Atribuciones de los jueces constitucionales de primer nivel

To study the flow structure and possible asymmetries on a flux surface, an array of CXRS diagnostics viewing the high-field side (HFS) of AUG has been installed. In contrast to the LFS CXRS diagnostics, which use a heating beam that provides the neutrals for the CX reactions, the CXRS diagnostics at the HFS utilize a deuterium (D) gas puff and collect the light, that is emitted after the impurity undergoes a CX reaction with a thermal D particle. Due to the penetration of the thermal D gas puff, the HFS measurements are restricted to the outermost region of the plasma.

The HFS array consists of a toroidal [108] and a poloidal view (see figure 4.5) and each view is equipped with two f /4 optical heads. For the toroidal system, one optical head views the gas puff directly, while the other views the background plasma parallel to the ‘active’ optical head (_∼10 cm above the ‘active’ view) and thus, collects the corresponding background spectra. Both optical heads are installed at the LFS and view through the plasma edge at the outer midplane, intersect either the diagnostic gas puff at the HFS or the plasma edge at the inner wall for the ‘background’ view, respectively. Then the views pass through the plasma edge at the other side (again LFS). Hence, the ‘background’ optical head collects passive emission from three separate regions along the LOS, while the ‘active’ optical head collects additional emission from the HFS region when the gas puff is switched on. Similarly, the poloidal system has one optical head viewing the gas puff directly, while the second views the background plasma at the same poloidal location, but at a different toroidal location (sector 3, see figure 4.5) to avoid the active CX signal. These are both installed at the HFS and thus, collect light from the plasma edge at the inner wall.

Due to the installation of two optical heads for each view, the background signals, i.e. the passive emission, can be subtracted from the active spectra. One issue associated with the

-2 -1 0 1 2 x[m] 0.0 0.5 1.0 1.5 2.0 2.5 y[m] 15 14 16 1 2 3 4 5 gas puff NBI #3 tor/polLFS CXRS tor/polHFS CXRS

Figure 4.5: Top down view of AUG showing the collection regions (crosses) of the edge CXRS diagnostics. The optical heads are marked by rectangles.

4.2 CXRS measurements at the high-field side of AUG 39 494 495 496 Spectral radiance [10 ph/m /sr/nm/s] 17 2 1.4 1.2 1.0 0.8 0.6 0.4 0.2

0.0 LOS 4 LOS 5 LOS 6

LOS 3 LOS 2 LOS 1 BKG view ACT view #28093, B5+ 494 495 496 494 495 496 1.0 0.8 0.6 0.4 0.2 0.0 λ_[nm] λ_[nm] λ_[nm] BV (n=7 6) BV (n=11 8) fit

Figure 4.6: Relative calibration: Spectra obtained with the active (ACT) and background (BKG) toroidal HFS diagnostic shortly before the gas puff is applied.

background subtraction is the emission from molecular D2, which is injected from the HFS,

and might disturb the fitting procedure. However, the molecules radiate in a very limited radial range in the near scrape-off layer. In the present work, the analysis is limited to the edge of the confined plasma where the D2molecules are already disassociated.

Each optical head is equipped with six LOS allowing a radial profile to be measured. In the focal plane the spot size of the LOS is 8.8 mm for the toroidal HFS system and 5 mm for the poloidal HFS diagnostic, while the channel spacing is 1.25 cm. The radial resolution can be increased by applying the radial plasma sweep technique (see section 3.2) and moving the plasma position towards the inner wall. All optical heads are aligned such that they view the plasma as tangential to the flux surfaces as possible. The collected light is transmitted to a high throughput f /4 Czerny-Turner spectrometer which has identical properties as the one employed for the toroidal edge CXRS system at the LFS.

To obtain measurements at the HFS thermal D is puffed from the inner wall for a certain time period. A few milliseconds before the gas puff is applied, a refinement of the calibration on the background channels relative to the active channels is performed. Relative calibration factors with respect to the corresponding active channels are calculated. Using this method the relative brightnesses of the active and background views, which arise due to imperfections of the calibration, are taken into account.

Figure 4.6 shows the measured spectra from the active and the background view, shortly before thermal D is injected. The spectra are fitted by a Gaussian function (solid line in figure 4.6) and the fitting region is restricted to the B5+(𝑛 = 7_→6) spectral line. The relative calibration factors for each LOS are calculated from the ratio of the heights of the fitted Gaussians of the active and background LOS. For the example given in figure 4.6 the calibration factors for the different LOS vary between 0.67 and 1.05. To isolate the active CX line, the background spectra measured during the gas puff are multiplied by the corresponding calibration factors and subtracted from the active spectra. The Doppler shift and Doppler width of the measured active spectral line give information on the rotation velocity and on the ion temperature. The direct evaluation of the HFS impurity density profile from the measured radiance of the spectral

40 4. The edge CXRS diagnostics at ASDEX Upgrade

line is not yet available at AUG, as a neutral density measurement after the application of the gas puff is missing.

Corrections to the measurements due to the CX cross-section effects [109] (see section 4.5.1) are negligible since at thermal energies the cross-section has a weak dependence on the energy and hence, the temperature dependence of the effective CX reaction rate is small. The HFS measurements, combined with the data obtained from the LFS diagnostics, enables localized CXRS measurements at two different poloidal locations (see figure 4.1(a)) on a flux surface and allows for studying possible asymmetries on a flux surface (see chapter 5).