T. Ohgo

Experiments with tungsten limiters in TEXTOR-94

Abstract The release of tungsten and light impurities from tungsten limiters exposed into the plasma edge of TEXTOR-94 has been measured by spectroscopic methods. Absolute effective tungsten sputtering yields are compared with model calculations on physical sputtering. The agreement is reasonable: however the observed strong decrease of tungsten release with increasing density cannot be fully explained. Erosion areas are clearly separated from carbon deposition zones. Surface analysis found neither carbon nor deuterium on the shiny metallic areas: A very sharp transition from “clean” metallic areas to carbon deposition zones within about 2–4 mm is found, instead. The carbon deposit is about 200–300 nm thick and contains deuterium with a D/C ratio of 0.05–0.1.

Operation of TEXTOR-94 with tungsten poloidal main limiters

In TEXTOR-94, experiments have been performed with the upper and lower poloidal limiter blocks made of vapour sprayed (VSP) tungsten (about 0.5 mm) deposited on graphite with a rhenium interlayer. A series of discharge conditions have been performed (density scan, scan of the auxiliary heating power, radius scan). There has been found no restriction for operation at any density with auxiliary heating. For Ohmic conditions the same density with testlimiters could be reached. Under siliconized conditions no severe accumulation of tungsten in the plasma centre could be detected. The blocks could in general stand surface temperatures below 1700 K. Most of them survived also temperatures above 3000 K without exfoliation. However, some blocks showed severe damage by melting or exfoliation probably due to insufficient contact of the tungsten layer with the graphite.

Comparison of impurity production, recycling and power deposition on carbon and tungsten limiters in TEXTOR-94

Abstract Impurity production, hydrogen recycling and power deposition on carbon and tungsten limiters have been investigated in TEXTOR-94 using a C–W twin test limiter. Considerable differences have been observed on W and C surfaces, which can be explained by the different particle and energy reflection coefficients of hydrogen on these surfaces. The measurements show in addition that the majority of the carbon release is from recycled carbon and that only a small part (below 10%) is due to net-erosion from the bulk carbon material. The heat deposition on C and W sides differs under the same plasma conditions significantly and is typically about 30% larger on the carbon surface. The behaviour of the impurity production, recycling and power deposition for various discharge conditions is presented.

Simulation calculations of mutual contamination between tungsten and carbon and its impact on plasma surface interactions

Mutual contamination between C and W, resulting from the simultaneous use of these materials as plasma facing components, is simulated by means of a computer simulation code, Erosion and Deposition based on Dynamic model (EDDY). W deposition on C rapidly increases the reflection coefficient for D and C impurity. In comparison between the calculation and a C-W twin test limiter experiment in TEXTOR-94, C release from the C side of the limiter is dominated by reflection of C impurity from the W deposits, in addition to physical sputtering of C; chemical erosion is strongly suppressed. Due to the dynamic effect which makes C-W mixed layer, C deposition on W gradually changes the reflection coefficient and sputter yields. Formation of a sharp boundary between erosion and C deposition zones on the W side of the limiter is well reproduced by simulation. Local redeposition patterns of C and W on the limiter surface are also calculated.

Deuterium release and microstructure of tantalum–tungsten twin limiter exposed in TEXTOR-94

Ta--W twin limiter experiments had been carried out in TEXTOR-94. D β and D γ intensities were observed during discharges. As a result, the intensities from the Ta surface were less than those from the W surface. This is attributed to the release ratio of molecules and atoms from the surface, i.e. less atomic release and more molecular release from Ta. Concerning long term D retention in Ta, D distributed uniformly over the limiter and the content was about 4 × 10 5 D per Ta atom. The total amount in the Ta limiter was estimated 3 × 10 20 D in the half limiter. After exposure, the Ta bulk showed a significantly modified surface: grain growth, wrinkle like deformations and local recrystallization. However, open cracks were not observed in Ta exposed in the same condition in which W created big open crack.

Material mixing on W/C twin limiter in TEXTOR-94

Abstract In order to investigate the effect of mutual contamination between tungsten (W) and carbon (C) and its influence on the plasma, a W–C twin test limiter, half made of W and the other half of C, was inserted into the edge plasma of TEXTOR-94 under ohmic and NBI heating conditions. The contamination process was observed by spectroscopy, and the intensity distribution of WI showed migration of W onto the C side by the successive cycles of sputtering and prompt redeposition. On the other hand, the deposition of C on the W surface was not obvious. Most of the hydrogen (deuterium) on the limiter was found to be retained in the deposited layers and that in the deposited C layer much higher than that in the deposited W layer. This indicates that tritium retention is smaller in metallic deposits above 500 K. The AES analysis conducted after the exposure of the test limiter showed that W deposited on C reacted with the substrate to form carbides at higher temperatures. The thickness of carbide layer, and/or the content of W in C were influenced by the temperature and flux distributions, and no carbide layer was formed at the limiter edge where the temperature was relatively low.

High-heat-flux-exposure-experiments of a tungsten-test-limiter at TEXTOR-94

Abstract To test the performance of tungsten as a plasma facing material, a small test limiter made of solid pure W was immersed in a TEXTOR plasma. A surface temperature in excess of 3000 K and the rate of temperature increase close to 2000 K/s was measured for the auxiliary heating conditions. These values corresponded to a power flux density of more than 40 MW/m 2 based on the calculations of a heat transfer model. The test limiter showed no observable degradation after exposure to this high heat flux density. The oxygen flux from the test limiter showed an increase during intense plasma exposure. However, no enhancement in the normalized W yield was observed when the W release was normalized to the C and O flux to the limiter, and the W release appeared to be dominated by physical sputtering.

Inhomogeneous heat loading to high-Z test limiters depending upon the limiter materials

Abstract The change in the temperature distribution was investigated for different target materials by inserting a Ta/W twin limiter into TEXTOR edge plasmas. Depending upon the limiter material and the limiter location, the position of the maximum temperature on the limiter surface changed. A three-dimensional heat transfer calculation taking the radiation loss from the surface into account was conducted to find out the factors governing the heat flux onto the test limiter. The result for the Ta test limiter shows that the experimentally observed temperature asymmetry in the toroidal direction can be explained if the plasma heat flux decreases exponentially from the last closed flux surface with the e-folding length of the ion drift side about 50% longer than that of the electron drift side. The intensity of D β light around the ion drift side of the limiter was larger than that at the electron drift side by about 30%, corresponding with a larger particle flux to the ion drift side.

Simulation and experimental studies of impurity release from tungsten exposed to edge plasmas in TEXTOR-94

Abstract Particle release from a W test limiter exposed to deuterium edge plasmas, containing impurities C and O, in the tokamak TEXTOR-94 is studied both experimentally and theoretically by means of Monte-Carlo simulations. The simulation model combines the dynamic composition change in the surface layer with the transport of released particles in the plasma. The release of W impurities from the limiter is strongly suppressed by prompt redeposition, whereas low- Z impurities (C and O) and D are much less influenced. Experimentally observed radial distributions of WI and CII line intensities in front of the limiter can be explained by physical sputtering of W and deposited C, respectively, in addition to high-energy reflection of impurity C ions. On the other hand, the CII line emission observed in front of graphite as a limiter is attributed to many low-energy (∼0.1 eV) atoms probably due to chemical sputtering by impact of D ions, which is strongly suppressed by deposition of high- Z impurities, such as W. The observed Dγ line emission indicates an important contribution of high-energy reflection to the release of D atoms from the W limiter, in addition to a large contribution of low-energy re-emission.

Neutral Fe atom measurements by laser induced fluorescence spectroscopy in Heliotron-E

Abstract In order to investigate the production mechanism of iron atoms from the vacuum vessel wall, a coaxial laser induced fluorescence spectroscopic system, which uses only one diagnostic port, has been developed at Heliotron-E. The neutral iron atoms near the wall were measured in currentless plasmas initiated by ECH and heated by neutral beam injection, and compared to the charge exchange neutral fluxes. The generation mechanisms and the relation between iron outflux form the wall and iron density in the plasma are discussed.