Session GO4: DIII-D Tokamak

Overview of Recent DIII-D Experimental Results

Recent DIII-D experiments have addressed critical ITER issues and improved the physics basis for steady state operation, making use of new diagnostic and hardware capabilities. The support mechanism of one beamline has been modified to allow off axis neutral beam injection (NBI) up to 16.5 deg from the horizontal. Off-axis NBI shows clear off-axis deposition and current drive at $\rho \sim 0.5$, consistent with calculations, and experiments to optimize the $q$-profile for steady-state operation, will be discussed. Progress in other major areas of research will be presented: understanding transport induced by 3-D field perturbations; determining reliable alternative techniques for ELM control that do not require internal coils; systematic tests of existing models of pedestal structure through comparisons with experimental results; evaluation of profile stiffness and critical gradients in the electron transport; improved control and dissipation of runaway electrons; development of plasma scenarios with low torque or rotation; improved characterization and control of sawteeth and neoclassical tearing modes.   

Initial Off-Axis Neutral Beam Checkout and Physics Experiments on DIII-D

Two of the eight neutral beam sources on DIII-D have been modified to allow vertical steering, with the injection angle varying from horizontal to downward at an angle of 16.5 degrees for off-axis deposition. Initial experiments to assess the basic beam functionality, geometry, and confinement were carried out. D$_\alpha$ images of beam into gas and plasma yield beam neutral profiles and are key in assessing beam shape and clipping. Neutron and fast-ion D$_\alpha$ (FIDA) diagnostics verify classical behavior of the off-axis beam ions in MHD-quiescent conditions. An initial physics experiment takes advantage of the downward steered beams to vary the fast-ion gradient $\nabla\beta_f$ from centrally peaked to hollow. Systematic scans determine the stability and impact of reversed shear Alfv\'en eigenmodes and toroidal Alfv\'en eigenmodes as a function of $\nabla\beta_f$.   

Off-Axis NBCD Experiments in DIII-D

Experiments on off-axis neutral beam current drive (NBCD) in DIII-D have clearly demonstrated off-axis NBCD using the new tilted beamline. The local NBCD and beam ion density profiles were measured in H-mode plasmas under a range of beam injection and discharge conditions, including on/off-axis, parallel/perpendicular injections, beam energy, injection power, toroidal field direction, plasma beta, and ratio of beam energy to electron temperature. For the off-axis injection, the magnetic pitch angles measured by the motional Stark effect diagnostic show clear evidence of off-axis NBCD when compared with the on-axis injection at the same electron temperature and density. The beam-stored energy estimated by equilibrium reconstruction, neutron, and fast-ion D$_\alpha$ data indicate no large anomalous losses of NBCD and fast ions. The measurements are compared with the classical model calculation using NUBEAM for validation of the off-axis NBCD physics.   

High $q_{min}$ Steady State Scenario Development Using Off-axis Neutral Beam Injection on DIII-D

Initial high power DIII-D experiments using off-axis neutral beam injection have produced plasmas with broader pressure and current density profiles and higher core safety factor than in similar plasmas employing only on-axis NBI. Such changes are expected to increase the ideal $\beta_N$ stability limits, avoid 3/2 and 2/1 tearing modes when $q_{min}>2$, thus enabling access to a high $\beta_N$, high bootstrap fraction steady state scenario. The maximum achieved $\beta_N$ in $q_{min}>2$ plasmas using off-axis neutral beam injection and the calculated resistive and ideal stability limits will be shown, as well as the transport properties and noninductive current drive fraction. Predictive modeling of the equilibrium profiles, stability and noninductive current will assess the need for additional auxiliary current drive power and flexibility.   

Testing Pedestal Models for Joint Research Target on DIII-D

As part of the FY2011 DOE Joint Research Target on Pedestal Physics, tests are being performed on DIII-D for gyrokinetic modes that have been proposed as physics mechanisms for controlling the H-mode pedestal structure. These modes include kinetic ballooning modes (KBM), candidates for limiting the total pedestal pressure gradient, ion temperature gradient modes (ITGM), candidates for limiting the $T_i$ gradient at the pedestal top and electron temperature gradient modes (ETGM), candidates for limiting the $T_e$ gradient both in the pedestal and on top of the pedestal. The theoretical control parameters for these modes are $\alpha_{MHD}$ for KBM, $\eta_e$ for ETGM and $\eta_i$ for ITGM. Experiments are being performed in DIII-D to determine if measured values of these parameters are close to the theoretical threshold values for the linear onset of these modes.   

The Dynamics of Turbulence and Shear Flow Approaching the L-H Transition

Comprehensive 2D turbulence and turbulent flow measurements have been obtained before, during, and after the L-H transition during an ion gyro-radius scan in DIII-D to understand if the strong threshold dependence on $B_T$ is connected to turbulence behavior. Other non-dimensional parameters ($\nu^*$, $q_{95}$, $\beta$) were kept nearly constant at the pedestal top. The amplitude of long wavelength density fluctuations, measured with the 2D BES array, is found to scale approximately with $\rho^*$. A mean shear flow layer is observed near $r/a \sim 0.92$ with a shearing rate exceeding the local turbulence decorrelation rate. Velocimetry shows that the GAM, which peaks near $r/a\sim 0.9$, appears a few hundred ms before the L-H transition, and decays in amplitude approaching the transition, while a lower-frequency flow structure increases in amplitude during this period. New measurements of the density dependence of the turbulence-zonal flow system will also be presented.   

Effects of Resonant Magnetic Field Perturbations on Density Profiles, Particle Transport, and Turbulence in \hbox{DIII-D}

First direct measurements of an increase in particle diffusivities and reduction in pinch velocities in DIII-D plasmas with applied resonant magnetic field perturbations (RMP) is presented for both L- and \hbox{H-mode} plasmas. Modulated gas puff combined with high-resolution profile reflectometry techniques are used and the results confirm particle transport enhancement with RMP. In H-mode, turbulence levels (measured by Doppler backscattering and BES) increase substantially with high I-coil current ($>$ 4 kA) RMP, consistent with the observed changes in transport. Initial TGLF analysis indicates increased growth rates in this wavenumber range. Finally, observations that core density changes due to RMP can be minimized via adjusting I-coil current are presented.   

RMP Effects on Pedestal Structure and ELMs

Plasma toroidal rotation can prevent reconnection of resonant magnetic perturbation (RMP) fields on rational surfaces and hence magnetic island formation and stochasticity. However, the magnetic flutter induced by RMP fields off the rational surfaces induces a radial electron heat diffusivity $\chi_e^{RMP} \sim (\delta B^2/B_pB_T)\chi_1$ in which $\chi_1 \propto \nu_{Te}^2/\nu_e$ is an effective parallel electron heat diffusivity. The flutter also diffuses electrons radially and causes a factor of about 30 smaller increases in density diffusion, the radial electric field and plasma toroidal rotation. Model effects that may correlate with low collisionality DIII-D RMP experimental data [1] in the pedestal top region include $q_{95}$ resonance windows, importance of bootstrap current in reducing magnetic shear there, and effects of increasing the I-coil current ($\propto\,$$\delta B$): broadening of the resonance window, progressively stronger reductions in the $T_e$ gradient that may stabilize peeling-ballooning instabilities and thereby prevent ELMs, smaller density gradient reductions and slight increases in the radial electric field and plasma toroidal rotation.\par \vskip6pt \noindent [1] T.E.\ Evans, et al., Nucl.\ Fusion {\bf 48}, 024002 (2008).   

The relation between upstream radial widths of $n_e$ and $T_e$ and outer target power width for H-mode discharges in DIII-D

For H mode discharges in DIII-D, the relation between the power width at the outer target, $\lambda_{q_{target}}$, and the radial profiles near the outside midplane is in somewhat better agreement with flux-limited (weak collisionality) electron parallel heat conduction, $q_\parallel \propto n_e T_e^{3/2}$, i.e. $\lambda_{q_{target}}^{flux-lim} = [3/(2\lambda_{T_e}^{up}) + 1/ \lambda_{n_e}^{up}]^{-1}$, than Spitzer (collisional) electron parallel heat conduction, $\lambda_{q_{target}}^{Spitzer} = (2/7) \lambda_{T_{e}}^{up}$. It appears that cross-field transport is the basic controlling process of SOL widths, manifesting itself most directly in upstream widths, with parallel transport and volumetric losses in the SOL/divertor then controlling the relation between upstream and target widths. For an initial data set of three discharges it was found that $\lambda_{q_{target}}^{measured}/ \lambda_{q_{target}}^{flux-lim} = 0.93$, 0.90, 1.00 while $\lambda_{q_{target}}^{measured}/ \lambda_{q_{target}}^{Spitzer} = 1.32$, 0.71, 1.21. Further results will be reported for discharges in upcoming experiments. We find that for DIII-D H-mode shots, the strongest dependence for $\lambda_{q_{target}}^{measured}$ is $I_p^{-1}$. The separate contributions of $\lambda_{n_e}^{up}$, $\lambda_{T_e}^{up}$ to the observed $I_p$ scaling is assessed.   

C$^{2+}$ Flow Measurement in DIII-D Using Coherence Imaging Spectro-Polarimetry

Recently, C$^{2+}$ flows have been measured at DIII-D using a recent, advanced technique in coherence imaging spectroscopy [1]. The diagnostic yields a two dimensional C$^{2+}$ flow measurement over of the entire lower divertor, with approximately $\pm$1~cm resolution in space, and approximately $\pm$2~km/s resolution in velocity. Preliminary analysis of data from a plasma shot during the 2010 campaign yields high flow ($\sim$30~km/s) near the X-point and a flow reversal between the low and high field side of the scrape-off-layer. Results are in good agreement with predictions from the fluid plasma code, UEDGE. A discussion on the fundamental physics behind the flow as indicated by experimental and UEDGE results will be included. Data will likely prove useful in efforts to understand impurity transport processes critical in tokamak operations.\par \vskip6pt \noindent [1] J.~Howard, et~al., Rev.\ Sci.\ Instrum.\ {\bf 81}, 10E528 (2010).   

Comparison of 3-D Modeling With Experimental Results on Fast Wave Antenna Loading in DIII-D

In DIII-D and other tokamaks, with a fixed system voltage limit, the parameter that limits the ICRF power that can be coupled to H-mode plasmas is the antenna loading resistance $R_L$. For a fixed antenna geometry and excitation (phasing), $R_L$ is determined by the electron density profile in the antenna near-field region. Quantitative understanding of the coupling physics is obtained by comparing the resistive ($R_L$) and reactive components of the antenna loading, without and with plasma, to predictions of 3-D models of the antenna and the edge plasma (Microwave Studio and TOPICA). When measured density profiles from reflectometers are used, good agreement between predicted and measured values of $R_L$ is obtained without any adjustable parameters in the model. The improved understanding is applied to enhancement of $R_L$ in advanced scenarios in DIII-D to increase the coupled fast wave power.   

Simulation of Observed EGAM Induced Beam-ion Losses in DIII-D

In DIII-D a non-perturbative beam-driven instability, the energetic particle geodesic acoustic mode (EGAM) is often observed when counter-going beams are injected in the current ramp-up phase of the discharge. The EGAM is an $n=0$ mode in the plasma core with a frequency about 50\% below the ideal GAM and it is an excellent mode to validate codes that are used to predict fast-ion losses in burning plasmas. When the EGAM appears the 2.5 MeV neutron emission drops 10\% to 15\%. This signal is dominated by beam-ion collisions with the thermal plasma and indicates a decrease in the fast-ion population. Recently, accurate measurements have been made of EGAM induced losses with a fast-ion loss detector (FILD) mounted below the mid-plane in DIII-D. We report on state-of-the-art full-orbit following fast-ion loss simulations for the first 50 ms of beam injection using the 3-D beam birth deposition from TRANSP. A very good qualitative agreement between simulation and experiment was found.   

Measuring Kinetic Contributions to Resistive Wall Mode Stability Using Active MHD Spectroscopy

Recent \mbox{DIII-D} experiments showcase the usefulness of active MHD spectroscopy for understanding resistive wall mode (RWM) stability. Plasma response measurements made using this technique have manifested evidence of wave-particle interactions with the trapped ion population now believed to play a critical role in influencing RWM stability above the ideal MHD no-wall limit. The beam ion distribution function can now be modified in DIII-D by tilting the injection angle of one of the neutral beams downward from the magnetic axis. When the toroidal field direction is chosen so that the field line pitch is closely aligned with the trajectories of the off-axis beam neutrals, the trapped ion fraction is reduced. Measurements and comparisons with theory of the effect of off-axis neutral beam injection on RWM stability will be reported.   

Attributes of argon pellet fast shutdowns in DIII-D

Observable spatial and temporal attributes of the initial phase (0-2 ms) of 140 fast plasma shutdowns effected with injection of small ($2\times 10^{20}$ atoms) solid argon cryopellets into lower-single-null and inner-wall-limited ``target" plasmas are surveyed and compared in detail using a database approach. Both the immediate injection effect and subsequent runaway electron (RE) generation are assessed. The resulting data correlations elucidate similarities and differences owed to target configuration and heating method and provide insight into the origins of both early RE generation and loss and subsequent development of RE current plateaus with initial magnitudes of up to 650 kA. Comparison with pellet ablation and RE generation models and implications for fast shutdowns effected in future high-avalanche-gain tokamaks will be discussed.   

Advanced Inductive Plasmas with Low Torque Startup

Experiments on DIII-D have been performed to investigate the access and performance of advanced inductive discharges produced with zero net torque for the full duration of the plasma. With low torque, the plasma is particularly susceptible to 2/1 neoclassical tearing modes, which typically slow and lock, spoiling confinement and ending the high beta phase. Attempts to reduce the residual error field did not appear to significantly modify the stability at low torque. The addition of a modest amount of electron cyclotron heating (1~MW), configured for current drive aimed at the $q=2$ surface, appeared adequate to reduce the mode amplitude sufficiently to allow stable operation at $\beta_N\sim 2.5$ with $q_{95}\sim 4.3$. Details of the stability, confinement and current profile, and comparison with typical high torque advanced inductive discharges will be presented.