D. Matveev

Modelling of impurity deposition in gaps of castellated surfaces with the 3D-GAPS code

The Monte-Carlo neutral transport code 3D-GAPS is described. The code models impurity transport and deposition in remote areas, such as gaps between cells of castellated plasma-facing surfaces. A step-by-step investigation of the interplay of different processes that may influence the deposition inside gaps, namely particle reflection, elastic neutral collisions, different particle sources, chemical erosion and plasma penetration into gaps, is presented. Examples of modelling results in application to the TEXTOR experiment with a castellated test limiter are provided. It is shown that only with the assumption of the presence of species with different reflection probabilities, do simulated carbon deposition profiles agree with experimental observations for side surfaces of the gaps. These species can be attributed to different particle sources, e.g. carbon atoms and hydrocarbon radicals. Background carbon ions and atoms have low and moderate values of the reflection coefficient (R ≤ 0.6), while some of the hydrocarbon radicals produced by chemical erosion of redeposited carbon layers have high reflection probability (R ≥ 0.9). Deposition at the bottom of the gaps cannot be adequately reproduced unless extreme assumptions on particle sources and reflection properties are imposed. Elastic neutral collisions and ionization of neutrals escaping the gaps have no significant influence on the results. Nevertheless, particle-in-cell simulations of plasma penetration into gaps are essential for estimating the incoming ion flux and leading to a better quantitative agreement with experimental observations.

ERO code benchmarking of ITER first wall beryllium erosion/re-deposition against LIM predictions

Previous studies (Carpentier et al 2011 J. Nucl. Mater. 415 S165–S169) carried out with the LIM code of the ITER first wall (FW) on beryllium (Be) erosion, re-deposition and tritium retention by co-deposition under steady-state burning plasma conditions have shown that, depending on input plasma parameter assumptions and sputtering yields, the erosion lifetime and fuel retention on some parts of the FW can be a serious concern. The importance of the issue is such that a benchmark of this previous work is sought and has been provided by the ERO code (Pitts et al 2011 J. Nucl. Mater. 415 S957–S964) simulations described in this paper. Provided that inputs to the codes are carefully matched, excellent agreement is found between the erosion/deposition profiles from both codes for a given ITER-shaped FW panel. Issues regarding the difficult problem of the correct treatment of Be sputtering are discussed in relation to the simulations. The possible influence of intrinsic Be impurity is investigated.

Modelling of local carbon deposition on a rough test limiter exposed to the edge plasma of TEXTOR

A Monte-Carlo code called SURO was developed to study the influence of surface roughness on the impurity deposition characteristics in fusion experiments. SURO uses the test particle approach to describe the impact of background plasma and the deposition of impurity particles on a sinusoidal surface. The local impact angle and dynamic change of surface roughness as well as surface concentrations of different species due to erosion and deposition are taken into account. Coupled with the three-dimensional Monte-Carlo code ERO, SURO was used to study the impact of surface roughness on 13C deposition in 13CH4 injection experiments in TEXTOR. The simulations showed that the amount of net deposited 13C species increased with surface roughness. Parameter studies with varying 12C and 13C fluxes were performed to gain insights into impurity deposition characteristics on the rough surface. Calculations of the exposure time needed for surface smoothing for TEXTOR and ITER were also carried out for different scenarios.

Particle-in-cell simulations of plasma interaction with shaped and unshaped gaps in TEXTOR

This paper presents particle-in-cell simulations of the plasma behaviour in the vicinity of gaps in castellated plasma-facing components (PFCs). The point of interest was the test limiter of the TEXTOR tokamak, a PFC designed for studies of plasma–wall interactions, in particular, related to impurity transport and fuel retention. Simulations were performed for various plasma conditions in the vicinity of the castellated surface, where the gaps can be either shaped or unshaped. It was observed that depending on plasma parameters the transport of plasma particles inside the gap can be either in potential- or geometry-dominated regimes. The mechanisms responsible for the formation of a potential peak inside the poloidal gap and its consequences on plasma deposition profiles are discussed. A study of gap shaping was performed in order to validate its effectiveness.

Determination of Be sputtering yields from spectroscopic observations at the JET ITER-like wall based on three-dimensional ERO modelling

Estimations of the ITER first wall (FW) lifetime, previously made using the three-dimensional Monte-Carlo ERO code (Borodin et al 2011 Phys. Scr. T145 014008), 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 experiments. Applying the same sputtering input data in ERO as those used before in the ITER-predictions, the ERO simulations for the Be light intensity (using up to date atomic data from ADAS and measured plasma conditions) reveal a factor of 2 overestimation in 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 necessity to correct (reduce) the respective estimates for the Be sputtering yield and, accordingly, re-visit the ITER FW lifetime predictions.

ERO modelling of local deposition of injected C-13 tracer at the outer divertor of JET

The 2004 tracer experiment of JET with the injection of (CH4)-C-13 into H-mode plasma at the outer divertor has been modelled with the Monte Carlo impurity transport code ERO. EDGE2D solutions for inter-ELM and ELM-peak phases were used as plasma backgrounds. Local two-dimensional (2D) deposition patterns at the vertical outer divertor target plate were obtained for comparison with post-mortem surface analyses. ERO also provides emission profiles for comparison with radially resolved spectroscopic measurements. Modelling indicates that enhanced re-erosion of deposited carbon layers is essential in explaining the amount of local deposition. Assuming negligible effective sticking of hydrocarbons, the measured local deposition of 20-34% is reproduced if re-erosion of deposits is enhanced by a factor of 2.5-7 compared to graphite erosion. If deposits are treated like the substrate, the modelled deposition is 55%. Deposition measurements at the shadowed area around injectors can be well explained by assuming negligible re-erosion but similar sticking behaviour there as on plasma-wetted surfaces.

Prediction of long-term tritium retention in the divertor of ITER: influence of modelling assumptions on retention rates

ERO modelling of long-term tritium (T) retention has been done for the divertor of ITER with graphite target plates assuming a certain beryllium influx into the divertor, eroded from the main chamber. The divertor beryllium (Be) influx relative to the deuterium ion flux has been fixed at 0.1% for the outer divertor and 1.0% for the inner divertor. In addition to the original B2-Eirene plasma background, the influence of variations of temperature and density in the divertor has been studied. Moreover, assumptions for enhanced erosion of redeposited carbon and effective sticking for hydrocarbons have been analysed. With graphite target plates, long-term tritium retention is dominated by T co-deposition in deposits. Within the studied parameter range, the modelling yields 200?500 possible ITER discharges without cleaning before reaching the safety limit of 700?g of in-vessel retained tritium. Surface temperature-dependent tritium amounts in carbon and beryllium deposits have been taken into account.

Estimation of the contribution of gaps to tritium retention in the divertor of ITER

An estimation of the contribution of gaps to beryllium deposition and resulting tritium retention in the divertor of ITER is presented. Deposition of beryllium layers in gaps of the full tungsten divertor is simulated with the 3D-GAPS code. For gaps aligned along the poloidal direction, non-shaped and shaped solutions are compared. Plasma and impurity ion fluxes from Schmid (2008 Nucl. Fusion 48 105004) are used as input. Ion penetration into gaps is considered to be geometrical along magnetic field lines. The effect of realistic ion penetration into gaps is discussed. In total, gaps in the divertor are estimated to contribute about 0.3 mgT s−1 to the overall tritium retention dominated by toroidal gaps, which are not shaped. This amount corresponds to about 7800 ITER discharges up to the safety limit of 1 kg in-vessel tritium; excluding, however, tritium release during wall baking and retention at plasma-wetted and remote areas.

Plasma-surface interaction activities in KSTAR

Selected topics of Plasma-Surface Interaction (PSI) activities in KSTAR are briefly introduced. SOL parameter measurements, particle balance and fuel retention, in-vessel dust research, and finally tungsten R & D are discussed. Some quantitative numbers from the initial phase of the operation are given for comparison with that of other machines.

Modelling of carbon deposition from CD4 injection in the far scrape-off layer of TEXTOR

To analyse the impurity transport in plasma-shadowed, remote areas, methane CD4 has been injected into the far scrape-off layer of TEXTOR through a cylinder equipped with a quartz micro balance (QMB). CD4 transport including break-up and resulting deposition on the QMB (shot-resolved) and on the cylinder top surface (shot-integrated) has been modelled with the codes ERO and 3D-GAPS. The modelling shows good agreement with the observations if reflection coefficients based on molecular dynamics simulations are used. In contrast to plasma-wetted areas, no enhanced erosion has to be applied.