Fast waves at frequencies far above the ion cyclotron frequency and approaching the lower hybrid frequency (also called 'helicons' or ‘whistlers’) have application to off-axis current drive in tokamaks with high electron beta. The high frequency causes the whistler-like behaviour of the wave power nearly following field lines, but with a small radial component, so the waves spiral slowly towards the plasma centre. The high frequency also contributes to strong damping. Modelling predicts robust off-axis current drive with good efficiency compared to alternatives in high performance discharges in DIII-D and Fusion Nuclear Science Facility (FNSF) when the electron beta is above about 1.8%. Detailed analysis of ray behaviour shows that ray trajectories and damping are deterministic (that is, not strongly affected by plasma profiles or initial ray conditions), unlike the chaotic ray behaviour in lower frequency fast wave experiments. Current drive was found to not be sensitive to the launched value of the parallel index of refraction n‖, so wave accessibility issues can be reduced. Use of a travelling wave antenna provides a very narrow n‖spectrum, which also helps avoid accessibility problems.

No abstract

No abstract

Second harmonic 60 GHz electron cyclotron (EC) pre-ionization using low field side (LFS) X-mode launchers has been effective in producing target plasmas for startup of the DIII-D tokamak with electron densities comparable to fundamental LFS pre-ionization plasmas using the same EC system. A visible Bremsstrahlung array showed that breakdown occurred at the 2 nd harmonic resonance location and after a few milliseconds the EC driven plasma filled the entire vessel, independent of the resonant location, which was varied from near the inner wall to the center of the torus. The power threshold for ionization of ~0.4 MW was observed in DIII-D for these 2 nd harmonic preionization experiments. An orbit following code calculated that cold (0.03 eV) electrons could be heated to energies above 20 eV where ionization of the neutral deuterium gas can occur. Scaling from DIII-D to ITER indicates that ECH 2 nd harmonic pre-ionization and initial plasma formation is possible if ITER operation at reduced toroidal field is desired, but additional experiments are required to extrapolate the entire ECH start-up scenario from DIII-D to ITER.

The first observation of //-mode confinement with electron cyclotron heating as the sole auxiliary heating method has been made in divertor discharges in the DIII-D tokamak. These discharges exhibit the usual characteristics of the H mode, including improved confinement of particles and energy, when electron cyclotron heating is added at a power level above 0.7 MW. The //-mode transition is accompanied by the development of an electron temperature pedestal of 0.25 keV and a dramatic steepening of the density gradient near the separatrix.PACS numbers: 52.55.Fa, 52.50.GjThe plasma energy-confinement time in tokamaks has generally been found to decrease when auxiliary power is added to the Ohmic heating. This phenomenon, L-mode scaling,' makes it difficult to heat a tokamak plasma to reactor temperatures. However, a regime of improved confinement, ^-mode scaling, has been observed in several tokamaks with neutral-beam injection (NBI) heating. ^"^ In the H mode, the energy confinement time may remain at the Ohmic level without degradation when auxiliary power is added, or the confinement time may degrade with power but remain a factor of about 2 larger than the confinement time under the L-mode scaling, as in the JET tokamak.^ Recently, the H mode has also been obtained with ion cyclotron heating,^ and, in the JFT-2M tokamak, with ion cyclotron heating,^ with a combination of NBI and electron cyclotron heating (ECH),^^ and with NBI in limiter discharges. ^^ The present work reports the first observation of the H mode with electron cyclotron heating as the sole source of auxiliary heating.These experiments have been carried out in the DIII-D tokamak operated in the expanded boundary divertor configuration under conditions which have facilitated the //-mode transition in discharges heated by NBI.^'^^ Typical plasmas had Ro = \.6S m, a-0.62 m, elongation = 1.8, with a single null divertor located in the direction of the ion \B drift, as preferred for the H mode. ^^ The gap between the separatrix flux surface and the nearest material wall was typically 3.5 cm. The plasma was developed from deuterium gas, and the plasma current was 0.48 MA.The ECH power was applied at a frequency of 60 GHz, for which the second-harmonic resonance occurs at a local magnetic field of 1.07 T. The antennas were located near the outside midplane of the plasma, and launched 85% of the incident power in the extraordinary mode and 15% in the ordinary mode. The total incident power, after allowance for waveguide losses, was up to 0.9 MW. The antennas launch power in the HEn mode at an angle ±17° to the radial with a Gaussian dispersion 11 ° in half width. When the resonance is at the plasma center, for the 17° launch angle, the local cutoff density for the extraordinary mode is 1.86x10^^ m"^, and for densities well below cutoff", the power is deposited in a volume with approximate minor radius of 20 cm in the midplane. The ECH power was typically a factor of 8 larger than the Ohmic power during the heating, and so the plasma power balance was domi...

To realize a compact spherical tokamak (ST) reactor, operation without the central solenoid (CS) must be demonstrated. In particular plasma current (I p) ramp-up from zero to a level required for fusion burn is crucial. Plasma initiation and I p ramp-up in ST by waves in the lower-hybrid (LH) frequency range were demonstrated for the first time on TST-2. A combline antenna was used to inject RF power of ~ 100 kW at 200 MHz. Formation of a low current (~ 1kA, mainly driven by pressure gradient) ST configuration can be accomplished by waves over a broad frequency range (21 MHz to 8.2 GHz in TST-2), but further I p ramp-up (to ~ 10 kA, mainly driven by RF) is most efficient with uni-directional traveling waves in the LH frequency range. I p ramp-up to 15 kA was achieved with 60 kW of net RF power. Soft X-ray emission in the direction of electron acceleration by RF wave was enhanced more strongly in the co-drive case (acceleration in the direction to increase I p) compared to the counter-drive case. Hard X-ray spectral measurements showed that the photon flux is an order of magnitude higher and the photon temperature is higher in the co-current-drive direction than in the counter-current-drive direction. These observations are consistent with acceleration of electrons by a unidirectional RF wave. The combline antenna excites vertical electric fields which match the polarization of the fast wave (FW). There is evidence that the LH wave (or the slow wave, SW) is excited nonlinearly, based on the frequency spectra measured by magnetic probes in the plasma edge region. The time evolution indicates the tendency of the pump wave to weaken when the sideband waves intensify. It is expected that the effectiveness of current drive would improve if the LH wave could be excited directly by the antenna. Two types of travelingwave LH antennas will be tested on TST-2, a dielectric-loaded waveguide array ("grill") antenna, and an array of capacitively coupled elements with the electric field polarized in the toroidal direction. During initial operation of the grill antenna, wavenumber components were measured by an array of magnetic probes. Results were qualitatively consistent with expectations based on dispersion relations for the FW and the SW.

Spherical tokamaks (STs) have the advantage of superior stability at high beta, but to realize a compact fusion reactor, the central solenoid (CS) must be eliminated. The plasma current could be maintained mainly by the self-driven current during the steady-state burning phase, but there is no established method of non-inductive current ramp-up from zero to a high enough level required for fusion burn. The lower-hybrid wave (LHW) is commonly used in tokamaks, but this technique is believed to be unusable in ST plasmas with very high dielectric constant. Wave excitation, propagation and damping were evaluated by numerical modeling for the TST-2 spherical tokamak (R = 0.68 m, a = 0.25 m, B = 0.3 T, I = 100 kA). Wave excitation is calculated by the RF antenna simulation tool based on the finite element method (COMSOL). The plasma is modeled as a medium with cold plasma dielectric response and artificially enhanced loss. Excitation of a traveling fast wave (FW) by the TST-2 combline antenna was confirmed by COMSOL calculation. However, the FW must be mode converted to the LHW to achieve efficient current drive. The TORLH full-wave solver, which can treat diffraction effects properly, was applied to the TST-2 LH current drive experiment. It is shown that core current drive by LHW is possible in the low density, low current plasma formed by ECH. It is important to keep the density low during current ramp-up, and the wavenumber must be reduced as the current increases in order to maintain core current drive. The results of these calculations and the initial experiment on TST-2 will be used to design an optimized LHW antenna with appropriate polarization and wavenumber spectrum controllability for the current ramp-up experiment. Taking a conservative value for the current drive figure of merit, the steadystate driven current by 200 kW of RF power is estimated to be 150 kA, which should be adequate for evaluating the usefulness of this technique up to a current level of 100 kA in TST-2. This technique should be usable during the low-density, low-current, non-burning start-up phase of a reactor to reach a sufficient current level needed for further heating. Initial results from the TST-2 experiment demonstrated that initial current formation and ST equilibrium formation can be achieved by RF power in the LH frequency range. Further current ramp-up by directly excited LHW will be investigated on TST-2. The success of the TST-2 experiment would provide a scientific basis for quantitatively evaluating the required CS capability for a low aspect ratio reactor.

No abstract