T. Tanabe

Tritium retention of plasma facing components in tokamaks

Abstract The areal distribution of tritium retention in tiles from TEXTOR, TFTR, JT-60U and JET has been measured via the imaging plate technique and the results are discussed from the perspective of carbon–hydrogen chemistry. It is found that the observed tritium distribution clearly shows asymmetries in poloidal and toroidal directions and also reflects the local temperature history of the analyzed tiles. We show the first clear evidence of the loss of high energy tritons by toroidal magnetic field ripple. We distinguish three different contributions to tritium retention in tokamaks with carbon plasma facing components: high energy tritons escaping from the core plasma, low energy ions and neutrals from the edge plasma, and molecular tritium from gas fueling. These components are retained at different depths and with different concentrations. Tritium from the edge plasma dominates the retained inventory but could be reduced if the surface temperature was higher. We propose tokamak operation with plasma facing components above 1000 K as a possible way to reduce the tritium inventory.

Tritium distribution in JT-60U W-shaped divertor

Abstract Detailed tritium profiles on the JAERI Tokamak-60U (JT-60U) W-shaped divertor and the first wall tiles were examined by a tritium imaging plate technique (TIPT) and full combustion method. The highest tritium level (60 kBq/cm 2 ) was observed at the dome top tiles. The tritium level of the divertor target was lower (2 kBq/cm 2 ). The result of the triton deposition simulation using orbit following Monte-Carlo code was consistent with the tritium distribution obtained by TIPT and full combustion method. These results indicate that the tritium distribution of the JT-60U W-shaped divertor reflects mainly the distribution of the energetic triton impinging on the wall. According to the simulation, the tritium atoms produced by D–D nuclear reaction in JT-60U are not loosing completely their initial energy of 1 MeV and around 1/3 of them are implanted into the wall.

Application of tungsten for plasma limiters in TEXTOR

Abstract Three different types of W limiters were exposed in the TEXTOR plasma and the response of the plasma and materials performance of the limiters were investigated. 1. A W bulk limiter operated with preheating above 800 K withstood a plasma heat load of about ∼20 MW/m 2 for a few seconds with some slight surface melting during the highest heat load shot. However, it was severely damaged when operated at around 500 K. 2. A C/W twin test limiter, half made of bulk W and the other half of graphite (EK-98) gave very useful information on how low- and high-Z materials behave under conditions of simultaneous utilization as PFM such as cross-contamination and the influence of a large mass difference on hydrogen reflection and deposition. 3. Two sets of main poloidal W limiters made of vacuum vapor sprayed (VPS)-W deposited on graphite (IG-430U) with a Re interlayer could absorb about 60% of the total convection heat and the ohmic plasma with a density as high as 5xa0×xa010 13 cm −3 was sustained. Most of the VPS-W coated limiters tolerated a heat load of ∼20 MW/m 2 . nThis series of W limiters experiments in TEXTOR has shown that W is applicable as a PFM, if its central accumulation is avoided by NBI and/or ICRH heating. Nevertheless, some concerns still remain, including difficulty of plasma start-up, W behavior in higher temperature plasmas, and materials selection.

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.

Tritium distribution on the surface of plasma facing carbon tiles used in JET

Tritium surface profiles on divertor tiles used in JET were successfully determined applying imaging plate (IP) technique. The tritium intensities measured by IP were quite consistent with the previous tritium analysis made by full combustion measurements. Present results are summarized as follows. Most of the tritium in the divertor tiles was retained in co- or re-deposited layer and did not move because of their temperature was rather low. In addition tritium produced by D–D reaction are implanted in subsurface layers rather homogeneously, which seems the main tritium source for the inner divertor tiles with some thermal modification. Still there is another but very small tritium retention observed at the back side of the tiles, of which profile reflects the 2-D CFC structure, indicating preferential absorption and migration of tritium. 2003 Elsevier Science B.V. All rights reserved.

Nondestructive measurement of surface tritium by β-ray induced X-ray spectrometry (BIXS)

Abstract Applicability of a newly developed β-ray induced X-ray spectrometry (BIXS) has been examined to measure nondestructively tritium retained on/in the graphite samples. Examination was carried out by using the graphite plates irradiated with tritium ions and an ALT-II limiter tile exposed to D-plasmas in TEXTOR. For the former samples, a sharp intense peak and a broad weak peak appeared clearly in the spectra; the former peak was attributed to the characteristic X-rays from argon used as a working gas, and the latter peak was assigned to the bremsstrahlung X-rays from sub-surface layers of graphite. On the other hand, for the latter sample, a rather weak characteristic X-ray peak was observed along with a diminutive bremsstrahlung X-ray peak. Although the intensities of those X-rays differed from spot to spot, the tritium levels retained on the limiter tile were determined to be 58–132 Bq / cm 2 . It was concluded, therefore, that valuable information on the amount and the distribution of tritium retained on/in the wall materials can be nondestructively obtained by using the BIXS.

Comparison of Tokamak Behaviour with Tungsten and Low-Z Plasma Facing Materials

Graphite wall materials are used in present day fusion devices in order to optimize plasma core performance and to enable access to a large operational space. A large physics database exists for operation with these plasma facing materials, which also indicate their use in future devices with extended burn times. The radiation from carbon impurities in the edge and divertor regions strongly helps to reduce the peak power loads on the strike areas, but carbon radiation also supports the formation of MARFE instabilities which can hinder access to high densities. The main concerns with graphite are associated with its strong chemical affinity to hydrogen, which leads to chemical erosion and to the formation of hydrogen-rich carbon layers. These layers can store a significant fraction of the total tritium fuel, which might prevent the use of these materials in future tritium devices. High-Z plasma facing materials are much more advantageous in this sense, but these advantages compete with the strong poisoning of the plasma if they enter the plasma core. New promising experiences have been obtained with high-Z wall materials in several devices, about which a survey is given in this paper and which also addresses open questions for future research and development work.

Co-permeation of deuterium and hydrogen through Pd

Experiments on permeation of deuterium through a palladium membrane, which was accompanied by co-permeation of hydrogen, were performed in the pressure range below 1 Pa. Permeation of deuterium exhibited features of surface limited regime (SLR) both with and without hydrogen co-existence. Release of D2, HD, and H2 molecules from the downstream side of the membrane was observed in presence of hydrogen. No visible blocking of deuterium permeation by hydrogen appeared. The net permeation rate of deuterons in D2 and HD forms in the co-permeation experiments with hydrogen was found to be approximately the same as in the experiments without hydrogen, if the effective deuterium pressure p(D2)+p(HD)/2 remained the same irrespective of the partial pressure of hydrogen p(H2). The experimental data were described by a steady-state permeation model, which took into account D2, HD, and H2 absorption and desorption.

Surface Distribution of Tritium on Graphite Tiles of Divertor Area in JT-60U

ABSTRACT Tritium distributions on the graphite tiles used as plasma facing tiles in divertor tiles, dome units, and the baffle plates of JT-60U were successfully measured. The highest tritium level was found at the top of the dome or the private region and the outer baffle plates, where the plasma did not hit but the distance from the plasma was the shortest. For the divertor tiles, the tritium retention was very small. Such tritium distribution observed in JT-60U tiles can be well explained by the homogeneous implantation of rather high energy tritium and thermal release due to the heat load.

Hydrogen recycling study by Balmer lines emissions in linear plasma machine TPE

We have investigated the influence of target materials and temperatures on Balmer series emission in a linear plasma apparatus, Tritium Plasma Experiment (TPE). The intensities of the Balmer series emission in front of the target were higher for heavier mass target and also for lower target temperature, showing rather linear relationship between the emission intensity and hydrogen reflection coefficient. For exothermic hydrogen occluders of Ti and Ta, the intensity ratio of Dβ/Dα increased with the target temperature markedly, whereas the intensity ratio stayed rather constant for endothermic hydrogen occluders of Ni, Cu and W. This is a clear demonstration that the target materials and temperatures modify the boundary plasma. In addition the intensity ratio Dβ/Dα is not simply a function of plasma temperature but has clear target temperature dependence.