T. Sunn Pedersen

Dipole equilibrium and stability

A plasma confined in a dipole field exhibits unique equilibrium and stability properties. In particular, equilibria exist at all beta values and these equilibria are found to be stable to ballooning modes when they are interchange stable. When a kinetic treatment is performed at low beta, a drift temperature gradient mode is also found which couples to the MHD mode in the vicinity of marginal interchange stability.

Pedestal profiles and fluctuations in C-Mod enhanced D-alpha H-modes

High resolution measurements on the Alcator C-Mod tokamak [I. H. Hutchinson et al., Phys. Plasmas 1, 1551 (1994)] of the transport barrier in the “Enhanced Dα” (EDA) regime, which has increased particle transport without large edge localized modes, show steep density and temperature gradients over a region of 2–5 mm, with peak pressure gradients up to 12 MPa/m. Evolution of the pedestal at the L-H transition is consistent with a large, rapid drop in thermal conductivity across the barrier. A quasi-coherent fluctuation in density, potential, and Bpol, with f0∼50–150u200akHz and kθ∼4u200acm−1, always appears in the barrier during EDA, and drives a large particle flux. Conditions to access the steady-state EDA regime in deuterium include δ>0.35, q95>3.5, and L-mode target density ne>1.2×1020u200am−3. A reduced q95 limit is found for hydrogen discharges.

Diagnosing pure-electron plasmas with internal particle flux probes

Techniques for measuring local plasma potential, density, and temperature of pure-electron plasmas using emissive and Langmuir probes are described. The plasma potential is measured as the least negative potential at which a hot tungsten filament emits electrons. Temperature is measured, as is commonly done in quasineutral plasmas, through the interpretation of a Langmuir probe current-voltage characteristic. Due to the lack of ion-saturation current, the density must also be measured through the interpretation of this characteristic thereby greatly complicating the measurement. Measurements are further complicated by low densities, low cross field transport rates, and large flows typical of pure-electron plasmas. This article describes the use of these techniques on pure-electron plasmas in the Columbia Non-neutral Torus (CNT) stellarator. Measured values for present baseline experimental parameters in CNT are phi(p)=-200+/-2 V, T(e)=4+/-1 eV, and n(e) on the order of 10(12) m(-3) in the interior.

Experimental demonstration of a compact stellarator magnetic trap using four circular coils

An experimental demonstration of a compact stellarator magnetic trap created from four circular coils is presented. The coil manufacturing and assembly tolerances were on the order of 0.5–1%, far less stringent than most other stellarators. The simplicity, loose mechanical tolerances, and low cost of the trap design makes it feasible for stellarators to be used for a variety of novel physics experiments, in addition to their present use for magnetic confinement fusion. The experiment, the Columbia Non-neutral Torus, has several other desirable features such as no significant internal island chains and the lowest aspect ratio, A⩽1.9, of any stellarator built to date.

Experiments and modelling of external kink mode control using modular internal feedback coils

We report on recent advances in modelling and experiments on resistive wall mode feedback control. The first experimental demonstration of feedback suppression of rotating external kink modes near the ideal wall limit in a tokamak is described [1]. This was achieved using an optimized control system employing a low latency digital controller and directly coupled modular feedback coils. The magnitude of plasma dissipation affecting kink mode behaviour has also been experimentally quantified for the first time using measurements of the radial eigenmode structure of the poloidal field fluctuations associated with the rotating kink mode. New capabilities of the VALEN code [2] are also reported. These include the ability to simulate multiple plasma modes and mode rotation in the model of the feedback control loop. Results from VALEN modelling of resistive wall mode feedback control in ITER are also presented, showing a significant improvement in performance with internal coils. Evidence for a lack of mode rigidity in HBT-EP is given, and plans to address this and other issues related to coil coverage and coil modularity are presented.

A digital control system for external magnetohydrodynamic modes in tokamak plasmas

A feedback system for controlling external, long-wavelength magnetohydrodynamic activity is described. The system is comprised of a network of localized magnetic pickup and control coils driven by four independent, low-latency field-programable gate array controllers. The control algorithm incorporates digital spatial filtering to resolve low mode number activity, temporal filtering to correct for frequency-dependent amplitude and phase transfer effects in the control hardware, and a Kalman filter to distinguish the unstable plasma mode from noise.

Measurements of large poloidal variations of impurity density in the Alcator C-Mod H-mode barrier region

Simultaneous high resolution measurements of the soft x-ray emission at the top and outboard edge of the Alcator C-Mod tokamak [I. H. Hutchinson et al., Phys. Plasmas 1, 1511 (1994)] are presented. Impurity density profiles are derived from these measurements. The impurity density varies considerably on a flux surface in the high confinement mode (H-mode) pedestal region. The soft x-ray and impurity density pedestals are not on the same flux surface at the top and outboard edge. The pedestal widths are consistently larger at the outboard edge in the Enhanced D-alpha (EDA) H-mode, suggesting a ballooning-like character of the EDA quasi-coherent mode. The usual separation of time scales between radial and parallel impurity transport does not hold in the H-mode pedestal region of Alcator C-Mod. Thus, one should not necessarily expect that the impurity density be constant on a flux surface in this region.

Studies of the neoclassical transport for CNT

The original optimization of the Columbia Nonneutral Torus (CNT) considering only volume (and error field resilience) was also successful in optimizing the stored energy. To assess the general confinement properties of a device, studies of the 1/ν neoclassical transport (effective ripple eff) are important. For CNT the field line tracing code NEO [1] is used to compute eff. NEO is used by the code SORSSA [2, 3] for computation of the total stored energy based on neoclassical transport.

Numerical studies of transport in the Columbia Non‐neutral Torus

The confinement of pure electron plasmas in the Columbia Non‐neutral Torus (CNT) stellarator is limited by the presence of unconfined orbits. The existence of a very large electric field across magnetic surfaces should preclude such unconfined orbits. However variations in the electric potential on magnetic surfaces, inherent to the CNT equilibrium, add to the complexity of the trajectories and lead to bad orbits. We have written a code using magnetic coordinates to integrate the electron drift trajectories in the electric and magnetic fields expected in CNT equilibria. Results of such calculations are presented showing that there exists unconfined orbits in CNT if the potential is not constant on surfaces.

Confinement of pure electron plasmas in the CNT stellarator

The Columbia Non‐neutral Torus is a stellarator devoted to non‐neutral and electron‐positron plasma research. Confinement and transport processes have been studied in some detail now, and an understanding of these processes has emerged. Transport is driven in two ways: The presence of internal rods, and the presence of neutrals. Both transport processes are clearly distinguished experimentally, and a model of the rod driven transport has been developed, yielding very good agreement with experimental data. The neutral driven transport is faster than originally expected and indicates the presence of unconfined orbits in CNT. Numerical modeling of the electron orbits in CNT confirms the existence of loss orbits and shows that a flux surface conforming electrostatic boundary will greatly improve confinement. Such a boundary has now been installed in CNT, with initial results showing an order of magnitude improvement in confinement.