Page 2 - Dual wavelength imaging of a scrape-off layer in an advanced beam-driven field-reversed configuration
P. 2

Bolte et al.
Rev. Sci. Instrum. 87, 11E520 (2016)
FIG. 1. Movable FIDA view (yellow) is nominally perpendicular to south- east beam 5 and tangential to fast-ion orbits. Tilting the collimator from shot to shot maps out the FIDA signal’s spatial dependence. Inset: Movable optic mount has two axes of rotation.
throughput to truly measure FIDA spectra or to rule out CX lines for the shots in question. This is left for future work. However, data-mining of survey spectra has produced promising results. Using beam modulation or termination for background subtraction, purely radial views—where any fast- neutral Doppler shift is averaged out—are found to have equal spectra before and after beam turn-o↵ (in the range 653- 655 nm) while tangential views show a rise during the beam-on phase. Both views have su cient access to the beam volume and therefore should detect CX lines, however, only the view with the Doppler shift favorable for FIDA light detected a di↵erence spectrum.
FIG. 3. Experimental FIDA signal. X0 is the x-coordinate where a line- integrated LOS passes through the y = 0 (horizontal) plane.
The same bandpass filter was used with a camera to image beam emission. The camera7 is a Sanstreak Edgertronic, typically operated using a 192 ⇥ 192 frame size at 9527 fps, mounted with a wide-angle radial view near the C-2U midplane. The camera orientation is largely toward the oncoming neutral beam 5. This produces the large Doppler shift required to shift beam emission into the filter passband while rejecting thermal halo emission. Figure 4 shows the camera image taken at 1 ms.
FIDASIM models neutral beam deposition and produces synthetic Balmer-alpha radiation and neutral particle analyzer signals. Simulated Balmer-alpha light is calculated by a collisional radiative model and is produced by neutral atoms in the form of injected beam protium neutrals, thermal halo deuterium neutrals resulting from multiple charge-exchange events between beam neutrals and thermal plasma ions, and finally from fast neutrals, which form when confined fast-ions CX with beam and/or halo neutrals.
Until recently, FIDASIM required the input of a TRANSP8-like guiding-center fast-ion distribution function, making assumptions of field uniformity that are inconsistent with an FRC field topology. Therefore, the recently released FIDASIM 1.0.06 has been developed to accept full ion-orbits and correspondingly uses local field values.
The inputs for FIDASIM for this work are produced using a mixture of experimental data and modeling values. Experimentally obtained plasma radial profiles at the midplane
FIG. 4. Experimental camera image of beam emission Doppler shifted into the passband of the filter at 1 ms (98 μs exposure time). Beam is coming toward the camera and to the left. The coordinate system shown for reference is the same as that in Fig. 1 and appears curved due to the fisheye lens distortion.
FIG. 2. Density and temperature measured for the 26 discharges used in the FIDA study as well as the values used for the simulation. Temperatures are averaged inside the separatrix at the midplane.

   1   2   3   4