Plasma surface interactions - Operations

Aims and scientific program

The PSI-Operations group is responsible for construction and operation of the new linear plasma device Magnum-PSI. This device will operate in steady-state with a 3-T superconducting magnet with a 10-cm diameter plasma beam. For hydrogen plasmas, flux densities of ~10²⁴ m^-2s^-1 should be reachable at low electron temperatures (<10 eV). Basic operation of the device will start in fall 2010. This includes operation of several techniques for plasma and surface diagnostics.

The research program of the group involves technical studies relevant for construction and operation of the device and for plasma diagnostics. In addition, hydrogenic retention in refractory metals exposed to high-flux plasmas is studied; this research line will include experiments with novel plasma-facing materials.

Personnel

Research Highlights

Modeling and experiments on differential pumping in linear plasma generators operating at high gas flows 

We have used neutral gas simulations and done experiments to show that differential pumping can be used effectively in linear plasma generators operating at high gas flows. The neutral gas dynamics of the linear plasma generator Magnum-PSI has been modeled with the DSMC code developed by Bird.¹ This code was chosen because Magnum-PSI will operate in the transitional gas flow regime, with local Knudsen numbers well above 0.1. An efficient way to reach low pressures with large gas flows is differential pumping, where the vacuum vessel is divided by skimmers into separate chambers that are individually pumped. In a two stage differentially pumped system, the optimum shape and position of the first skimmer has been determined. For a good performance of the skimmer, it was found that the tip of the skimmer should be inside the low density region of the expansion since the neutral density increases in the shock region. Therefore the skimmer should be able to penetrate the shock with a minimum influence on the flow. The optimum position thereby depends on the operating conditions of the source (e.g. atomic mass number and the gas flow). The simulation results agree with experimental data obtained on the linear plasma devices Pilot-PSI and PLEXIS. The angle between the skimmer and the gas flow must be kept shallow enough as to not interfere with the expansion, but a skimmer that is too shallow will form a flow restriction for the plasma beam. The optimum inner angle of the skimmer was found to be around 53 degrees. It is shown that differential pumping works in large linear plasma generators operating in the transitional regime. In Magnum-PSI the distance between the source and the skimmer can be varied. This makes it possible to place the skimmer before the shock position in different operating conditions (e.g. gas flow, atomic mass number, background pressure). In the Magnum-PSI operating conditions, a factor 4 pressure reduction in the case of H₂ can be achieved with a two stage differential pumping scheme. This factor increases for heavier gasses (e.g. D₂ and Ar). In Magnum-PSI a 3 stage differentially pumped vacuum system will be used to keep the neutral pressure in the target chamber below 1 Pa, the limit set by the ITER relevance of PSI studies.

¹G. A. Bird, Molecular gas dynamics and the direct simulation of gas flows (Clarendon, Oxford, 1994).

Fraction of particles which travel through a 10 cm diameter sampling area with its center on axis, as a function of the distance from the source (lines) and fraction of particles crossing the opening of a 10 cm diameter skimmer at different positions (open symbols). A lower fraction indicates a higher skimmer performance.

Pressure plot of the DSMC calculation where 40 slm D₂ gas expands in a three stage differentially pumped vacuum system. Some flow lines are shown for clarity. Neutral pressure below 1 Pa in the target chamber is reached.

Hydrogenic retention in tungsten exposed to ITER divertor relevant plasma flux densities

To better understand the effect of high-flux plasmas on refractory metals, tungsten (W) targets were exposed to high density, low temperature deuterium plasmas in Pilot-PSI. This investigation measured the amount of plasma-implanted deuterium that was trapped and retained in the tungsten target for a range of plasma exposure times (4 – 160 s). The plasma conditions were similar to what is expected in the ITER divertor (n_e ~ 10²⁰ m^-3, T_e ~ 2 eV, heat load ~ MW∙m^-2) and the W target surface temperature was ~1600 K at the center of the target and decreased to ~1000 K at the edges. Deuterium retention was measured locally in the first 3 µm of the surface by nuclear reaction analysis (NRA). A 2-D NRA scan of the surface revealed significantly higher retention at the cooler edges (6 mm off center, T_w ~ 1000 K) of the target as compared to the center of the target. This indicated that surface temperature was playing a dominant role in determining hydrogenic retention properties as compared to plasma flux density or plasma fluence. Thermal desorption spectroscopy (TDS) measured the global retained deuterium inventory in the exposed W targets. TDS analysis showed very low retained fractions (10^-5-10^-7 D_retained/D_incident) and overall D inventory (D_retained = 0.5-1.5 x 10¹⁶ D). TDS also revealed that polishing the surface of the target enhances retention by a factor of ~2 while annealing the target at 1300 K for 30 minutes has little effect on hydrogenic retention under these conditions. Both TDS analysis and NRA showed no clear dependence of retained D on incident plasma fluence possibly indicating the absence of plasma-driven trap production under these exposure conditions. These results indicate that when operating at surface temperatures of 1000-1600 K, the W strikepoints of the ITER divertor will not retain significant amounts of deuterium (or tritium) due to the bombardment of the surface by the high flux of low energy plasma hydrogenic ions.

The left graph shows the results from the 2-D NRA scan of the W target exposed to D plasma for a total of 80 s (~10²² D total fluence). This clearly shows a minimum in retention at the hot center of the target and the highest retention at the cooler edges. The low retention seen at some of the 8 mm off center locations may be a shadowing effect from the target clamping ring. The right graphs show a) the total retained fluence as a function of incident fluence integrated across the entire exposed surface, and b) the retained fraction as a function of total incident fluence. The steep decrease and then flattening of the retained fraction in b) may indicate saturation in the W target.

Read more in: G.M Wright et al., J. Nuc. Mat., accepted for publication

Recent publications

‘Pre-design of magnum-PSI: A new plasma-wall interaction experiment’, H.J.N van Eck et al., Fusion Engineering and Design 82 (2007) 1878

‘Hydrogenic retention in tungsten exposed to ITER divertor relevant plasma flux densities’, G.M. Wright et al., Journal of Nuclear Materials 390–391 (2009) 610

'Modeling and experiments on differential pumping in linear plasma generators operating at high gas flows', H.J.N. van Eck et al., Journal of Applied Physics 105 (2009) 063307

‘Carbon film growth and hydrogenic retention of tungsten exposed to carbon-seeded high density deuterium plasmas’, G.M. Wright et al., Journal of Nuclear Materials 396 (2010) 176

‘Hydrogenic retention of high-Z refractory metals exposed to ITER divertor-relevant plasma conditions’, G.M.Wright et al., Nucl. Fusion 50 (2010) 055004 (9pp)

 ‘Hydrogenic retention in irradiated tungsten exposed to high-flux plasma’, G.M.Wright et al., Nucl. Fusion 50 (2010) 075006 (8pp)

'Collective Thomson scattering for ion temperature and velocity measurements on Magnum-PSI: a feasibility study', H.J. van der Meiden, Plasma Physics and Controlled Fusion, 52, 045009, March 2010