C. M. Greenfield

Alfven instabilities excited by energetic ions are used as a means to reduce the central magnetic shear in a tokamak via redistribution of energetic ions. When the central magnetic shear is low enough, ballooning modes become stable for any plasma pressure gradient and an internal transport barrier (ITB) with a steep pressure gradient can exist. This mechanism can sustain a steady-state ITB as demonstrated by experimental data from the DIII-D tokamak. It can also produce a shear in toroidal and poloidal plasma rotation. Possible application of this technique to use the energetic alpha particles for improvement of burning plasma performance is discussed.

Measurements of high-frequency oscillations in JET [Joint European Torus], JT-60U, Alcator C-Mod, DIII-D, and TFTR [Tokamak Fusion Test Reactor] plasmas are contributing to a new understanding of fast ion-driven instabilities relevant to Advanced Tokamak (AT) regimes. A model based on the transition from a cylindrical-like frequency-chirping mode to the Toroidal Alfven Eigenmode (TAE) has successfully encompassed many of the characteristics seen in experiments. In a surprising development, the use of internal density fluctuation diagnostics has revealed many more modes than has been detected on edge magnetic probes. A corollary discovery is the observation of modes excited by fast particles traveling well below the Alfven velocity. These observations open up new opportunities for investigating a ''sea of Alfven Eigenmodes'' in present-scale experiments, and highlight the need for core fluctuation and fast ion measurements in a future burning-plasma experiment.

Advanced Tokamak (AT) research in DIII-D [K. H. Burrell for the DIII-D Team, in Proceedings of the 19th Fusion Energy Conference, Lyon, France, 2002 (International Atomic Energy Agency, Vienna, 2002) published on CD-ROM] seeks to provide a scientific basis for steady-state high performance operation in future devices. These regimes require high toroidal beta to maximize fusion output and poloidal beta to maximize the self-driven bootstrap current. Achieving these conditions requires integrated, simultaneous control of the current and pressure profiles, and active magnetohydrodynamic stability control. The building blocks for AT operation are in hand. Resistive wall mode stabilization via plasma rotation and active feedback with nonaxisymmetric coils allows routine operation above the no-wall beta limit. Neoclassical tearing modes are stabilized by active feedback control of localized electron cyclotron current drive (ECCD). Plasma shaping and profile control provide further improvements. Under these conditions, bootstrap supplies most of the current. Steady-state operation requires replacing the remaining Ohmic current, mostly located near the half radius, with noninductive external sources. In DIII-D this current is provided by ECCD, and nearly stationary AT discharges have been sustained with little remaining Ohmic current. Fast wave current drive is being developed to control the central magnetic shear. Density control, with divertor cryopumps, of AT discharges with edge localized moding H-mode edges facilitates high current drive efficiency at reactor relevant collisionalities. A sophisticated plasma control system allows integrated control of these elements. Close coupling between modeling and experiment is key to understanding the separate elements, their complex nonlinear interactions, and their integration into self-consistent high performance scenarios. Progress on this development, and its implications for next-step devices, will be illustrated by results of recent experiment and simulation efforts.

Practical tokamak reactors will need to maintain high ion temperature and nD approximately equal to nT in the core for long durations. These conditions will necessitate low energy transport, low impurity concentrations, and high bootstrap current. Large extrapolations from present experiments are needed to predict performance. A number of approaches are being used for these extrapolations including: (1) dimensionless scaling arguments, (2) empirical scaling, and (3) physics-based simulations of anomalous transport.

Results from recent experiments on the DIII-D tokamak have revealed many important details on transport barriers at the plasma edge and in the plasma core. These experiments include: (a) the formation of the H-mode edge barrier directly by pellet injection; (b) the formation of a quiescent H-mode edge barrier (QH-mode) which is free from edge localized modes, but which still exhibits good density and radiative power control; (c) the formation of multiple transport barriers, such as the quiescent double barrier (QDB) which combines an internal transport barrier with the quiescent H-mode edge barrier. Results from the pellet-induced H-mode experiments indicate that: (a) the edge temperature (electron or ion) does not need to attain a critical value for the formation of the H-mode barrier, (b) pellet injection leads to an increased gradient in the radial electric field, E_r, at the plasma edge; (c) the experimentally determined edge parameters at barrier transition are well below the predictions of several theories on the formation of the H-mode barrier, (d) pellet injection can lower the threshold power required to form the H-mode barrier. The quiescent H-mode barrier exhibits good density control as the result of continuous magnetohydrodynamic activity at the plasma edge called the edge harmonic oscillation (EHO). The EHO enhances the edge particle transport whilst maintaining a good energy transport barrier. The ability to produce multiple barriers in the QDB regime has led to long duration, high-performance plasmas with β_NH₈₉ values of 7 for up to 10 times the confinement time. Density profile control in the plasma core of QDB plasmas has been demonstrated using on-axis electron cyclotron heating.

The quiescent double barrier (QDB) regime is a high performance regime recently identified in DIII-D and characterized by a double transport barrier structure (core and edge) that can be maintained for several seconds, often limited only by the pulse length capabilities of the DIII-D hardware. The QDB regime has been sustained for up to 25 τ_E with fusion performance of up to β_NH₈₉≈7. The edge barrier is ELM-free, but modulated by low frequency MHD activity that allows density control via an external cryopump. The core barrier is similar to those seen in previous internal transport barrier experiments, but is maintained without complete stabilization of turbulence. Instead, the turbulence correlation lengths become very short so as to minimize the transport length scales. The two barriers are separated by a region of high transport that is a consequence of a zero-crossing in the E×B shearing rate. These discharges typically possess highly peaked density profiles. This has several implications: narrow bootstrap current profile, reduced beta limit and increased impurity retention. We will report on studies of each of these issues.