J.D. King

An ion Doppler spectrometer instrument for ion temperature and flow measurements on SSPX.

A high-resolution ion Doppler spectrometer (IDS) has been installed on the sustained spheromak plasma experiment to measure ion temperatures and plasma flow. The system is composed of a 1 m focal length Czerny-Turner spectrometer with a diffraction grating line density of 2400 lines/mm, which allows for first order spectra between 300 and 600 nm. A 16-channel photomultiplier tube detection assembly combined with output coupling optics provides a spectral resolution of 0.0126 nm/channel. We calculate in some detail the mapping of curved slit images onto the linear detector array elements. This is important in determining the wavelength resolution and setting the optimum vertical extent of the slit. Also, because of the small wavelength window of the IDS, a miniature fiber-optic survey spectrometer sensitive to a wavelength range 200-1100 nm and having a resolution of 0.2 nm is used to obtain a time-integrated spectrum for each shot to verify specific impurity line radiation. Several measurements validate the systems operation. Doppler broadening of C III 464.72 nm line in the plasma shows time-resolved ion temperatures up to 250 eV for hydrogen discharges, which is consistent with neutral particle energy analyzer measurements. Flow measurements show a sub-Alfvenic plasma flow ranging from 5 to 45 kms for helium discharges.

Feedback-assisted extension of the tokamak operating space to low safety factora)

Recent DIII-D and RFX-mod experiments have demonstrated stable tokamak operation at very low values of the edge safety factor q(a) near and below 2. The onset of n = 1 resistive wall mode (RWM) kink instabilities leads to a disruptive stability limit, encountered at q(a) = 2 (limiter plasmas) and q95 = 2 (divertor plasmas). However, passively stable operation can be attained for q(a) and q95 values as low as 2.2. RWM damping in the q(a) = 2 regime was measured using active MHD spectroscopy. Although consistent with theoretical predictions, the amplitude of the damped response does not increase significantly as the q(a) = 2 limit is approached, in contrast with damping measurements made approaching the pressure-driven RWM limit. Applying proportional gain magnetic feedback control of the n = 1 modes has resulted in stabilized operation with q95 values reaching as low as 1.9 in DIII-D and q(a) reaching 1.55 in RFX-mod. In addition to being consistent with the q(a) = 2 external kink mode stability limit, the...

Hybrid-like 2/1 flux-pumping and magnetic island evolution due to edge localized mode-neoclassical tearing mode coupling in DIII-D

Direct analysis of internal magnetic field pitch angles measured using the motional Stark effect diagnostic shows m/n=2/1 neoclassical tearing modes exhibit stronger poloidal magnetic flux-pumping than typical hybrids containing m/n=3/2 modes. This flux-pumping causes the avoidance of sawteeth, and is present during partial electron cyclotron current drive suppression of the tearing mode. This finding could lead to hybrid discharges with higher normalized fusion performance at lower q95. The degree of edge localized mode-neoclassical tearing mode (ELM-NTM) coupling and the strength of flux-pumping increase with beta and the proximity of the modes to the ELMing pedestal. Flux-pumping appears independent of magnetic island width. Individual ELM-NTM coupling events show a rapid timescale drop in the island width followed by a resistive recovery that is successfully modeled using the modified Rutherford equation. The fast transient drop in island width increases with ELM size.

A digital lock-in upgrade of the motional Stark effect diagnostic on DIII-D.

The use of lock-in amplifiers for phase sensitive detection of motional Stark effect (MSE) diagnostic signals is of critical importance to real-time internal current profile measurements in tokamak plasmas. A digital lock-in (DLI) upgrade utilizing field programable gate array firmware has been installed on the MSE system of the DIII-D tokamak for the eventual replacement of largely obsolete analog units. While the new digital system has shown a small reduction in electronic noise over the analog, the main advantages are reduced cost, hardware simplicity, compact size, and phase tracking during plasma operations. DLI recovery of MSE polarization angles was accomplished through use of reference processing to produce only photoelastic modulator (PEM) second harmonic frequencies and electronic signal processing to maximize the fidelity of the recovered signal. A simplified discrete analytical solution was found that accurately describes the new DLI hardware. The DLI algorithm was found to cause a prohibitively large oscillating artifact atop the demodulated signal. The artifact was caused by the accumulator interval not containing an exact integer number of PEM multiplier periods. Successful MSE measurements require the minimization of this oscillating artifact amplitude. The analytical solution was used to select an appropriate accumulator interval that both reduces the artifact and maintains the greatest temporal resolution possible. Sample EFIT equilibria reconstructions and corresponding safety factor profiles showed very close agreement between the analog and digital lock-ins.

Magnetohydrodynamic interference with the edge pedestal motional Stark effect diagnostic on DIII-D.

Accurate measurement of internal magnetic field direction using motional Stark effect (MSE) polarimetry in the edge pedestal is desired for nearly all tokamak scenario work. A newly installed 500 kHz 32-channel digitizer on the MSE diagnostic of DIII-D allows full spectral information of the polarimeter signal to be recovered for the first time. Fourier analysis of this data has revealed magnetohydrodynamic (MHD) fluctuations in the plasma edge pedestal at ρ ≥ 0.92. By correlating edge localized mode fluctuations seen on lock-in amplifier outputs with MSE spectrograms, it has been shown that edge pedestal tearing mode fluctuations cause interference with MSE second harmonic instrument frequencies. This interference results in unrecoverable errors in the real-time polarization angle measurement that are more than an order of magnitude larger than typical polarimeter uncertainties. These errors can cause as much as a 38% difference in local q. By using a redundant measure of the linear polarization found at the fourth harmonic photo-elastic modulator (PEM) frequency, MHD interference can be avoided. However, because of poorer signal-to-noise the fourth harmonic signal computed polarization angle shows no improvement over the MHD polluted second harmonics. MHD interference could be avoided in future edge pedestal tokamak polarimeters by utilizing PEMs with higher fundamental frequencies and a greater separation between their frequencies.