and with rotating n RMP = 3 RMPs. n RMP = 4, 6 fields cause less rotation braking whilst the power to access H-mode is less with n RMP = 4 than n RMP = 3, 6. Refuelling with gas or pellets gives plasmas with mitigated ELMs and reduced peak heat flux at the same time as achieving good confinement. A synergy exists between pellet fuelling and RMPs, since mitigated ELMs remove fewer particles. Inter-ELM instabilities observed with Doppler backscattering are consistent with gyrokinetic simulations of micro-tearing modes in the pedestal. Meanwhile, ELM precursors have been strikingly observed with beam emission spectroscopy (BES) measurements. A scan in beta at the L–H transition shows that pedestal height scales strongly with core pressure. Gyro-Bohm normalized turbulent ion heat flux (as estimated from the BES data) is observed to decrease with increasing tilt of the turbulent eddies. Fast ion redistribution by energetic particle modes depends on density, and access to a quiescent domain with ‘classical’ fast ion transport is found above a critical density. Highly efficient electron Bernstein wave current drive (1 A W−1) has been achieved in solenoid-free start-up. A new proton detector has characterized escaping fusion products. Langmuir probes and a high-speed camera suggest filaments play a role in particle transport in the private flux region whilst coherence imaging has measured scrape-off layer (SOL) flows. BOUT++ simulations show that fluxes due to filaments are strongly dependent on resistivity and magnetic geometry of the SOL, with higher radial fluxes at higher resistivity. Finally, MAST Upgrade is due to begin operation in 2016 to support ITER preparation and importantly to operate with a Super-X divertor to test extended leg concepts for particle and power exhaust.

Research on the Alcator C-Mod tokamak has emphasized RF heating, self-generated flows, momentum transport, scrape-off layer (SOL) turbulence and transport and the physics of transport barrier transitions, stability and control. The machine operates with P-RF up to 6 MW corresponding to power densities on the antenna of 10 MW m(-2). Analysis of rotation profile evolution, produced in the absence of external drive, allows transport of angular momentum in the plasma core to be computed and compared between various operating regimes. Momentum is clearly seen diffusing and convecting from the plasma edge on time scales similar to the energy confinement time and much faster than neo-classical transport. SOL turbulence and transport have been studied with fast scanning electrostatic probes situated at several poloidal locations and with gas puff imaging. Strong poloidal asymmetries are found in profiles and fluctuations, confirming the essential ballooning character of the turbulence and transport. Plasma topology has a dominant effect on the magnitude and direction of both core rotation and SOL flows. The correlation of self-generated plasma flows and topology has led to a novel explanation for the dependence of the H-mode power threshold on the del B drift direction. Research into internal transport barriers has focused on control of the barrier strength and location. The foot of the barrier could be moved to larger minor radius by lowering q or B-T. The barriers, which are produced in C-Mod by off-axis RF heating, can be weakened by the application of on-axis power. Gyro-kinetic simulations suggest that the control mechanism is due to the temperature dependence of trapped electron modes which are destabilized by the large density gradients. A set of non-axisymmetric coils was installed allowing intrinsic error fields to be measured and compensated. These also enabled the determination of the mode locking threshold and, by comparison with data from other machines, provided the first direct measurement of size scaling for the threshold. The installation of a new inboard limiter resulted in the reduction of halo currents following disruptions. This effect can be understood in terms of the change in plasma contact with the altered geometry during vertical displacement of the plasma column. Unstable Alfven eigenmodes (AE) were observed in low-density, high-power ICRF heated plasmas. The damping rate of stable AEs was investigated with a pair of active MHD antennae.

We report on impurity transport modeling using the DIVIMP code which is able to qualitatively reproduce the poloidal variation of non-recycling impurity penetration factor (PF NR ) found in C-Mod experiments: a lower PF NR is computed at the inboard (3.6%) and divertor target locations (0.7%) than at the outboard (11%). By artificially increasing the modeled inner SOL plasma flow to correspond to measured values, a better quantitative agreement between modeled and measured PFs is achieved. We have also roughly reproduced the observed penetration factor for recycling impurities both in time dependence and magnitude. The model has shown that under attached conditions, the majority of recycling impurity ions flow into the confined plasma through the outboard side separatrix. For detached conditions the impurity influx across the separatrix is more concentrated near the divertor.

Use of carbon in tokamaks leads to a major tritium retention issue due to co-deposition. To investigate this process a low power L-mode experiment was performed on DIII-D in which 13 CH 4 was puffed into the main vessel through the toroidally-symmetric pumping plenum at the top of lower single-null discharges. Subsequently, the 13 C content of tiles taken from the vessel wall was measured. The interpretive OEDGE code was used to model the results. It was found that the 13 C deposition pattern is controlled by: (a) source strength of 13 C + , (b) Δ r s , radial location of the 13 C + source, (c) D ⊥ , (d) M ∥ , the scrape-off layer parallel Mach number. Best agreement was found for (a) ∼50% conversion efficiency 13 CH 4 → 13 C + , (b) Δ r s , ∼+3.5 cm (outboard of separatrix) near 13 CH 4 injection location, (c) D ⊥ ∼ 0.3 m 2 /s, (d) M ∥ ∼ 0.4 toward inside.

As part of a study of carbon–tritium co-deposition, we carried out an experiment on DIII-D involving a toroidally symmetric injection of 13CH4 at the top of a LSN discharge. A Monte Carlo code, DIVIMP-HC, which includes molecular breakup of hydrocarbons, was used to model the region near the puff. The interpretive analysis indicates a parallel flow in the SOL of M∥ ∼ 0.4 directed toward the inner divertor. The CH4 is ionized in the periphery of the SOL and so the particle confinement time, τc, is not high, only ∼5 ms, and about 4X lower than if the CH4 were ionized at the separatrix. For such a wall injection location, however, approximately 60–75% of the CH4 gets ionized to C+, C2+, etc., and is efficiently transported along the SOL to the inner divertor, trapping hydrogen by co-deposition there.

The ITER divertor will be more collisional than divertors in existing tokamaks and the neutral modeling tools used for predictive ITER modeling need to be validated for similar divertor conditions in order to increase confidence in the results. The EIRENE neutral code has been applied to the Alcator C-Mod divertor, where collisional effects such as Lyman series photon trapping and neutral viscosity have been observed. The plasma solution used as input to EIRENE agrees well with the available diagnostic data for the divertor and was generated using an empirical Onion-Skin Method interpretive code that requires a large amount of experimental data as input. The calculated divertor pressure is 11 ± 3 mTorr, a factor ∼2 lower than the measured value of 25 ± 3 mTorr. The neutral solution is sensitive to photon trapping, viscosity, elastic collisions between plasma ions and molecules, and the amount of leakage through the outer divertor substructure.

Recently a number of major, unanticipated effects have been reported in tokamak edge research raising the question of whether we understand the controlling physics of the edge. This report is on the first part – here focused on the outer divertor – of a systematic study of the simplest possible edge plasma – no ELMs, no detachment, etc. – for a set of 10 repeat, highly diagnosed, single-null, divertor discharges in DIII-D. For almost the entire, extensive data set so far evaluated, the matches of experiment and model are so close as to imply that the controlling processes at the outer divertor for these simple plasma conditions have probably been correctly identified and quantitatively characterized in the model. The principal anomaly flagged so far relates to measurements of Te near the target, potentially pointing to a deficiency in our understanding of sheath physics in the tokamak environment.

The EIRENE Monte-Carlo neutral code is used to calculate neutral pressures throughout the divertor region of the Alcator C-Mod tokamak. A semi-empirical onion-skin method (OSM) is used to calculate the hydrogenic background plasma for a moderate density discharge, 〈 n e 〉=1.5×10 20 m −3 , with a partially detached inner target and an attached outer target. The model parameters are adjusted until there is agreement with the Langmuir probe and spectroscopic data for the divertor region. The measured divertor pressure is 30 mTorr, and for C-Mod divertor dimensions this implies the transition flow regime between free molecular and viscous neutral transport. Therefore, a non-linear BGK-model Boltzmann collision term is included in EIRENE to allow for interactions between neutral particles. For the standard model parameters, the calculated divertor pressure is 7 mTorr. The model parameters are adjusted to reduce the discrepancy, and the plausibility of each variation is discussed.

OSM analysis provides, in principle a method for establishing the 2-D edge `fields' of n e , T e , T i , etc., which is the prerequisite for analyzing the physics processes occurring in the edge, including impurity behavior. In order to further test this method, an OSM analysis of an extensive edge database for an L-mode DIII-D discharge has been carried out, the first part of which is reported here. Consistency of the OSM results with Langmuir probe, Dα, and edge Thomson scattering measurements encourages further development of the method.

Simulations of carbon transport using the DIVIMP code [P.C. Stangeby, J.D. Elder, J. Nucl. Mater. 196–198 (1992) 258] are compared with 2D toroidal images of CII and CIII radiation near the external divertor plates in TdeV ohmic plasmas ( I p =170 kA, n e =3 × 10 19 m −3 , B T =1.4 T ). The main plasma parameters in the SOL and divertor are calculated by the onion skin model (OSM) [K. Shimizu et al., J. Nucl. Mater. 196–198 (1992) 476] included in DIVIMP, the neutrals being calculated by EIRENE [D. Reiter, Internal Report, KFA, Julich, 1947 (1984), 2599 (1992)] in an iterative loop. The results show that the carbon is mainly created by chemical sputtering, with a considerable fraction coming from the external oblique plate. By interpreting experimental CII and CIII distributions, it is found that carbon is affected by the biasing (−125 to +125 V) through a combination of at least three processes: the ion flux to the plates, the E × B drift velocity, and the cross field diffusion.

The tokamak boundary plasma is inherently 2D/3D, which impedes the detailed validation of transport models due to the limited spatial coverage of most diagnostics. A potential method for determining the 2D profiles of n(e) and T-e in the plasma volume is to resolve spatially the emission ratios of atomic helium, principally the lines at 667, 706 and 728 nm. Unfortunately, there are several challenges associated with this approach: the traditional use of line-of-sight spectrometer data; crowding by other impurity lines; low signal levels; reliance on accurate atomic physics; and He I meta-stables. Several of these issues are explored in the present study, which utilises tomographic reconstruction of filtered CCD camera images to measure He emission throughout the divertor. Plasma gradients are resolved and compared with results from the OSM-EIRENE code. The near-target n(e) and T-e values agree with Langmuir probe measurements to within a factor similar to 2.

Electrical currents in the scrape-off layer (SOL) of MAST are modelled using an interpretive Onion-Skin Model (OSM) constrained with experimental data from MAST diagnostics. The model was extended to include the effects of the magnetic mirror force, which has a strong influence on the particle and momentum balance in spherical tokamaks, such as MAST [1] . These modifications serve to more accurately model the parallel electric fields present in the MAST SOL, which can alter plasma dynamics via the E × B drift. Simulations show that the electrical current at the divertor targets is predominantly thermoelectric, whereas Pfirsch–Schluter currents have a greater contribution to the total current in the bulk of the SOL plasma.

The design, operation, and initial results of a novel tracer-gas injection system is described. This system has been developed to diagnose local impurity transport and screening properties as a function of depth into the scrape-off layer (SOL) of Alcator C-Mod. A fast-scanning Langmuir–Mach probe has been outfitted with a capillary gas-feed and an inertially activated valve. Impurity injection plumes lasting 5–10 ms can be formed at any depth into the SOL, up to the separatrix location. The local dispersion of selected charge state impurities in the SOL can be followed by camera imaging and Doppler spectroscopy. In principle, the local screening properties of the SOL can also be directly assessed by injecting gas at different depths into the SOL and observing the resultant impurity concentrations in the core.

A series of experiments has been carried out to determine the relative screening of the low recycling gaseous impurities, nitrogen and methane, in Alcator C-Mod tokamak [Phys. Plasmas 1, 1511 (1994)] discharges. The impurity density in the core plasma is directly proportional to the rate of impurity injection. Screening was found to vary weakly with plasma density but to be markedly dependent on the poloidal position of injection. Detachment of the plasma results in significantly less screening both in Ohmic and in high confinement (H-mode) discharges, particularly when the gas is injected into the divertor or at the inner vessel wall. Evidence of impurity transport from the scrape-off layer into the divertor using bolometry and visible spectroscopy is presented and modeling of the transport using a two-dimensional Monte Carlo code is described.

The behavior of argon and nitrogen, injected into single null diverted Ohmic discharges, has been compared. The screening of the divertor and scrape-off layer plasma is studied by comparing the number of impurity atoms in the core with the number injected by a calibrated gas puff. The effect of the poloidal position of injection on screening has been investigated. The core nitrogen density is dependent of an poloidal position of injection while the argon density is not. The spatial distribution of low charge states in the divertor has also been studied using a multichord visible spectrometer. Initial results of nitrogen modeling with the Monte Carlo code DIVIMP are presented.

A capillary injection system has been installed on Alcator C-Mod which allows the localized introduction of gaseous species at a variety of poloidal locations. An experimental program has been undertaken to observe impurity puffs with a CCD camera through appropriate optical filters. The comet-like shape of an ion line emission plume formed in the region of an injection graphically displays the direction of background deuterium flow as well as that of the poloidal impurity ion drift. Parallel and cross-field one-dimensional fluid models are used to characterize the plume shapes and to deduce the background plasma temperature, the parallel flow velocity and the poloidal drift velocity of the impurity ion. Model results are benchmarked against simulations of DIVIMP, a Monte Carlo code, and fast scanning probe measurements. The experiments present a novel diagnostic technique to elucidate local edge impurity transport.

Estimations of the ITER first wall (FW) life time, previously made by the 3D Monte- Carlo ERO code, depend strongly on the assumptions of the physical sputtering yield for beryllium (Be). It is of importance to validate the respective model and data at existing devices including the JET ITER- Like Wall (ILW) as most ITER- relevant experiment. Applying the same sputtering input data in ERO as before in the ITER- predictions, the ERO simulations for the Be light intensity (up- to- date atomic data is used) near the solid Be limiter in the JET- ILW case reveal a factor 2 overestimation of the assumed yield even if the low estimate assuming 50% D surface content is used. This result indicates the preference of this assumption for plasma- wetted areas. It points to a possible (after eliminating of uncertainties) necessity to correct (reduce) the respective estimates for the Be sputtering yield and accordingly, re- visit the ITER FW life time predictions.

The neutral beam heating and current drive system in ITER consists of 3 beam lines (2 present plus one future upgrade) with each beam line designed to deliver 40 A of accelerated deuterium beams at 1 MeV with a 25% duty cycle. The beam line is coupled to the vacuum vessel port of the tokamak through a series of front end components and a connecting duct. The edge of the beam line and the walls of the vacuum vessel up to the blanket aperture are lined with duct liners to protect them from heat loads from the direct and re-ionised beam interception during the transport of the neutral beam. The direct interception of the beam is due to the inherent divergence of the beam or its halo component. The re-ionised beam consists of ions born due to the interaction of the accelerated neutral beam with the back ground gas all along the beam line, after the neutraliser exit. The motion of these ions is also affected by the electric field of the residual ion dump (RID) and the magnetic field from the tokamak during its various phases of operation. A systematic study to assess the heat loads during the neutral beam transport on the different front end components, the various regions of the duct and the blanket modules is necessary to ascertain the proper thermo-mechanical design of these components. The beam transmission code "BTR" has been used for that purpose. Simulations have been carried out of the gas profile along the neutral beam line considering gas flux from the ion source, the neutraliser, the RID (due to the dumped ion beams) and the flow of the gas from the tokamak to the duct. The re-ionisation losses have been estimated to be 13.8 % for the region between the exit of the neutraliser and the blanket module edge. The magnetic fields for the various operating scenarios of the tokamak like the start of the burn (SOB), end of burn (EOB), X point formation (XPF), XPF + 20 s, EOB + disruption have been simulated for the 15 MA DT scenario. The beamlet divergence has been considered to range between 3 - 7 mrad for the main beam component and 30 mrad for the halo fraction which has been taken as 15% of the main beam. The simulations have been performed for the neutral beam axis vertical inclination of 49 mrad with an additional 10 mrad vertical tilt, which is required for off-axis current drive and avoidance of beam excited toroidal Alven eigenmodes in the ITER plasma. The results of these simulations will be presented and discussed.

Thick free-standing boron-doped diamonds were prepared by microwave plasma assisted chemical vapour deposition. Samples with a final thickness close to 5 mm and with lateral dimensions 25 × 25 mm were produced. The thermal shock resistance of the material was tested by exposure in the JUDITH electron gun to reproduce the conditions expected in a fusion reactor during plasma instabilities. Boron-doped diamond is shown to exhibit thermal shock resistance far better than current material of choice for a fusion reactor like tungsten, with no surface damage after exposure to 100 cycles at 2.5 GW m−2 for 5 ms.

Microcrystalline undoped and heavily boron-doped polycrystalline diamond layers have been deposited on various substrates by hot filament CVD and exposed to hydrogen plasma in a linear plasma reactor (Pilot-PSI, The Netherlands) that simulates the high flux and high density plasma conditions of tokamak divertors, as well as in the DIII-D tokamak (US). Pre- and post-exposure analysis by SEM and Raman spectroscopy characterised the surface appearance and the sp3 and sp2 components of the diamond films respectively. Surface roughness variation was measured by AFM. Hardness and Young's modulus were assessed by nanoindentation in order to characterise the effect of the plasma on the mechanical properties. HRTEM and EELS have been used to evaluate the nature of the modification induced at the diamond surface by the plasma exposure. The measurements have shown that, despite some surface amorphisation of the exposed layers, further long-term exposure studies are warranted as neither delamination, dramatic film failure nor entire erosion of the film were observed. © 2010 Elsevier B.V. All rights reserved.

First wall material erosion and migration after fresh Be evaporation in the JET tokamak were studied in a series of consecutive identical L-mode discharges. The evolution of Be and C wall and divertor sources towards steady state conditions after deposition of a finite amount of Be at the carbon first wall of JET by the beryllium evaporation procedure provides a data set for benchmarking impurity transport simulations. Furthermore the experiment serves as a reference case for comparison of Be erosion to that of the planned ITER-like wall experiment with mainly Be plasma facing components (PFCs) in the main chamber. The experimental results confirm the migration pattern obtained by campaign integrated accounting of impurity sources and sinks, which is characterised by the main chamber wall and outer divertor as erosion dominated zones and migration of eroded material predominantly to the inner divertor where the material is finally deposited at the target plates.

Nanocrystalline diamond layers were deposited by hot-filament CVD on several substrates including molybdenum tiles and foils, graphite blocks and silicon wafers. The structural characterisation of the deposited layers, achieved by SEM, AFM, Raman and X-ray photoelectron spectroscopies, showed the typical features of nanocrystalline diamond. A nanocrystalline diamond coated molybdenum tile was exposed to a fusion grade deuterium plasma in the MAST tokamak, with a plasma current of 700 kA, core electron temperature of about 0.6 keV, and total exposure time of about 1 s. The diamond coating showed good resistance to the extremely harsh conditions of the fusion plasma edge.

We have studied the chemical erosion of different carbon composites in Pilot-PSI at ITER-relevant hydrogen plasma fluxes (∼1024 m -2 s-1) and low electron temperatures (Te∼1 eV). Optical emission spectroscopy on the CH A-X band was used to characterize the chemical sputtering. Fine grain graphite (R 6650, SGL Carbon Group), ITER-reference carbon fiber composite material (SNECMA NB31 and NB41; Dunlop 3D), nano- and micro-crystalline diamond coatings on molybdenum and SiC (Silit® SKD Reaction-Bonded, Saint-Gobain Ceramics) were compared. The chemical sputtering was similar for the different composites under comparable plasma conditions, except for SiC, which produced a ten times lower rate. The CH emission was constant at electron temperatures Te>1 eV and ion fluxes ranging between 1023 and 1024 m- 2 s-1, but decreased at lower temperatures. This decrease is possibly due to changes in the excitation of CH and not due to a change in the chemical erosion rate.

A study of the evolution of the filaments observed during Type I ELMs on ASDEX Upgrade and MAST is presented. The filaments start off rotating toroidally/poloidally with velocities close to that of the pedestal. This velocity then decreases as the filaments propagate radially. On both devices the ion saturation current e-folding lengths of the filaments show a weak, if any, dependence on the size of the ELM (δWELM/Wped). On MAST the measured radial velocities of the filaments also show at most a weak dependence on δWELM/Wped.

Stark broadening analysis of high‐n Balmer series transitions (n=10→2) is benchmarked against electron density measurements from a Thomson scattering system in the divertor region of the DIII‐D tokamak. The time‐averaged Thomson data show ne = 6.0±0.5×1020 m−3 in the low Te region (∼1 eV) of a detached divertor plasma. Analysis of the measured Stark broadened Balmer series line widths using conventional theory gives ne = 6.7±1.3×1020 m−3. A more advanced analysis technique, which is referred to here as generalized theory and includes additional Stark broadening effects related to the dynamics of ion electric field [E. Oks, Stark Broadening of Hydrogen and Hydrogenlike Spectral Lines in Plasmas: The Physical Insight, Alpha Science International, Oxford, UK, 2006], was also applied and produces ne = 5.3 ± 0.5 × 1020 m−3. In addition, electron temperature information of Te ≈ 1 eV is also determined (conventional theory does not allow for a temperature dependence). The improved diagnostic results suggest that these additional mechanisms should be included when analyzing Stark broadening data from tokamak divertors.

A series of experiments on the effect of divertor baffling on the Alcator C-Mod tokamak provides stringent tests on models of neutral gas transport in and around the divertor region. One attractive feature of these experiments is that a trial description of the background plasma can be constructed from experimental measurements using a simple model, allowing the neutral gas transport to be studied with a stand-alone code. The neutral-ion and neutral-neutral elastic scattering processes recently added to the DEGAS 2 Monte Carlo neutral transport code permit the neutral gas flow rates between the divertor and main chamber to be simulated more realistically than before. Nonetheless, the simulated neutral pressures are too low and the deuterium Balmer-alpha emission profiles differ qualitatively from those measured, indicating an incomplete understanding of the physical processes involved in the experiment. Some potential explanations are examined and opportunities for future exploration a re highlighted. Improvements to atomic and surface physics data and models will play a role in the latter.

New diagnostic, modelling and plant capability on the Mega Ampère Spherical Tokamak (MAST) have delivered important results in key areas for ITER/DEMO and the upcoming MAST Upgrade, a step towards future ST devices on the path to fusion currently under procurement. Micro-stability analysis of the pedestal highlights the potential roles of micro-tearing modes and kinetic ballooning modes for the pedestal formation. Mitigation of edge localized modes (ELM) using resonant magnetic perturbation has been demonstrated for toroidal mode numbers n = 3, 4, 6 with an ELM frequency increase by up to a factor of 9, compatible with pellet fuelling. The peak heat flux of mitigated and natural ELMs follows the same linear trend with ELM energy loss and the first ELM-resolved Ti measurements in the divertor region are shown. Measurements of flow shear and turbulence dynamics during L-H transitions show filaments erupting from the plasma edge whilst the full flow shear is still present. Off-axis neutral beam injection helps to strongly reduce the redistribution of fast-ions due to fishbone modes when compared to on-axis injection. Low-k ion-scale turbulence has been measured in L-mode and compared to global gyro-kinetic simulations. A statistical analysis of principal turbulence time scales shows them to be of comparable magnitude and reasonably correlated with turbulence decorrelation time. Te inside the island of a neoclassical tearing mode allow the analysis of the island evolution without assuming specific models for the heat flux. Other results include the discrepancy of the current profile evolution during the current ramp-up with solutions of the poloidal field diffusion equation, studies of the anomalous Doppler resonance compressional Alfvén eigenmodes, disruption mitigation studies and modelling of the new divertor design for MAST Upgrade. The novel 3D electron Bernstein synthetic imaging shows promising first data sensitive to the edge current profile and flows. © 2013 IAEA, Vienna.

In this contribution, we report on experiments performed in MAST to investigate dust creation, transport and influence on plasma performance. The exceptional diagnostic access of MAST allows stereoscopic imaging of dust particles' motion in both the divertor and the main chamber, utilizing fast infrared cameras. This technique allows the 3D trajectory of the particles to be reconstructed. Infrared imaging of dust creation during disruptions revealed an isotropic release of dust particles from the surface with very high velocities (up to 350 m s−1). Stereoscopic imaging has been used to study, for the first time, the mobilization and transport in the divertor plasmas of carbon and tungsten particles with known size distributions, which were introduced into the vessel through a divertor probe. A correlation between the carbon particle size and acceleration by the plasma is observed. Tungsten particles are found to move with lower velocities and experience lower acceleration and are found to be more prone to vertical motion towards the core plasma. In the case of large particles this can lead to early disruptions. Modelling of the dust injection experiments has been conducted using the DTOKS code in an attempt to validate the transport equations employed in the simulation.

The impact of resonant magnetic perturbations (RMPs) on the structure of the edge transport barrier has been studied. A model for the density pump-out mechanism during the stochastization of the plasma edge is proposed. The observed phenomena are explained as a result of the impact of the ambipolar electric field, which is modified during RMP, on the particle fluxes in the pedestal region. It is demonstrated that the rise of the particle fluxes inside the transport barrier leads to the pump-out effect on density, while the pedestal temperature increases in spite of the big electron heat conductivity in the stochastic magnetic field. The latter is not sufficient to change significantly turbulent heat conductivity in the barrier region and only compensates the rise of the pedestal temperature caused by the density drop for constant heating power. The analytical approach is supported by results of simulations with the B2SOLPS5.2 2D transport code which uses a full description of particle sources and transport phenomena in the pedestal region. Simulations are performed for ASDEX-Upgrade and MAST configurations for various values of electron stochastic conductivity. The radial electric field with RMPs is predicted to be less negative than without RMP. The density drop and temperature rise in the pedestal region are observed in accordance with the experimental results. Generation of toroidal rotation in the co-current direction is predicted. Extrapolations to ITER are discussed.

Several improvements to the MAST plant and diagnostics have facilitated new studies advancing the physics basis for ITER and DEMO, as well as for future spherical tokamaks (STs). Using the increased heating capabilities P NBI ⩽ 3.8 MW H-mode at I p = 1.2 MA was accessed showing that the energy confinement on MAST scales more weakly with I p and more strongly with B t than in the ITER IPB98(y, 2) scaling. Measurements of the fuel retention of shallow pellets extrapolate to an ITER particle throughput of 70% of its original designed total throughput capacity. The anomalous momentum diffusion, χϕ, is linked to the ion diffusion, χi, with a Prandtl number close to P ϕ ≈ χϕ/χi ≈ 1, although χi approaches neoclassical values. New high spatial resolution measurements of the edge radial electric field, E r , show that the position of steepest gradients in electron pressure and E r (i.e. shearing rate) are coincident, but their magnitudes are not linked. The T e pedestal width on MAST scales with rather than ρpol. The edge localized mode (ELM) frequency for type-IV ELMs, new in MAST, was almost doubled using n = 2 resonant magnetic perturbations from a set of four external coils (n = 1, 2). A new internal 12 coil set (n ⩽ 3) has been commissioned. The filaments in the inter-ELM and L-mode phase are different from ELM filaments, and the characteristics in L-mode agree well with turbulence calculations. A variety of fast particle driven instabilities were studied from 10 kHz saturated fishbone like activity up to 3.8 MHz compressional Alfvén eigenmodes. Fast particle instabilities also affect the off-axis NBI current drive, leading to fast ion diffusion of the order of 0.5 m2 s−1 and a reduction in the driven current fraction from 40% to 30%. EBW current drive start-up is demonstrated for the first time in a ST generating plasma currents up to 55 kA. Many of these studies contributed to the physics basis of a planned upgrade to MAST.

A new B2SOLPS5.2 transport code has been developed and implemented for the simulation of H-mode shots. A new equation system is proposed, which is equivalent to the system which was used in B2SOLPS5.0 previously. The main idea is to replace the major part of the large radial ∇B driven convective fluxes by poloidal fluxes with the same divergence both in the particle balance and in the energy balance equations. This is of special importance for the H-mode where the diffusion coefficient is strongly reduced inside the barrier and large radial convective flows are strongly undesirable from the numerical point of view. The H-mode shots of ASDEX-Upgrade and MAST have been simulated with the new version with reasonable time steps and convergence. It is demonstrated that the radial electric field inside the edge transport barrier and in the pedestal region is close to the neoclassical electric field as in previous simulations of Ohmic shots. The toroidal rotation is co-current directed as in L-mode but is significantly larger in absolute value. It is shown that the shear of the poloidal drift at the inner side of the barrier is close to the value of the shear before the transition, while inside the barrier the value of the shear is significantly bigger. This fact determines self-consistently the width of the edge transport barrier. It is demonstrated that to match the experimental density and temperature radial profiles the drop in the diffusion coefficient within the barrier needs to be significantly larger than the drop in the electron heat conductivity coefficient. For the H-mode the pedestal region usually corresponds to the collisionless regime, so several corrections were introduced into the transport coefficients to extend the applicability of the code to the plateau and banana regimes in the inner regions of the simulation domain.

A comparison of the spatial and temporal evolution of the filamentary structures observed during type I ELMs is presented from a variety of diagnostics and machines. There is evidence that these filaments can be detected inside the LCFS prior to ELMs. The filaments do not have a circular cross section instead they are elongated in the perpendicular (poloidal) direction and this size appears to increase linearly with the minor radius of the machine. The filaments start off rotating toroidally/poloidally with velocities close to that of the pedestal. This velocity then decreases as the filaments propagate radially. By comparing the results from all measurements and from comparison with models it is most likely that the filaments have at least their initial radial velocity when they are far out into the SOL and before they have interacted with the nearest limiter surface. There is a general consensus that the dominant loss mechanism in the separated filaments is through parallel transport and that the transport to the wall is through the radial propagation of these filaments. Measurements of the filament energy content show that each filament contains up to 2.5% of the energy released by the ELM at the time it separates from the LCFS, assuming Ti = Te. The parallel flux e-folding length measured on DIII-D, AUG and MAST has a weaker scaling with normalised ELM size than appears to be necessary to explain the deficit in the ELM energy arriving in the divertor on JET, assuming a purely exponential decay of the filament energy with time.

Trace C13H4 injection experiments into the main scrape-off layer (SOL) of low density L-mode and high-density H-mode plasmas have been performed in the DIII-D tokamak [J. L. Luxon, Nucl. Fusion 42, 614 (2002)] to mimic the transport and deposition of carbon arising from a main chamber sputtering source. These experiments indicated entrainment of the injected carbon in plasma flow in the main SOL, and transport toward the inner divertor. Ex situ surface analysis showed enhanced C13 surface concentration at the corner formed by the divertor floor and the angled target plate of the inner divertor in L-mode; in H-mode high surface concentration was found both at the corner and along the surface bounding the private flux region inboard of the outer strike point. Interpretative modeling was made consistent with these experimental results by imposing a parallel carbon ion flow in the main SOL toward the inner target, and a radial pinch toward the separatrix. Predictive modeling carried out to better understand t...