[en] The key problem for the use of carbon in a future fusion device is the formation of tritium containing a-C:H layers. ASDEX Upgrade offers the possibility to investigate these layers at ITER relevant divertor conditions. Long term probes show, that the layer growth under the structure of the new divertor (Div IIb) is very similar to Div II. Additional quartz microbalance monitors offers measurements on the layers growth on a shot-to-shot base. The layers are found to grow continuously during the campaign. Using similar shots, a proportionality of the growth rate to the divertor neutral pressure is found. A Langmuir probe, installed below the divertor structure, measured a strongly variable plasma with electron densities up to 1x10¹⁸ m^-3. The density of this parasitic plasma is correlated to the divertor radiation and neutral density, which point to photo ionisation or photo emission as the origin of the plasma

[en] In present tokamaks nitrogen seeding is used to reduce the power load onto the divertor tiles. Some fraction of the seeded nitrogen reacts with hydrogen to form ammonia. The behaviour of ammonia in ASDEX Upgrade is studied by mass spectrometry. Injection without plasma shows strong absorption at the inner walls of the vessel and isotope exchange reactions. During nitrogen seeding in H-mode discharges the onset of a saturation of the nitrogen retention is observed. The residual gas consists of strongly deuterated methane and ammonia with almost equal amounts of deuterium and protium. This confirms the role of surface reactions in the ammonia formation. The results are consistent with findings in previous investigations. A numerical decomposition of mass spectra is under development and will be needed for quantitative evaluation of the results obtained

[en] The ITER fusion reactor which is under construction will use a deuterium–tritium gas mixture for operation. A fraction of this fusion fuel remains inside of the machine due to various mechanisms. The evaluation of this retention in present fusion experiments is of crucial importance to estimate the expected tritium inventory in ITER which shall be limited due to safety considerations. At ASDEX Upgrade (AUG) sufficiently time-resolved measurements should take place to extrapolate from current 10 s discharges to the at least intended 400 s ones of ITER. To achieve this, a new measurement system has been designed that enables accuracy of better than one per cent

[en] The radiation field of a small antenna immersed i a magnetized plasma is, for certain frequency bands, resonantly enhanced on the ''resonance cone'' (r.c.), which is aligned with the magnetic field, and whose apex is at the transmitting antenna. It is well known that, in a warm plasma, electron density and temperature can be deduced from the r.c. maximum angle and from an associated interference pattern. Because the maximum angle depends on the frequency, one expects an immediate influence of the Doppler effect ω'=ω-k·v_d. In the case of a field aligned electron drift, the r.c. pattern is apparently convected downstream, which leads to an upstream-downstream asymmetry of the cone-angle, whereas cross-field drifts bend the r.c. sidewards. Since this asymmetry is easily detectable, it was proposed to use this asymmetry as a diagnostic technique for measuring the electron drift velocity. However, the cited theoretical approaches result in different functional dependences as well as different coefficients for deducing the drift velocity from the measured asymmetry. Moreover, up to now only the case of field aligned drift has been treated rigorously. For exactly perpendicular drift, a cold-plasma approximation is available. For arbitrary drift orientation, we had suggested recently to use an interpolation formula, which reduces to the kinetic theory result for field-aligned drift, and uses the cold-plasma coefficients for the angular dependence. In this contribution we present experimental observations from the r.c. instrument ''COREX-1'', which investigated plasma parameters and field aligned drifts in the mid-latitude ionosphere. In addition, laboratory results are presented for a critical comparison of field-aligned drifts with the competing models. (author) 11 refs., 5 figs

[en] The growth of codeposited layers has been studied with long term samples below the divertor IIb and in a pump duct of ASDEX Upgrade from March to August 2001. The composition of redeposited layers and their optical properties were analyzed with ion beam techniques and ellipsometry. The deposition in the sub-divertor area showed a complicated deposition pattern with a maximum deposition of about 1.3 μm. All deposits form soft hydrocarbon layers which consist mainly of deuterium and carbon with D/C from 0.7 to 1.4. Only a small deposition was observed in the pump duct, with a maximum of about 2.5x10¹⁵ D-atoms/cm² at the duct entrance. The observed deposition pattern in the duct is compared with simulation calculations assuming neutral hydrocarbon radicals as precursors for film deposition. The deposition pattern can be explained by two different radical species with surface loss probabilities β<10^-3 and 0.1≤β≤0.9. The most likely species are CD₃, C₂D₅ and C₂D₃ radicals

[en] Underneath the ASDEX Upgrade divertor two different types of amorphous hydrocarbon layers were found. Transparent soft layers with a typical ratio of D/C ≅ 1 and brownish hard layers with a ratio D/C ≅ 0.5: The amount of carbon deposited at the inner divertor is a factor ≅ 3 larger than at the outer divertor. Typical 0.3% of the total deuterium input during one experimental campaign are found in these layers in the divertor and only about 8 x 10^-5 of the D input in the pumping ducts. From the deposition pattern it is concluded, that the layers are mainly build up by species with high sticking probability 0.1 < β < 0.9. Neutral collisions and a parasitic plasma below the divertor are involved in the layer formation. Discharge resolved measurements of the layer deposition reveal a proportionality of the layer growth on the neutral density below the divertor

[en] An alternative for low-Z materials in the main chamber of a future fusion device are high-Z materials, but the maximal tolerable concentration in the plasma core is restricted. A step by step approach to employ tungsten at the central column of ASDEX Upgrade was started in 1999. Meanwhile almost the whole central column is covered with tiles, which were coated by PVD with tungsten. Up to now 9000 s of plasma discharge covering all relevant scenarios were performed. Routine operation of ASDEX Upgrade was not affected by the tungsten. Typical concentrations below 10^-5 were found. The tungsten concentration is mostly connected to the transport into the core plasma, not to the tungsten erosion. It can be demonstrated, that additional central heating can eliminate the tungsten accumulation. These experiments demonstrate the compatibility of fusion plasmas with W plasma facing components under reactor relevant conditions. The erosion pattern found by post mortem analysis indicates that the main effect is ion sputtering. The main erosion of tungsten seems to occur during plasma ramp-up and ramp-down. (author)

[en] To investigate the retained D in a device different techniques are used. For the sample technique tiles are removed after a campaign and investigated using surface analysis techniques. For gas balances the amount of gas retained during a shot is measured as the difference between the puffed and pumped gas. This allows distinguishing different scenarios and wall conditions. From probe investigation, it is known that the majority of the retained deuterium is chemical bound as a:CH layers, which exist only for C first wall. AUG made a stepwise transition to full W PFCs, so carbon still remained in the vessel due to deposited layers from previous campaigns. For C PFCs the whole divertor shows carbon deposition. The majority is found at the inner divertor with 5 mg/s deposition rate. During the transition to the full W PFC deposition at the outer divertor vanishes. For full tungsten, almost no deposition is found at the divertor with the exception of the private flux region. The D inventory is mostly found in C and B dominated layers, formed by deposits remaining from earlier campaigns. For extrapolation to ITER we have to investigate the gas balance during a high density discharge in detail. Most of the puffed gas was retained during the plasma build up. For a carbon device retention of 70 mg D is needed to saturate the wall and reach steady state conditions. Due to the limited accuracy of dynamic gas balance a retention rate of 9 ± 12% of the puffed gas is observed during this phase. After the wall is saturated no additional D is retained. Comparing high density discharges with C and W wall material, first investigations show a higher dynamic retention for W materials. For highest puffing rates, no retention of D is observed after wall saturation. It seems that higher gas puff rate is needed in W to reach wall saturation, in comparison with C wall. D implantation was measured at the outer divertor strike point area. Only 50 mg was found for the outer divertor VPS coatings, whereas for polycrystalline Langmuir probes the inventory is less by a factor of 10. (author)

[en] The suitability of tungsten as plasma facing material for main chamber and baffle components was investigated in ASDEX Upgrade. In an initial step four W-coated graphite test tiles were employed at different poloidal positions of the inner column heat shield structure for a full experimental campaign. Tungsten erosion rates were determined by ion beam analysis of the test tiles before and after the campaign. The results are compared to predictions assuming CX sputtering as main erosion mechanism. In a second step the machine was operated with two toroidal rows of W-coated graphite tiles. The experimental campaign was started without wall conditioning by siliconization to ensure a clean tungsten surface exposed to the plasma. Tungsten influx and plasma concentration were measured spectroscopically. Tungsten migration was determined by ion beam analysis of collector probes exposed in the outer midplane and the outer divertor plate. In all important discharge scenarios the W-concentration stayed below the detection limit of ∼5x10^-6. Consequently no influence on plasma performance parameters was found, which encourages the use of W for even larger surface areas in the ASDEX Upgrade main chamber

[en] One of the main disadvantages of carbon as first wall material in a fusion device is the co-deposition of hydrogen with the eroded carbon. These layers will contain a significant amount of the tritium inventory of a fusion reactor. After venting brownish layers and flakes were found under the divertor structure of ASDEX Upgrade. First investigations were made on these flakes. Due to the complicated structure beyond the divertor, shadowing effects occur indicating that the brown layers are deposited by ionised particles. The flakes were analysed using SEM and ion beam techniques. Two different types of hydrocarbon layers were found: The brownish hydrogen poor layer (D/C = 0.4) and transparent hydrogen rich layer (D/C = 1) The total amount of carbon beyond the divertor could be estimated to 1.5 g, deposited in 3000 s of plasma discharges. First measurement of the layer growth using a quartz crystal microbalance instrument is presented