Page 2 - High sensitivity far infrared laser diagnostics for the C-2U advanced beam-driven field-reversed configuration plasmas

11E128-2 Beall, Deng, and Gota
Rev. Sci. Instrum. 87, 11E128 (2016)
FIG.1. DensityresolutiondataoftheCO2systembefore(a)andafter(b)the common mode noise subtraction, as well as that of the FIR system magnified by 100 (c) for comparison.
other hand, low noise proprietary FIR lasers8 are used for the FIR interferometer, and superior resolution is achieved, as shown in Fig. 1(c). Even without a 2-color system to subtract system vibrations in the FIR, we still see a 70-fold decrease in noise, allowing for density resolution down to 1 ⇥ 1016 m 3. This exceeds the 40-fold increase in wavelength from the 10.6 μm CO2 to the 432.8 μm FIR. The system is also significantly less susceptible to contaminating noise from nearby high-power systems firing, which are visible in the CO2. As a result, the system can maintain a spectral bandwidth better than 1 MHz.
III. IMPROVED DENSITY PROFILE MEASUREMENTS
By placing the new chords above the machine axis in the midplane, the FIR system solves the problem of incomplete spatial sampling with the two-color CO2/HeNe interferometer alone. Full spatial sampling of the plasma core is achieved with CO2/FIR chord coverage from 35 to +45 cm in the confinement vessel, which has a wall radius of 70 cm. The system does not have a complete sampling of the very low density Scrape-O↵ Layer (SOL) region outside of the FRC separatrix but does thoroughly cover the FRC (typical radius 40 cm or less) even in the event of large (20–30 cm) plasma radial displacement by n = 1 wobble motion.
With improved interferometer coverage, accuracy of den- sity profiles from Abel inversion is significantly improved. To produce radial density profiles, the measured line-integrated densities are interpolated using Reinsch smoothing splines9 and then passed through an inverse Abel transform.10 The use of smoothing splines allows for density profile analysis without presumption of the plasma structure. An example is shown in Fig. 2. In Fig. 2(a) the plasma is shifted upwards, and the CO2 interferometer only samples the plasma gradient region. In this case, the FIR interferometer data are essential in providing the core plasma profile information. Shown in Fig. 2(b) is the opposite case when the plasma is shifted downwards, and the FIR data provide the plasma gradient data. In addition, the error bars of the FIR data are much smaller so that they are weighted much more than the CO2 data in the smoothing spline interpolation. This helps weed
FIG.2. SmoothingsplinefitoftheCO2(squares)andFIR(circles)measure- ments for plasmas (a) moving up and (b) moving down.
out possible excess errors in the CO2 interferometer data and consequently leads to further improvement in density profile measurements.
Once the FIR chords have been included, and full spatial data are available, centroid fitting can be used to calculate the vertical center position of the plasma, as indicated by the vertical dashed lines in Fig. 2. This new method is not a↵ected by the neutral beams, allowing it to be used as a cross-check of the bolometer data. In addition, the plasma position data from interferometer measurements allow for observations of the n = 1 wobble mode.1 One of the most common and easily identified instabilities of the C-2U plasma, the wobble mode involves the FRC revolving around the machine axis as a single mass. While potentially destructive, the bulk motion of the plasma is fairly well stabilized by line-tying of the end-on plasma guns located inside each divertor. Generally, it is often seen when the plasma guns are not used or weakly coupled; or, as seen in Fig. 3, as a small o↵set is generated during the initial plasma merging process around t = 0.1 ms. Due to minor asymmetries of the plasma formation process, some FRCs are formed and translated slightly o↵-center of the machine and begin to revolve before the mode is damped. In Fig. 3, only the early phase of a long lifetime high performance FRC (HPF) discharge is plotted to show the stabilization of the wobble mode by plasma guns and neutral beams. Clearly, the FIR/CO2 density traces are peaking when the plasma goes up/down. It should be noted that the derived plasma position information
FIG. 3. CO2 (a) and FIR (b) interferometer time traces and plasma center position (c) derived from the interferometer revealing the stabilization of the n = 1 wobble mode in the early phase of C-2U HPF discharge.