R.T. McGrath

Erosion/redeposition modeling and calculations for carbon

Abstract Several key aspects of carbon erosion/redeposition in fusion devices have been examined. These include the effects of an oblique magnetic field geometry plasma sheath on ion impingement and sputtering, and the effect of surface hydrogen buildup on sputtering. The calculations use a version of the MEASTRI hydrogen implantation code and a cloud-in-cell type sheath code. Time dependent erosion calculations have been made for the TFTR bumper limiter, using the REDEP erosion/redeposition code with updated sheath and sputtering models. The computed plasma carbon content during a TFTR limiter “deconditioning” run and for a “supershot” beam heating case is similar to measured values. The calculations predict that depletion of the near surface graphite hydrogen concentration enhances the self and hydrogen sputtering coefficients. This mechanism provides an explanation for the larger Z- effective values observed on TFTR immediately after limiter conditioning.

ALT-II toroidal belt pump limiter performance in TEXTOR☆

The Advanced Limiter Test (ALT-II) is a toroidal belt pump limiter in the TEXTOR tokamak. ALT-II is composed of 8 blade segments which form an axisymmetric toroidal belt of 3.4 m2 exposed surface area, located on the outside of the torus at 45° below the horizontal midplane. Ohmic plasma operation with ALT-II as the main limiter is characterized by a line-averaged density range of 5 × 1012 to 5.5 × 1013cm−3 at BT = 2 T and IP = 340 kA, Zeff = 1.1 to 2 and typically 40 to 95% of the power radiated depending on the plasma density. ICRH heating of the plasma with up to 2.6 MW of incident power has been achieved, which modifies the scrape-off layer (SOL) and the pump limiter performance. The recycling coefficient in TEXTOR is normally close to one, but helium RG conditioning and baking of the limiter at 400 ° C is found to lower the recycling coefficient to 0.8 for the order of 10 shots. Measurements by arrays of probes in the SOL and thermocouples in the limiter tiles indicate the flow to the limiter is toroidally symmetric (taking field ripple into account) and poloidally asymmetric. The asymmetries result in different power and particle fluxes to the ion and electron drift sides of the limiter. The density and power scrape-off lengths are on the order of 1 cm and significantly longer on the outside of the torus (electron drift side). In spite of the flow asymmetry favoring the ion drift side near the tangency point, the longer e-folding lengths on the electron side in the SOL result in equal or higher particle collection by the electron side. The probe arrays indicate that during ohmic heating a total of 15 to 20% of the core efflux is incident on the neutralizer plates located in scoops beneath the blades. More particles are collected during ICRH auxiliary heating due to changes in the SOL profiles and shorter particle confinement times. Based on particle re moval experiments with pumping on one blade, the total exhaust efficiency of the limiter if pumped at all eight blades is 5 to 10%.

Impurity deposition on ALT-I limiter heads☆

Abstract Detailed spatial profiles of impurities deposited on the surface of the ALT-I pumped limiter heads used in TEXTOR have been measured using proton induced X-ray emission in the external ion beam analysis system. The TiC coated, ATJ graphite ALT-I limiter head was analyzed after each of two plasma exposure periods. Measurements were also made on the uncoated ATJ limiter head. For limiter operation at plasma radii ranging from 40 to 46 cm we find large accumulations of metal impurities on the ion and electron sides of the limiter. The composition implies that the source of these impurities is predominantly the stainless steel main limiters in TEXTOR. Surface concentrations of iron as high as 10 18 /cm 2 have been measured. On each of the limiter front faces a relative maximum in metal concentration occurs near the tangency point and relative minima are found near the toroidal extremities. An analysis of impurity transport in the edge region of TEXTOR using Redep is also presented. The calculations predict impurity atom migration patterns that strongly resemble the measured metal atom deposition profiles. The analysis provides considerable insight into impurity transport in the tokamak edge. Benchmark calculations of this type are valuable for the design of future limiter or divertor systems.

E×B transport in the DIII-D boundary plasma

We have measured the electrostatic turbulence and associated particle transport in the DIII-D boundary plasma using a fast reciprocating Langmuir probe array located on the outboard midplane. Both the normalized rms fluctuation levels (density and floating potential) and the fluctuation-driven particle transport are altered by the L-H transition in the SOL. At the separatrix, the density fluctuation level is reduced a factor of 2, consistent with reflectometry results. There is a corresponding decrease in the turbulent particle flux. Deeper in the SOL, the turbulent particle transport in H-mode exceeds the L-mode value. The perpendicular diffusion coefficient D ⊥ and particle confinement time τ p have been estimated, assuming that the transport is purely turbulent and uniform on a flux surface. We find D ⊥ =0.7 D B (L) and 0.04 D B (ELM-free H), and τ p =54 ms (L) and 480 ms (ELM-free H).

Scrape-off layer measurements in DIII-D

In this paper, scrape-off layer measurements in DIII-D are presented as a function of the main discharge plasma parameters. A systematic study is under way to understand and predict the behavior of the edge and divertor plasma in DIII-D and this scaling behavior will be crucial for the design of ITER. To facilitate the studies, a fast reciprocating Langmuir probe incorporating five graphite tips was installed at the midplane of DIII-D which has the capability of performing multiple plunges 1 cm inside the separatrix during 5 MW of NBI. Density and temperature profiles in the midplane (reciprocating probe), near the top (Thomson scattering) and at the lower divertor plate (fixed Langmuir probe array) are compared by mapping the measurements into magnetic flux coordinates. Local pressure measurements are compared on different parts of a flux surface. The three different local measurements also indicate the spatial evolution of plasma conditions as plasma approaches the divertor plate. Ohmic and L-mode discharges exhibit similar (exponential) density and temperature decay in the scrape-off layer. H-mode discharges, however, display a faster spatial decay reflecting at least a factor of 3 decrease in the perpendicular diffusion coefficient. Consistency of the magnitude and scaling behavior of the edge profile parameters with models of the scrape-off layer is examined.

Transport of sputtered impurities in the vicinity of the TFTR inner bumper limiter

Abstract The generation and transport of metallic and carbon impurities on the inner bumper limiter of TFTR have been investigated in detail. The spatial profiles for impurity deposition and deposited layer thickness predicted by the analytical model are compared to the extensive postplasma exposure measurements of surface conditions that have been made on the TFTR inner bumper tiles. Agreement between the analytical model and the experimental data is good. The work provides considerable insight into the dynamics of particle transport in the edge region.

Engineering and materials issues in designing a cold-gas divertor

Abstract One of the key challenges facing the International Thermonuclear Experimental Reactor (ITER) Project is the development of plasma-facing components (PFCs) that can withstand the severe environmental conditions at the plasma edge. The most intensely loaded element of the PFCs is the divertor. The divertor must handle high fluxes of energetic plasma particles and electromagnetic radiation without excessive impurity buildup in the plasma core. The “cold-plasma-target” mode of divertor operation proposed for ITER expands the divertor design window to include several alternate heat sink and armor materials that were not available for the previous “high recycling divertor” approach. In particular, beryllium armor can now be considered with copper, niobium and vanadium heat sink materials; and helium or liquid metal coolants are feasible in addition to water. This paper presents material properties and compatability assessments for these materials and coolants along with parametric studies of thermal and mechanical performance. A viable design window is found for copper and niobium heat sinks with beryllium armor, but not for vanadium unless thin (∼ 1 mm) coolant structures can be accomodated mechanically.

Thermographic and power deposition measurements on ALT-II blades

Abstract On the ALT-II toroidal belt pump-limiter in TEXTOR thermographic and thermocouple measurements have been performed. The heat distribution on the blades shows some inhomogeneities which are attributed to different effects: (a) The leading edges at the ends of the blades are heated more than the rest, partly due to alignment errors, (b) poloidally there is an asymmetry because of differences in the directed particle fluxes, (c) the magnetic field ripple causes a modulation of the power flux and (d) finally the toroidal flatness of each tile generates non-uniformities toroidally. From the observed data the total energy flux to the limiter, the time resolved power flux to the limiter and the radial power decay length are derived.

Plasma influence on throat conductance of the textor pump limiter ALT-I

Abstract On the TEXTOR pump limiter ALT-I conductance measurements for the backstreaming of gas from the pump limiter vessel through the pump limiter entrance have been performed. In these experiments neutral gas has been injected into the pump limiter plenum during a short pulse. The influence of the instreaming plasma results in a reduction of the conductance of the outstreaming gas. For helium the conductance is reduced to about 40% of the molecular conductance when a plasma flux of 0.8 A / cm 2 ( T e = T i = 11 eV ) is streaming into the pump limiter throat. The reduction of the conductance for backstreaming hydrogen and deuterium under the same plasma conditions is smaller; about 70% of the molecular conductance is obtained. This reduction can be explained by an increased recycling of ions which have been produced in the throat back to the neutralizer plate. The experimental results can be reproduced by Monte Carlo neutral transport code calculations if the recycling coefficient is about 0.85 for hydrogen and deuterium and about 0.95 for helium ions. Processes causing these high recycling coefficients are discussed and their influence is estimated.

TFTR tritium inventory analysis

Abstract The Tokamak Fusion Test Reactor (TFTR) is scheduled to begin DT operation in 1990 with the on-site tritium inventory limited to 5 grams. The physics and chemistry of the in-vessel tritium inventory will impact safety concerns, and also the entire operating schedule of the tokamak. We have investigated plasma-material interaction processes that will affect this first tritium-fueled tokamak. Tritium inventory estimates for TFTR are derived from: (1) laboratory simulation, (2) in-situ plasma measurements, (3) post-run surface analysis, and (4) modeling. This paper presents the results of these investigations, the derivation of a tritium inventory estimate and its uncertainties, and a discussion of its impact. A particular discharge-by-discharge operating schedule has been developed and evaluated. The major source of in-vessel tritium inventory will be codeposition of tritium and eroded carbon onto surfaces. We find that the on-site limit may be approached unless specific inventory reduction techniques are invoked, e.g., discharge cleaning.