Table of contents

The International Workshop on Plasma-Facing Materials and Components is a successor of the International Workshop on Carbon Materials for Fusion Applications series. Between 1985 and 2003 ten 'Carbon Workshops' were organised in Jülich, Stockholm and Hohenkammer. Then it was high time for a change and redefinition of the scope of the symposium. To reflect the requirements of ITER and the ongoing evolution of the field it was decided to change the name of the workshop from 'International Workshop on Carbon Materials for Fusion Applications' to 'International Workshop on Plasma-Facing Materials and Components for Fusion Applications'. This workshop provides a discussion forum for experts from research institutions and industry dealing with materials for plasma-facing components in present and future controlled fusion devices. The operation of ASDEX-Upgrade with tungsten-coated wall, the fast progress of the ITER-Like Wall Project at JET, and the beginning of ITER construction emphasize the importance of plasma–wall interactions and component behaviour and impart enhanced momentum to the field. In this context, the properties and behaviour of beryllium, carbon and tungsten under plasma impact are research topics of foremost relevance and importance.

The 11th International Workshop (PFMC-11) was held in Greifswald, an old charming hanseatic and university town on the Baltic coast. The meeting was hosted by the Greifswald branch of the Max-Planck Institute for Plasma Physics where the large stellerator, Wendelstein-7X, is being constructed. A key objective for the symposium was to provide timely input for crucial decisions on the selection of first-wall materials for ITER. Our community is aware of the advantages and serious drawbacks of all candidate materials. As a result, a strong discussion is in progress whether to use carbon or to eliminate it and replace it by metal components. It is well known to the community that carbon—with its excellent power-handling capabilities but also strong 'addiction' to hydrogen—is an 'enfant terrible' in the family of candidate materials for first wall components. However, we know less about the behaviour of tungsten and beryllium especially in an environment containing carbon. There are several appealing issues of material mixing and fuel retention that call for robust and comprehensive studies. In this sense, the aim of the Workshop was not only to discuss hot topics but also to identify the most important research areas that need urgent resolution. Another topic of foremost relevance for ITER is the development of plasma-facing components which are able to withstand the extreme power fluxes, in particular those during transient phases. Materials and production methods for high-heat-flux components have to be further developed and industrialized. A key requirement is the development of non-destructive testing methods for qualification of methods and quality assessment during production.

A main goal of this workshop is to achieve an integration of physics and technology related aspects and phenomena. Therefore, invited talks and contributed presentations dealt with aspects of fundamental processes, experimental findings, advanced modelling, and technology of fusion reactor components. 23 oral and 65 poster presentations were given by participants coming from research laboratories and industrial companies. 155 researchers from 23 countries from all over the world participated in the lively and intense exchange of knowledge and ideas. The presentations covered several major topics selected as the key topics of the workshop.

• Materials for the ITER-divertor (erosion, re-deposition, fuel retention)

  • carbon-based materials (including C removal)

  • tungsten

  • beryllium

  • mixed materials (intentional and unintentional).

• The ITER-like Wall Project.

• High heat flux component development.

• Transient heat loads (ELMs, disruptions).

Conditio sine qua non for future success and progress in fusion science and engineering is the recruitment of young, talented scientists. Our responsibility is to get them interested in the topic and acquainted with major issues and challenges. For that reason, the Workshop was preceded by a series of tutorials on plasma–wall interactions and properties and testing of relevant materials. The response was beyond our expectation: nearly eighty young colleagues registered for the course and joined the workshop.

The workshop was financially supported by the Max-Planck Institute for Plasma Physics. We are very grateful to the staff of the Max-Planck Institute who helped with the organisation. Our most cordial thanks and gratitude go to Mrs Anne Eggeling for all her kindness and efficiency in the organisation that helped all of us to enjoy the meeting.

The present proceedings of this workshop contain manuscripts covering the contents of most of the invited presentations and a number of poster contributions which were pre-selected by the programme committee. In total, 48 manuscripts covering all topics of the workshop document the development and actual status of the field. We thank all the participants for their contributions and we thank the referees of the submitted papers. Thank you all for your hard work and co-operation. We look forward to the next meeting.

Wolfgang Jacob Max-Planck Institute for Plasma Physics, Garching, Germany Marek Rubel Royal Institute of Technology (KTH), Stockholm, Sweden

By the end of September 2006, the engineering commissioning of the Experimental Advanced Superconducting Tokamak (EAST) was completed and the first H₂ plasma was achieved with stainless steel as the plasma-facing material (PFM). In the following phases, with gradual increase in the heating power, the EAST divertor targets have to handle expected heat loads up to ∼10 MW m⁻². The PFMs to be employed include doped graphite and W coatings. The former will be used to cover the whole plasma-facing surface in the first phase, and then the latter applied to replace the graphite tiles gradually in the second phase, and finally a whole W plasma-facing surface will be expected. Graphite has been studied for many years and contributed greatly to the HT-7 to achieve long pulses of some 100 s. The W coatings are being developed and the latest R&D development is reported in the paper.

Recent data on deuterium retention in carbon fibre composites and tungsten both irradiated with D ions and exposed to D plasmas are presented. Deuterium depth profiles measured up to depths of 7–14 μm allow understanding of the mechanism which is responsible for the hydrogen isotope trapping in these materials. In the CFC materials the amount of retained deuterium increases with the ion fluence at all irradiation temperatures in the range from 323 to 723 K. No saturation is reached as observed in pyrolytic graphite. Depth profiles show that saturation occurs only within a near surface layer corresponding to the ion range. The increase in total retention at near-room temperature is accompanied by an increasing of the long profile tail extending beyond 14 μm with the D concentration of about 10⁻¹ at.% at a depth of 10 μm for fluences above 10²⁴ D m⁻². The depth at which deuterium is retained in tungsten (W) can be divided into three zones: (i) the near-surface layer (up to a depth of 0.2–0.5 μm depending on ion energy), (ii) the sub-surface layer (from ∼0.5 to ∼2 μm), and (iii) the bulk (>5 μm). Low-energy D ion irradiation modifies the W structure to depths of up to about 5 μm, both for W single crystals and polycrystalline W. The high D concentration (0.1–0.3 at.%) at depths of 1–3 μm relates to accumulation of D₂ molecules in vacancy clusters and voids. These defects are supposed to be generated due to plastic deformation of the W surface caused by deuterium supersaturation within the near-surface layer.

Wall conditioning is carried out in today's fusion machines to control the plasma impurity content. In ASDEX Upgrade boronization is routinely used to optimize the plasma start-up in the limiter phase and the performance of low-density, high power scenarios. We present data on the impurity content in the scrape-off layer (SOL) immediately after a boronization monitored at eight consecutive, nearly identical H-mode discharges. Discharge-resolved measurements of the deposition of impurity species on a collector probe were carried out in the outer midplane of ASDEX Upgrade. While the amount of collected boron decreases slowly the impurity concentration of most elements increases slowly and steadily. The collected amount of the first wall material tungsten was below the detection limit. Spectroscopic data observe an increase at a very low level. In contrast Ca and Fe, possible precursors of dust formation, increase at a significantly higher rate.

FOM-Rijnhuizen is building, in cooperation with its Trilateral Euregio Cluster (TEC) partners, a PSI-laboratory to study plasma surface interaction (PSI) under extreme, ITER relevant plasma conditions. The largest linear plasma generator of PSI-lab is Magnum-PSI, and is designed to deliver up to 10 MW m⁻² power over a 10 cm diameter hydrogen plasma beam with an electron density (n_e) up to 10²¹ m⁻³ and electron-temperature (T_e) between 1–5 eV. Magnum-PSI is presently under construction and its predesign is presented. Its forerunner is Pilot-PSI, in which record plasma parameters of n_e=4×10²¹ m⁻³ at T_e=2 eV in a ∼1 cm wide hydrogen beam confined by a magnetic field (B) ⩽1.6 T were measured at 40 mm downstream the source nozzle. At 17 mm in front of a target (located at 0.56 m distance from the source nozzle), n_e>10²¹ m⁻³ and T_e⩽ 4 eV have been demonstrated. Initial experiments on exposing fine-grain carbon samples are presented that showed up to 20 μm s⁻¹ erosion as a demonstration of the extreme plasma conditions. Spectroscopy was applied to compare chemical erosion yields for flux densities up to 5.0×10²⁴ m⁻² s⁻¹.

Carbon–carbon fibre composites (CFCs) have a unique combination of high conductivity, low Z and resistance to damage due to high heat loads. Given the demanding environmental requirements in the ITER divertor, specially developed CFCs are required as plasma facing materials. However, the manufacture of the plasma facing components, copper tubes surrounded by the CFC, and their inspection also places demands on the thermo-mechanical properties of the CFCs and their uniformity. This paper will describe, in detail, these demands and the work being undertaken to identify the appropriate thermo-mechanical properties of the CFC, since the three main demands, stability under ITER conditions, manufacturing ability of the plasma facing components and inspect-ability are not necessarily complementary. Due to the nature of CFC there are significant variations in its properties; these variations need to be limited, taking also into account cost and delivery requirements. The aim of the study undertaken is to identify the necessary CFC material properties, define their allowed range and select suitable CFC grades of material in 2008 for the European contribution to ITER.

A strong effect of a moderately elevated surface temperature on net carbon deposition and deuterium co-deposition in the DIII-D divertor was observed under detached conditions. A graphite DiMES sample with a 2 mm wide, 18 mm deep gap lined with silicon catcher plates was exposed to lower-single-null (LSN) L-mode plasmas first at room temperature, and then pre-heated to 200 °C by a built-in electrical heater. At the elevated temperature, deuterium co-deposition in the gap was reduced by an order of magnitude. At the plasma-facing surface of the pre-heated sample net carbon erosion was measured at a rate of 3 nm s⁻¹, whereas without pre-heating net deposition is normally observed under detachment. In a related experiment three sets of molybdenum mirrors recessed 2 cm below the divertor floor were exposed to identical LSN ELMy H-mode discharges. The first set of mirrors exposed at ambient temperature exhibited net carbon deposition at a rate of up to 3.7 nm s⁻¹ and suffered a significant drop in reflectivity. In contrast, two other mirror sets exposed at elevated temperatures between 90 and 175 °C exhibited practically no carbon deposition and their optical reflectivity in the wavelength range above 500 nm was essentially preserved.

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 2×10²⁵ 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.

Hydrocarbon injection experiments with in situ calibration of spectroscopic signals have been carried out to determine the hydrocarbon flux and the chemical erosion yield under detached plasma conditions in ASDEX Upgrade. A plasma scenario was developed to detach the outer divertor in L-mode and to provide a cold (T_e≃1–2 eV) and volume recombining deuterium target plasma. A significant decrease of carbon-related intrinsic photon fluxes, in particular of the CH Gerö and of the C₂ Swan band was observed in detachment. Moderate methane/ethene injections were performed through a single gas valve and an increase of carbon-related intrinsic photon fluxes was detected, which can be attributed to the amount of injected molecules. The experimentally determined effective inverse photon efficiencies are slightly larger than the ones deduced in attached plasmas. Nevertheless, the low intrinsic CH and C₂ light observed under detached conditions can still be attributed to a reduction of the hydrocarbon flux of about one order of magnitude. The impact of detachment on the chemical erosion is discussed.

Castellated structures are proposed for the first wall and divertor area in ITER. However, there is a concern that impurities and fuel will be deposited in the gaps of castellation leading to the undesirable tritium inventory. Optimization of the shape of cells in the castellated structures appears to be a possibility to reduce the impurity transport and fuel accumulation.

A new tungsten castellated limiter has been exposed in TEXTOR. The castellation cells of the rectangular and of the roof-like geometries were exposed under the same plasma conditions, allowing a direct comparison of the carbon transport and fuel inventory in the gaps. The gaps of shaped cells contained a factor of 2–3 less carbon and deuterium, in comparison to those of rectangular cells.

We present experimental chemical sputtering results for D⁺, D₂⁺ and D₃⁺ ions incident on ATJ graphite in the energy range 5–60 eV D⁻¹, and compare them with simulations for deuterated amorphous carbon impacted by neutral D, D₂ and D₃. The measured methane yields/D for the different species compared at the same energy/D diverge below about 60 eV D⁻¹, the incident triatomic molecular ions leading to the largest yields/D, and the atomic ions to the smallest, reaching a factor of two difference at 10 eV/D. The measured yields/D are in reasonable agreement with molecular dynamics simulations over the entire calculated energy range. The model surfaces were prepared by D, D₂ and D₃ impacts in a way that mimics the experiment. For D₂ incident at energies below 15 eV/D, the simulations show a strong dependence of the sputtering yields on the vibrational state of the incident projectile.

Here, we present new findings for the oxidative erosion of codeposits on divertor tiles removed from the DIII-D tokamak (General Atomics, San Diego). We have extended the previously (Ochoukov R et al 2006 Phys. Scr. T 124 27) studied temperature and O₂ pressure ranges to 673 K (400 °C) and 79 kPa (600 Torr), respectively. We have also examined the dependence of the amount of deuterium removed by oxidation on the impurity concentration in the specimen.

Finely dispersed Ti- and Zr-doped isotropic graphites have been manufactured using three different starting raw materials. The aim is to obtain doped fine grain isotropic graphites with reduced chemical erosion, high thermal shock resistance and low cost, which aim to be competitive with present carbon-based candidate materials for next step fusion devices. First ITER relevant thermal shock loads were applied on test specimens of these materials. The brittle destruction behaviour of graphite is greatly improved by doping with Ti or Zr, most probably due to a significant increase of thermal conductivity related to the catalytic effect of TiC and ZrC on the graphitization. Doped graphites manufactured with the synthetic mesophase pitch 'AR' as raw material showed the best performance from the three investigated raw materials due to its higher graphitability. The eroded surfaces of doped graphites exhibit a thin solidified carbide layer, probably caused by the segregation of liquid carbide during the thermal shot.

The use of carbon-based materials, i.e. graphite and carbon fibre composites (CFCs), as plasma facing materials in experimental nuclear fusion devices is widely spread. For the strike point area of the ITER divertor, a European candidate material was NB31, a three directional (3D) CFC material consisting of ex-pitch, ex-PAN and needled ex-PAN fibres, which will be replaced in the future by the newly developed material NB41. This composite material is used for the fabrication of divertor components for Wendelstein 7-X. In the frame of an extensive characterization on three different material batches of NB31 called 'pilot', 'serial' and 'additional' productions for W7-X and ITER, thermo-physical measurements and tensile tests were performed. The results were correlated to density and microstructure. The latter varied due to difficulties in the manufacturing process, in particular in the needling process. Therefore the specification of a minimum density and unit cell width (combined thickness of an ex-pitch and ex-PAN layer) is necessary to meet the ITER specifications for tensile strength and for thermal conductivity in the ex-pitch direction.

Bulk hydrogen retention and the analysis of absorption kinetics have been studied on graphite irradiated with neutrons at various conditions. Two kinds of hydrogen trapping sites may exist and be additionally produced during irradiation: interstitial cluster loop edge sites (trap 1) and carbon dangling bonds at edge surfaces of crystallites (trap 2). Neutron irradiation preferably creates trap 2 sites at lower fluences and trap 1 sites at a higher fluence. Trap 2 tends to be annealed out at high temperatures, although trap 1 is hardly decreased even at 1873 K. The activation energy of hydrogen diffusion is found to be increased from 1.04 to 1.60 eV by neutron irradiation.

Ultra-fine grained (UFG) W–TiC with a high purity matrix of low dislocation density is expected to exhibit improve resistance to irradiation with neutrons and helium ions and the room temperature mechanical properties. Aiming at such UFG W–TiC with the desired microstructure, powders of W with 0.25–0.8 wt% TiC additions were subjected to mechanical alloying (MA) and hot isostatic pressing (HIP), where purified H₂ and Ar were used as the MA atmosphere. Microstructural observations and room- and high-temperature mechanical tests were performed for UFG W–TiC before and after neutron irradiation to a fluence of 2×10²⁴  n m⁻² at 873 K. It is shown that the MA atmosphere significantly affects grain refinement, room-temperature strength and high-temperature tensile plasticity of UFG W–TiC. W–0.5TiC with H₂ in MA (W–0.5TiC–H₂) shows a larger strain rate sensitivity of flow stress, m, of 0.5∼0.6 at temperatures from 1673 to 1973 K, which is a feature of superplastic materials. Whereas W–0.5TiC–Ar shows a smaller m value of approximately 0.2. No radiation hardening is recognized in UFG W–0.5TiC–H₂ and W–0.5TiC–Ar.

The paper provides an overview of processes and underlying physics governing tungsten melt erosion in the fusion plasma environment. Experiments with three different bulk tungsten test-limiters were performed in TEXTOR: (i) thermally insulated solid plate fixed on a graphite roof-like limiter heated up by the plasma to the melting point, (ii) macro-brush of the ITER-relevant castellated structure and (iii) lamellae structure developed for the JET divertor. The main objectives were to determine the metal surface damage, the formation of the melt layer and its motion in the magnetic field. PHEMOBRID-3D and MEMOS-1.5D numerical codes were used to simulate the experiment with the roof-like test-limiter. Both experiments and simulation showed that the melting of tungsten can lead to a large material redistribution due to thermo-electron emission currents without ejection of molten material to the plasma.

The strain rate dependence of the brittle-to-ductile transition (BDT) temperature gives the activation energy controlling the BDT. Until recently, data were only available for a limited number of materials. Experimental data on the BDT of tungsten and other bcc metals have recently become available. We have compared all the data from different materials and sources, finding a distinctive relationship between the BDT temperature (T_BDT) and the activation energy for BDT (E_BDT) which holds over a wide range of materials, temperatures and activation energy values. The ratio E_BDT/kT_BDT gives approximately the value 25 for all the materials considered.

Deuterium retention in the implantation-induced defects in polycrystalline tungsten has been studied. Deuterium was implanted with different energies and concentrations of retained D were analysed with secondary ion mass spectrometry and nuclear reaction analysis. Annealings were carried out at four pre-determined temperatures corresponding to four different defect types that can trap deuterium. A quantitative number of each defect type produced by 5, 15 and 30 keV D implantation with a dose of 5.8×10¹⁶ cm⁻² was obtained.

Four kinds of tungsten samples were exposed to 76 eV D₂⁺ ions (38 eV D⁻¹) at a flux of 10²² D m⁻²s⁻¹. For plasma exposure at 315 K, blistering occurred more significantly on the unrecrystallized and single crystal W samples than the partially and fully recrystallized W samples. The un-recrystallized W sample showed the largest retention ratio at the same fluence. For samples exposed to higher fluences (up to 10²⁷ D m⁻²), the phenomena of bursting release of deuterium and blister bursting were clearly observed. Preliminary positron annihilation measurements indicated that the vacancy concentration in the near-surface region of tungsten increased after the deuterium plasma exposure. The results of electron back-scattering diffraction showed a strong dependence of blistering upon grain orientation, suggesting the possibility of alleviating blistering by selective texturing of tungsten for future fusion reactors.

Self passivating tungsten-based alloys may provide a major safety advantage in comparison with pure tungsten (W) which is presently the main candidate material for the plasma-facing protection of future fusion power reactors. Films of binary and ternary tungsten alloys were synthesized by magnetron sputtering. The oxidation behaviour was measured with a thermo balance set-up under synthetic air at temperatures up to 1273 K. Binary alloys of W–Si showed good self passivation properties by forming an SiO₂ film at the surface. Using ternary alloys the oxidation behaviour could be further improved. A compound of W–Si–Cr showed a reduction of the oxidation rate by a factor of 10⁴ at 1273 K.

The erosion of tungsten and carbon marker layers was studied in the outer divertor of ASDEX Upgrade. The outer strike point area and a large fraction of the outer baffle are net erosion areas for both materials. The net erosion rate of carbon is about 10–20 times larger than the net erosion rate of tungsten. The erosion is strongly inhomogeneous due to surface roughness, with a large erosion on plasma exposed areas of the rough surfaces, and deposition in recessions and pores.

Beryllium is planned as a plasma-facing material for ITER covering most of the inner wall of the plasma vessel. It is thus subjected to intense fluxes of escaping hydrogen ions from the plasma, which are implanted and retained in the plasma-facing material. As beryllium reacts very quickly with oxygen and especially with water, forming a surface oxide layer even under good vacuum conditions, previous studies of the interaction of deuterium and beryllium had to deal with oxygen contamination as a factor of uncertainty. To rule out the influence of a BeO-covered surface, the retention and release of 1 keV deuterium ions implanted in clean beryllium at room temperature are investigated by temperature programmed desorption (TPD). The surface composition is measured by x-ray photoelectron spectroscopy (XPS). The outline of a retention mechanism is developed by discussing the thermal release behaviour as a function of increasing deuterium fluence.

Beryllium-seeded deuterium plasma is used in PISCES-B to investigate mixed-material erosion and redeposition properties of ITER relevant divertor materials. The beryllium containing plasma simulates the erosion of first wall material into the ITER sol plasma and its subsequent flow towards the carbon divertor plates. The experiments are designed to quantify the behaviour of plasma created mixed Be/C and Be/W surfaces. Developing an understanding of the mixed material surface behaviour is crucial to accurately predict the tritium accumulation rate within the ITER vacuum vessel. The temporal evolution of the plasma interactions with the various mixed surfaces are examined to better understand the fundamental mechanisms in play at the surface and to allow scaling of these results to the conditions expected in the ITER divertor. A new periodic heat pulse deposition system is also installed on PISCES-B to simulate the transient temperature excursions of surfaces expected to occur in the ITER divertor during edge localized modes (ELMs) and other off-normal events. These periodically applied heat pulses allow us to study the effects of transient power loading on the formation, stability and tritium content of mixed-material surfaces that are created during the experiments.

The annealing of magnetron-sputtered carbon films doped with up to 8.5 at.% W, Ti, V, and Zr prior to the erosion experiments led to the formation of carbide cyrstallites of several nm. The total erosion yield by 30 and 200 eV D⁻¹, impacting as D₃⁺ projectiles, at 300 K and ∼750 K was determined from film thickness changes measured by ion beam analysis. Doping with W, Ti, V, and Zr always reduces the total erosion yield by a factor of up to 20 compared to a pure C film. The temperature dependency of the metal erosion was observed. The methane production yield was studied by mass spectrometry. The CD₄ production yield decreases less than the total yield or even increases. This implies that the distribution of the chemically eroded species is changed by the dopants. It can be speculated that the re-deposition of carbon is reduced by doping.

Mitigation of carbon chemical erosion by seeding of Be in the plasma is observed at the divertor simulator PISCES-B. The 3D Monte-Carlo ERO code is applied to model this effect. The time evolution of the process is modelled based on different surface models. The influence of various effects such as the plasma parameters, target biasing and carbide formation on the Be transport and the mitigation of carbon chemical erosion is investigated.

The intermetallic Be–W system is investigated by analysing both a W film on polycrystalline Be and the inverse system, Be films on polycrystalline W. The films are annealed up to 1070 K and the alloy formation is investigated by a combination of Rutherford backscattering spectroscopy (RBS) and x-ray photoelectron spectroscopy (XPS). For the structure analysis and identification of the formed alloys, x-ray diffraction (XRD) measurements are used. In the case of W films on Be, we identify Be₁₂W within the diffusion depth, whereas Be films on W show alloy formation restricted to the film–substrate interface. Both XPS and XRD measurements indicate the formation of Be₂W.

Work is in progress to completely replace, in 2008/9, the existing JET CFC tiles with a configuration of plasma facing materials consistent with the ITER design. The ITER-like wall (ILW) will be created with a combination of beryllium (Be), tungsten (W), W-coated CFC and Be-coated inconel tiles, with the material depending on the local anticipated heat flux and geometry. It is part of an integrated package of JET enhancements whose aim is to develop an understanding of the ITER materials issues and develop the techniques required to operate with inductive and advanced scenarios as close as possible to ITER parameters. Over 4000 tiles will be replaced and the ILW will accommodate additional heating up to at least 50 MW for 10 s. This paper describes the scientific background to the project, the technical objectives, the material configuration selected, the R&D behind the practical realization of the objectives and the generic problems associated with the Be tiles (power handling capacity and disruption induced eddy currents). One of the objectives is to maintain or improve the existing CFC tile power handling performance which has been achieved in most cases by hiding bolt holes, optimizing tile size and profile and introducing castellations on plasma facing surfaces.

The ITER-like wall project has been launched to design, manufacture and test all the necessary components in view of their installation in a dedicated shutdown 2008–09. One of the R&D activities in the frame of the project is the development of a concept of an inertially cooled bulk W component for the JET divertor. Two options were studied as the concepts, (i) W/CFC brazing design which consists of W blocks on a carbon fibre composite (CFC) tile; (ii) W lamellae design, which is a mechanical assembly of several stacks of W blades. Both concepts were realized as test tiles and tested under high heat flux loading. The lamellae test tile showed a high performance fulfilling the requirements. Finally, the lamellae concept was chosen to be the primary solution and comprehensive numerical analyses were performed for the further optimization of the concept.

In the frame of JET's ITER-like wall (ILW) project tungsten coatings on carbon fibre reinforced carbon substrates will be used in the divertor and highly loaded areas in the main chamber. Fourteen different types of samples were produced by physical or chemical vapour deposition and vacuum plasma spray (VPS) with coating thickness of 4, 10 and 200 μm. Similarly, three different VPS W coatings (200 μm) on two different graphite substrates, were produced for use at the strike-point regions of ASDEX Upgrade. All coatings were subjected to thermal screening and thermal cycling tests in the ion beam facility GLADIS. Additionally, the coatings intended for the ILW project were exposed to edge localized mode (ELM)-like thermal loads in the electron beam facility JUDITH. A general failure mode with the CFC substrate is crack formation upon cool-down, whereas the coatings on graphite do not show any crack formation. Additionally, metallographic investigations, x-ray diffraction measurements, adhesion testing as well as measurements on the contents of light impurities were performed.

Preparatory study for the operation of the JET tokamak with a full metal wall (ITER-like wall project) also comprises several activities aiming at the development of thin beryllium coatings. The purpose is 2-fold: (i) to coat Inconel® tiles of the inner wall cladding; (ii) to develop methods for production of films for so-called marker tiles in order to enable monitoring of Be erosion from limiters. Properties of the marker film must match, as closely as possible, those of bulk Be. The first step in the R&D process was to assess coating methods and the quality of layers deposited on test coupons. Smooth, dense Be films of high purity and good adhesion to the substrate were deposited with an average deposition rate of 5±0.5 nm s⁻¹ to a thickness of 7.5 μm. A marker structure consisting of a 7.5 μm Be film on top of a 2.5 μm Ni interlayer deposited on a bulk Be block has been developed and characterized by means of material analysis methods. An overview of manufacturing processes and properties of the marker coatings is presented.

Vaccum plasma sprayed tungsten coatings with an evaporated sandwich Re–W interlayer on graphite limiter blocks were studied after the experimental campaign in the TEXTOR tokamak. The coating morphology was modified by high-heat loads and co-deposition of species from the plasma. Co-deposits contained fuel species, carbon, boron and silicon. X-ray diffractometer phase analysis indicated the coexistence of metallic tungsten and its carbides (WC and W₂C) and boride (W₂B). In the Re–W layer the presence of carbon was detected in a several micrometres thick zone. In the overheated part of the limiter, the Re–W layer was transformed into a sigma phase.

In order to perform a fully integrated material test, JET has launched the ITER-like wall project with the aim of installing a full metal wall during the next major shutdown. The material foreseen for the main chamber wall is bulk Be at the limiters and Be coatings on inconel tiles elsewhere. R&D process comprises global characterization (structure, purity etc) of the evaporated films and testing of their performance under heat loads. The major results are (i) the layers have survived energy loads of 20 MJ m⁻² which is significantly above the required level of 5–10 MJ m⁻², (ii) melting limit of beryllium coating would be at the energy level of 30 MJ m⁻², (iii) cyclic thermal load of 10 MJ m⁻² for up to 50 cycles have not induced any noticeable damage such as flaking or detachment.

Combined magnetron sputtering and ion implantation (CMSII) is a deposition technique involving simultaneous magnetron sputtering and high energy ion bombardment of the coating during its growth. A high voltage pulse discharge (U=40 kV, τ=20 μs, f=25 Hz) is superposed over the magnetron deposition and in this way, positive ions are accelerated to the components to be coated, bombarding initially the substrate and then the coating itself. In the framework of the ITER-like wall project this method was applied to produce nanostructured W coatings on the carbon fibre composite (CFC) substrate. These coatings have been characterized in terms of adhesion, thickness, structure and resistance to high thermal loads (up to 23.5 MW m⁻²). Based on the results of these tests, which are presented in this paper, CMSII technology was selected for coating about 1100 tiles with a 10 μm tungsten layer for the JET first wall and divertor.

The joining of tungsten and carbon-based materials is demanding due to the incompatibility of their chemical and thermophysical properties. Direct joining is unfeasible by the reason of brittle tungsten carbide formation. High-temperature brazing has been investigated in order to find a suitable brazing filler metal (BFM) which successfully acts as an intermediary between the incompatible properties of the base materials. So far only low Cr-alloyed Cu-based BFMs provide the preferential combination of good wetting action on both materials, tolerable interface reactions, and a precipitation free braze joint. Attempts to implement a higher melting metal (e.g. Pd, Ti, Zr) as a BFM have failed up to now, because the formation of brittle precipitations and pores in the seam were inevitable. But the wide metallurgical complexity of this issue is regarded to offer further joining potential.

The development of actively cooled plasma-facing components (PFC) represents one of fusion's most challenging engineering efforts. In this frame, a high-quality bonding between the refractory armour and the heat sink is essential to ensure the heat removal capability and the thermal performances of PFC. Experience gained during manufacturing of Tore Supra actively cooled PFC led to the establishment of a qualification methodology and provided a large experience of acceptance criteria using an active infrared thermography (système d'acquisition de traitement infra-rouge, SATIR). This paper presents the application of this qualification process to the W7-X pre-series components, with the objective of assessing and defining workable acceptance criteria that enable reliable predictions of performance at the nominal heat flux requirements in W7-X. Finally, to check the reliability of the non-destructive examination (NDE) method by transient infrared thermography, the newly defined acceptance criteria were applied to W7-X pre-series target elements (batch #3). The SATIR results, benchmarked with HHF tests performed on the GLADIS ion beam facility were discussed to assess the ability to detect critical defects at the interface between tiles and heat sink.

Plasma facing components (PFC) for modern steady state fusion devices have to withstand heat fluxes up to 20 MW m⁻². In case of internal defects in the components, a critical heat flux event can occur at an incident heat flux lower than expected in a defect-free component. It is therefore essential to develop in-vessel means for on line safety monitoring to prevent PFC destruction, and health monitoring during shutdown to control PFC damage. Two methods are developed in this paper, covering these topics.

The assembly of the highly-loaded target plates of the WENDELSTEIN 7-X (W7-X) divertor requires the fabrication of 890 target elements (TEs). The plasma facing material is made of CFC NB31 flat tiles bonded to a CuCrZr copper alloy water-cooled heat sink. The elements are designed to remove a stationary heat flux and power up to 10 MW m⁻² and 100 kW, respectively. Before launching the serial fabrication, pre-series activities aimed at qualifying the design, the manufacturing route and the non-destructive examinations (NDEs). High heat flux (HHF) tests performed on full-scale pre-series TEs resulted in an improvement of the design of the bond between tiles and heat sink to reduce the stresses during operation. The consequence is the fabrication of additional pre-series TEs to be tested in the HHF facility GLADIS. NDEs of this bond based on thermography methods are developed to define the acceptance criteria suitable for serial fabrication.

In this paper, an optimized active metal cast copper to carbon composite (C/C) joint is examined. Mechanical characterization by means of shear and tensile tests as well as fracture toughness measurements at ambient temperature are employed, which show that the obtained values are above those of the reference joint that has been used so far. The phases occurring at the interface region and the fracture surface of the optimized joint are investigated in detail to explain the improved mechanical behaviour. The composition and the morphology at the interface region are analysed by means of scanning electron microscopy. In addition, the crack path is analysed by stereomicroscopy. The improved properties can be explained by the circumstance that the fracture of the optimized joint is not restricted to the interface between the occurring carbide and the C/C but changes between C/C, the carbide to C/C interface and the copper, depending on the local structure in the C/C composite.

Carbon fibre composite (CFC) armours have been successfully used for actively cooled plasma facing components (PFCs) of the Tore Supra (TS) tokamak. They were also selected for the divertor of the stellarator W7-X under construction and for the vertical target of the ITER divertor. In TS and W7-X a flat tile design for heat fluxes of 10 MW m⁻² has been chosen. To predict the lifetime of such PFCs, it is necessary to analyse the damage mechanisms and to model the damage propagation when the component is exposed to thermal cycling loads. Work has been performed to identify a constitutive law for the CFC and parameters to model crack propagation from the edge singularity. The aim is to predict damage rates and to propose geometric or material improvements to increase the strength and the lifetime of the interfacial bond. For ITER a tube-in-tile concept (monoblock), designed to sustain heat fluxes up to 20 MW m⁻², has been developed. The optimization of the CFC/Cu bond, proposed for flat tiles, could be adopted for the monoblock concept.

In experimental fusion reactors the copper alloy CuCr1Zr is a widely used alloy for heat sinks. The thermal conductivity at room temperature of this alloy measures 370 W m⁻¹K⁻¹. Its room temperature mechanical properties with a tensile strength of 400–470 MPa and a yield stress of 280–380 MPa are based on a dispersion hardening and an aging treatment. The long-term temperature capability is however limited due to an overaging effect taking place in the temperature range of roughly 350 °C up to an aging temperature of 480 °C. A possibile way to improve the properties at elevated temperatures is by embedding stiff, strong fibres, e.g. SiC-fibres. In the present study, the mechanical behaviour of SiC-fibre reinforced CuCr1Zr is determined at 550 °C and compared with the room temperature properties. The thermal conductivity is considerably reduced by embedding SiC-fibres. From measured values of the thermal conductivity of the composite material the axial thermal conductivity of the SiC-fibre can be roughly estimated to be 16 W m⁻¹K⁻¹.

Plasma facing components (PFC) in Tore Supra and W7X adopt the flat tile concept using carbon fibre composite (CFC) material for the plasma facing material. As the cooling structure is made of a copper alloy material (CuCrZr), the bonding technique between CFC tiles and CuCrZr is critical. Currently, a soft metallic compliant layer is interposed between the two; in such a way the significant thermal expansion mismatch between carbon and copper can be accomodated. The development of a reliable non-destructive inspection technique (NDT) for the bond, to be performed during the manufacturing process, is obviously of great importance. The SATIR (infrared thermography) test bed operating at Commisariat à l'Energie Atomique (CEA) Cadarache performs this function using transient infrared thermography: the thermal excitation is realized in the cooling channel and the presence of a faulty tile is detected in the form of a delayed thermal response. With this technique, the evolution of the surface temperature of an inspected element was compared to that of a defined free-defect element, using the so-called DTref criterion (maximum of the transient temperature difference). The defect detection capability of the SATIR test bed can be improved using signal processing methods. A first treatment based on spatial image autocorrelation allows a better localization of the bond defect. Moreover, the problem of detection and classification of random signals (like the thin defect signature) can be solved maximizing the signal-to-noise ratio (SNR). Two filters maximizing this ratio were optimized: the stochastic matched filter (SMF) aims at defect detection, while the constrained SMF aims at defect classification. These methods assume that the second-order properties of the process at play are known, through covariance matrices. All these methods process the SATIR signal utilizing any free-defect element as reference signal. The tile temperature signal is either processed by itself or compared to a numerical temperature evolution evaluated by finite element calculations. In view of ITER PFCs qualification, the possibility to extend all these methods from the flat tile concept to the monobloc target element concept is proved.

The power load limits of the WENDELSTEIN7-X divertor target elements were experimentally evaluated with heat loads considerably exceeding the expected operating conditions. The water-cooled elements are designed for steady-state heat flux of 10 MW m⁻² and to remove a power load up to 100 kW. The elements must allow a limited operation time at 12 MW m⁻² steady-state and should not fail for short pulses of up to 15 MW m⁻² for cooling conditions in the subcooled nucleate boiling regime. In the framework of the qualification phase, pre-series target elements were loaded up to 24 MW m⁻² without loss of CFC tiles. A critical heat flux at the target of 31 MW m⁻² was achieved. The paper discusses the results of the tests performed at the high heat flux test facility GLADIS. The experimental results compared to transient nonlinear fine element method (FEM) calculations confirm a high thermal safety margin of the target design sufficient for plasma operation in W7-X.

New experimental results on transient loads during ELMs and disruptions in present divertor tokamaks are described and used to carry out a extrapolation to ITER reference conditions and to draw consequences for its operation. In particular, the achievement of low energy/convective type I edge localized modes (ELMs) in ITER-like plasma conditions seems the only way to obtain transient loads which may be compatible with an acceptable erosion lifetime of plasma facing components (PFCs) in ITER. Power loads during disruptions, on the contrary, seem to lead in most cases to an acceptable divertor lifetime because of the relatively small plasma thermal energy remaining at the thermal quench and the large broadening of the power flux footprint during this phase. These conclusions are reinforced by calculations of the expected erosion lifetime, under these load conditions, which take into account a realistic temporal dependence of the power fluxes on PFCs during ELMs and disruptions.

Carbon-fibre composite (CFC) and tungsten macrobrush armours are foreseen as PFC for the ITER divertor. In ITER the main mechanisms of metallic armour damage remain surface melting and melt motion erosion. In the case of CFC armour, due to rather different heat conductivities of CFC fibres a noticeable erosion of the PAN bundles may occur at rather small heat loads. Experiments carried out in the plasma gun facilities QSPA-T for the ITER like edge localized mode (ELM) heat load also demonstrated significant erosion of the frontal and lateral brush edges. Numerical simulations of the CFC and tungsten (W) macrobrush target damage accounting for the heat loads at the face and lateral brush edges were carried out for QSPA-T conditions using the three-dimensional (3D) code PHEMOBRID. The modelling results of CFC damage are in a good qualitative and quantitative agreement with the experiments. Estimation of the droplet splashing caused by the Kelvin–Helmholtz (KH) instability was performed.

Surface temperature control of actively cooled targets is part of the machine protection in long pulse experiments. The protection of the target is as good as the surface measurement and the thermal model of the target. This paper deduces thermal parameters necessary to correct the measured temperature for surface effects. The temperature measurement error due to the intrinsic CFC structure is discussed. Layers on top of the bulk material result in a change of the temperature response and in an overestimation of the calculated heat flux. All discussed effects did not reduce the machine safety but can reduce the operational space of the experiment. The thermal surface properties have to be checked in situ by changing the heat load to the target and calculating the temperature response.

The behaviour of a preheated tungsten target under repetitive pulsed plasma impacts of the energy density 0.75 MJ m⁻² with the pulse duration of 0.25 ms was studied with the quasi-stationary plasma accelerator (QSPA) Kh-50. Two identical samples of pure sintered tungsten have been exposed to numbers of pulses exceeding 100. One sample was maintained at room temperature and the other sample preheated at 650 °C. The experiments demonstrated that on the cold surface some macro-cracks dominate, but on the hot surface they do not develop. However, in both cases some fine meshes of micro-cracks are observed. With increasing the number of exposures, the width of the micro-cracks gradually increases, achieving 0.8–1.5 μm after 100 pulses. In addition, the SEM shows some cellular structure with the cell sizes about 0.3 μm, and after large numbers of exposures some blisters of sizes up to 100–150 μm appear.

The need for non-destructive inspection of the interface between carbon fibre composites (CFC) and copper is met by different thermographic inspection procedures. The most established one is transient thermography where the component is heated via the cooling system by introducing hot water. This method is applied on finished components containing a cooling system. A promising set-up is pulsed thermography where the component is heated externally by a single powerful flash of light. This method can be applied in any stage of production which enables the recognition of critical production steps. Results obtained from test bodies containing artificial defects demonstrate the defect resolution and the good reproducibility of this method. Since no reference element is needed the efficiency of inspection is raised to a level which is appropriate for industrial application. Finally, the rating of the first test series of W7-X divertor components shows excellent coincidence with the results obtained from high heat flux tests.

Two fusion-relevant carbon fibre composites (CFCs), NB31 (3D) and DMS704 (2D), have been thermally exposed in five different orientations to simulate transient events (disruptions and ELMs). In these experiments with intense transient heat pulses in the electron beam JUDITH and the ion beam facility RHEPP-1, CFCs showed a preferential erosion of the fibre bundles aligned in the surface-plane. This phenomenon also occurred when these fibre bundles were the ones with the highest thermal conductivity. First results show that reduction of the number of fibres aligned parallel to the surface-plane strongly contributes to limiting the preferential erosion by fibre segment ejection (brittle destruction) in CFCs. Strong mechanical attachment is essential to avoid fibre segment ejection during disruption. The selection of an optimum material orientation of the available CFCs is of relevant importance and will be further investigated.