Session CO1: DIII-D

Overview of Recent DIII-D Experimental Results

DIII-D experiments have demonstrated the effectiveness of recent upgrades in assessing key fusion science and ITER physics issues. These upgrades include: 1)~the reorientation of a neutral beam to allow co-, counter-, and balanced injection, 2)~the modification of the lower divertor to allow particle exhaust in high triangularity, double-null (DN) configurations, 3)~modification of the current feeds for the toroidal field and 4)~high-bandwidth power supplies for controlling the internal asymmetric coil set. Using these tools, experiments have demonstrated the capability to maintain near zero toroidal rotation, density control in a wide range of plasma shapes, and a reduction in both the intrinsic and corrected error fields. H-mode confinement is observed to decrease slightly as toroidal rotation decreases, yet the best cases with near-zero rotation is better than the scaled confinement needed for ITER ($H_{98y2}=1.2$). Advanced Tokamak experiments have shown the benefit of DN operation in achieving high $\beta$, and edge localized mode suppression in the ITER shape at low collisionality using resonant magnetic perturbations was demonstrated.   

Progress Toward Fully Noninductive, High Performance Conditions in DIII-D

The Advanced Tokamak (AT) program on DIII-D is aimed at developing the scientific basis for steady state, high performance operation in future devices. Recent experiments (since the Long Torus Opening Activities) produced AT discharges using the new high triangularity, lower divertor double-null configuration. The normalized beta exceeded 4.0 transiently and 3.7 (well above no-wall limit) was sustained for over 0.5 seconds. Scanning magnetic configurations shows a clear advantage to a double-null configuration for high performance AT discharges, and significant sensitivity to details of the plasma shape. Optimization of $q$ profiles ($q_{min}$ and $q_0 - q_{min}$) will be carried out to improve MHD stability, bootstrap and noninductive current fractions. Theory based modeling, benchmarked against the experiment, predicts in-principle steady-state conditions with high normalized beta using the new DIII-D capabilities and predicts to a steady state scenario with $Q\geq 5$ in ITER.   

Modeling of DIII-D Discharges With Feedback Control of the Safety Factor Profile Evolution

Closed loop control of the $q$ evolution is implemented by modifying the rate of relaxation of the current density through changes in the conductivity using neutral beam or electron cyclotron power for electron heating. Comparisons between the experimental $q$ evolution and transport code (ONETWO, TRANSP, CRONOS, CORSICA) modeling are being used to develop the physics model of the control process. We find that the neutral beam current drive profile in the simulation must peak near the mid-radius, instead of near the axis as predicted by NBCD theory, in order for the predicted $q$ evolution to match the experiment. Alfv\'en eigenmode type fluctuations observed in the experiment could have provided a mechanism for fast ion transport. A simplified physics model is being developed for use in design of a model-based $q$ evolution controller which will be tested in CRONOS simulations and in the experiment.   

Density Control Using the New Divertor Pumping Configuration in DIII-D

The lower divertor of the DIII-D vacuum vessel has been modified to improve density control in symmetric double-null (DN), high triangularity plasmas. Achieving good density control is an important component in the effort to demonstrate fully noninductive current drive in DN ``AT'' plasmas. With three independently-controlled divertor cryopump systems, particles can be exhausted from both outer divertor targets (lower and upper), as well as from the upper inner target. Particle pumping thus can be maintained, even as the up/down balance in the DN is varied. Experimental results using the new divertor show an additional 7~Torr-l/s is exhausted at moderate density in a balanced DN configuration. This result is in good agreement with a design prediction of 6~Torr-l/s at a similar density. Examples of density control in several plasma operating regimes are presented.   

Study of RWM Stability and RWM Feedback Control in Reactor Relevant Low-Rotation Discharges in DIII-D

Advanced tokamak operation in ITER is expected to require active control of the resistive wall mode (RWM) [1]. After the re-orientation of 1 of 4 neutral beam injectors DIII-D is now ideally suited to address the stability of non-rotating high-beta plasmas. In discharges with balanced NBI heating the RWM becomes unstable at the no-wall stability limit. Only a modest increase of the plasma rotation is sufficient to remain stable beyond this limit suggesting a lower rotation threshold than anticipated from previous experiments, which used non-axisymmetic magnetic fields to lower the plasma rotation. Feedback experiments using an upgraded audio-amplifier system to stabilize the RWM in low-rotation plasmas have begun and comparisons with feedback modeling will be shown.\par \vskip6pt \noindent [1]~Y.\ Liu, et al., Nucl.\ Fusion {\bf 45}, 1131 (2005).   

Decoupling the Effects of Momentum and Power Input From Neutral Beam Injection

At least one reason that plasma rotation remains relatively poorly understood is the limited opportunity to decouple the effects of power and torque, which are simultaneously input into the plasma by neutral beam injection. With the newly rotated counter neutral beam line on DIII-D, we now have the possibility to separate out the two effects and begin to systematically study momentum transport. In particular, three torque scans at approximately fixed ion temperature were performed, by using the DIII-D plasma control system to hold beta constant. We investigate the dependence of the ratio of momentum confinement time to ion thermal confinement time on NBI power and torque. Experiments conducted with balanced beam injection show that immediately after the beams turn on, the rotation is generally in the counter direction, presumably due to the larger prompt torque associated with the counter beams. However, after approximately 350 ms, the rotation changes to the co-direction.   

DIII-D Quiescent H-Mode Experiments With Co Plus Counter Neutral Beam Injection

In many ways, quiescent H-modes are the ideal H-mode plasma. They exhibit H-mode confinement for long duration ($>$4~s or 30~$\tau_E$) with constant density and radiated power. The absence of edge localized modes (ELMs) means no pulsed divertor heat loads. The quiescent edge is also quite compatible with core transport barriers. To utilize QH-mode in future devices, the goals of our recent QH-mode experiments are to develop an improved physics understanding of the QH-mode, especially the ELM stabilization, and to broaden the QH-mode operating space. During the 2006 campaign, we utilized DIII-D's new co plus counter NBI capability to determine how much counter injection is necessary for QH-mode operation. As plasma triangularity is increased, increasing amounts of co-injection can be used while still maintaining the quiescent state. This is consistent with expectations based on peeling-ballooning mode theory. Further experiments are planned to explore the co-counter boundary more thoroughly.   

RMP Assisted ELM-Free H-modes in DIII-D Plasmas With ITER Shapes and Pedestal Collisionalities

Resonant magnetic perturbations (RMP) eliminate ELMs in DIII-D with ITER Similar Shapes (ISS) and pedestal collisionalities. Significant differences in the RMP plasma response are seen in ISS compared to lower average triangularity (LAT) shapes. In ISS, ELM suppression is obtained by increasing the the RMP coil-current/$B_T$ by 50\% which may result from a slight increase in the ISS plasma-coil separation. The divertor recycling increases ~40\% rather than decreasing with $n_{e\, ped}$. No increase in carbon sources (typical of LAT) are seen in ISS and the divertor remains in a flux limited rather than a sheath limited regime although $n_{e\, ped}$ in ISS and LAT are essentially identical. The effectiveness of RMP assisted divertor target plate shielding improves significantly in ISS due to increased particle recycling and a $\sim$100\% increase in the lower divertor radiated power. $\tau_E$ in both ISS and LAT were the same.   

Pedestal Turbulence Evolution After L-H Transition and During ELM Cycles on DIII-D

High-resolution quadrature reflectometer density fluctuation measurements ($\tilde n_e$) have been obtained on DIII-D during H-mode pedestal studies. Initial results indicate that turbulence evolution after the L-H transition and prior to the first ELM appears dependent on pedestal beta ($\beta_{ped}$); in high $\beta_{ped}$ discharges, there is a significant increase in both local $\tilde n_e$ and magnetic turbulence (as observed on magnetic loops) leading up to the first ELM, while the turbulence in low $\beta_{ped}$ discharges shows little change. This $\beta_{ped}$ dependence suggests electromagnetic effects on pedestal turbulence. Later during Type-I ELM cycles two cases are observed: a low level of magnetic turbulence together with an increased $\tilde n_e$ level just after the ELM crash, and high magnetic fluctuation levels with a decrease in $\tilde n_e$ after the ELM crash. Comparisons will be made with electromagnetic turbulence models.   

Understanding the High Confinement Properties of Hybrid Plasmas in DIII-D

Hybrid discharges on DIII-D have remarkably good transport properties, with an energy confinement time that is up to 1.65 times higher than the IPB98(y,2) scaling relation in stationary, high-$\beta_N$ discharges. In hybrid plasmas with a dominant 4/3 NTM, the experimental ion thermal conductivity is equal to the neoclassical value across the plasma cross-section. The electron heat conduction dominates the energy loss process, which is consistent with GYRO simulations that show the ETG mode and TEM cause the majority of transport. Hybrid discharges have been created with the controlling X-point either towards or away from the grad-B drift direction, with similar values of $H_{89P}$ achieved. Using the new ability to inject either co or counter neutral beams into DIII-D, the effect of toroidal rotation on the transport properties of hybrid plasmas is being investigated. Preliminary experiments show that the energy confinement time decreases as counter-NBI is added to an established co-NBI hybrid discharge while keeping $\beta_N$ fixed, which is likely due to the reduced $E\times B$ velocity shear as the toroidal rotation decreases.   

Integrated Scenario Modeling for Steady State and Hybrid Scenario in DIII-D and ITER

Integrated scenario modeling and simulation are carried out for steady-state Advanced Tokamak (AT) and hybrid discharges in DIII-D, and then applied to ITER using the parallelized ONETWO/GLF23 code with particle transport and fast ion diffusion. Time-dependent simulations with GLF23 model for thermal particle transport reveal the complex interactions with the energy transport leading to the strong influence on the current profile evolution in DIII-D AT discharges. Modeling of current profile with ad-hoc assumed fast ion diffusion successfully reproduces the experimentally measured broad current profile with $q_0>1$ in a stationary phase of DIII-D hybrid discharges. The integrated modeling tools validated against DIII-D experiments are applied to ITER, indicating existence of fully noninductive operations at $Q\sim 5$ with \mbox{Day-1} hardware capabilities. Simulations also suggest that high fusion performance with an extended burning duration at $Q\sim 10$ can be achieved with fully penetrated current profile and $q_0>1$ for an ITER hybrid scenario.   

Disruption and Runaway Electron Mitigation With MGI in DIII-D

Past and ongoing disruption mitigation studies in DIII-D employing massive gas injection (MGI) are reviewed and compared with theoretical expectations. Emphasis in the review will be placed on 1)~the gas hydrodynamic delivery considerations that determine the rate of impurity and electron delivery to the plasma edge, 2)~the role of MHD instability and internal reconnection in effecting edge-to-core mixing of the edge-deposited impurities, and 3)~assessment of the mechanism(s) whereby MGI mitigates divertor energy deposition, reduces halo current magnitude and asymmetry and avoids runaway electron production and/or Coulomb-avalanche multiplication. Selected considerations for application of DIII-D MGI results to ITER (wherein time scales for impurity delivery are relaxed relative to DIII-D and other present experiments) will also be addressed.   

Modification of Sawteeth Periods By Trapped Fast Ions in DIII-D

The main auxiliary heating methods for ITER are neutral beam and ion cyclotron wave heating. Sawtooth physics is very important in optimizing the heating efficiency for ITER. This requires understanding of the interaction between fast ions and fast Alfv\'en wave (FW) on MHD stability. Experimentally, the DIII-D discharges have demonstrated strong acceleration of deuterium beam ions above the injected beam energy from measurements of enhanced neutron emissions during FW heating. Theory predicts that high pressure from fast ions in the center of plasma may act as a stabilizing kinetic effect on ideal internal kink mode. However, the DIII-D experimental results showed that sawteeth characteristics strongly depend on a combination of plasma and wave conditions. We apply a Monte-Carlo orbit code (ORBIT-RF) and ideal MHD code (GATO) to model existing DIII-D experiments and explore the triggering and stabilization mechanisms for sawteeth. The analytical model by Bussac and Porcelli will be compared with NOVA-K calculations.   

Neoclassical Tearing Mode Control With Modulated Electron Cyclotron Current Drive in DIII-D

The neoclassical tearing mode (NTM) is a helical island structure in the magnetic field of a high beta tokamak plasma that can degrade confinement and lead to disruption. The DIII-D control system has suppressed the NTM by driving continuous-wave (cw) current at the island using electron cyclotron current drive (ECCD). This requires accurate alignment of the ECCD deposition with the rational q-surface of the island. Methods have been developed to achieve and maintain good alignment. Modulation of ECCD with the island has not previously been attempted since cw ECCD is effective when the deposition region is narrow. However, in ITER the region will be relatively wide. The control system has been recently upgraded to modulate the ECCD so that current is driven only when the island passes by the deposition point. The intent is to test effectiveness of modulated ECCD with a wide deposition region. Key issues are real-time identification of island phase and commands of gyrotron pulses. Validation of the upgraded system and possibly experimental data will be presented.   

Neutral Beam Driven Current During MHD Activity in DIII-D Hybrid Plasmas

The hybrid regime on DIII-D has demonstrated stationary sawtooth free operation for several current diffusion times. However the mechanism for sustaining $q(0)>1$ is still unclear. One possibility for maintaining $q(0)>1$ is the redistribution of neutral beam ions by the helical magnetic perturbation of the mode. The structure and magnitude of MHD activity has been obtained on DIII-D from radially resolved ECE measurements in low density Hybrid plasmas. The measured temperature perturbations are consistent with an off axis 3/2 island together with an even parity component of finite amplitude on the magnetic axis. An effective fast ion radial diffusivity will be inferred from transport analysis of the plasma using the TRANSP code and from ORBIT analysis of the beam ion distribution using a model magnetic perturbation matched to the experimental data.